AN MKEALYSES QF APECAL PRQLEFEWAUCN
[N THE FGRELIMB REGENERATEON ELAEYEMA
OF THE AXOLOTL AMEYSTGMA MEXKANUM

The“: for Hm Degree 0% pk. D.
MlCHEGAN STATE UNIVERSITY

Lester L. Hearson
i966

{meets

 

LIBRA ”"14

Michigan State
University

 

This is to certify that the

thesis entitled
An Analysis of Apical Profileration in the

Forelimb Regeneration Blastema of the

Axolotl, Ambystoma mexicanum.

 

presented by

Lester Leroy Hearson

has been accepted towards fulfillment
of the requirements for

Ph. D. Zoology

degree in

I , ”/M/éfl glared/zflm

Major professor

 

Date March 29L 1966

 

0-169

ABSTRACT

AN ANALYSIS OF APICAL PROLIFERATION IN THE FORELIMB
REGENERATION BLASTEMA OF THE AXOLOTL
AMBYSTOMA MEXICANUM

by Lester L. Hearson

It has been demonstrated in this study using the
precise experimental methods of mitotic index calculations
and statistical analyses that an apical dominance in mitotic
activity exists among the most distally located mesenchyma-

tous cells of the forelimb blastema of the axolotl Ambystoma

 

mexicanum. The high apical mitotic rate was evident through-

 

out the rapid growth phase of the blastema (10 to 16 days of
regeneration) and produced a dense apical mass of mesenchy—
matous cells. As proposed by Faber (1960, 1965) this apical
dense mass of proliferating mesenchymatous cells may be re- '
garded as the "apical proliferation center."
Regional patterns in mitotic activity and cell

density were also looked for using regenerating asymmetrical
blastemata of the axolotl and regenerating aneurogenic blas-

temata of Ambystoma gpacum. An apical peak in the mitotic

 

activity of the mesenchymatous cells was also found for

these regenerating limbs.

Lester L. Hearson

In the axolotl forelimb blastema a striking temporal
relationship which might represent a causal relation, was
found to exist between the mitotic activity of the wound epi-
thelium and the initial accumulation and regional distribution
of mitoses of the underlying blastemal cells. A similar rela—
tionship in the mitotic activities between the wound epithelium
and the mesenchymatous cells in the asymmetrical and aneuro-
genic blastemata was observed.

An additional part of this study was a check on the
mitotic distributions of blastemal cells with relation to the
major regenerating nerve trunks of the limb stump. No signifi-
cant differences in mitotic activity between those regions sur-
rounding the maJor nerve trunks and those which did not could

be demonstrated.

AN ANALYSIS OF APICAL PROLIFERATION IN THE FORELIMB
REGENERATION BLASTEMA OF THE AXOLOTL
AMBYSTOMA MEXICANUM

By

Lester L. Hearson

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Zoology

1966

ACKNOWLEDGMENTS

The author wishes to express his sincere thanks
to Dr. Charles S. Thornton for suggesting the problem and
also for his interest, direction, and encouragement during
the course of its investigation.

Thanks are also extended to Dr. G. B. Wilson and
Dr.J.R. Shaver for their helpful suggestions and for serving
as committee members. Acknowledgment is made to Dr. R. N.
Band for service as a committee member.

Special thanks are extended to Dr. w. E. Cooper

.for his interest and assistance in the statistical analysis

and to Mrs. Mary Thornton for her technical assistance.
Special recognition and appreciation is made to
my wife, Patricia, whose encouragement and sacrifices have
made the completion of this study possible.
This research was supported by a Grant (GB-2618,
National Science Foundation) administered by Dr. Charles S.

Thornton.

ii

ACKNOWLEDGMENTS
LIST OF TABLES

LIST OF FIGURES

INTRODUCTION

TABLE OF CONTENTS

MATERIALS AND METHODS

RESULTS

Series I:

Series II:

Series III:

Series IV:

Series V:

‘DISCUSSION
SUMMARY . .

LITERATURE CITED

General Morphology and Histology
“of the Regenerate . . . . . .

The Wound Epithelium

Analysis of Blastema Cell
Proliferation along the
Longitudinal Axis

Analysis of Mitotic Activity
within Regions of the Blastema
Studied in Cross Section .

Analysis of Mitotic Activity
within Asymmetrical Blastemata

Analysis of Mitotic Activity
within Aneurogenic Limbs

iii

Page
ii

iv

20
3A

3A

38

Al

53

6O

65
69
81
98

LIST OF TABLES

Table

1 Changes in mitotic activity within the
epidermis of the forelimb of the
axolotl for three regeneration periods

2 Two-way analysis of variance between-
regeneration times and zones along
the vertical axis of the forelimb
blastema of the axolotl .

3 Two-way analysis of variance between

regeneration times and regions of
the forelimb blastema of the axolotl
studied in cross section . . . . .

iv

Page

All

57

Figure

1a

1b

10

LIST OF FIGURES

Diagrammatic drawing of the forelimb
blastema illustrating the five sample
levels for the cell counts . . . . .

Diagrammatic drawing of a center section
of the forelimb blastema showing the
four sample areas, A, B, C, and D for
the cell counts . . . . . . . . . .

Twelve day aneurogenic blastema showing
numerous apical mesenchymatous cell
divisions . . . . . . . . . . . . . .

The apical wound epithelium of a 10 day
axolotl regenerate possessing numerous
epidermal cell divisions . . . . . . .

A 13 day axolotl regenerate showing the
intense mitotic activity of the apical
mesenchyme . . . . . . . . . -z- . .

The prominent apical epidermal lobe of a
16 day axolotl regenerate . . .1. . .

Longitudinal section of the "early bud
stage" (10 day blastema) illustrating
the arbitrary vertical zones . . . . .

Longitudinal section of the "mound stage
blastema" (13 days) showing the verti-
cal zoning . . ... .p. . . . . . . .

Longitudinal section of the "cone blas-

tema" (16 days of regeneration) showing

the vertical zoning . . . . . . . . .
Longitudinal section of the 19 day re—

generate "paddle stage" showing the

vertical zoning . . . . . . . . .

Cross section near the distal tip of a
13 day blastema (early bud stage)

V

Page

83

83

85

85

85

85

87

87

87

87

89

Figure Page

11 Sixteen day asymmetrical blastema illus-
trating the sample areas for the cell
counts . . . . . . . . . . . . . . . . . . 89

12 Cross section near the center of a 13
day blastema showing the 5 regions
for the cell counts . . . . . . . 89

13 Cross section near the base of a 13 day
blastema O O O O O O I O O O O O O O 0 O O 89

14 Graphic representation of the mitotic
index and cell density distributions
of the A longitudinal zones for.the
A regeneration stages . . . . . . . . . . 914

15 Graphic representation of the mitotic
index and cell density distributions
of the 5 cross sectional regions for
the A regeneration stages . . . . . . . . 93

16 Graphic representation of the mitotic
index and cell density distributions
of the A sample areas for the 3 re—
generation stages of the asymmetrical
limb blastema . . . . . . . . . . . . . . 95

1? Histogram showing the mitotic index and
cell density distributions for the
base and tip areas for the A aneuro-
genic blastemata . . . . . . . . . . . . . 97

vi

INTRODUCTION'

The epimerphie regeneration of an amputated limb
in aredele amphibians is dependent on the formation and dif-
ferentiation of a regeneration bud or blastema. A signifi—
cant problem in blastema formation has been the source of
origin of its constituent cells. Several origins have been
proposed and investigated during the last several decades.
These have included the concept of an epithelial contribution
to the underlying blastema (Godlewski, T928; Rose, i9483;

-Hay,,%95£r-and~fiese7~Qaaetlen_and—ReeeT—l9537—l955+; a hema-
togenic contribution, (Hellmich,-i93U:“l93l+—Kazanee¥7_lgah;

.and_Ide—Rozae7~l936); a connective tissue contribution,

fMettetalT—i939e'Duther7”19#87—tiebman7~l9491 Manner, T953;
and—GlaéeT—l9639; a reserve cell contribution, fBrunst—and

—Cherometiena,_l936¢=Weiss,.l939+—and~Needham7'T9A2+1 and a
general contribution by a "dedifferentiation" of the stump
tissues, (Thornton, l938a7~$938b,_19A3+—Butler—ané—G4Brion,
1-9424, Chalkley , WNW
Bodemer*and~EvanetLVal959;mHaymandeisehmanyw&96OT“T96i).

his?“ (Erma-£3 \w

Godlewski observed for thefgxeletl limb a reserve
of indifferent cells located in the—basal—layer—of-the epi-

dermis. In his opinion, these cells, following amputation,

were stimulated to migrate internaliy’and give rise to the

blastema.

Rose} €i9fifiaj described~tengaes~of wound epithelial
cells entering the developing blastema of the amputated am-
phibian limb and believed they Sfifiiiafiigéiy transformed into
mesenchymhbous cells of regeneration. He came to this con-
clusion as a result of observing that-supravital staining of
the wound epidermis subsequently gave rise to stained mesen-
chymgtous cells. Furthermore, measurements of cell numbers
in the epidermis and subepidermal regions at the time of
young blastema formation revealed a sudden decrease in epi-
dermal cells and the corresponding rapid increase in mesen-
chymatous cells. Finally, when the limb ef-the—newt was
eXposed to x-rays, subsequent regeneration of the limb was
reported after replacing the irradiated limb skin by unirra—
diated epidermis fRose7xQuastlerymand—ReseT—l953)w 'Theyiu-
concluded that only epidermal cells could have contributed
the blastema cells of regeneration.

The e idence of Hay,/[1952) for an epidermal origin

of mesenchymatous issue was derived from the identification

of cells in the earl blastema which were apparent migrants

 

from heteroploid skin ansplants. In later blastemal
stages, however, heteropIoid cells could not be distinguished
in the undifferentiated par \of the blastema nor the differ-
entiating internal tissues. \
A number of investigators have re-evaluated the
epidermal hypothesis and have concluded that such a contri-

bution is highly unlikely. Heath, €l95347a£er-example, con-
sidered the epidermis a negligible source of blastemal

3 Y

elements, basing his conclusion upon evidence from differen-
tial growth ratesof eh;gzzgie limbs created by exchanges of
epidermal and subepidermal limb-bud tissues between embryos
with markediy-different growth potentials. “Here, the experi-
mentai deaign provided evidence that the control of the
growth rate+of the organ resides chiefly in the limb meso-
derm and gziegzegainet the participation of the wound epider-
mis in blastemal formation.

The hypothesis of an epidermal origin of the
developing blastema is also not supported by the work of
.Maufier, ($353). Through histological study and cell counts
Manner was unableto‘reveal any evidence for an epidermal
ingression but indicated\instead a major contribution to the
early blastema by fibrobyasts. He emphasized a cytological

/
similarity of the fibpoblasts\and the cells of the epidermal

stratum germinativum and noted thisésimilarity of character

I
.r

might easily engender confusion as o blastemal cell source.

Chalkley has~pre¥ided_anmexcellentflsystemetie—study,
(l954)—and‘reView, {l959)-of the mechanismeof“tho—preduetion
and~aceumulation“of"cells during~blastemaefermation. He
could detect no loss-gain relationship between epidermal and
subepidermal components. The 2£$§;§ rise in the number of
subepidermal cells and the comparable drOp in the number of
cells in the overlying epidermis, as reported by Rose for
the early blastema, could not be egggghggigede ~Ghaikieyrhas
suggestedmthat-losses in epidermal-cell~number5mwere~probably

awfunction~efedesquamation rather-thanmcellulariingressions.

<9
.\
U

The most convincing evidence against the epidermal hypothesis
can be derived from Chalkley's discovery that dedifferentia—
ting proximal mesodermal tissues become mitotically active
during the very early phases of regeneration and at an
equally early time cells produced by this mitotic prolifer-
ation migrate distally and accumulate to form the regenera-
tion bud. Fgggsphenomenon can easily account for the increase
in blastemal cell numbers.

Chalkley'fiQéfl) has interpreted epidermal intru-

sions nto the underlying blastema as simply mechanical

   
  

/
infoldin 3 resulting from amputation.f

f

; Scheuing and Singer, 1957; Bodemer, 1958;

Other investigators,
(Taban, l9
Kamrin and Singer, 1959; Singer amd Salpeter, 1961) have
observed internaleovements in the form of long chains or
cords and in somechees<dis- -associations of epidermal cells.
For the majority of these workers and especially Singer and
Salpeter (1961) such responses are believed to be a measure

‘x. 1"]

of the sensitivity of the wound epidermis to the topographi-
cal and chemical composition.of the wound tissues. Indeed,
epidermal cells, singly or in hiusters,have been observed to
encapsulate and engulf cellular\debris and foreign matter
and to subsequently extrude these through the skin. It is
the opinion of Singer amd Salpeter (1961) that reactivity
of this type, although common to all epithelia, is a more

highly developed feature of the wound epidermis.

I
5 \

is also of interest to ndte that Luther (1948),

/

rradiation of intact tails, observed that

    
 
       
  
 

following distal
the dying irradiated idermal/cells were replaced by unir-

radiated proximal epiderma cells which were accompanied by

subepidermal cells. A similar

//
I

account for the results of Rose,

henomenon could very well
stler, and Rose (1953).
Similarlysisuch a phenomenon would provide an al-
ternative explanation to the results of Hay (1952). Indeed,
Hay and Fischman (1961) have considered this possibility

since, as they indicate,/polyplQid fibroblasts were components

if
J

of the proximally located skin gra s used in Hay's 1952 work.

Perhaps the most damaging evidence against the epi-
dermal origin of blastemal cells is that of Riddiford—(lgfifl)
-and—Hay—and_Eischman_Ll96%i1 Ridgiford exchanged labeled
epidermis for unlabeled epidermis between labeled and un-
labeled regenerates. The unlabeled epithelium covering the
labeled blastema gradually became labeled g::::mably by
movements of cells and_cellular—deerie-into the epidermis
from below. In no instance did the refieree movement of
label from the wound epithelium into the blastema take place.

$::;e:;m-there appears to be little evidence in
support of the epidermal hypothesis.

A second hypothesis cencerning the origin of the
blastema.whiehrchronelegieallywprecededwthe—epithelialueen-
-eept, was the idea of a hematogenic contribution (Colucci,
IGGAT—eitedwinwehalkley;“1959%T**Such‘arsearee~wasmalsonsug—
gestedebnyellmichj“(1930; 1931)erazancevrrti93A)waneu

Ide—RezaeT~(l9369. Hellmich @1936) distinguished en—punely

.histologieal~greunds two general kinds of cellular components
within the blastema--hematogenic and histogenic. For

Hellmich, these types: erythrocytes, lymphocytes,.eos1
ififiphils,.speeialrleucocytes, plasma cells, fibrocytes,
wandering cells, mast cells, pigment cells, giant cells,
and pericytes were all indispensable components for the

restoration of the missing structure. atever the regener-

  
   
 

\ \
ating part, it was thought by that regeneration was

possible only through the physiol ical COOperation of the
\\
entire organism.
The idea that leucocytes and other elements of the

general circulation were the source of blastemal cells was

not supported by the irradiation experiments of Butler 6&933,

l93§l;_Bnunst.and_Cheremetievewti936iruButlerrand~O+Brten
Re
ClgAZJFandflBurnStM+}950+?m Batler 6%9337—l935) was able to
9/0 (£3117, 55“.

show that larval Ambystoma limbs situated in shielded regions

 

during x- ray exposure of the remaining body retained their
”(LCL'CL (EALW
capaélty to regenerate. bikewieeg unirradiated limbs trans-

planted homOplastically to irradiated hosts, entirely en-

Wtr

A pv.~r.x.‘.‘—'.; 'fir’

capable of regeneraviee, retained the ability to regenerate
following amputation.

