NUCLEAR CHANGES ASSOCIATED WITHregeneration, limb regeneration and activation of division in blood lymphocytes. The cytological changes in nuclear structure could be correlated with

J. Cell Set. a8, 87-105 (1977) 87Printed in Great Britain © Company of Biologists Limited igy7

NUCLEAR CHANGES ASSOCIATED WITH

CELLULAR DEDIFFERENTIATION IN

PEA ROOT CORTICAL CELLS CULTURED

IN VITRO

LEWIS J. FELDMAN* AND JOHN G. TORREYfDepartment of Biology, Harvard University, Cambridge, Massachusetts 02138, U.S.A.

SUMMARY

Root cortical tissues explanted from seedling roots of the garden pea, Pimm sativiun L.,cv. Little Marvel cultured in sterile medium containing indoleacetic acid and kinetin werefixed, sectioned and studied with the electron microscope. Nuclear changes over 60 h of culturewere examined during the events of dedifferentiation of mature parenchyma cells into sub-divided newly meristematic derivatives. Table 1 summarizes the events which are evident in7 population classes designated I—VII. The initially small, round nucleus with a single,unvacuolated nucleolus and highly condensed and disperse chromatin showed marked volumeincrease, nucleolar enlargement of almost 20-fold, development of diffuse and then clumpedchromatin and then subdivision of the nucleoli into a number (5-10) of small nucleoliimmediately preceding cell division. Mitoses involved cells with diplochromosomes and daughternuclei with doubled chromosome number, either 471 or 8n. Subdivided cortical cells showedan increased ratio of nucleus to cytoplasm although the nuclei were reduced in volume to thatof the original nuclear population. Nucleoli in subdivided cells were small and unvacuolated andchromatin changed from an initially condensed state through a diffuse and disperse conditionto a condensed form. Cortical cells divided at least twice in rapid succession before initiatingnew events leading to redifferentiation. Nuclear changes in dedifferentiation in pea were com-parable in many respects to the similar process studied in animal systems including eye lensregeneration, limb regeneration and activation of division in blood lymphocytes. The cytologicalchanges in nuclear structure could be correlated with documented changes in DNA, RNA andprotein levels in these systems. By 60 h cells presumed to be ready for redifferentiation andcell specialization were observed in the subdivided population but structural evidence forcommitment to a new course of cytodifferentiation was not obtained.

INTRODUCTION

The early events associated with the differentiation of a totipotent cell into a par-ticular cell type are only vaguely understood. In part, this problem arises because ofthe difficulty in specifying precisely when a cell is set on the course of cyto-differentiation. For this reason, it is often impossible to determine exactly when andto which cytological and biochemical events a particular cell owes its origin. Ininvestigations of cytodifferentiation therefore, clear advantage often can be gained by

• Mailing address: Department of Botany, University of California, Berkeley, California,94720, U.S.A.

f Permanent address: Cabot Foundation, Harvard University, Petersham, Massachusetts,01366, U.S.A. (Please send reprint requests to this address.)

88 L. J. Feldman and J. G. Torrey

employing a system in which one induces an already determined cell type to differ-entiate into an entirely different cell product.

In both plants and animals, examples exist in which an already determined cell canbe induced experimentally to dedifferentiate and subsequently to redifferentiate intoa new cell type. During the course of this cytodifferentiation numerous cytologicaland metabolic processes must occur. Included in these processes are the loss of thecharacteristics which distinguish a particular cell type, and the assumption of processes(cytological, biochemical) which mark the cell as ^differentiating into a new cell type.Thus, one is dealing with 2 distinct, but not necessarily separable events, the reversionof a mature differentiated cell to a meristematic state and the reactivation of thededifferentiated cell into an entirely new course of cytodifferentiation.

In plants one of the most carefully studied systems employs cultured pea rootcortical explant tissue. In this experimental system a large proportion of the popu-lation of already mature, differentiated cortical cells is induced to dedifferentiate andsubsequently redifferentiate into tracheary elements. The trigger for these processesare hormones: i.e., auxin and cytokinin, which, when present in the tissue culturemedium, initiate a set of events culminating in the formation of tracheary elementsfrom subdivided cortical cells.

Some of the biochemical and cytological processes associated with the differ-entiation of tracheary elements in pea root cortical explant tissue have been documentedin recent papers with special emphasis on DNA synthesis and related nuclear events(Torrey & Fosket, 1970; Libbenga & Torrey, 1973: Phillips & Torrey, 1973). At thelight-microscope level the time course of this cytodifferentiation involves a number ofdistinct phases. The cortical cells with interphase nuclei at 2C and 4 C levels of DNA,initially all diploid, begin DNA synthesis 24 h after subculture. As a result of thissynthesis, several discrete populations, most of which are polyploid, are producedprior to the first observed mitoses. The first cells to divide are usually tetraploid,although thereafter the level of ploidy may vary in populations of dividing cells, i.e.,4« and 8n. Five to seven days after culturing the explant, secondary cell wallscharacteristic of tracheary elements are observed in derivatives of recently dividedcortical cells as the first morphological evidence of redifferentiation.

At the electron-microscope level Bowes & Torrey (1976) examined early ultra-Structural changes in these cell populations, including in particular, changes in the cellwalls. In explants as old as 72 h in culture they were unable to observe any directevidence that divided cortical cells were yet committed to differentiate into trachearyelements.