Wmanner BrunstMeremetieWSH
Butler and O'Brien £1942)1*anfl—Brun§t"TI9503 demonstrated
that failure to regenerate occurs only in that region of the

telly
.appendage actually exposed to x- rays. Butler—and‘0*BrIen

 

6%9429, for examplen,destroyed by local x- ray treatment the

regeneration capacity in the knee-joint region of the Amby-

- «’1‘! ' 3 L“:

stoma larval limb while shielding the other of the

 

limb. No regeneration occurred following amputation through
the exposed part while normal regeneration was observed in

IJQ $19.35 (72‘1“

the proximal—end-distal'parts of the limb when subsequent
amputations were performed in these regions. Conversely,

Brunst and Cheremetieva (1936) destroyed by localized ir-
radiation the regenerative capacity in both the proximal
and distal parts of thexlimb, although the regeneratiVe
ability of a small intermediate protected area of the same
limb was unaffected. Brunst (I95Q) found, as did Butler
earlier (1933, 1935) for transplanted\limbs, that normal
regeneration buds when transplanted to tFeated body parts

could regenerate normally.

     
    
 

Earlier, He wig ( 927) obtained similar results

by the simple technique f transplanting haploid limbs to
limg stumps of diploid . The resultant blastema was
formed mainly from ha oid el ments.

The 5g;e;;;ng experiments provide proof beyond
reagenahlefdoubt that blastemal cells must be the products
of tissue components in the immediate vigigity of the ampu-
tation surface. §%:;:?6he evidence is clearly contrary to
the idea that body parts in;general contribute cellular ele-
ments to the regeneration site. How is it then that local
stump tissue components give rise to blastema cells? Weiss
6E§39);has suggested that "reserve cells“bapparently "left

i
*1 If"? 4‘ 3’1.

over' from embryonic development and situated somehow among

the formed tissues of the stump, are stimulated by amputa—
tionel trusma to preliferate—and—form the blastema. Tesdate,
numerous detailed histological surVeys have failed to reveal
any such store of reserve elementSAe=Their existence in
AMphihia remains purely hypothetical.

The most widely held nlliin theory of blastema!
cell origin is that of fiedifferentiation of stump tissues.
The injured cells of the mesodermal tissues of the amphibian
limb, Sula-llullc, undergo loss of histological structure
during the first few days following amputation and 'spill
out‘ cells, WW,
which can be traced into the blastema.

Thus, David (193A) fqllowed distal loss of struc-
ture in the injuredxiimb cartilage and described migration
of cells, released by thexhistolysis of their capsules, into
the developing blastema of the\axolotl limb. These cells
were subsequently thought to redonstitute the missing skele-
tal structures.

Thornton 61938a7Wl938b) in ancarefut"and*detailed
histologieal—investigation of Ambystoma limb regeneration

 

described a multiple origin for the blastema involving the
gedifferentiation of cartilage, nerve sheath, perishendrium,
muscle and general connective tissues. .Fer-exempleT—he—deSr
eribedramdisselu onmofvthevdyétalwmusele~fibersmwith—separa-
tion of muscle cell n clei/shrrounded by scanty cytoplasm

which could be followed 9, light microscope observation)
1

 

/
into the developing blastema. LIn subsequent years there

followed a variety of ' focusing upon the mesodermal

  
    
  

hypothesis (Thornton, 194 ; Chalkley, 1954; Hay, 1958, 1959;

Holtzer, et_a1., 1957; odem r and Everett, 1959; Hay and
Fischman, 1960, 1961).
Thornton 6&949i gained-eea£ipmatory-evidence for
his eafiIEEr conclusions on fiedifferentiation by implanting
unirradiated limb and tail muscul‘éeure (mm
connectiv€"ti§§fie7 into irradiated limb stumps of adult
Triturus. The transplanted muscurétupe not only underwent
partial dedifferentiation to form blastemal cells but was
shown—torhave directed the develOpment and differentiation
of the formed blastema. It sh d be noted, however, that

not only muscle e s , but als connective tissue elements

took part in regenera ion.

  
   
     
   
   
  
 

r- fl‘ It
w.~¢.4_»y '- .\.

(eell~eeante)~ as prepeeed a tissue ecific origin for the
blastema. He in *substanti cellular contribution
from the general con 0 ive tissue/Yfibroblast, etc.) and
otherfyesodermal components such

as muscle, cartilage, per

lesser contributions
hond ium, etc. In his opinion,

. 1
limited dedifferentiation, qiiferation, and migration of
proximal stump elements provi ed for the early growth of

the blastema. Simi fin ngs were reported by Manner

(195“). Hay (1958, 19

   
  
 

,Has con irmed and extended by

 

electron microscop the e rlier light microscope

v],
analys
work of Thornto. (1938a3. Her studie show, for cartilage

and muscel at least,fthat differentiated_cell types can

pm...
.5 \

lO

lose their characteristic pgtological properties and become

mesenchymatous like regeheFate components.

Criticisms (see—MannerT—1953) that tissue "dedif-

Carlie-.15, ,1.
ferentiation" is really only tissue degeneration are.met by

the experiments of‘Bodemer‘and—Everett—(l959)wand~Hay and
Fischman @196I71 Using isotope labeling they demonstrated
that as limb stump tissue! dedifferentiategthe resulting
cellilar units enter into a phase of protein and nucleic
acid synthesis:;reparat$;y to active cell proliferation,
which, ef—eourse1 dying cells would not be expected to do.
An dditional technique w lch has proved to be a
valuable tooI§>h<kjnvestigating bl stemal cell origins is
that of fluoresce t\antibody lab ling of specific components
/

of stump tissues. Uiing f1uoreScent anti-myosin as label,

Holtzer, Marshall and inck, (3957) were able to trace de—

  
  
   
   
   
 
    
  
 

differentiated muscle c lsfinto the blastema. Unfortunately,
/

the label was not detect? re during the period of blastemal

cell proliferation (frquap oximately the 8th to the 12th

day after amputationL/ after ich the fluorescent label of
;

differentiating mgécle cells ould again be detected. The

4 day gap in labeiing thus preven one from establishing

the possibilityjof metaplasia in re neration but does
clearly answer the question of muscle articipation in blas-

tema origin. ‘(

[hie-3?"

It is now generally accepted by eheimajorityrof

investigators that tissue uedifferentiation is a mandatory

prerequisite to the appearance of a blastema. Thus, all

ll

stump tissues with—the—exeeptionwef—the”epidermisembleed—ele-
mante"afld“ner¥emtissue~proper provide cells of general.mosen-
chymateus morphology for early blastemal growth.- The—genetic
nfllfinllalitie6~e£mthessicells.isistillwianuestionuanduthus
the‘problemmefealpossiblewmetaplasiamiseyetmteebemestgglished.
Whatever the spfifizfiic origin of the blastema may
be, it is clear that some kind of ”blastema-forming; activity

“tr; b @1049...“

must-be—in—eperatien if regeneration is to occur. Thus,

stump tissue dedifferentiation and the appearance of mesen-
chymatous cells are not_in—téggsilvee—suffieient to achieve
the regeneration of a salamander limb. These events all
occur in the non-regenerating limb of the adult frog-—
histolysis of stump tissues and appearance of mesenchymétous
cells--yet no blastema ever forms and eeheeQHently regenera-
tion fails. Even the salamander limb can be prevented from
blastema formation and, thus, regeneration by the—simple
experiment—ef—substituting head skin for the limb skin.
Even though after amputation threugh—the—skin—graft—tissue,
dedifferentiation may occur, no blastema forms, and regenera-
tion fails mpg-"limb WWW
remWMMnr—Héfiw

In both the frog limb and the salamander limb with
head skin, wound healing ef-the~aumputation_sun£ace is ab—
normal in that the epidermal cells involved do not lose
their fully developed histological characteristics. No
wound epithelium consisting of epidermal cells of "dedif—

ferentiated" morphology is produced. As a result*of~a—long~

_, '1
/‘

‘..~, 'l.
12

series’df—Investigations, Thornton (l957-i965)7—has developed

42!;
the theory that the wound epithelium in some way exerts—an-

. influenee'on the aggregation and development of a blastema.

From experiments performed on tadpoles(Rana) and
GAMMA < A.»

Ambisteme-larvae, Thornton‘(I95#~—l956)~hae described a dis-

LflM&Ad
tinct thickening of the wound epithelium, which he has_tenmed

the "apical cap," overlying the area where the dermis has
0* mfip.geh

been interrupted. Subsequeneueéjthis thickening and immedi-
)2fl19WV

ately underlying it develops the early accumulation of blas-

temal cells. The temporal relationship between these two
fig, " lira. W
phenemena appeared significant. ,*.

.iumviu‘

Insight into the indTSpeneeble nature of the ‘apical
cap“=was gained through two simple bet—eleer—eet experiments
(TherntenT-19577*&958). in.§;§ee experiments there was a
total failure of blastemal cell accumulation following con-

éitiens_o£aeither "cap" removal or "cap" treatment with

  
   
    
 
 

localized ultraviolet irradiation. Also, y experimentally
inducing the formatign\of eccentric api a1 caps in bath
normal andiexeeseivelycrggreeeed-larv l limbs, Thornton
(l960f'produced eernospendingly ec entric blastemata.
Eccentric apical caps, howeve ay be innervated by eccen-
tric sensory nerve fibers whi cduld serve as migration
pathways for mesenchymatous cells. Itxfollows from this
that aneurogenic limbs w h eccentric apTeal\caps should
show no eccentric blas ema formation. Howeverg\Thornton

and Steen (1962) dem nstrated that aneurogenic lfihQ:::d pro-

duce eccentric blaétemata directly correlated with ec entric

13

directing cells bene

A‘SO
Fifia£iy, Thornton and—Thornton‘6i96fifl transplanted

epithelial caps autepéaeticaliy to the base of developing

blastemata. Successfully grafted caps were found to support

apical caps. Therefo ac nerve axons are not necessary for
a2h\the epidermal cap.

secondary blastemal develOpment and supernfimerary-limb parts

:14er
were differentiated. Thus, the wggndzakiAMZGth~in~th94nonmal

-and4eneunogenic+regeneratienwmilieuwxgs specifically been
indicated as having an essential role in blastemal formation.

In addition to the concept that the wound epider-

{
x“
mis functions to “direct“ the accumulation Sp blastemal cells,

a number of other theories concerning its noie in regenera-
A\’Y‘\Cfl"3 fihnc avg;

tion have been suggested. These4ine&fide histolysis, mechan-

ical regulation and the control of morphogenesis. Of par-

 

ticular interest is the recent suggestion

by Faber 6i966§ thaE;:he apical wound epidermis might play a

role in the estgbggshment of an "apical proliferation center"
vnorfi disianf

among the.mostmdistally~&ocated-mesenchymatous cells of the

blastema.

Once the regenerating blastema is formed, its
further growth has been generaiiy considered to be by means
of the rapid proliferation of its constituent cells. Not
much attention has been focused on the patterns of mitosis
within the various regions of the blastema, although studies
-o£—mitotic.pnoli£eration*of the blastema as a whole have
been made, éfiitWiIlér?“19397“?oiezajew“andrfiinsburg7w%9#3§

Forsythjwl9uégwManner;“i953,mGhalkley,Wl9fihfr However, some

14

studies of regional growth patterns in embryonic limb buds

} »1.-_‘,

(117.23 ' —. . .44.-
have been completed. Thus, the anaren hindlimb tnui-(Tsehumi,

W???“ and the avian limb bud (W
W WW
,4H&357’I9627 are believed to grow by apical proliferation.
The_phenemenon"of“apI551”proIiferatibn”pointsrupwanotherr
parailel~in—tha—modewefwgrowthwwhich4existswhetmeen~regenanar
.tien—biastemata'and'embryonic-and-larvalwiimb buds. A most

Maui, .sifliin-r ,(MAx
striking feature in all of these systems is the presence at

fiausfi/l‘ __IZL.

the apex of the growing rudiment of a thickening of the epi—
thelium.

The eharaeteristic properties of this thickening
has been warieasiy described by a number of investigators.
Thornton Q195hi-has described an "apical cap" for regenera—
ting lim?s, while Tschumi“{i9553, Saunders (lghfléwandn%wii-
~iingiilgaéiuhavéflgbserved a thickening in the form of an
“*apical ectodermal ridge‘r for develOping limb buds. An»
ectodermal_ridge.haS—as well been demonstrated4fer the limb
~buds~efwsome reptilian and mammalian forms (Hinrichsen,4l956;
Milaire,‘l956, 1957; O'Rahilly,!Gardner and Gray, 1956).
Lastly, Faber 0&966) has deggrgged afidistinetive ’epidergal
lobe‘~at the apex of the regenerating forelimb :h1.Ag§::;;ma

mexicanum4 He has noted a possible causai relationship

 

between the appearance of the "apical epidermal lobe" and

the actiVe proliferation of the underlying-mesenchymeie:e

15

is necessary to note here that t e experiments
of Tschumi (195;), Saunders (19MB), Faber ( 960) and others

are suggestive rather than conclusive co cerning the exis-

‘

tence of a specific\proliferation center. In all cases,

1

/

‘\
both in regenerating blastemata a:3/deve10ping limb systems,

the criterion of apical growth ha been the localization of

charcoal particles which were\~6rced into the mesenchyme at
/‘\
the apex of the bud. Since_individual cells, and therefore

mitotic cells,are not mar ed by this\technique, it cannot
represent a precise measure of the act al patterns of cellu-
lar activity within the developing struct re, a matter of
considerable impo tance, for, as suggested by\Faber (1960),

such a proliferation center may constitute an autonomous

organizatio 1 center which controls in some way \he regional

s.