In the work presented here, ultrastructural observations were extended to includea detailed study of nuclear morphology. The distinct nuclear populations as revealedby the electron microscope were evaluated in the light of autoradiographic and micro-spectrophotometric evidence concerning the onset and magnitude of DNA synthesisin differentiating cortical cells and in studies correlating ultrastructural images ofnuclei with those at the light-microscope level specially stained to show nucleolarstructure.

Nuclear changes in cultured pea root cells 89

MATERIALS AND METHODS

Seed of Pisum sativum, cv. Little Marvel, were surface sterilized in commercial Chlorox(sodium hypochlorite solution), rinsed 3 times in sterile distilled water and allowed to imbibe indistilled water in the dark for 16 h. The seeds were then placed aseptically with the radicle endup in Petri plates containing o-8 % agar dissolved in distilled water, and germinated in the dark.From roots 15-25 mm in length segments 1 mm thick were cut 10—11 mm proximal to theroot tip. The vascular cylinder was removed by the technique described by Libbenga & Torrey(1973) and the cortical explants were transferred aseptically to Petri plates containing SzMmedium supplemented with 10 mg/1. kinetin and o-8 % agar. The explants were cultured inthe dark at 25 °C.

Electron microscopy

Segments were fixed in either 3 % glutaraldehyde for 25 h at room temperature or ato °C in 10 % acrolein for 48 h, washed and subsequently postfixed in 2 % osmium tetroxide for2 h at room temperature or for 24 h at o CC; all solutions were prepared with 0025 M sodiumphosphate buffer, pH 6-8. Segments were dehydrated initially through a graded acetone seriesand then through a graded absolute acetone-propylene oxide mixture and embedded in Epon-Araldite (Phillips & Torrey, 1974). Sections were cut on a Reichert ultramicrotome and werestained in 2% uranyl acetate for 30 min and in lead citrate for 4 min (Reynolds, 1963), thendried and examined with a Philips 300 microscope.

Symbols used on electron micrographs

cCC

cpcwcyerf8is

chromatincondensed chromatincell platecell wallcytoplasmendoplasmic reticulumflbrillargranularintercellular space

/mnewnmnonunvV

lacunaemitochondrionnew cell wallnuclear membranenucleolus organizernucleolusnucleolar vacuolevacuole

Calculations of nuclear and nucleolar volumes from squashes prepared for light mict oscopy

Segments were fixed for 24-48 h in Craf II, washed overnight, hydrolysed in I N HC1 at60 °C for 8 min, stained in Schiff reagent in the dark for 35 h and rinsed 3 times, 10 min each,in SO2 water. The explant tissue was then rinsed and treated for 3 h at room temperature with°'3 % pectinase in 02 M sodium acetate buffer, pH 45. Segments were then squashed onalbumin-coated slides, heated slightly and the coverslips floated off in 50% ethanol. Coverslips,to which the squashed segments were affixed, were rinsed briefly in water, air dried, andcounterstained in 015 % methylene blue (aqueous) so that nuclear and nucleolar detail couldbe distinguished. In calculating nuclear and nucleolar volumes, nuclei and nucleoli were takenas spherical, their volumes being calculated as 4/3 nrs. The radius, r, was obtained by measuring2 diameters at right angles and halving the average value, since neither the nucleus nor nucleoluswas perfectly spherical. The values obtained are therefore somewhat approximate.

\O

Tab

le I

. Su

mm

ary of

nucl

em e

vent

s du

ring

cul

ture

o

Po

stu

late

d

DN

A a

mo

un

ts

Dim

ensi

ons

and

sta

te o

f (b

ased

on

T

ime

r A

-

data

fro

m

Pop

ula-

in

N

ucle

us

Nuc

leol

us

Dis

tinc

tive

L

ibb

eng

a &

Il

lust

rate

d

tion

cu

ltur

e, ,

A 7 7

A

.-

,

Nuc

lear

u

ltra

stru

ctu

ral

To

rrey

, in

fig

ure

no.

h D

* vt

D

+

vt

appe

aran

ce

1973

) no

. va

cuol

es

Sm

all

elon

gate

, ap

prea

sed

to

cell

wal

l

Sti

ll f

lat,

ap

pres

sed

to

wal

l. D

oubl

ed

size

Lar

ge,

appr

esse

d to

wal

l

Nuc

leus

to

cen

tre

of

cell

. L

arge

, di

ffus

e, o

void

< 2

17

S

ingl

e,

Sm

all

or

Ch

rom

atin

hig

hly

con-

2

C -+

4

C

I, 2

; I

5

rou

nd

, ab

sen

t;

dens

ed a

nd

dis

per

se i

n sm

all

smal

l nu

cleu

s. N

ucle

olus

sh

ow

s P

la

cuna

e b

oth

gra

nula

r an

d f

ibri

llar

4

zone

s. N

ucle

olus

org

aniz

er

m %

evid

ent

g

4' 1

29

9 S

ingl

e S

mal

l or

C

hro

mat

in m

ore

dif

fuse

; D

NA

3

;16

,17

5

enla

rged

ab

sen

t;

gran

ular

an

d f

ibri

llar

sy

nthe

sis

nu

mer

ou

s zo

nes

inte

rmin

gled

. be

gins

8

lacu

nae

Nuc

leol

us o

rgan

izer

24

h

evid

ent

. .