{U 13.! lkx'rzwj

organization of distal limb parts. Through

 

experiments performed with forelimb blastemata of the axolotl

 

it has been shown , , ,
Jkflarr7~l9615 that the mesenchyme of therblastema isolated
becom°B {3.53”i’““""r=v~..”*cai

from the stump acquiresidiflflerentiation—tendencies for struc-
tures which are markedly more distal than those which they
normally would have produced. Thus, blastemata at various
stages of development afiew61Imas‘proximai—andrd1stairhatves
of’palette—etage-blastemata-and-proximal“haives“o£~thewpalette
stagewblastemata with a reversal of the~proxime—distai
-poiarity, were transplanted to a-neutral site épocket4—on

the back of the animal. In all of these transplants,wand

especially~thewnoninvertedwand inverted proximal~halves of

l6

therpalettewstagey-there occurred dedifferentiation and dis-
organization of the developing tissues of the regenerate.
Subsequently, aswdeterminedrinmmost“instances~by*carbonr
*marking, apical proliferation was initiated and the grafts
produced normally-oriented limb outgrowths. The striking
feature of these experiments was that the grafts péééggii
neatly formed distal structures Ldigitso for which, at
least in the case of the early blastema stages}, and—proximal
haIVESywno rudiments were present. Thus, as described by
Faber,0l9691 the distal predominance must have resulted
from the action of an "apical organization center" 99—99
orgaaézing—prifinipie which was responsible for the self-
organization of the mesenchyme. Faber,w(l960) haSywin~
addition, proposed  that the "apical organizationficenter"
might be identical with  the_ ”apical proliferation center";
thus, cells formed in the Apical proliferation center by
division of dedifferentiated cells\would intrinsically
possess distal differentiation tendencies.

An imp tant feature in all of these experiments
ion of the transplanted

was the fact that th self-organizat

        
   
   

mesenchymatous materia was alway associated with the
occurrence of apical prol ferat on. Mesenchymatous cell
proliferation, in turn, app ed to depend upon the cover-
ing of the wound surface by th wound epidermis, which was

supplied either from the transpla ted limb section itself

 
 

or from the back skin the transpl ntation site. There-
,/
fore, as has previously been demonstra _d by Thornton

l7

(l956al958), a wound spithelial covering devoidmofmtheflnnder-
lyingndermis,iwasinecessarynformregenerationwtewpreeeedemln
viewwofwthisufact, Faber tl9660 haé:5;0posed that the wound
epidermis, in—seme-unknownumanner5 initiates the appearance
from among the distal mesenchymgteus cells of a center 0f pro—
liferation and organization which functions to produce and

organize,xatmleastwferwthe‘distal~structuresw the materials

of the developing limb.

    
 
    
      
   
 

In view 0 this hypothesis 1 is apparent that a

re-examination of apica proliferati is needed using more
precise methods than charc al mark ng, so that a clear and
unequivocal picture of the b mal cell proliferation
patterns can be obtained. Al gh there have been a num-
ber of studies of the mitot c prol eration in the blastema,
none have focused upon t e patterns 0 the mitotic activity.
The most significant work.towdute on the regional
distributions of mitoses was presented by Chalkley £l95#)
for the newt, butfithiswetudy~wasmprimerilywconeerned—with*
the; problem-‘Uf’ti‘s’ 's’”“""é‘"”d‘e fiff-wmtatmmm
“biastemalwceiis: Thus, Fbr both epidermal and mesodermal
tissues, Chalkley has described a gradual distal shift from
the limb stump to the blastema in both cell number and mitoiwp
Itie—indax. It was not until the 19th day¢ however, that
the peak of miotic activity was totally giggéggg within the
blastema. At this time the mesenchymal mitotic index was

highest in-roughly the distal half of the blastema with the

18

peak of mitotic distribution still extending into the stump
portion.

In earlier study (Litwiller, 1939) of the blas—

 
  

’1
tema of the Japane e newt, a peak/of mitotic activity was

demonstrated for the ase of the young blastema which

shifted distally in older egenerates. The gradual shift
in mitotic intensity within the blastema was associated
with and preceded différentiation,\ Only in a vague manner

/
were recurrent proximo-distal mitot patterns between re—

generation stages made evident in this work. /

PolezaJe nd Ginsburg (l9h3), /§6;:yth (19u6)
and Manner (1953) have own that mitosis of the mesen-
chyme cells is infrequent in the very early blastema, but
that it increases rapidly and\\reaches a high degree of
intensity during the period of\{apid blastemal growth..
Although no attempt was made to rgveal mitotic patterns
within the blastema, Polezajew and\ginsburg (19H3) demon-
strated for the axolotl a maximum in\cell divisions at the
cone stage, which then decreased andxrfmained relatively
constant until digital indentations weréKformed.

The present investigation, therefore, was under-
taken to examine the regional distributions of mitoses
within the blastema and to test experimentally by mitotic
index calculations the hypothesis proposed by Faber (1960)
that there exists within thefdistal mesenchyme in the limb

blastema of the axolotl an ”apical proliferation center."

19

The investigation was attempted in addition to determine if
a correlation might, as suggested by Faber (1960), exist
between the appearance of the apical epidermal lobe and the

mitotic proliferation patterns within the blastema.

MATERIALS AND METHODS

Larvae of the Mexican axolotl Ambystoma mexicanum

 

and larvae of Ambystoma opacum were used in this investiga-

 

tion. Axolotls were reared from eggs or newly hatched
larvae kindly supplied by Dr. R. R. Humphrey of the Depart-

ment of Zoology of the University of Indiana. The Ambystoma

 

opacum larvae were reared from eggs obtained from Mr. Glenn
Gentry of the Tennessee Conservation Commission.

The axolotls were fed brine shrimp and Enchytrae

 

worms during their early growth phases and were later hand
fed beef liver 3 times weekly. Animals were maintained in
individual plastic containers in a Model 805 Precision In-

cubator at 20 i .5 degrees C. during the experimental tests.

The Operations

 

Axolotl larvae of A to'6 centimeters snout-anal
length were anesthetized in a 1:2,000 concentration of MS
222 (Tricaine methanesulphonate, Sandoz), and both forelimbs
were amputated through the elbow under the binocular dis-
secting microscope. Following the initial amputation,
which resulted in the retraction of the soft tissues, the
protruding cartilage was reamputated so as to provide as
nearly as possible an even amputation surface perpendicular

to the limb axis. The remaining humerus stump comprised

20

21

approximately three fourths.of its original length. Since
it is possible that the mitotic activity of regenerating
amphibian limbs might undergo daily periodic fluctuations
all amputations were carried out on predetermined dates at
12:00 M. Indeed, Litwiller (19u0) while studying regenera-
tion in the forelimb of the newt (Triturus pyrrhggaster),
discovered a daily rhythm in the mitotic rate which corres-
ponded with specific lighting conditions. He found, for
example, high mitotic peaks at 12:00 M. under the conditions
of either constant darkness or normal day-night illumination.
In contrast, the 12:00 M. mitotic peak was depressed con-
siderably and a new peak occurred at mid-night under constant
illumination. The light source in this experiment was a 60
watt light bulb suspended at a distance of 5 feet above the
animal containers.

Cameron (1936), on the other hand, found no cor-
relation between mitotic indices and the time of day in
Ambystoma larvae and gang tadpoles.

The larvae used in this study were kept under
conditions of low intensity, constant illumination in order
to obtain maXimal feeding and growth. The light source
used was a 10 watt bulb located 30 inches above the animals
which were maintained in plastic containers housed on,and
covered with,18" x 24" metal trays. Under these conditions,
those animals closest to the light source (at the periphery
of the trays) were exposed to l footcandle of light as

measured by a Weston Model “16 light meter. This situation

22

cannot be compared to the conditions reported by Litwiller
(1940) since he did not describe his experimental method.
Measurements in this laboratory indicate the light inten-
sity of a 60 watt bulb at a distance of 5 feet to be 6.5
foot candles of light. As will be explained below, the
average mitotic index at the apex of rapidly growing blas-
temata of the axolotl (16 days regeneration) is 3.0 percent
under low constant light conditions. These results indicate
either that light has no inhibiting effect on the mitotic
activity of axolotl blastema cells or that the light source
utilized was of an insufficient intensity to cause any
alteration intnnemitotic patterns. To test the possibility,
however, that the.light intensity employed might have af-
fected the mitotic activity at 12:00 M., a small group of

6 animals was kept in constant darkness (except for feeding
periods) for a 16 day regeneration interval. The mitotic
indices of these animals were then compared with 16 day re-
generates maintained under one foot candle of constant
illumination. No-differences in the pattern of mitotic
activity could be detected between the two groups allowing
for normal range of variability.

For the present study, u series of regenerating
axolotl limbs were prepared for histological examination.
Group I comprised axolotl forelimbs of post—amputation
periods between one and 25 days. Daily camera lucida draw-
ings were made of the regenerating limbs in order to provide

records of the gross morphological changes for this series.

23

Limbs were fixed in Bouin's fluid and sectioned at 10 microns
parallel to the long axis. Sections were stained for general
histological features by Heidenhain's iron-hematoxylin
method.

Series 11 included 10, 13, 16, and 19 day forelimb
regenerates. Limbs were fixed in Zenker's (acetic acid)
fluid and differentiated for chromosome detail by means of
the Feulgen method. Longitudinal sections of the paraffin
mounts were prepared at 10 microns. For sectioning, the
limbs were not deliberately oriented with respect to the
dorso—ventral, anterior-posterior axes.

In-series III regenerated axolotl forelimbs were
amputated and fixed beginning at 10 days and thereafter at
3 day intervals up to 19 days post-amputation. Limbs were
fixed in either Pienaar's or Zenker's fixatives and sec-
tioned at 10 microns. Sectioning was at right angles (cross
sectioned) to the dorso—ventral axis of each limb. Prepar-
ations were stained with Schiff's reagent (Feulgen technidue)
for the localization of DNA.

For series IV, axolotl limbs were allowed to re-
generate for l“, 16, and 18 day periods. During the regena
eration period, the established epidermal cap of these
limbs was shifted to an accentric position according to the
technique of Thornton (1960). The cap was shifted from its
normal position at the apex of the limb stump to the anterior
edge of the initial amputation surface by removing a narrow

strip of skin (dermis and epidermis) approximately 1.5 mm.

24

wide and .6 mm. in length running parallel and adjacent to
the anterior border of the amputation surface of the limb.
Operated limbs were approximately 9.5 mm. in diameter.

The first skin removal was at 7 days post-amputation when
the apical cap first becomes established and before a blas-
tema has formed, and the procedure was repeated once, 2
days later to insure an adequate cap shift in these large
limbs. Correspondingly eccentric blastemata were subsequently
formed at the tips of these limb stumps. Zenker's fixative
and Schiff's reagent were used for these limbs. Longitu-
dinal sections were prepared at 10 microns, the. plane of
sectioning being as nearly parallel to the long axis of the
eccentric blastema and limb stump as possible.

In series V, (Ambystoma opacum), aneurogenic fore-
limbs were produced (see Thornton and Steen, 1962) and at
the A digit stage were amputated at the elbow and fixed
from 10 to 16 days post-amputation in Zenker's fluid and
stained for DNA localization with Schiffis reagent. These

limbs were sectioned longitudinally at 10 microns.

The Cell Counts

Since the;primary focus of this study is on blas-
temal cell proliferation, considerable care has been devoted
to the method of counting cells and mitotic figures. For
counting nuclei and as a basis for cell density determinaa
tions, an ocular grid ruled into 100 square units was used.

Each grid unit, at 225x, was found to be .027 mm. on a side.

 

 

25

When the grid was superimposed over a field of the blastema
(see below) all nuclei, with the exception of small nuclear
fragments, red blood cells, distorted nuclei and identifi-
able leucocytes lying within each small grid unit surveyed,
were scored as either interphasic or mitotic cells. Those
nuclei with one half their volume within the outer boun-
daries of the total area surveyed were included in the
counts. A total of 100 (interphasic + mitotic) nuclei were
counted for each area sampled and the number of small grid
units covered to obtain this total were also recorded. It
should be noted that except for large aggregates, those
units occupied by cell fragments, red cells, etc., were

not subtracted from the total grid units recorded. For
these reasons, density recordings represent approximate
values only. They are, however, sufficiently accurate for
the purposes of this study.

‘ In order to determine if 100 cells was a suffi-
ciently high figure for detecting differences in mitotic
rate between regenerating limbs, pilot counts of 100 cells
at randomly selected sites for six 16 day regenerates were
made. These counts produced results indicating an average
deviation of i .3“ in the number of mitotic figures counted.
From a statistical point of View, the small variability in
the number of mitoses between limbs obviated the necessity
for counting a larger number of cells. The mitotic cate-
gories identified were mid and late prOphase, metaphase,

anaphase, and early telOphase figures. Late telophase

26

figures were counted as one mitosis. Early prOphase and late
telophase figures were scored as interphasic cells. Where
there was doubt concerning the identity of a scored mitotic
figure, the cell in question was re-examined at a magnifica—
tion of A30 X for positive identification. Visual periodic
checks were made to determine if the same mitotic figure

was being recorded in the different sections. In no in-
stance was this found to be the case.

An inherent variable in a study of this type is
the degree of subjectivity on the part of the scorer with
reference to the criteria being used. Indeed, early counts
undertaken to establish a consistent sampling technique
showed a progressive decrease in density scoring and a
slight increase in the number of mitotic figures recorded.
These earlier counts served to standardize the more precise
counts recorded in this thesis. Nevertheless, as a check'
on the accuracy of the scoring, periodic recounts of sample
areas selected at random were made. Such counts revealed

an error of i 3 percent in the recorded figures.

The Blastema Areas Sampled

For those limbs sectioned longitudinally, mitotic
index determinations were carried out on two sections each
interspaced by one section at 5 different levels within the-
blastema. Sample level #1 was at the center of the blas—
tema and usually corresponded to the median plane through

the humerus stump while levels #2 and #3 were located

27

parasagitally through the right and left halves of the blas-
tema. Sample levels #4 and #5 were located at the lateral
edges of the blastema lateral to sample levels #2 and #3
respectively (Figure la).

Based upon histological topography, the regenera—
ting portion of the forelimb was arbitrarily divided into
2 zones; a distal blastema and a proximal dedifferentiation
area.‘ The dedifferentiation area comprised the cellular
elements located between the base of the blastema proper
and the level of the differentiated tissues of the stump.
The base (the approximate primary amputation plane) of the
blastema was determined using the more proximal edge of the
dermis as a landmark. This method provides a base level
which corresponds fairly accurately with the base line used
for blastemata cut in cross section. An occasional level
difference between the lateral dermal elements was practi-
cally unaviodable. It should be noted in this regard that
because of unequal stump tissue masses which apparently
cause differential tissue retraction and probably differen—
tial tissue dedifferentiation, dermal structures never in-
dicate exactly the primary level of amputation but serve as
an approximate indication.

In the longitudinally sectioned blastemata a
maximum of 4 sample areas (Figure lb) was selected for mi-
totic density calculations. Sample area A was at the tip
of the blastema immediately beneath the epithelial cap;

area B was at the center at the base of the blastema;

28

area C was at the left periphery at two thirds of the dis-
tance from the base line through the blastema; area D was

at the right periphery at one third of the distance from

the base line. The positions of sample areas varied
slightly for the smaller blaste-ata. In the 10 day regener-
ates, for example, the sample areas were located within H
equal zones along the length-of the blastema as illustrated
in Figure u.

To avoid the possibility that the areas counted
in the smaller blastemata (10 and 13 day regenerates) might
overlap one another, density and mitotic counts were spread,
in some cases, over several sections. The levels for the
counts were kept constant between each of the sections
scored.