5'1

434

Sin

gle,

lar

ge

Pre

sen

t,

Ch

rom

atin

clu

mp

ed a

nd

S

om

e 4

; 18

-20

'

larg

e an

d

peri

pher

al ; n

ucle

olus

ca

ses

cent

ral

larg

e an

d v

acuo

late

d. N

o

invo

lvin

g nu

cleo

lus

orga

nize

r se

en

end

ore

du

pli

-

7 9

cati

on

2.5

354

Fro

m s

ingl

e M

ult

iple

C

hro

mat

in c

lum

ped

an

d

8C

an

dlo

r 5

-8;

to m

ult

iple

p

erip

her

al;

nucl

eolu

s I 6C

21

-23

and

smal

l w

ith

lar

ge o

ute

r g

ran

ula

r (5

-10)

to

an

d s

mal

ler

inn

er f

ibri

llar

di

sapp

eari

ng

zone

s, w

ith

vac

uole

s.

Nuc

leol

us o

rgan

izer

ass

oci-

at

ed w

ith

per

iph

ery

of

the

nucl

eolu

s

+ D

iam

eter

, pm

. t

Vol

ume,

pm

a.

Niulem changes in cultured pea root cells


RESULTS

Electron microscopy of the nucleus

Nuclear morphology was examined at time o immediately after explanting and at42 and 60 h after explanting. At 42 h the population as a whole showed peak[3H]thymidine incorporation activity (in parallel studies, 60% of the cortical cellnuclei showed incorporation after a 6-h-label period); at 60 h the population showeda peak mitotic index (~ 6-5 % of the cells were in mitosis. See also fig. 7 in Phillips &Torrey, 1973). For clarity in discussions seven populations of cells were designated,related to the time in culture and to a presumed sequence of change. The observationson each population are described below and summarized in Table 1.

Table 2. Nuclear dimensions in the 7 cortical populations of Table 1 determined fromlight-microscopic measurements of fixed squash preparations stained with Schiff's reagentcounterstained with methylene blue

Popula-tion

III

IIIIV

VVI

VII

Timein

culture, ,h

0

42

4260

606060

Each

Radius

High

6-6io-o1 2 5

1 2 9

5-87 98-3

sample is

of nucleus, /tin

Low

4-2

7 91 0 8I I - 2

^Average

5-i9 4

" • 51 2 4

First mitosis; 2 nuclei4-26 26 2

based on

5 i7-17-1

Volume of nucleus,1

High

1200

41198200

9000

8152065

2420

A

Low

3 1 2

206552805880

3 1 010001000

measurements of 5-6 nuclei.

/im3

Average

542342063708100

5561500

1500

Time o (population I nuclei)

In the cultured explant at time o, the cortical cells are large, averaging 40 x ioo/tmand possessing an extensive vacuole and a thin parietal cytoplasm (cf. Bowes &Torrey, 1976). The nuclei are small, flattened and elongate, appressed to the cell wallwith an average diameter of about 5 /tm (Table 2). The nuclear membrane is smoothand unlobed. As noted by Lafontaine & Lord (1974) and by Chaly & Setterfield(1975), nuclei in cortical cells of pea are composed of an extensive network of highlycondensed chromatin* often attached to the nuclear membrane and intermingled withless-dense regions in the nucleoplasm. Each nucleus contains a single, round nucleolus,1-6-2-0/(m in diameter, only rarely associated with the chromatin (Figs. 1, 15). Thegranular zone of the nucleolus is peripheral and fairly extensive, comprising approxi-mately one-half of the nucleolar volume. Internal and surrounded by the granularzone is a distinct and large fibrillar zone, occupying usually one third to one half of

• Because of problems of definition (see Chaly & Setterfield, 1975), the terms hetero-chromatin and euchromatin will not be used.


the nucleolar volume (Fig. 2). Small lacunae are occasionally noted within the fibrillarzone. Nucleolar vacuoles, that is, lighter-staining regions within the granular material,are lacking. The nucleolar organizer region is usually evident, most often near theperiphery of the nucleolus.

In their measurements of relative DNA amounts in nuclei of time o pea rootcortical cells, based on total fluorescence of DNA-specific dye in whole nuclei studiedin sectioned material, Libbenga & Torrey (1973) found 2 distinct classes of nucleiwith respect to their DNA content. One class was interpreted as 2C and the other as4C. No nuclear dimensions were measured in these experiments. We have used thesquash method described in Methods to prepare time o cortical cells to try to dis-tinguish 2 nuclear classes with respect to nuclear radius or volume. Measurements ofover one hundred nuclei gave a single bell-shaped curve with a continuous range ofnuclear radii varying from 3-6 to \2-o fim, average 7-6/jm. This value, which rep-resents one class discernible by radial dimensions, agrees well with the dimensionsderived from the much smaller samples given in Table 2. It must be presumed there-fore that DNA values ranging from 2C to 4C are subsumed within this single volumeclass.

Nuclei in explants cultured for 42 h {populations I-III)

In explants cultured for 42 h 3 distinct types of nuclei, designated populations I,II and III were observed. In most respects nuclei of population I resembled the time onucleus and are assumed to be unresponsive to culture. Nuclei of population II haveenlarged considerably with an increase in nuclear and nucleolar diameters approxi-mately twice that observed in the time o nuclei (see Figs. 16, 17 and Tables 1 and 2).Such nuclei are more lobed, but still adjacent to the cell wall. Chromatin is dispersedand no longer found in dense clumps as at time o. Occasional connexions are observedbetween the nucleolus and the chromatin. A peripheral, not very compact granularzone is present in the nucleoli (Fig. 3). Internally, fibrillar and granular materials havebecome intermingled. An increase in the number of small lacunae, almost exclusivelyassociated with the fibrillar zone are observed in population II (Fig. 3). Occasionalsmall vacuoles are noted within the granular zone and nucleolar organizer regions arenoted frequently, usually in a somewhat more interior position than in the time o nuclei.

Nuclei of population III are readily distinguished by the presence of a largenucleolar vacuole (Figs. 4, 18-20). The nuclei are larger than those of population II(Table 2) and less regular in shape. The chromatin, occasionally seen as large denseclumps, is peripheral in distribution, sometimes attached to the nuclear membrane,but is rarely in direct contact with the nucleolus. Nucleoli are round to slightly ovoidwith extensive granular and fibrillar zones. Occasionally the granular zone extends asa continuous band from the exterior to the interior of the nucleolus. Small lacunae aredistributed throughout the fibrillar material. Within the central vacuole, which com-prises approximately one third to one half of the diameter of the nucleolus, arenumerous small granular particles, sometimes appearing connected in chains. Nu-cleolar organizer regions are not observed in nucleoli containing a large central vacuole.

L. J. Feldman and J. G. Torrey

A7/7if

Fig. i. Nucleus at the time of explanting (population I). The chromatin is highlycondensed. The nucleolus is relatively small, x 7850.Fig. 2. Nucleolus of a population I nucleus. Note that there is little interminglingbetween peripheral granular and interior fibrillar zones, x 30200.Fig. 3. Nucleus 42 h after explanting (population II). Note the numerous lacunae inthe nucleolus and the highly intermingled nature of the granular and fibrillar zones.X6170.

Fig. 4. Nucleus 42 h after explanting (population III). Note the well developed nucleo-lar vacuole surrounded by lacunae and the highly intermingled nature of the fibrillarand granular zones in the nucleolus. x 9050.


. nm vnm

8Fig. 5. Nucleolus of an undivided cell 60 h after explanting (population IV). Notethe greatly enlarged nucleolus and nucleolar vacuole. The chromatin is typicallyclumped and peripheral. X5150.

Fig. 6. Enlarged portion of Fig. 5 showing detail of the nucleolus organizer. Note theextensive peripheral granular zone of the nucleolus. x 27650.

Fig. 7. Nucleus of an undivided cell, 60 h after explanting. Note the multiple nucleoli,each with a halo of granular particles and a compact central fibrillar zone. Thechromatin is more uniform in distribution; compare with Fig. 5. This nucleus isprobably preparing for its first mitosis, x 6700.

Fig. 8. Enlarged portion of the nucleus of an undivided cell. Arrows point to possibleremnants of the nucleoli. x 11 200.


Nuclei in explants cultured for 60 h [populations IV-VII)

This tissue is comprised of several discrete populations of cells including somefitting the descriptions of populations I—III and additional cell types, designatedpopulations IV-VII. Included are cells which have not yet divided, but with markednuclear changes, cells which have divided once, and cells which have divided twice,subdividing the original cortical cell into 4 smaller cells. By comparing the thickness ofcell walls, it is possible to determine how many divisions preceded the formation ofany given cell as was shown by Bowes & Torrey (1976).

Within the populations of undivided cells which responded to the culture treat-ment, population IV is a broad category which includes cells with a single largenucleolus, cells with multiple small nucleoli, and cells about to enter cell division withaccompanying disappearance of the nucleolus. Nuclei of some of these cells havereached their greatest measured size (population IV, Table 2), increasing 15-16 timesin nuclear volume as compared to the time o nuclei (Figs. 21, 23). The nuclei are nolonger appressed to the cell wall, but rather are located in a more central positionwithin the cell (Fig. 5). The chromatin is seen as discrete, dark-staining massesperipheral in distribution, often in contact with the slightly lobed and undulatenuclear membrane. The enlarged nucleolus consists primarily of numerous zones ofinternal fibrillar material bounded by a peripheral granular zone. Within the granularzone, which comprises approximately two-thirds of the nucleolus, are several largeunequal-sized vacuoles. Eachvacuole contains dark-staining granular material (Fig. 6).Numerous small lacunae are evident within the fibrillar zones. Nucleolar organizersare evident as spherical bodies in direct contact with the nucleolus proper (Figs. 5, 6).Material comprising the nuclear organizer is more fibrillar in nature when comparedto the distinctly granular zones around the large nucleolar vacuoles and appears to bepermeated by small channels.

In the undivided, responsive cell at a slightly later stage but still in population IVthe large nucleolus has dissolved into 5-10 small nucleoli, with rough, irregularedges, and apparently composed mainly of fibrillar material, bounded by a halo ofgranular particles (Figs. 7, 22, 23). The nucleus with its nuclear membrane still intactbecomes slightly ovoid and more or less lobed. The chromatin condenses into scattered,dark-staining patches and is no longer peripheral in distribution. In some cells, afterthe nucleoli have completely disappeared, small circular bodies can be observed(Fig. 8), which may be remains of the nucleoli.

In Fig. 13 is seen a cortical cell in early prophase of its first division. The nuclearenvelope has broken but is still apparent at the periphery of the nuclear area. Thecondensed chromosomes are cut in section and show a striking paired arrangement,probably demonstrating their diplochromosomal nature. No specialized structuresrelated to this endoreduplicative origin are evident other than the pairing. Atmetaphase, these chromosomes will separate, giving rise to the tetraploid chromosomenumber. Here, about 25 of the total number of 28 chromosomes are seen in section.