For cell counts at the tip (area A) the ocular
grid was superimposed over the blastema and oriented under
the tip curvature so that its distal edge or corners, de-
pending upon the blastemal morphology, Just touched the
base line of the wound tip epithelium. For the base calcula-
tions (sample area B) the ocular grid was superimposed over
the blaStema so that the base of the grid coincided with
the base line of the blastema. The mid-line of the grid
was located so as to correspond with the medial point of
the blastema. At sample areas 0 and D, the grid was super-
imposed over the blastema so that the lateral grid border
at its center touched the bordering wound epithelium at the
level of the one third or two thirds base-distal measurements

(Figure lb).

29

This manner of selection of the sample levels and
areas thus provided an accurate 3 dimensional survey of the
developing blastema. As-a result of the random orientation
of those limbs sectioned longitudinally several points on a
vertical (proximo-distal) axis and on a 360 degree circum-
ferential plane for each regeneration time were sur-
veyed, thereby allowing comparison of mitotic intensity and
cellular density distributions between external and internal
locations as well as between the base and the tip.

Cell counts on limbs cut in cross section were
made at 3 levels: At the top; at the center; and at the
base of the blastema (Figures 10, 12, and 13). The base line
for the blastema was determined as the most proximal cross
section from the tip possessing a continuous dermis. If in
any case differentiated stump tissues were contained within
the section, selected in this manner, the first distal sec-
tion not possessing differentiated stump tissue was used as
the base line. Two sections, each interspaced by one sec-
tion, were sampled at the center and base level. At the tip
of the blastema cross sections each interspaced by one sec-
tion were surveyed until a total of 200 cells had been
counted. With the exception of the tip, each section to be
scored was divided into 5 regions (an axial center and A
quadrants-—dorsal, ventral, anterior, and posterior (Figure
12). With the aid of the ocular grid, each region was
examined for mitotic activity and cell density. At the

center and base of the blastema, the ocular grid was

3O

superimposed over the cross sections of the blastema within
the A quadrants, (dorsal, anterior, ventral and posterior)
near the center region of the limb. Care was taken to avoid
any overlap in scoring between adjacent quadrants. Thus,
the cross sectional material added still another dimension
to the survey of the blastema providing information on the
internal distribution of cell density and cell division at

2 proximo-distal levels.

For surveying the asymmetrical blastemata, sample
levels at 3 different locations were selected. The center
level was selected as the best possible section based upon
size through the eccentric blastema which usually corres-
ponded to the medial section through the humerus of the
limb stump. The other levels studied were located 8 sec-
tions bilateral to the center level. The sample areas were
located within the sections as illustrated in Figure 11.
Sample area #1 was located at the tip; area #2 was at the
base of the blastema at its center and adjacent to the re-
maining humerus stump; area #3 was located laterally adja-
cent to the wound epithelium and distal to the most distally
located dermis; area #U was located laterally opposite to
and at the same level from the tip as area #3.

Because of their small size, aneurogenic limbs
were sampled at two locations only--at the tip and at the
base of the blastema. The base sample was located at the
center of the blastema immediately distal to the cartilagi-

nous humerus. Beginning at the center of the blastema and

31

moving bilaterally, several sections, each interspaced by
one section, were surveyed for mitotic activity until a
total of from 150 to 250 cells had been scored. More than
half of each blastema was surveyed by this method.

Mitotic index determinations were completed for
each limb series by dividing the total number of cells into
the total number undergoing mitosis and multiplying by
1,000. Densities were determined for all limb groups as
cells per unit surface area, the unit of area being the
2).

The number of cells per unit area was determined by dividing

area under one small square of the ocular grid (.027 mm.

the total number of cells by the total number of ocular grid
units.

The relationship between mitotic activity and
"vertical and horizontal" position within the regenerating
limb bud as a function of time from initiation of develop-
mental activity was demonstrated by a standard two-way
analysis of variance with equal replication. Measurements
from which the statistical treatment effects for longitu—
dinal sections were derived have been made between A time
periods and A physical locations on a disto-proximal axis.
Identical measurements were made for the cross sectional
material, but between A time periods and 5 physical (cen—
tral, dorsal, ventral, anterior, and posterior) locations.
There was a relationship between variance and mean within
the cells analyzed. The mean points of the cells, however,

were distributed in a significant linear relationship so

32

that the estimated error term was not biased. An analysis
of this type obviated, especially for limbs cut longitudin-
ally, the consideration of geometrical variation of blaste-
mal morphology which characteristically developed.

H3 thymidine treatment has not been used in the
present analysis for several reasons. Paramount among these
was that isotOpe labeling does not provide specific infor-
mation on the exact number of cells undergoing division.
This fact was made clear from the autoradiography studies
of Messier and Leblond (1960) who found that the radioactive
index (percent labeled cells) is not a true measure of pro-
liferation rate since it is also influenced by the rate of
incorporation of the label into the intermediate compounds
leading to DNA synthesis and by the duration of the period
of DNA synthesis.

Although tritiated thymidine is known to be incor-
porated into the DNA molecule, there remains the possibility
that it might be taken up by non—dividing cells such as
phagocytes, etc., and metabolized via a variety of enzyme
pathways. The availability of the tritium label to cold
limb tissues has been demonstrated by the experiments of
Riddiford (1960), in which she exchanged newt blastema epi—
thelium between labeled and unlabeled blastemata. Subse-
quently, in the blastemata provided with H3 thymidine, the

isotope was identified in the formerly unlabeled skin.

33

Barr (1963) has in addition noted for Hela cell
populations an alteration in the duration of the mitotic
cycle following exposure to exogenous H3 thymidine. Thus,
under some conditions, the thymidine treatment caused a
prolongation of metaphase. Therefore, it is possible that
this phenomenon could give a false impression of a high
degree of mitotic activity.

A specific problem in using tritiated thymidine
for the study of proliferation rate among mesenchymal cells
of the Amphibia is the time span between DNA synthesis and
cellular division. As shown by Hay and Fischman (1961)
this time lapse may constitute 3 to A days, a factor which
probably allows for wide spread cellular displacement
before mitosis. This characteristic of the label might
well be useful in studies of the relation of DNA synthesis
in developing blastema cells and their eventual locus of
division. For_the purpose of the present study, however,
in which specific mitotic indices are compared between
various levels of the blastema and between various stages
of blastemal development, the classical, if.more laborious,

methods promise the more clear—cut interpretations.

RESULTS

Series I:

General Morphology and Histology
of the Regenerate

 

 

Although aspects of the regeneration of the fore—

 

limb of the axolotl, Ambystoma mexicanum, have been previously
described (Kazancev, 193A; Faber, 1960), no general review

of the major morphological and histological features of the
process is available. Since the detailed analysis of cel-
lular proliferation patterns in the blastema is meaningful
only in relation to the regeneration process as a whole,

the following description seems necessary.

Following amputation through the distal one third
of the humerus, there is early closure of the wound and at
2A hours post amputation, the amputation surface is covered
by epidermal cell layers equal in thickness (6 to 7 cells)
to the epidermis of the normal stump skin. During the first
3 days there occurs a period of phagocytosis of cellular
debris resulting from the trauma of amputation. Numerous
leucocytes (phagocytes) are present and cellular fragments
and debris begin to appear in the wound epithelium. Tissue
debris and pycnotic nuclei are evident in the wound skin in

some limbs as long as 10 to 13 days after amputation,

3U

35

apparently arising from the general histolysis of stump
tissues during the period of dedifferentiation. Debris of
various sorts within the intercellular spaces of the epi-

dermis has previously been reported for Ambystoma larvae

 

(Thornton, 1938a) and in the newt (see review by;Singer and

Salpeter, 1961).

By A to 5 days after amputation, sarcolysis of the

 

transected muscle is apparent and general histolytic changes

 

of the cartilage are beginning. This period constitutes,
therefore, the beginning of the "dedifferentiation" phase,
(Thornton, 1938a; Hay, 1959), a process which may continue
in some limbs as late as 10 to la days after amputation.
These observations are similar to those of Polezajew and
Ginsburg (19u3) who noted in the forelimb of the axolotl
that tissue dedifferentiation was detectable even on the
lflth to 16th day of regeneration.

Blastema cell aggregation is first seen at ape

proximately 6 to 7 days after amputation and is accompanied

 

by the accumulation of a mesh of connective tissue fibers
between the wound epithelium and humerus stump. In subse-
quent stages the connective tissue mesh is obscured except
at the periphery by the aggregation of the blastemal cells.
It is possible, but has not been demonstrated, that the
connective tissue network may serve as a substructure upon

which the blastema cells migrate.

36

An interesting feature of limb regeneration in
the axolotl is the appearance at an early stage (8 to 10
days) of free melanocytes among the accumulating cells of
the blastema. For the most part these cells tend to remain
in topographical relation to their former (dorsal and an-
terior) location, and thus are found primarily in the dor-
sal and anterior regions of the regenerate immediately
beneath the wound epidermis (Figure 13). In many cases,
the.pigment granules from melanocytes are identifiable?
within the wound epidermis (Figure 5).

During the period of early blastema accumulation
(6 to 7 days) mitotic figures although not numerous, are
found among the distally aggregating mesenchymatous cells.
Infrequent mitotic figures are also observed among the de—
differentiating tissues proximal to the amputation plane.
Thus, although no quantitative data on the early regenera-
tion stages was obtained in this study, blastemal formation
must be, in part at least, the product of early cell pro—
liferations. These observations on the existence of cell
division during the early regenerative phases agree with
the data of earlier experiments (Polezajaw and Ginsburg,
1943) for the axolotl and for the newt (Chalkley, 195A;
Hay and Fischman, 1961).

Once mesenchymatous cell aggregation is well
linderway, it is possible to classify the developing blastema
into several morphological stages which are depicted in

.figures 6 to 9. The stages--"early bud," "mound," "cone,"

37

and "paddle"-—are equivalent to the Stages I, II, III, and
IV previously described for the axolotl forelimb by Faber
(1960). Under the conditions of this study (controlled
temperature, 200 C., uniform age and size of animals used,
and optimal feeding) the regeneration times for these
stages were 10, l3, l6, and 19 days respectively. The
shape of the early bud stage (10 days) is roughly conical
and its constituent mesenchymatous cells are relatively
homogeneous in distribution and not very densely aggregated
(Figure 6). Based upon calculations from 6 limbs, the
average length for a blastema at this stage is .31 mm. as
measured from amputation plane to distal tip of the mesen-
chymatous mass.

By 13 days the blastema has enlarged considerably
(average length = .60 mm.) and is mound shaped. As can be
seen from Figure 7, a blastema at this stage possesses a
definite pattern of blastemal cell density with the distal
one half having a considerably more densely aggregated mass
of cells than the proximal half. At 16 days, the regener-
ate is cone shaped and begins to show some anteroposterior
flattening at the distal tip. The blastema at this stage
has grown to an average proximo-distal length of 1.12 mm.
The number of mesenchymatous cells is considerably higher
distally and this high cellular density extends over ap—
loroximately two-thirds of the blastema. In the more ad-
‘Vanced 16 day blastemata there is some evidence of axial
Inesenchymatous cell condensation in conjunction with the

Stump humerus (Figure 8).

38

Continued antero-posterior flattening and distal
growth produce by 19 days a paddle shaped regenerate whose
average length is 1.41 mm. As can be seen from Figure 9,
a considerable proximal differentiation of the humerus ru-
diment has occurred and mesenchymatous cell condensation
has produced distinct patterns indicating the development
of the radius, the ulna, and digits #1 and #2. During

later stages of development (days 22 and 25) digits #3 and

 

#4 will arise from the mesenchymatous mass (arrow) at the
posterior periphery of the 19 day regenerate (Figure 9).
Visual observations on the distal digital condensations,
and also those destined to produce digits 3 and 4, revealed
numerous mitotic figures for these regions. It would seem
likely, although it remains to be demonstrated, that these

regions grow by apical cellular proliferation.

The Wound Epithelium

 

With this brief general account of limb regenera-
tion in the axolotl as background, we can now focus on one
of its aspects which has received little attention in the
past. This is the activity of the wound epidermis which
covers the amputation surface. During the first few days
after amputation, a gradual thickening of the wound epider-
mis occurs and by 6 to 7 days an "apical cap" 10 to 12 cell
layers thick has been produced. Apical proliferation within
the wound epidermis becomes especially prominent at the 7,

10, and 13 day stages studied (Table l) and there exists a

39

Table l.--Changes in mitotic activity at three levels within
the forelimb epidermis of the axolotl for three regeneration

 

 

 

 

periods.
Levels
*Site 1 Site 2 Site 3
Days Regeneration **MI ' MI ' ' MI
7 Days 1.0 .7 .6
10 Days 2.5 1.0 .8
13 Days 1.8 .8 .9

 

* Site 1 = Apical epithelium; Site 2 = Stump epithelium
just proximal to the amputation plane; Site 3 = Stump
epithelium 1 mm. proximal to the amputation plane.

** The mitotic index (MI) represents 3,000 (interphasic
+ mitotic) cells for each sample site. ’

striking difference in the mitotic rate in the apical cap as
compared to the 2 proximal stump levels of the normal skin.
As can be seen in Figure 3, there are numerous cell divisions
within the apical wound skin of a 10 day regenerate. It
should be noted that the cell counts reported in Table l

were obtained by counting 1,000 cells for each sample loca-
tion on 3 different limbs near their center for each regener-
ation time. Thus, a total of 3,000 cells were scored for
each limb. It would appear from the data of Table 1 that

it is mainly through the proliferation of the wound epithe-
lial cells that the apical cap increases in thickness from
10 to 12 layers of cells at 7 days, to 17 to 20 layers at

10 days of regeneration (Figure 6) Continued epithelial

4O

proliferation apparently is the cause of the formation of a
unique and prominent "epidermal lobe" which is first seen

at the 13 day stage, (Figure 7), and which usually reaches

a maximum size (35 to 40 cells, (cap + lobe) at approximately
16 days of regeneration, (Figure 5). Epidermal tongues have
been reported for the regenerating anuran limb stumps (Van
Stone, 1955) but the axolotl seems to be the only urodele
reported to possess this interesting structure. It should
be noted that the lobe as a distinct entity does not develOp
on every regenerating limb, but limbs lacking the lobe
possess an apical cap—like thickening, (Figure 8). It is
important to note that no significant difference other than
normal variability in mitotic activity or cell density could
be discovered between those limbs lacking the lobe and those
possessing it.

In later developmental stages, after 16 days of
regeneration, the apical lobe normally decreases in size and
gradually disappears. It may still be evident in some cases
as late as 22 days of regeneration and in these instances is
usually associated with the first digit projection. Thornton
(1962) has reported for a variety of salamanders other than
the axolotl, that Leydig cells are not found in wound epi-
thelia which support limb regeneration. He has found, on
the other hand, that wound epidermis which fails to support
limb regeneration is heavily laden with Leydig cells. It
is of great interest, therefore, to find that Leydig cells

are very numerous in the wound epithelium and apical lobe

41

of the regenerating axolotl limb (Figures 3 and 5). Kazancev
(1934) observed Leydig cells in axolotl wound epithelium and
Faber (1960) did also. The latter author reported a degener-
ation of these Leydig cells, however, and did not find them
ever to be very numerous in wound epithelium. Degeneration
of Leydig cells in the wound epithelium has never been ob-
served in the present study. Indeed, considerable mitotic
activity has been found in these cells in the wound epithe-
lium throughout regeneration.