Nuclear changes in cultured pea root cells

Fig. 9. Nucleus of a once-divided cell, 60 h after explanting (population V). Note theenlarged nucleolus, and highly condensed, discrete clumps of chromatin. x 3520.Fig. 10. Sister nuclei, the result of the first division of a cortical cell (population V).Note the highly compact nature of the nucleoli and the absence of vacuoles or lacunae.X7650.Fig. 11. One of the daughter nuclei of a once-divided cortical cell (population VI).Compare this nucleus with those in Fig. 10. x 4670.Fig. 12. Daughter nucleus of a once-divided cortical cell. It may be preparing toundergo a second mitosis (population VII). Compare with Fig. 7. x 5900.

L. J. Feldman and J. G. Torrey

I nm 13

cwcw

V

14Fig. 13. Metaphase in a nucleus undergoing its first mitosis. This section cuts portionsof about 25 chromosomes, x 6050.Fig. 14. Sister nuclei typical of a subdivided cortical cell after completion of the newcell wall. The condensed chromatin suggests that these nuclei are relatively inactive,x 4670.

Cells at 60 h which Itave divided [populations V-VII)

In the once-divided cells (population V) the nucleus is more or less circular insection and remains centrally located in the cell adjacent to the newly formed cell wall(Figs. 9, 10). The nuclear diameter and volume may have been reduced to that of theoriginal size seen in population I (Table 2) or may be intermediate in size. In suchnuclei the chromatin is distributed at the periphery of the nuclear envelope as dark,

Nuclear changes in cultured pea root cells m

26Figs. 15-26. Squashes of whole nuclei from explants cultured 48 h, stained with theFeulgen method and methylene blue. All x 1450. In Figs. 22 and 23 note the multiplenumber of nucleoli per nucleus. The 2 sister nuclei in Fig. 26 (arrows) are products ofeither the first or second mitosis of a polyploid cortical cell.

i oo L. J. Feldman and J. G. Torrey

highly condensed masses. Occasional contact is observed between the chromatin andthe nuclear membrane and/or peripheral zones of the nucleolus. The nucleoli,generally one per nucleus, are lobed or somewhat irregular in shape and may occupya relatively large portion of the nucleus. A peripheral granular zone is usually observed,occasionally completely encircling the more internal fibrillar matrix. In the interiorof the nucleoli the granular and fibrillar zones may or may not be highly intermingled.Nucleolar lacunae are small and numerous and often contain a lighter-staining materialwhich may be chromatin.

Such cells may enter promptly into cell division again so that in 6o-h samples,cortical cells may have divided once or twice (cf. figs. 11,13 m Bowes & Torrey, 1976).Nuclei immediately after such divisions give the appearance described as population V.In Fig. 10 are shown 2 nuclei with the new cell plate just forming between them. Eachcontains a single, non-vacuolated nucleolus and chromatin-rich nucleoplasm. Althoughit is difficult to be certain from static sampling (as compared to time-lapse cine-micrography, for example), it is probable that such nuclei in just-divided cells undergothe sequence described in Table 2, going on to form cells of populations VI and VII.The nuclei undergo a diameter and volume increase (population VI) and are oftencharacterized by the presence of a single large nucleolus with a large central vacuole(Figs. 11, 24, 25). Chromatin becomes more diffuse and disperse. Nucleoli appearcomposed largely of fibrillar material. The new cell wall is well formed and the ratioof nucleus to cytoplasmic volume is relatively high.

Immediately prior to the next cell division, these new meristematic cells arecharacterized by the presence of multiple nucleoli (Fig. 12), a phase which appears toprecede nuclear division, repeating the sequence observed in populations III and IVprior to the first division. If the cells do not divide further, they arrest at the stage VIcondition with nuclei containing only a single large nucleolus with a large vacuole(Fig. 14). Fig. 26 may be compared to the cells from a first division (Fig. 10) or mayresult from a second division, a fact not determinable from squash preparations.Fig. 14 represents subdivided cortical cells after either the first or second division inwhich the nuclei give the appearance of relative inactivity, i.e., condensed chromatinand large vacuolated nucleoli. Such cells are reminiscent of population I cells but maybe cells prepared now to redifferentiate. Evidence for such a sequence remains to beobtained.

DISCUSSION

Cortical cells before division

In undivided cortical cells, a pronounced enlargement of the nucleus and itsnucleolus is one of the first and most obvious indications that a cell is responsive tohormonal stimulus. Phillips & Torrey (1973) showed that approximately 60% of theinitial explant population responds to excision and culture by undergoing DNAsynthesis. This observation, which has been confirmed in detail repeatedly, cor-responds well with what is noted at the ultrastructural level. In the nuclei of enlarging,undivided cells the abundant, condensed chromatin of a typical time o nucleus


disappears with a concomitant increase in the regions of diffuse chromatin. Variousworkers (Tokuyasu, Madden & Zeldis, 1968) have shown that it is within the loosened,highly dispersed state of chromatin that the most active DNA synthesis occurs. Inseveral plant systems (Deltour & Bronchart, 1971; Jordan & Chapman, 1971, 1973;Lafontaine & Lord, 1974) as well as in a variety of animal systems (Hay, 1959;Tokuyasu et al. 1968) the decondensation of the chromatin is associated with earlyand rapid synthesis of DNA. Thus, it appears likely that renewed DNA synthesis incortical cells is at least in part a consequence of the dispersal of the chromatin, as hasalready been suggested in other reactivated tissues (Dumont & Yamada, 1972; Harris,1967).