It has recently been demonstrated (Kelley, unpub—
lished) that Leydig cells are a characteristic component of
larval skin and probably function in storing mucin materials.
The Leydig cells are thought to reach a peak in numbers at
the mid-larval stage and then gradually disappear, and are
totally absent in the adult skin. The significance of the
Leydig cells, if any, in the regeneration process is not
apparent. It is evident, at least for the axolotl, that
they do not function to inhibit the regeneration process
since they were very numerous in the wound skin during rapid

growth phase of the blastema.

Series II:

Analysis of Blastema Cell Proliferation
Along the Longitudinal Axis

The object of this experimental series was to
determine the mitotic index and cell density distribution

along the disto—proximal axis within the blastema.

42

Twenty-four axolotl forelimbs, 6 for each of the regeneration
stages of 10, 13, 16, and 19 days, were used in this experi—
ment.

For the purpose of demonstrating differences in
mitotic activity and cell density along the longitudinal axis
of the blastema and as an aid in the statistical analysis,
the blastema was arbitrarily divided, using a center section
(sample level #1) of the blastema as the base for the divi-
sions, into 4 disto-proximal zones which corresponded to the
4 disto-proximal sample areas at that level, (see Materials
and Methods, page 26, Figure lb and Figures 6 to 9). The
division times were drawn half way between the points of the
4 sample areas. The cell counts from the 4 sample areas of
each longitudinal section surveyed, thus from the 5 sample
levels (10 longitudinal sections), were then tabulated into
the appropriate zones (I, II, III, and IV) previously estab-
lished for the entire blastema. For example, because of the
geometrical design (roughly cone shaped) of the blastema,
which causes the longitudinal section at sample levels #2
and #3 (Figure 1a) to be shorter than those at the center of
the blastema, the tip counts for these levels were grouped
with Zone 11 of the blastema. Also the cell counts from the
other sample areas at these levels, for example B, C, and D
and also the sample areas of sample levels #4 and #5 were,
in like manner, tabulated into the blastemal zones which

corresponded to their individual lengths.

43

The statistical treatment of the data collected by
the method described above is represented in Table 2. As
can be seen, each block cell of the table contains the ob-
served mean of mitotic activity, the expected mean of the
mitotic activity, assuming that zones are independent of
time, and the corresponding increments of the sums of
squares for the interaction term. For example, in the block
cell at the upper left corner of the table (zone 1, day 10)

the figures represented are:

Observed mean = x = 3.723
Expected mean = Y = 2.849
Sums of squares = SS = .764

Thus, the magnitude and the sign of the deviation between
the 2 Y's or the SS value indicate the degree of deviation
from the assumed lack of interaction. The interaction term
may be defined as the expression of the relationship between
the different zones which changes as a function of time.
Thus, Zones 1, II, and III at 10 days of regeneration and
Zones 1, II, and III at 19 days of regeneration show a sig~
nificant interaction. Expressed in a slightly different
manner, it can be stated that at 10 days of regeneration,
Zone I possessed a greater number of mitotically dividing
cells than could be explained by the row and column Y's or
the expected mean value. The exact reverse of this situa-
tion can be noted at 19 days regeneration whereas in Zone I
fewer than expected mitotically dividing mesenchymatous cells

were observed. Therefore, for these zones and times there

44

Table 2.—-Two-way analysis of variance between regeneration
times and zones along the vertical axis of the forelimb blas-
tema of the axolotl.

 

 

Days of Regeneration

 

 

Zones 10 13 16 19
I X = 3.723 X = 3.362 X = 3.070 X = 1.514 x = 2.917
x = 2.849 X = 3.260 X = 3.074 x = 2.486
ss= .764 ss= .010 ss= .000 ss= .945
11 Z = 1.945 X = 2.702 X = 2.472 X = 2.602 x = 2.430
X = 2.362 x = 2.773 x = 2.587 x = 1.999
ss= .174 ss= .005 ss= .013 ss= .364
111 X = 1.280 X = 2.158 X = 2.019 X = 1.514 E = 1.742
X = 1.674 x = 2.085 x = 1.899 x = 1.311
ss= .155 ss= .005 ss= .014 ss= .041
IV X = .945 X = 1.312 X = 1.229 X = .805 i = 1.073
x = 1.005 x = 1.416 x = 1.230 x = .642
ss= .004 ss= .011 ss= .000 ss= .027
x = 1.972 x - 2.383 x — 2.197 I = 1.609 x = 2.040
Source df ‘ MSS F
Sites 3 15.556 25.090*
Days _3 2.665 4.298*
Interaction 9 1.688 2.723
Error 80 .620
Total 95

* Significance at the l per cent level

The Expected Mean (Day 10, Zone I) = The Grand Total Mean
adjusted for days [i + (idlO - i)] or [2.040 = (1.972 —
2.0400] and then adjusted for zones [i + (Yd10 - Y) +
(Izl - i)] or [2.040 - .068 + (2.917 - 2.040)]. The ex-
pected mean therefore equals 2.040 - .068 + .877 = 2.849.

45

has been an effect upon the mitotic rate which has caused a
deviation from the expected mitotic incidence, indicating
that interaction has occurred between time and zone of the
regenerating blastema.

With this general information as background, it
is now possible to consider the mitotic patterns which are
apparent for the individual regeneration times, 10, 13, 16,
and 19 days. It should be reemphasized here that the analy-
sis presented is an experimental test of the hypothesis for
apical proliferation proposed by Faber (1960) and therefore
regeneration stages equivalent to those used in his experi-
ments (Stages 1, II, III, and IV) were selected for this
study.

At the "early bud stage" (10 days regeneration),

 

which represents the first measurable accumulation of mesen-
chymatous cells, a striking pattern in the mitotic activity
has already been established. It can be seen (Figure 14

and Table 2) that at this early time there exists a definite
linear difference in blastemal cell mitotic activity, and
the mitotic index is obviously highest at the tip (Zone 1).
The proliferation rate at the tip represents a 300 percent
higher rate than the base index. As stated above, the
interaction term (SS = + .764) of Zone I at 10 days of re-
generation (Table 2) indicates that a much higher than ex—
pected rate of mitotic activity was found. Significant,
however, is the fact that just the reverse effect can be

noted for Zones II and III where lower than the expected

46
rates in proliferation were observed. Thus, there exists an
effect between these 3 zones (sites) and the time in regener—
ation” No clear explanation for this deviation from the
expected pattern is evident, but it is interesting to note
that the tip region is particularly close to the epidermal
cap. All mesenchymatous cells in Zone I (tip region) are
more closely associated with the apical wound epithelium
than are the majority of cells which occupy the other zones
of the 10 day blastema. This observation has been borne
out by specific measurements of these zones with the aid of
the ocular micrometer. To illustrate this point, all of the
mesenchymatous cells in the tip zone of the blastema shown
in Figure 6 lie within .067 mm. of the wound epithelium
while, on the other hand, more than half of the cells
counted in Zone 11 were more than .067 mm. from the nearest
part of the wound epithelium.

What is again striking for the 10 day blastema is
that the apical wound epithelium which is associated with
Zone I shows an intense rate of cell proliferations which
is not found, as determined by cell counts, for the more
proximal levels of the wound epithelium adjacent to Zones
II, III,and IV of the blastema. Again using the blastema
represented in Figure 6, a mitotic index of 2.7 per cent was
observed for the tip epithelium while the proximal wound
epithelium possessed a rate of 1.2 per cent which represents,
therefore, a 150 per cent difference in mitotic activity.

It is interesting to speculate on the possibility that the

47

observed temporal relationships between the intense mitotic
activity of the apical wound epithelium and the underlying
apical mesenchymatous cells may also represent a causal re—
lationship (see also Faber, 1965).

In contrast to the proliferation patterns at the
10 day stage is the distribution of the mesenchymatous blas-
tema cells. No apparent pattern is evident as the cells
are distributed relatively equally at all disto-proximal
levels and the density is not very great (Figure 14). This
feature of homogeneity in cell density at this stage seems
to indicate that the high rate of proliferation, as found
for the tip zone, may just have begun.

Three days later at the "mound stage" blastema

 

the mitotic rate still shows a disto-proximal gradient and
there are substantial increases in the proliferation rates
within Zones II and III while Zone IV shows a small in-
crease (+ .367 per cent) in the mitotic rate. For Zone I

a slight decrease (— .361 per cent) in the mitotic rate from
the high rate of the previous stage was noted, but the rate
was still considerably higher (over 200 per cent) than that
for the base zone. The most striking feature for the "mound
stage" blastema is the observed increases in proliferation
rates for Zone 11 and III. It would appear that the stimulus
for cell proliferation, whatever it may be, has by this
stage, "diffused" into the more proximal areas (Zones II and
III) of the blastema. It is most interesting to note, in

this regard, that the majority of the mitotic figures recorded

48

for Zone III were found within the dense mass of mesenchy-
matous cells which extended proximally from Zone 11 into
this zone (Figure 7).

As can be seen from Figure 14 and illustrated by
the photomicrograph of Figure 7, the density of the mesen-
chymatous cells at this stage is greatest in roughly the
distal half of the blastema; thus Zones I and II are more
densely populated than are Zones III and IV. However,
there is no sharp transition in density but rather a gradual
disto-proximal decline in the number of mesenchymatous cells
across Zone 111. These findings seem to indicate a corre-
lation between the density and the active proliferation of
the mesenchymatous cells. Such a correlation seems to be
borne out by the fact that the greatest difference and thus
the greatest decline in mitotic divisions along the disto-
proximal axis of the blastema has occurred between Zone III
and Zone IV (Table 2). It should be noted, while considering
the "mound stage" regenerate, that Faber (1960) has identi-
fied this stage (Stage II in his studyL due to the dense
accumulation of mesenchymatous cells at the distal tip, as
the approximate time for the beginning of apical prolifera-
tion.

It is significant to note here that the wound epi-
thelium at 13 days regeneration does not possess the striking
difference in mitotic activity between the tip and the base
that was observed for the 10 day stage. Indeed, cell counts

involving a total of 1,000 cells within the apical wound

49

epithelium adjacent to Zones II, III, and IV for the blastema
shown in Figure 7, revealed no difference in the mitotic
index between these 2 levels. Each location possessed a
2.0 per cent rate in cell divisions. In view of this fact
it is interesting that there has been an overall increase in
mitotic proliferation within the blastema with very notice-
able increases for Zones II and III. It is also interesting
that with the decrease in mitotic division in the apical
wound epithelium (— .7 per cent) there has also been a de-
crease in proliferation in the apical mesenchyme of Zone I
(- .292 per cent).

The gradient in the mitotic rate along the disto-
proximal length of the blastema is still very much evident

at the "cone stage" blastema (16 days regeneration). A11

 

zones at this time have undergone slight decreases in the
number of mitotic divisions, but there still remains a two-
fold difference in the proliferation rates between the tip
and the base regions of the blastema. A very significant
aspect of the observed decreases in mitotic activity at
this stage is that Zones III and IV decreased less in mito-
tic activity as compared with the previous stage than did
Zones I and II. A most interesting parallel to this situa-
tion was the observed decreases in the mitotic rates within
the wound epithelium. Although decreases were found in
both the apical and proximal regions of the wound epidermis,
a larger decrease was noted for the apical part. Indeed,

cell counts of 1,000 cells for both these regions showed a

50

1.4 per cent mitotic index for the tip epithelium and an
index of 1.6 per cent for the proximal part. Both regions
possessed rates of 2.0 percent at 13 days regeneration.
Thus, a striking correlation between the mitotic activities
of the wound epithelium and the underlying mesenchymatous
cells has, as it was for the 10 and 13 day regenerates,
again been revealed.

The mesenchymatous cells at the "cone stage" con—
tinue to form dense accumulations, particularly throughout
the distal two-thirds of the blastema. The most pronounced
increases in cell density were evident for Zones I and II
which apparently reflect the high degree of mitotic activity
observed for those zones at the previous stage (13 day re-
generation). Increases in cell density were also found for
Zones III and IV and the base increase (Zone IV is apparently
a measure of the axial "condensation" of mesenchymatous cells,
which is beginning at this stage (Figure 8). It should be
pointed out that the term "condensation" as used in this
thesis does not imply cell differentiation but rather is
used to describe the dense aggregations of the mesenchymatous
cells which form the preliminary patterns of the future car-
tilage elements of the regenerating limb. The inference that
these cells are not differentiating has been strengthened by
the repeated observations that the cells in these early con—

densations are undergoing divisions.

51

At 19 days regeneration, the "paddle stage," pro-

 

found alterations have occurred in the mitotic patterns of
the developing regenerate. This is especailly evident for
the tip zone, where the proliferation rate is only one half
as great as the mitotic value found for this zone at 16 days
of regeneration. Zone II, on the other hand, shows a slight
increase in mitotic rate over the preceding stage, while
Zones III and IV both have undergone decreases in the number
of mitotic divisions. Similar to the changes in mitotic
patterns are the alterations in the cell density distribu-
tions at the paddle stage. A drop in cell density was re-
corded, for example, in Zone I and an increase was deter—
mined for Zone II. Zones II and IV, however, both were
found to be more densely p0pu1ated than the equivalent

zones of the 16 day stage. As can be noted from Figure 9,
considerable cartilage differentiation is underway for

Zone IV.

Referring to the "interaction" term, or the sums
of squares (Table 2), it is apparent that Zones 1, II, and
III of the 19 day blastema have deviated considerably from
the expected mitotic rates. The mitotic activity for these
zones represents the exact reversal of the mitotic activity
found for the "early bud stage," (10 days regeneration).
Thus, at Zone I the mitotic rate is considerably lower than
the expected value and Zones II and III are slightly lower
than the expected mitotic rates. These differences between

the actual and expected values are perhaps explainable on

52

the basis of the histological alterations which have occurred
within the developing regenerate by 19 days of regeneration.
The tip region, for example, no longer possesses a homOe
geneously distributed dense mass of mesenchymatous cells
which was evident at the cone stage, but rather shows "con-
densation" patterns for the future digits (#1 and #2). The
position of the sample area for Zone I at 19 days of regener-
ation, for experimental consistency, was the same as that
used for the other stages; that is, at the tip and near the
center of the blastema which was therefore between the two
digital "condensations." It was thought possible that a
high rate in mitotic activity might be associated with the
digital "condensations." To test this possibility, 1,000
cells from the rudimentary digital masses of the 19 day
regenerate represented in Figure 9 were scored. The result
of this count revealed a significant 3.3 per cent mitotic
index for these regions which is a striking 180 per cent
higher than the rate found for Zone I at 19 days regeneration
(Table 2).

The higher mitotic rate (+ .6) than the expected
rate for Zone II is due perhaps to the fact that this zone
at 19 days of regeneration includes not only the distal
portion of the mesenchymatous condensation for the future
radial and ulnar cartilages but also the dense mass of
mesenchyme (arrow, Figure 9) which is destined to give rise
to the cartilages of digits 3 and 4. Thus, the calculated

mitotic rate for Zone II is a measure of the intense mitotic

53

activity of the mesenchymatous cells of the future limb
skeleton. The mitotic rate for Zone III, which is slightly
higher than the expected rate, can be attributed to the pro-
liferation of blastemal cells accumulating to form the proxi-
mal parts of the develOping autOpodium (radius and ulna).

In conclusion, it is clear (Table 2) that a lon-
gitudinal gradient of mitotic activity is present in the re-
generating axolotl forelimb. The mitotic rate is highest
at the blastemal tip at 10 days following amputation and,
although it progressively narrows, this mitotic differential
is maintained until the 19 day regenerate. In the latter
stage, also, if one includes the digital mesenchymal aggre-
gations, the tip still maintains a higher rate than Zone II

in actively dividing cells.

Series 111:

Analysis of Mitotic Activity within Regions
of the Blastema Studied in Cross—Section

 

 

In a recent investigation of blastemal growth in

the newt, Triturus viridescens, it has been suggested,

 

(Singer, Ray and Peadon, 1964) that the extent of cellular
movement and multiplication of mesenchymatous cells, during
the period of most rapid growth of the blastema is not the
same everywhere in the blastema but is greatest near the
regenerating nerve trunks of the limb. The evidence for
this was derived from experiments in which cross—sectional

regions (central, dorsal, anterior, ventral, and posterior)

54

of an early regeneration stage, which Singer and his co-
workers have termed "moderately early" were infused with
small quantities of Nile blue sulphate. Subsequently, dur-
ing the rapid growth phase of the blastema, it was observed,
by following the displacement of the dye, that some regions,
(ventral, posterior and the adjacent central region), gave
rise to large areas of the new limb, whereas the other
regions (dorsal and anterior) contributed very little
material to the developing structure. According to these
investigators, cells derived from the ventral, posterior,
and central areas of the early regenerate produced the hand
and fingers and most of the lower arm. In other words, the
cells of these regions formed the anterior and dorsal parts
of the regenerate as well as the ventral and posterior
structures.

Histological studies (Singer et_al,, loc. cit.,
1964) revealed that the 3 major nerve trunks, in the distal
part of the forelimb, are located mainly ventrally and pos-
teriorly within the soft tissues near the bone. They also
discovered that although regenerated nerve sprouts were
present throughout the blastema, the fiber concentrations
were more abundant in those areas overlying the regions
(mainly ventral and posterior) possessing the major nerve
trunks. In view of this fact it seemed apparent that the
contribution of blastemal cells from the individual cross-
sectional regions of the early blastema to the formation

of the new limb was correlated with the position of the

55

nerve fibers and trunks. This was further emphasized when
nerves were deviated preceding amputation from their normal
positions (ventral and posterior) into the dorsal and anterior
regions of the limb, with subsequent increased cellular con—
tribution to the blastema from these areas.

Although it was not the initial intent of the
present study to test for a correlation between cellular
proliferations and nerve fiber concentration, in view of the
experiments reported above and especially since the axolotl
blastema was found to grow by apical proliferation, it seemed
important to re-examine the blastema for mitotic activity
using sampling patterns which would approximate the dye in—
fusion regions studied by Singer, Ray and Peadon, (1964).

For this purpose axolotl regenerates, 2 for each
time period, were sampled at 10, l3, l6 and 19 days after
amputation. Cell counts were made on cross-sections at the
center and at the base of the blastema within 5 cross-
sectional regions, (center, dorsal, anterior, ventral, and
posterior), (see Materials and Methods and also Figure 12).
It is important to note that the precise dorso-ventral and
antero=posterior orientation of the cross-sections was
checked using the pigmentation patterns of the blastema and
also by the identification within proximal cross-sections

of the positions of the major nerve trunks and other stump

tissues.

56

The statistical treatment of the data (mitotic acti-
vity between the 5 regions for each regenerate stage) is pre—
sented in Table 3 and the mitotic indices and cell density
distributions are illustrated in Figure 15. For the individual
regeneration times it can be observed from Figure 15 that
there exists no significant difference between the regions,
especially those regions which overlie the major nerve trunk
and those which do not. Thus, for day 10 a relatively high
rate of mitotic activity is apparent for the dorsal (1.1 per
cent) and anterior (1.4 per cent) regions and the central
1.0 per cent) and posterior (1.0 per cent) regions are slightly
lower. What is most significant is the fact that the pro—
liferation rate of the ventral region which overlies the
greatest concentration of nerve fibers is considerablylower
(.6 per cent) than either of the other regions and especially
the dorsal and anterior regions.

At the 13 day stage, which represents the most
rapidly growing period of the blastema, no clear pattern in
mitotic activity nor cell density between the cross-sectional
regions has develOped. The anterior region was found to be
the most active (2.5 per cent) and was followed in activity
by the ventral region (2.0 per cent), then the posterior
region (1.9 per cent), the dorsal region (1.8 per cent) and
lastly the central area (1.0 per cent). Thus, the average
rate of mitotic activity (2.1 per cent) for the anterior and
dorsal regions is considerably higher than the other 3 regions,
(overlying the nerve trunks) which collectively possessed an

average rate of 1.3 per cent.

57

Table 3.--Two-way analysis of variance between regeneration
times and regions of the forelimb blastema of the axolotl
studied in cross section.

 

 

 

Source df MSS F
Time 3 2.193 9-37*
Quadrant 4 .808 3.543
T x Q 12 .135 .577
Error 20 .243

Total 39

* Significance at the l per cent level.

 

Regions Center Ventral Dorsal Posterior Anterior

X's 1.083 1.375 1.406 1.625 l.938**

 

** The continuous line indicates a lack of significance be-
tween the mean mitotic values of the regions which over-
lie it.

 

Test for Significance according to Duncan's New Multiple
Range Test.

At 16 days regeneration a drop in the proliferation
rate was noted for all regions but the magnitude in mitotic
activity remained essentially the same between the 5 regions.
Thus, the anterior region was highest in mitotic divisions
and was followed by the ventral and posterior (which were
equal), then the dorsal and central areas. The recorded drop

in the mitotic activity for all regions at this stage (Figure 15)

58

might be due to changes in cell distribution, especially at
the base of the blastema. For example, the peripheral re—
.gions of the blastema (dorsal, anterior, ventral, and pos-
terior) have become sparsely populated in comparison with
the previous stage while the center region has increased in
density. This effect is apparently the result of the axial
blastemal cell condensation which is beginning by 16 days
of regeneration.

For the 19 day stage, which is after the rapid

 

growth phase of the blastema, the patterns in mitotic activ-
ity appear to reflect the orientation of the mesenchymatous
cells which are aggregating to form the rudimentary patterns
of the future skeletal parts. The anterior and posterior
regions at this stage are equal and highest in mitotic ac-
tivity, and it should be noticed (Figure 15) that they are
quite densely populated with mesenchymatous cells. These
results are apparently due to the antero-posterior flatten-
ing of the regenerate by this time. The dorsal and ventral
regions are also equal in cell proliferation but are con-
siderably lower as compared to the anterior and posterior
regions. The center region is the lowest in mitotic pro-
liferation but considerably more dense than the other
regions.

In summary, as determined from the statistical
analysis of the mean mitotic values (lower row of Table 3)
there exists no significant differences between the indi-

vidual regions of the growing blastema. However, significant

59

differences were noted between two combinations of the 5
regions. For example, the dorsal, ventral, central, and
posterior regions, which are not significantly different
from one another, are collectively different from the an-
terior region. The posterior region, however, is not sig-
nificantly different from the anterior region. The most
significant feature to be noted here is the fact that the
mitotic indices of the regions overlying the major nerve
trunks are not significantly different from those which do
not. What is particularly important to note is the fact
the mean mitotic value for the ventral region (overlying
the major nerve trunks) of the blastema is considerably
lower than the other regions, except the central region,
and also that the anterior region which apparently does not
overlie any major nerves possesses the highest mean mitotic
rate. Thus, no correlation between the trunk nerve fibers
and the multiplication and density of cells within the
various regions of the blastema could be found. Indeed,
just the reverse was observed, with more mitotic divisions
recorded for those regions which did not surround the nerve
trunks. This was not an unexpected result since it was
previously demonstrated that following the early accumula-
tion of the blastema cells the apical proliferation of these
cells produces the major part of the growing blastema.

It is likely that the observed differences be-
tween the findings of the present study and those of Singer,

Ray and Peadon (1964) are due mainly to species differences

60

between the animals used for the experiments (Singer and his
co—workers used the adult newt) and also to differences in
the regeneration responses between adult and larval forms.
Other experiments, for example, especially those involving
the phenomenon of regression (the larval limb regresses
following amputation and denervation while adult limbs do
not) have made it common knowledge that there are important
differences between adult salamander limbs and larval sala-

mander limbs in their responses to injury.

Series IV:

Analysis of Mitotic Activity
within Asymmetrical Blastemata

 

 

It has been demonstrated in a preceding section
(Series II) that the axolotl blastema grows by apical pro-
liferation and since a striking relation between the mitotic
activity of the wound epithelium, especially the apical
portion, and that of the underlying blastemal cells was in-
dicated, a further investigation of this relationship using
asymmetrical regenerating blastemata seemed important.

Asymmetrical blastemata were produced by the re—
moval of a strip of skin (dermis and epidermis) running
parallel and adjacent to the border of the original ampu-
tation plane of the limb, (see Materials and Methods). It
should be noted that to produce asymmetrically oriented
blastemata on the large limbs used in this study, 2 skin

removal operations were necessary. The first operation was

61

undertaken at the time of apical cap production (6 to 7 days
of regeneration) and was followed by one additional operation
2 or 3 days later at 9 or 10 days after amputation. It
should be noted that in some limbs there was a tendency, with
time, for the angle of asymmetry of the blastema with the
limb stump gradually to be reduced. The asymmetry of the re-
_generates was expressed as a deviation in the direction of
outgrowth of the blastema so that the regenerate formed an
angle with the limb stump which varied from approximately
45 to 75 degrees.

Three asymmetrical blastemata for 3 different time
periods (l4, l6, and 18 days of regeneration) were sampled
(3 sample levels) for mitotic activity and cell density at
4 different sample areas (1, 2, 3, and 4) as illustrated in
Figure 11. It should be emphasized that the sample areas
were specifically selected to determine the changes in mito—
tic cell distributions with time and with relation to the
orientation of the apical wound epithelium. Thus, counts
were made at the tip and the base to determine the axial
distribution ofmitotic cells and at the anterior and pos-
terior borders of the blastema (sample areas #3 and #4) to
determine the lateral distributions of mitotic cells.
Sample area #3 was selected as representing the approximate
site of the blastema (apical to the limb stump) before shift—
ing the apical cap, and sample area #4 was selected for the
purpose of detecting any increases in mitotic activity and

cell density which might develOp in relation with the

62

extensive wound epithelial surface which existed on the an—
terior border (operated side) of the asymmetrical regenerate
(Figure 11).

It should be pointed out before considering the
mitotic patterns and cell density distributions for the
three regeneration times that because of the 2 operations
for skin removal and also the several exposures to the anes-
thesia (MS 222), both of which were found to slow the re-
generation process, the regeneration stages examined (14,
16, and 18 days) are not comparable to those used in the
preceding section (Series II) but they do represent the
approximate period for the rapid growth of the blastema.

As can be noted from Figure 16 for the first
asymmetrical regeneration stage studies (14 days) there
exists a high incidence of mitotic activity (2.9 per cent)
at the tip and significantly lower rates at the base area
and also area #4. If a comparison were made between the
14 day asymmetrical blastema and the regeneration stages
of Series II, based upon size, the 14 day anymmetrical
state is approximately equivalent to the 13 day undeviated
regenerate. Histologically, however, the two stages are
not strictly comparable, since the densely aggregated mass
of mesenchymatous cells of the asymmetric blastema was
found to be more closely associated with the stump tissues
than is the case in the undeviated blastema. This is sup-
ported by the observation that each of the proximal sample

areas, (2, 3, and 4) possesses a much higher mitotic rate

63

and cell density than could be found at equivalent distances
from the stump tissues in normally oriented 13 day blastema.
These unexpected results in mitotic activity and cell divi—
sion were apparently a function of displacing the apical
epithelial cap. What is most interesting, however, for the
14 day asymmetrical blastema, is that sample area 3 (apical
to the humerus stump) was found to have a strikingly high
mitotic rate of 2.3 per cent which is not greatly different
from the apical blastemal rate (area #1). It was also
noted by histological examination that the mesenchymatous
cells were more closely aggregated along the distal wound
epithelial border of the blastema than they were next to

the wound epithelium at the anterior border (Operated side)
of the limb. It is possible that the unusually high rate
of divisions at area #3 and the peculiar cell density dis—
tribution in the 14 day asymmetrical regenerate may repre—
sent the old patterns of blastemal cell orientation which
developed before the apical epithelial cap was displaced.

At 16 days of regeneration considerable alteration
in the mitotic and cell density patterns has become evident
(Figure 16). There remains, however, a high rate of pro-
liferation at the tip (sample area 1), but decreases in the
mitotic rate were observed for areas 2, 3, and 4. The
interesting feature here is the fact that sample area 3
(apical to the humerus stump) underwent the greatest de-
crease (- 1.4 per cent) while area 4 (adjacent to the wound

epithelium at the anterior border of the regenerate) showed

64

the smallest drop (-.7 per cent). It appears, therefore,
that there has been a reorientation of the dividing mesen—
chymatous cells toward the tip and anterior (operated side)
border of the regenerate. Further support for this can be
drawn from the observed changes in cell density (Figure 16).
It can be seen, for example, that decreases in cell density
have occurred for both sample areas #2 and #3 while increases
were found for areas #1 (tip) and #4 (anterior border). The
changes in density distributions are clearly apparent in
Figure 11, which also demonstrates the unusually thick
nature of the wound epithelium which has developed at the
anterior border (operated side) of the regenerate.

At 18 days of regeneration the mitotic divisions
at the tip have decreased (— .9 per cent) but are still more
numerous than at any of the proximal regions. The most sig—
nificant feature at this stage, as can be seen in Figurel6,
is that the mitotic rate (+ .5 percent) and cell density
have increased at area 4 (anterior border next to wound epi-
thelium) while no change in mitotic activity' and only a
slight increase in density were apparent for area 3.

It would appear, therefore, that the asymmetrical
disposition of the apical epithelium may somehow influence
a corresponding asymmetry of blastemal cell proliferation,
particularly since the change in proliferation patterns of
the blastemal cells follows in time the change in location

of the thickened apical epithelium.

65

Series V:

Analysis of Mitotic Activity
within Aneurogenic Limbs

Since it has been demonstrated in a preceding
section of this study (Series II) that innervated regenera-
ting axolotl forelimbs grow by apical proliferations, the
question was posed as to whether aneurogenic limbs also
might regenerate by a similar mechanism. Because axolotl
embryos were not available for this study, aneurogenic or
sparsely innervated larvae were produced using Ambystoma
opacum embryos instead. However, due to a high mortality
rate, only 10 aneurogenic limbs were available for this
experiment.