Since it is known from [3H]thymidine incorporation data that DNA synthesis incortical explant cells does not commence until approximately 24 h after subculturing(Phillips & Torrey, 1973), any modification of the chromatin reticulum which occursfrom o to 24 h must not take place during the S period. This suggests that maturecortical cells are normally stabilized in the Gx period of the cell cycle. Similar con-clusions have been reached with regard to adult iris epithelial cells involved inWolffian lens regeneration (Reese, Puccia & Yamada, 1969).

In the Wolffian lens regenerating system, marked RNA synthesis precedes activationof DNA synthesis. In the pea cortical explant system we have not yet examined indetail the occurrence of biosynthetic events which may precede DNA synthesis.Shininger & Polley (1977) reported that pea root cortical explants showed a 2- to 4-fold enhancement of the rate of RNA synthesis compared to controls lacking hormone.Stimulation of RNA synthesis could be detected as early as 9 h after explanting. Inthe same system, Simpson & Torrey (1977) showed that in the presence of hormones,protein accumulation began after 2 days, with a steady increase through day 4. It wasinferred that during the first 2 days, protein degradation occurred, thereafter netsynthesis of protein was apparent, leading to cell division. Thus, in pea, preliminarywork indicates that protein synthesis is greatly enhanced subsequent to explantculture, suggesting either a more accelerated activity of, or a marked increase in thenumbers of ribosomes. Observations on the ultrastructure of the nucleolus, whereprecursors of ribosomal RNA are believed synthesized (Brown & Gurdon, 1964;Perry, 1967; Birnstiel, 1967), support the latter proposal, that is, that increasedrRNA synthesis occurs in the activated cells.

In responsive cortical cells in culture, nucleoli show an increase in the proportionas well as in the absolute amount of granular material, a breaking up of the extensive,compact fibrillar region into numerous smaller packets, and the formation of a centralenlarged vacuole, itself bounded by a granular zone. In parallel to these ultrastructuralchanges either the nucleolar volume increases or the number of nucleoli per cellincreases and the frequency of cells whose nucleus contains multiple nucleoli increases.These changes were also noted by Fowke & Setterfield (1968) in cultured cells ofJerusalem artichoke and by GifFord & Nitsch (1969) in tobacco pith cells. Each of thesenucleolar modifications is believed associated with an increased capacity for synthesisof precursors of rRNA (Rose, 1974). For example, evidence suggests that, in nucleoliactively synthesizing precursors to rRNA, the granular and fibrillar zones are highly


intermingled (Hyde, Sankaranarayanan & Birnstiel, 1965; Rose, 1974) and that these2 zones become segregated in relatively inactive nucleoli. In other systems (Johnson,1969) the presence of a large central vacuole was correlated with the incorporation ofrelatively more [3H]uridine as compared to nucleoli lacking vacuoles. In the Tetra-hymena nucleolar system Nilsson & Leick (1970) suggested that the dissociation oflarge nucleoli into more numerous smaller units may serve as an efficient mechanismfor synthesizing precursors of rRNA. Thus, Hyde (1967), Miller & Beatty (1969) andmany others have considered that nucleolar structure is a valid indicator of nucleolaractivity in rRNA synthesis.

As the cortical cell prepares to undergo its first mitosis after the S period is com-pleted (population IV), the large, single nucleolus is replaced by 5-10 small nucleoliin which separation between the granular and fibrillar regions becomes more or lessindistinct. Such nucleoli are probably no longer active in synthesis of rRNA pre-cursors. Nuclei with multiple nucleoli show chromatin in the form of large compactmasses peripherally distributed. Then the nucleoli disappear and the cell entersmitosis and cytokinesis, forming 2 new cells.

Cortical cells after division

From the experiments of Phillips & Torrey (1973) and Libbenga & Torrey (1973)we know that the first divisions occur in cortical cells whose nuclei contain DNAamounts higher than the normal diploid populations. Frequently such cells aretetraploid at the first mitosis or may even be octaploid. These cells may undergoanother round of DNA synthesis and then divide a second time at the same ploidylevel within 60 h of culture.

At the ultrastructural level, there is no clear indication of the ploidy level of thesecells. From Table 2 one might be inclined to relate the marked increase in nuclearvolume observed with the light microscope in population IV nuclei to the knownchanges in DNA values. But there appear to be no reliable volume changes which canbe used as a measure of the ploidy level since after the first division, nuclear volumesdiminish to essentially time o dimensions. Efforts to discern distinctive nuclear changesassociated with endoreduplication failed although evidence for polyploid chromosomenumber was found.

Each daughter cell of the first cortical division usually contains a relatively smallnucleus in which the chromatin is initially condensed, then becomes disperse. Thesingle nucleolus is large relative to the nuclear size and irregular in shape. Fibrillarmaterial is present but the granular zone is not easily distinguished. Observations ofthe ultrastructure of these nucleoli suggests they are not very active in rRNAsynthesis. However, the ratio of nucleus to cytoplasm has increased and the totalcytoplasm per cell has increased so that the cells take on the appearance of a moremeristematic state (see figs. 9-14 in Bowes & Torrey, 1976).