Aneurogenic limbs were sampled at 10, l2, l4, and
16 days after amputation and surveyed for mitotic activity
and cell density in the distal third and basal third of the
blastema. Two limbs each were scored for days 10 and 16
while 3 limbs each were sampled for days 12 and 14. The
distributions of mitoses and cell densities found in the
aneurogenicblastemata are shown in FiguretlJ; The number of
cells counted for each of the 2 areas sampled (tip and base)
depended upon the size of the blastema and ranged from 150
to 250 cells. Thus, for the small blastema (10 days of re-
generation) 300 cells were scored for each limb and for the
larger blastemata (l2, l4, and 16 days of regeneration) 500

cells.

66

As can be determined from Figure 17, the overall
pattern of growth for the regenerating aneurogenic limbs as
determined from the mitotic activity shows a close parallel
with the growth patterns established for the axolotl fore—
limb blastema. Even so, however, some differences are ap-
parent and should be considered. Thus, at 10 days of re—

generation, which is an "early mound" stage, the mitotic

 

rate of the blastema is relatively low and there exists
little difference in mitotic division between the tip and
the base.

Due to its size, measuring only .11 mm. in length
from the original amputation level to the tip of the apical
mesenchyme, the 10 day aneurogenic blastema (at 16 degrees
C.) apparently represents the initial accumulation of blas-
temal cells. This also seems to be borne out by the fact
that the mesenchymatous cells at this early time are equally
distributed throughout the blastema (Figure 17) and are not
very densely aggregated (2.9 cells per unit of the ocular
_grid for both the tip and the base).

For the 12th day of regeneration, which approxi—
mates the development in both mitotic activity and shape
that of the 13 day axolotl blastema, the mean mitotic rate
has increased markedly showing a high degree of mitotic
activity at the tip (mitotic index = 4.7 per cent) while
the base rate shows only a slight increase (.5 per cent)
over the 10 day stage. An interesting feature of the 12

day blastema is that the apical wound epithelium is much

67

more pronounced in its development than was observed for the
10 day stage. It is also to be noted that the apical wound
epithelium at the 12 day stage was found to be highly active
mitotically. This correlation between the mitotic divisions
of the apical wound epithelium and the underlying blastemal
cells is of interest since a similar relationship, although
beginning at an earlier stage, was noted for the axolotl
blastema. The mesenchymatous cell density distribution for
the 12 day regenerates, as can be noted in Figure 17, begins
to reflect the intense proliferation of the mesenchymatous
cells in the distal area of the blastema. The cell density
at the base of the blastema, on the other hand, has remained
relatively constant as compared to the 19 day blastema.

At the "cone stage," 14 days regeneration, the mean
mitoticzhnnncof the blastema has decreased by .3 per cent as
compared to the 12 day regeneration stage; and a very notice-
able drop in proliferation is evident for the distal third
of the blastema (— l. 2 per cent) but the proximal third
shows an increase (+ .4 per cent) over the preceding state
(Figure 17). It is important to note that even with these
changes in mitotic activity the distal rate is still con-
siderably higher (+ 1.3 per cent) than the rate at the base.
The cell density has increased in both the distal and proxi-
mal regions of the 14 day blastema but the cells are slightly
more densely aggregated at the base than at the tip. This
difference in cell density pattern is apparently a reflection

of the axial "condensation" which is underway at this stage.

 

68

For the 16 day aneurogenic blastema (early paddle

 

stage) the trend for a decrease in apical proliferation and
the decrease in the basal rate is continuing. It should be
noted that the base counts for this stage were—made just’
distal to the axial differentiation of cartilage which is
beginning at this time and represent, therefore, the high
proliferation rate (Figure 17) of the mesenchymatous cells
which are aggregating to form the future skeletal parts.

As can be noted (Figure 17) the mesenchymatous cell density
has increased for both the apical and basal areas.

In summary, the overall growth rate of the aneuro-
genic blastema shows a similar pattern to that found for the
axolotl blastema. Based upon the total mitotic activity
(tip and base) it can be observed from Figure 17 that the
aneurogenic blastema shows a rapid growth increase from 10
days with a peak at 12 days of regeneration and then a

gradual decline as develOpment proceeds.

DISCUSSION

In the axolotl during the initial days of regenera-
tion, there occurs at the apex of the limb stump a gradual
thickening of the wound epidermis which constitutes, at 6 to
8 days post-amputation, an "apical cap" 10 to 12 cell layers
thick. This apical thickening is apparently the product of
both apical epidermal cell divisions, (see page 39) and the
distal movements and divisions of cells from proximal epi-
dermal levels. Concomitant with the formation of this
thickening, and immediately underlying it, forms the initial
collection of mesenchymatous cells for the blastema. For the
axolotl then, as for other salamanders (see Thornton, 1960)
the develOpment of an apical epidermal thickening is cor-
related with blastemal cell aggregation.

Faber (1960) reported that preceding apical lobe
develOpment in the axolotl the thickening of the wound epi-
thelium was not vtry pronounced so that he was unable to
observe a distinct cap. The difference between Faber's
findings and the present observations on early wound skin
development is minor and might be resolved by a clearer
definition of what constitutes an "apical cap." According
to Thornton (1954) the epidermal cap thickening is at least

twice that of the remainder of the wound epithelium at the

apex of the regenerating limb. If this criterion is used,

69

70

a cap is easily discernible at the tip of the 6 to 8 day
axolotl regenerate. Perhaps, however, we might simply con-
sider the "apical cap" merely as the thickest part of the
wound epithelium and not importantly different in function
or structure (Thornton, personal communication).

At 10 days regeneration (early bud blastema) a
high degree of mitotic activity of the apical epidermis
was observed. Subsequent to this activity, and apparently
resulting from it, there developed a secondary wound epi-
dermal thickening--the "apical lobe." The possibility that
rapid apical epidermal proliferation was the major source
for apical lobe develOpment was strengthened by counts of
proximal and distal epidermal cells. The apical mitotic
index at the early bud stage was found to be an impressive
200 per cent higher than that for the stump epidermis.
Mitotic figures are found in the wound epidermis throughout
regeneration. Indeed, even Leydig cells, a common component
of both the wound epithelium and the apical lobe as deter-
mined in this inyestigation were repeatedly found in various
stages of division.

These findings are apparently at variance with

those reported for the adult newt Triturus viridescens

 

(Chalkley, 1954; Hay and Fischman, 1961). These investi-
gators observed a high degree of mitotic activity associated
with the proximal epidermis but bound a limited number of
divisions in the wound epithelium, especially the apical

region. Chalkley (1954) demonstrated by cellular counts a

71

distinct peak in mitotic activity in the proximal epidermis
at 7 and 13 days post-amputation. Only later, at 19 days
regeneration, did the peak of activity move into the blas-
tema. Even then the mitotic index remained closely related
to that of the stump, being highest at the base of the blas-
tema. A comparable stage of regeneration in the axolotl,

as far as can be determined, would be the early bud (10 day)
blastema, a stage in which considerable mitotic activity in
the apical epidermis is found. Consistent with the mitotic
index calculations of Chalkley are the autoradiographic
studies of Hay and Fischman (1961) for DNA synthesis in

limb regeneration in the adult newt. These workers reported
epidermal incorporation of thymidine at levels proximal to
the amputation surface, followed by distal migrations, and
only in a few instances were cellular divisions found within
the wound epithelium. Due to the dilution of the isotope
label, it was determined that those cells which divide api-
cally probably do so only once.

Although a systematic study of mitotic activity of
the limb epidermis was not a feature of the present investi-
gation, 3 periods of regeneration (7, 10, and 13 day regener-
ates) were surveyed for epidermal cell divisions. These
stages represent critical periods in wound epithelial and
blastemal development; at 7 days cap formation occurs; at
10 days the apical lobe is just about to begin its develOp-
ment; and at 13 days the intense mesenchymal proliferation

within the blastema is underway. For each of these stages

72

a high degree of proliferation was revealed for the apical
wound epidermis. The apical rate as illustrated in Table l
was considerably higher than the rate for either of the
proximal levels sampled. It is suggested that the incompat-
ability of the findings of this study on epidermal divisions
with those reported for the newt are due mainly to regenera-
tion differences characteristic of the species used. Another
possibility is that they represent regenerative differences
which might exist between adult and larval forms. Whatever
the underlying causes, it is apparent that the differences
in epidermal mitotic activity as reported above are signifi-
cant.

The influence of the wound epidermis on blastema
formation and development has been studied extensively over
the past several years and numerous theories have been
offered to explain its activities (cf. Introduction). Para-
mount among these considerations was the hypothesis that the
apical epidermis supports the accumulation of mesenchymatous
cells and operatgs in the establishment of proliferation
patterns within the blastema.

Thornton (1954, 1958) has described the apical
epidermal cap as a transitory sturcture in the regenerating

larval limbs of several species of Ambystoma. As an identi-

 

fiable entity the cap was present only during the phase of
limb regeneration which involves the accumulation of mesen-
chymatous cells for blastemal develOpment. As indicated

from cap deviation experiments (Thornton and Steen, 1962),

 

73

blastemal growth and differentiation may continue within the
confines of cell patterns established during the epidermal
cap phase of regeneration.

The concept that the wound epidermis may function
in establishing a proliferation phase for blastemal growth
has been proposed by Faber (1960). An increase in the apical
epidermal mass, described in this study as a secondary epi-
dermal thickening, or epidermal lobe, has been described by
Faber (1960), and he has correlated its appearance with the
onset of active proliferation within the developing blastema.
He has likened its function to that ascribed to the embryonic
ectodermal ridge (Tschumi, 1957) in the control of differen-
tiation of the apical mesenchyme. Faber noted, as a result
of the displacement of implanted carbon particles, that
growth of the limb blastema of the axolotl is noticeably
higher in distal than in proximal regions. In the early
"bud" [or conical blastema, as Faber (1960) has designated
it] the mesenchymal accumulation possesses a homogeneity in
distribution and is not very dense. In later stages of de-
velopment the density of the distal mesenchyme is consider-
ably increased, a fact which might indicate that growth is
by means of apical proliferation. Since the early bud
blastema exhibits a homogeneous cell density pattern, Faber
(1960, 1962, 1965) concluded that the "apical proliferation
center" had not yet been established at this stage but sub-
sequently appeared in relationship with the formation of

the apical epidermal lobe. The evidence presented in the

74

present paper clearly demonstrates, however, that distal pro-
liferation has already commenced by the early bud stage. The
proliferation phase, therefore, precedes apical epidermal

lobe development but is correlated with an intense mitotic
phase of the apical cap. This temporal relationship suggests
the possibility that distal cell division within the mesenchyme
might be influenced by factors initiated by the wound epi-

thelium. It also may indicate that the apical epidermal

 

lobe as a discrete entity has no Special role but is merely
an outgrowth formed by continued high mitotic activity of
the apical epidermis in general. In those cases in which
the lobe did not develop, (see page 39), no obvious effect
on regeneration could be observed.

It has been pointed out previously (Thornton,
1956-1958, 1960, 1962; Thornton and Steen, 1962) for both
normally innervated and aneurogenic limbs of several Aggy:
stgma species that the apical wound epithelium (the "epider-
mal cap") functions to "direct" the accumulation of mesen-
chymatous cells which produce the regeneration blastema.
The evidence presented in the present paper clearly demon—

strates for two Ambystoma species that within the mass of

 

mesenchymatous cells which accumulates in association with
the apical wound epithelium there develops at a very early
time (10 days for the axolotl) in the regeneration process
an apical dominance in the proliferation rate of the blaste-
mal cells. Indeed, the high apical rate in mitotic divisions

was evident throughout the rapid growth phase of the blastema

75

(between 10 — 16 days regeneration) and also for the 19 day
regenerate (paddle stage) provided the mitotic counts were
specifically focused upon the digital "condensation" area
(see page 52). Reference should be made to Figure 4 for a
view of the intense mitotic activity of the apical mesen—

chymatous cells of the mound blastema (13 days regeneration).

 

It was also made evident from cell density calculations in
. this study that throughout the rapid growth phase of the
blastema there was a continual increase in cell density in
the distal part of the blastema. This increase in cell
density might be regarded as additional evidence for the
observed high incidence of apical mitotic activity. It was
also noted that the increases in cellular proliferations at
more proximal levels (mainly Zones II and III of the blastema)
as regeneration progressed, were characteristically associ-
ated with the more proximal extension of this dense mass of
mesenchymatous cells.

Thus, the results of the present investigation
confirm the hypothesis of Faber (1960, 1965) that an "apical
proliferation center," here identified as the apical dense
mass of proliferating blastemal cells, is established within
the limb blastema of the axolotl. The present study further
provides evidence for the suggestion by Faber (1960) that
the "apical proliferation center" functions in producing
the cellular materials for the new tissue components
of the regenerating limb. Although the present study clearly

proves by precise mitotic index calculations and statistical

76

analyses that an "apical proliferation center" does exist,
whereas Faber's carbon marking experiments only suggested
the existence of the center; it was not designed to test
the theory that an "apical organizational center" (Faber,
1960) is also operative in limb regeneration (cf. Intro—
duction, p. 16).

The possibility that the "apical epidermal lobe"
might function to initiate the high degree of divisions
among the apical mesenchymatous cells has been ruled out,
since as was earlier indicated, a high rate in prolifera-
tions was found to precede the development of the epidermal
lobe. The question arises, however, as to the possibility
that the "epidermal cap" might in some way influence blas-
temal cell proliferation. A clue to this question might
be derived from the experiments of Michael and Faber (1962)
who found that for a renewal of growth and differentiation
of limb transplants to occur, a wound epithelial covering
(devoid 0f dermis) derived either from the transplanted
limb stump itself or from the skin of the back (graft site)
was required. A similar requirement for growth was noted
for the normal blastema by Thornton (1957, 1958). In the
experiments reported above the possibility that the wound
epithelium and especially the cap region might serve to
influence the underlying mesenchyme to become mitotically
active has been investigated and some interesting correla-
tions have been revealed. It was found, for example, that

a striking relationship exists between the mitotic activities

77

of the wound epithelium and the underlying mesenchymatous
cells. Indeed, for all blastemal stages, with the exception
of the 19 day regenerate, which was not examined for epi—
dermal mitosis, a striking parallel in the fluctuations in
the mitotic activity between these two tissues was observed.
Thus, intense mitotic activity in the apical wound epithlium
(day 10) was accompanied by a high rate of proliferation
within the adjacent region (Zone I) of the blastema. Subse-
quently, as the mitotic activity shifted into more proximal
wound epithelium regions and decreased in the apical part (:5
(days 13 and 16) a similar change in cellular divisions was
noted for the underlying mesenchymatous cells of the blas-
tema. Thus, Zones II, III, and IV showed increases while
Zone 1 underwent a slight decrease. Since high mitotic
indices were found throughout the proliferation stages of
regeneration in both the wound epidermis and the mesenchy-
matous cells of the blastema, the possibility must be con-
sidered that factors common to both tissues, and acting
simultaneously on both, influence their mitotic proliferation.
In other words, both wound epidermis and blastemal cells may
be independently reacting to the same mitotic stimulant.
This possibility, however, seems less likely in view of the
following facts. There is a high mitotic activity, although
not as intense as the 10 day stage, for the apical wound
epithelium at 7 days regeneration (apical cap formation)
which, interestingly enough, corresponds to the time for

the beginning of blastemal cell accumulation. Cell counts

78

performed on the apical wound epithelium of the 5 day re-
generate, on the other hand, revealed a much lower rate in
mitotic activity as compared to the 7 day limb. Thus,
beginning with the initial accumulation of blastemal cells
there is a very noticeable increase in the mitotic activity
within the apical wound epidermis (epidermal cap) which
shows a peak in activity at the 10 day stage (early bud)
commensurate with the burst of mitotic activity among the
most distally located mesenchymatous cells of the develOp-
ing blastema. Therefore, it can be shown that the heightened
mitotic activity of the wound epidermal cells precedes that
of the accumulating blastema cells and the relationship is
especially obvious in the first 10 days of regeneration.