The nature of dedtfferentiation

From several animal systems in which an already determined cell is induced todifferentiate into a new cell type (e.g., iris lens regeneration, Dumont & Yamada,


1972; limb regeneration, Jeanny & Gontch-Aroff, 1974; activation of peripheral bloodlymphocytes, Tokuyasu et al. 1968) investigators have shown that the initial steps indedifferentiation involve the following events: the decondensation and dispersal ofchromatin (Dumont & Yamada, 1972); the enlargement of nucleoli and an associatedincrease in the complexity and distribution of the fibrillar and granular zones (Hay,1959); an increase in RNA synthesis, usually preceding DNA synthesis (Hay &Fischman, 1961); an increase in DNA synthesis (Killander & Rigler, 1965); theacquisition of nuclear and cytoplasmic characteristics considered necessary for division(Jeanny, 1973); the activation of other biochemical processes (Darzynkiewicz, Bolund& Ringertz, 1969; Killander & Rigler, 1965).

According to Tokuyasu et al. (1968), in lymphocyte activation, RNA synthesisbegins about 24 h after initial stimulation. Hay & Fischman (1961) reported thatintracellular protein synthesis occurs early in the dedifferentiating cell in limbregeneration. Jeanny (1973) suggested that in limb regeneration in Desmognathus anincrease of cytoplasmic contents per cell is a necessary step in dedifferentiation leadingto mitosis.

Dedifferentiation in pea root cortical cells shows striking parallels to this sequenceof events. DNA synthesis begins approximately 24 h after subculturing explants(Phillips & Torrey, 1973). RNA synthesis occurs as early as 9 h after explanting andthen increases thereafter (Shininger & Polley, 1977). This timing would coincide wellwith the dramatic increase in nucleolar size and internal complexity observed in thisstudy which is suggestive of rRNA synthesis.

Bowes & Torrey (1976) noted an increase in the number of organelles, especially offree ribosomes and rough ER associated with the onset of DNA synthesis andthereafter. Simpson & Torrey (1977) reported the increase in net protein synthesisbeginning at about 48 h and continuing into the fourth day in culture. Such synthesisis presumed to be associated with the general increase in cytoplasmic density andcomplexity observed in these cells as they revert to the meristematic state.

Evidence of redifferentiation

In the literature on animal systems the events most often considered associated withredifferentiation of a new cell type are mitosis (Yamada & Roesel, 1971) followed bythe cessation of DNA synthesis (Dumont & Yamada, 1972), followed by the productionof cell-specific products. In cells which have divided, nuclear chromatin becomescondensed, symptomatic of the cessation of further DNA synthesis. In such cells, theredifferentiation process is suggested to have already begun (Dumont & Yamada, 1972).

In the pea root cortical cell system the relationship between DNA synthesis andredifFerentiation is complicated by the occurrence in many cells of DNA synthesis byendoreduplication. The evidence is strong that tracheary element formation occurs insubdivided endoreduplicated cortical cells. The role of polyploidization in this systemis unclear since tracheary elements may differentiate independent of specific DNAlevels (Phillips & Torrey, 1974). It is possible that polyploidization is not directlyrelated to tracheary element differentiation, but rather represents a mechanism forenhancing rRNA synthesis, a possibility quite consistent with the observation by


Bowes & Torrey (1976) of the marked increase in both free and polyribosomes incells which are undergoing endoreduplication. The present study documents thechanges which occur in nuclei of mature cortical cells induced to undergo dediffer-entiation by hormonal stimuli. Further studies will be necessary to gain insight intothe determinative events leading to ^differentiation.

The authors are indebted to the National Science Foundation for partial support of thisresearch under research grant BMS 74-20563.

REFERENCES

BIRNSTIEL, M. L. (1967). The nucleolus in cell metabolism. A. Rev. PI. Physiol. 18, 25-58.BOWES, B. G. & TORREY, J. G. (1976). Ultrastructural changes in cells of pea root cortical

explants cultured in vitro. Protoplasma 90, 99—118.BROWN, D. D. & GORDON, J. B. (1964). Absence of ribosomal RNA synthesis in the anucleolate

mutant of Xenopus laevis. Proc. natn. Acad. Set. U.S.A. 51, 139-146.CHALY, N. M. & SETTERFIELD, G. (1975). Organization of the nucleus, nucleolus and protein-

synthesizing apparatus in relation to cell development in roots of Pisum sativum. Can. J. Bot.53, 200-218.

DARZYNKIEWICZ, Z., BOLUND, L. & RINCERTZ, N. R. (1969). Nucleoprotein changes andinitiation of RNA synthesis in PHA stimulated lymphocytes. Expl Cell Res. 56, 418-424.

DELTOUR, R. & BRONCHART, R. (1971). Changements de l'ultrastructure des cellules radiculairesde Zea mays au d6but de la germination. Planta 97, 197-207.

DUMONT, J. N. & YAMADA, T. (1972). Dedifferentiation of iris epithelial cells. Devi Biol. 29,385-401.

FOWKE, L. C. & SETTERFIELD, G. (1968). Cytological responses in Jerusalem artichoke tuberslices during aging and subsequent auxin treatment. In Biochemistry and Physiology of PlantGrototii Substances (ed. F. Wightman & G. Setterfield), pp. 581-602. Ottawa, Canada:Runge Press.

GIFFORD, E. M., JR. & NITSCH, J. P. (1969). Responses of tobacco pith nuclei to growth sub-stances. Planta 85, 1-10.

HARRIS, H. (1967). The reactivation of the red cell nucleus. J. Cell Sci. 2, 23-33.HAY, E. D. (1959). Electron microscopic observations of muscle dedifferentiation in regener-

ating Amblystoma limbs. Devi Biol. 1, 555-585.HAY, E. D. & FISCHMAN, D. A. (1961). Origin of the blastema in regenerating limbs of the new

Triturus viridescens. An autoradiographic study using tritiated thymidine to follow cellproliferation and migration. Devi Biol. 3, 26-59.

HYDE, B. B. (1967). Changes in nucleolar ultrastructure associated with differentiation in theroot tip. J. Ultrastruct. Res. 18, 25-54.

HYDE, B. B., SANTCARANARAYANAN, K. & BIRNSTIEL, M. L. (1965). Observations on finestructure in pea nucleoli in situ and isolated. J. Ultrastruct. Res. 12, 652-667.

JEANNY, J-C. (1973). fitude cytophotom^trique des acides nucl&ques et des histones cellulescartilagineuses activ^es au cours de la regeneration du membre de Desmognathus fuscus.Amis Embryol. Morphog. 6, 25-41.

JEANNY, J-C. & GONTCH-AROFF, M. (1974). Synthese de l'ADN au cours des phases pr6-coces de la regeneration de membres posterieurs de Desmognathus fuscus. C. r. liebd. Sianc.Acad. Sci., Paris 278, 1513-1516.

JOHNSON, M. J. (1969). A study of nucleolar vacuoles in cultured tobacco cells using radio-autography, actinomycin D, and electron microscopy. J. Cell Biol. 43, 197-206.

JORDAN, E. G. & CHAPMAN, J. M. (1971). Ultrastructural changes in the nucleoli of Jerusalemartichoke {Helianthus tuberosus) tuber discs. J. exp. Bot. 22, 627-634.

JORDAN, E. G. & CHAPMAN, J. M. (1973). Nucleolar and nuclear envelope ultrastructure inrelation to cell activity in discs of carrot root (Daucus carota L.) J. exp. Bot. 24, 197-209.

KILLANDER, D. & RIGLER, R. (1965). Initial changes of deoxyribonucleoprotein and synthesisof nucleic acid in phytohemaglutinine stimulated human leucocytes in vitro. Expl Cell Res.39. 701-704.


LAFONTAINE, J. G. & LORD, A. (1974). An ultrastructural and radioautographic study of theevolution of the interphase nucleus in plant meristematic cells (AUium pomim). J. Cell Sci.14, 263-287.

LIUBKNGA, K. R. & TORREY, J. G. (1973). Hormone-induced endoreduplication prior tomitosis in cultured pea root cortex cells. Am. J. Bot. 60, 293-299.

MILLER, O. L., JR. & BEATTY, B. R. (1969). Nucleolar structure and function. In: Handbookof Molecular Cytology (ed. A. Lima-de-Faria), pp. 606-619. Amsterdam: North Holland.

NILSSON, J. R. & LEICK, V. (1970). Nucleolar organization and ribosome formation in Tetra-hymena pyriformis G i . Expl Cell Res. 6o, 361-372.

PERRY, R. P. (1967). The nucleolus and the synthesis of ribosomes. In Progress in Nucleic AcidResearch and Molecular Biology, vol. 6 (ed. J. N. Davidson & W. E. Cohn), pp. 219-257.New York: Academic Press.

PHILLIPS, H. L., JR. & TORREY, J. G. (1974). The ultrastructure of the quiescent center in theapex of cultured roots of Convolvulus arvensis. Am. J. Bot. 6i , 871-878.

PHILLIPS, R. & TORREY, J. G. (1973). DNA synthesis, cell division and specific cyto-differentiation in cultured pea root cortical explants. Devi Biol. 31, 336-347.

REESE, D. H., PUCCIA, E. & YAMADA, T. (1969). Activation of ribosomal RNA synthesis ininitiation of Wolffian lens regeneration. J. exp. Zool. 170, 259—268.

REYNOLDS, E. S. (1963). The use of lead citrate at high pH as an electron-opaque stain inelectron microscopy. J. Cell Biol. 17, 208-212.

ROSE, R. J. (1974). Changes in nucleolar activity during the growth and development of thewheat coleoptile. Protoplasma 79, 127-143.

SHININGER, T. L. & POLLEY, L. D. (1977). Early biochemical responses of pea root tissue tocytokinin: enhanced rates of RNA synthesis. PI. Physiol., Lancaster (in Press).

SIMPSON, S. F. & TORREY, J. G. (1977). Hormonal control of deoxyribonucleic acid and proteinsyntheses in pea root cortical explants. PL Physiol., Lancaster 59, 4-9.

TOKUYASU, K., MADDEN, S. D. & ZELDIS, L. J. (1968). Fine structural alteration of interphasenuclei of lymphocytes stimulated to growth activity in vitro. J. Cell Biol. 39, 630—660.

TORREY, J. G. & FOSKET, D. E. (1970). Cell division in relation to cytodifferentiation in culturedpea root segments. Am. J. Bot. 57, 1072-1080.

YAMADA, T. & ROESEL, M. (1971). Control of mitotic activity in Wolffian lens regeneration.J. exp. Zool. 177, 119-128.

(Received 18 March 1977)

NUCLEAR CHANGES ASSOCIATED WITHregeneration, limb regeneration and activation of division in blood lymphocytes. The cytological changes in nuclear structure could be correlated with

Documents