Evidence which would support the inference for a
relationship between the activities of the wound epithelium
and the underlying blastemal cells has been derived from
two additional experiments. A most interesting correlation
was found, for example, when the apical epidermis was
shifted from its normal position (apical to the limb stump)
to an eccentric position. It was highly interesting to
observe that following the removal of the skin (epidermis
and dermis), to produce asymmetry of the epidermal cap,
the wound epithelium which migrated to cover the extensive
wound surface became unusually thickened (Figure 11). As
regeneration proceeded in these asymmetrical blastemata,
the mesenchymatous cells of the blastema were found not

'only to aggregate beneath the thickened region of the wound

79

epithelium but a very noticeable increase (Figure 16) in the
number of mitotic divisions among these cells was also found.

A similar relationship between the activities of
the wound epithelium and the underlying blastemal cells was
found for the aneurogenic limb. Indeed, as the numbers of
cell layers of the apical wound epithelium in these regenera-
ting limbs increased and became more active mitotically, a
corresponding rapid increase in the mitotic activity among
the most distally located mesenchymatous cells was observed
(Figure 17, Figure 2). These results for the aneurogenic
limbs are also particularly interesting since they demon—
strate that blastemal cells aggregate distally (see Thornton
and Steen, 1962) and then undergo an intense proliferation
phase in the complete absence of nerve fibers. Thus, the
findings for both the asymmetrically oriented blastema and
the aneurogenic limb blastema, as well as the normal limb
blastema, support and extend the hypotheses of Thornton
and Faber that the wound epithelium functions,in.some un-
known manner, in both "directing" the accumulation of blas—
temal cells and in the "stimulation" of these cells to
proliferate.

In this regard, it is important to note that a
close association between the cells of the blastema and
the wound epithelium, especially the apical part, was main-
tained throughout the rapid growth phase of the blastema
(lO - 16 days of regeneration). As can be seen in Figure 4

a small space (30 to 40 microns across) separates the two

80

types of cells. By careful examination it was found that
this space is criss-crossed by cytoplasmic strands which
apparently originate from the mesenchymatous cells of the
blastema. Thus, the space itself is sufficiently small
enough to allow for the exchange of chemical materials

(see Grobstein, 1963) between the two types of cells, but
such an exchange could also be facilitated by the inter-
connecting cytoplasmic strands. I should like to suggest
that the wound epithelium in the presence of wound tissues
undergoes a "type" of dedifferentiation (see Thornton, 1962),
and becomes highly active mitotically, as has been observed,
and possibly highly active metabolically. As a result of
this alteration, the wound epithelium may in turn "direct"
and "stimulate," possibly through a chemically diffusible
substance, the accumulation and then the active proliferation
of the aggregated dedifferentiated blastemal cells. I should
also like to propose that once the "initial information" has
been supplied from the wound epithelium to the underlying
blastema (10 to 16 days of regeneration) the blastemal cells
continue their divisions until they are later organized into
the limb parts by other influences which may originate from
the "apical organization center" (see Faber, 1965) and also

the proximal stump regidns.

SUMMARY

It has been demonstrated in this study using the
precise experimental methods of mitotic index calculations
and statistical analyses that an apical dominance in mitotic
activity exists among the most distally located mesenchyma—
tous cells of the forelimb blastema of the axolotl. The
high apical mitotic rate was evident throughout the rapid
growth phase of the blastema (10 to 16 days of regeneration)
and provided a dense apical mass of proliferating mesenchy-
matous cells. As prOposed by Faber (1960, 1965) this apical
.dense mass of proliferating mesenchymatous cells may be re—
garded as the "apical proliferation center." An apical
peak in the number of mesenchymatous cell mitoses was also
demonstrated for the asymmetrically oriented blastema of

the axolotl and the aneurogenic limb blastema of Ambystoma

 

Opacum. No correlation as suggested by Faber (1960), be-
tween the development of the prominent apical epidermal

lobe and the rate of the proliferation of the mesenchymatous
. cells of the axolotl blastema could be found. A striking
temporal relation, which may represent a causal relation,
was found, however, to exist between the mitotic activity

of the wound epithelium and the initial accumulation and
regional distribution of bitoses of the underlying blaste-

mal cells.

81

82

Figure 1a.-—Diagrammatic drawing of the forelimb
blastema illustrating the five sample levels for the cell
counts.

Figure lb.--Diagrammatic draWing of a center sec-
tion of the forelimb blastema showing the four sample areas,
A, B, C, and D for the cell counts.

83

a...

 

84

Figure 2.--Twelve day aneurogenic blastema showing
numerous apical mesenchymatous cell divisions. 560x.

Figure 3.——The apical wound epithelium of a 10 day
axolotl regenerate possessing numerous epidermal cell divi-
sions. 560x.

Figure 4.--A 13 day axolotl regenerate showing the
intensive mitotic activity of the apical mesenchyme. 590x.

Figure 5.--The prominent apical epidermal lobe of
a 16 day axolotl regenerate. Note the excessive number of
Leydig cells and also the pigment granules within the wound
skin. 22OX.

Pitt‘-
is

l

~57"

g.
..
$31. 9
o, .

85

 

‘

86

Figure 6. —-Longitudinal section of the "early bud
stage" (10 day blastema) illustrating the arbitrary vertical
zones. Note the homogenous distribution of the mesenchyma-
tous cells. 150X.

Figure 7.—-Longitudinal section of the "mound stage
blastema" (13 days) showing the vertical zoning. Note the
dense apical mass of mesenchymatous cells and also the apical
lobe. 135x.

Figure 8.—-Longitudina1 section of the "cone blastema"

(16 days regeneration) showing the vertical zoning. Note the
axial aggregation of blastemal cells which is beginning at the
center and base of the blastema. 135x.

Figure 9.—-Longitudina1 section of the 19 day regen-
erate "paddle stage." Note the pattern of zoning and also the
mesenchymatous cell condensations for digits 1 and 2 at the
apex of the limb. The arrow indicates the mass of mesenchyme
destined to produce digits 3 and 4. 120X.

87

_; .

,. #251
. .

‘ ‘.

   
  

         
 
     
      

. '. -s-.
34’ $2
5);,
14"“ 05:1...
.49 Saxarfx‘
I x3}
L" ° ‘91!

.

as» '
' - its.

   

88

Figure lO.--Cross section near the distal tip of
a 13 day blastema (early bud stage). The upper arrow indi-
cates the dorsal side and the bottom arrow the ventral side
of the blastema. 380x.

Figure ll.--Sixteen day asymmetrical blastema il-
lustrating the sample areas for the cell counts. Arrow
indicates the normal axis of the limb. 110x.

Figure l2.—-Cross section near the center of a 13
day blastema showing the 5 regions for the cell counts. 220x.

Figure l3.--Cross section near the base of a 13 day
blastema. 150X.

89

 

90

Figure l4.--Graphic representation of the mitotic
index and cell density distribution of the 4 longitudinal
zones for the 4 regeneration stages.

91

Figure 14

ZONES

IV

III

II

6nU..U.826
33211

_II!: OmzmHeM I Ombbw wmw czHe.iul

60hm826
23211

6nwu826
33n/._l_l.o_

6nU..U.826
33211

 

 

‘1J_.u-4J4uuq

-—-—-------

  

W
m
..

ht_.__—__...

...d.fiqd.d‘..

   

 

.—.———L?. .F__

 

.dedd.—-_._d

d_._._+—..q_d

.
.
—
.
.
—
.

  

_.
v
~

 

 

P—sL._bPu—e.b

 

 

u—f—f—-btb._.

 

6nwu82rnw
33211...

0H

6nU.u.826
33211

 

ma

mezH UHBOBHE

60.“.826
3321.1.

 

0H

ZOHB<mmzmwmm mo mw<m

60.4826
33211

ma

 

92

Figure 15.--Graphic representation of the mitotic
index and cell density distribution of the 5 cross sectional
regions for the H regeneration stages.

93

Figure 15

REGIONS

6nwu826

........ umzmHem . ombrm wmm c2332---

 

 

 

 

 

 

 

 

 

60.4826 6nwu826 60u826
.. . C
33211 3321.]... 33211 332a].
.d-_.-.qdd1dqq..qq_._.qd_44J44J‘4_u_u_d ._._A-—1fiAJJ
.
— or
, .
v
r p
a a
a
’
w .
,. x
_ ,
. I
. I
. I
a
_ ,
. a
_ I
. a
. .
_ .
.
. .
.. _
.
_ .
.
-
.
. x
. \
fl \
, \
a
. \
— s
._._+_.—._.__ F—pL JF—TPL—+P.—-—_ .P.—p—.-P_._

 

 

60.“.826
33211

OH

 

meZH OHBOBHS

rnwnwu826
33211

 

ma

ZOHB¢mmzmwmm mo mw<o

Inwouazroo
33211

ma

 

94

Figure 16.-—Graphic representation of the mitotic
index and cell density distribution of the u sample areas
for the 3 regeneration stages of the asymmetrical blastema.

95

Figure 16

SAMPLE AREAS

lllll .lbmzmHe& I ombbm wmw CZHalllilu

 

 

 

 

 

 

 

 

 

:H

8&0628u. 8“”0628“. 8u0628u.
. C . . . ... C . . C . C
22211 22211 22211
..d‘fi4J ._4 fi+_ 1+J 4‘ qfil .._ ._._ ._q \Aq. _... _q.?.‘J‘w+

I I]

" ll
._.—F_._.|..—_—. .r.—»—-_._~b-b_ pbp—phrbr—pb_—b
8“».0628“ 840628ul 8A0628u.
222110—1 222.]..1 22211

XMQZH OHBOBHS
ma ma
ZOHB<mmzmcmm mo mw<o

 

96

Figure l7.--Histogram showing the mitotic index and
cell density distributions for the base and tip areas for the
M aneurogenic blastemata.

Figure 1?

LO
LO

5-5

97

DENSITY - CELLS PER UNIT

memo

O mom

anzrmmm Nt-it—I

omomomom
anrfi'MMNNr—l

XHGNI OILOIIW

W

1.0

 

Ln

 

.5

TIP BASE TIP BASE RIP BASE

BASE

TIP

1“ 16
DAYS OF REGENERATION

12

10

LITERATURE CITED

Barr, H. J. 1963 An effect of exogenous thymidine on the
mitotic cycle. J. Cell and Comp. Psysiol., Q:
119-128.

 

Bodemer, C. W. 1958 The develOpment of nerve-induced
supernumerary limbs in the adult newt, Triturus
viridescens. J. Morph., 102: 555. '

Bodemer, C. w. and Everett, N. B. 1959 Localization of
newly synthesized proteins in regenerating newt
limbs as determined by radioauto raphic locali-
zation of injected methionine-83%. Develop. Biol.,
1: 327-342.

 

Brunst, V. V. 1950 Influence of x-rays on limb regeneration
in Urodele Amphibians. Quart. Rev. Biol., £5: 1-
29.

Brunst, V. V. and Cheremetieva, E. A. 1936 Sur la perte
locale du pouvoir régénérateur ches 1e Triton et
1'axolot1 Causée par 1'irradiation avec les rayons
X. Arch. 2001. explt.ggener., 1g: 57-57.

Butler, E. G. 1931 X-radiation and Regeneration in Ambly-
stoma. Science, 15: p. 100.

1933 The effects of x-radiation on the regenera-
tion of the forelimb of Amblystoma larvae. J.
Exp. Zool.,Ԥ5: 271-315.

 

 

1935 Studies on limb regeneration in x-rayed
Amblystoma larvae. Anat. Rec., Qg: 295.

,” Butler, E. G. and O'Brien, J. P. 19u2 Effects of localized

x-radiation on regeneration of the Urodele limb.
Anat. Rec., g5: A07-413.

 

Cameron, J. A. 1936 The origin of new epidermal cells in
skin of normal and x-rayed frogs. J. Morph., 59:
327-3A9.

Chalkley, D. T. 195“ A quantitative histological analysis
of forelimb regeneration in Triturus viridescens.

J. Morphol., 93: 21-70.

 

98

99

1959 The cellular basis of limb regeneration. In
Regeneration in Vertebrates (C. S. Thornton, ed.),

 

Univ. of Chicago Press, Chicago, Illinois, 34-58.

Colucci, V. 1884 Intorno alla regenerazione degli arti e

David, L.

Faber, J.

della coda nei Tritoni. Studio sperimentale. Mem.
ric. accad. sci. inst. Bologna (ser. 9)., 5: 501—
566.

 

193” La Contribution du materiel cartilagineux
et osseux au blasteme de régénération des membres
ches les amphibiens urodeles. Arch. d'Anat. Microsc.,
39; 217-234.

1960 An experimental analysis of regional organi-
zation in the regenerating forelimb of the axolotl
(Ambystoma mexicanum). Arch.de Biol., 11: 1-72.

 

 

1961 The self-differentiation of the paddle-shaped
limb regenerate, transplanted with normal and re-
versed proximo-distal orientation after removal of
the digital plate (Ambystoma maxicanum). Arch. de
Biol., 12: 301-330.

 

1962 Additional experiments on the self-differenti-
ation of transplanted whole and half forelimb re-
generates of Ambystoma mexicanum. Arch. de Biol.,

15: 369-378.

 

Forsyth, J. w. 1946 The histology of anuran limb regeneration.

J. Morph., 15: 287-321.

Glade, Richard w. 1963 Effects of tail skin, epidermis, and

dermis on limb regeneration in Triturus viridescens
and Siredon mexicanum. J. Exp. Zool., 152: 169-19“.

 

 

Godlewski, E. 1928 Untersuchungen Uber Die Auslosung and

Hemmung der Regeneration beim Axolotl. Roux. Arch.,
11H: 108-1A3.

 

Grobstein, Clifford 1961 Passage of radioactivity into a

membrane filter from spinal cord pre-incubated with
tritiated amino acids or nucleosides. In La Culture
Organotypique, Editions,_Centre National de la
Recherche Scientifique, 15: 169—182.

 

Haas, Hermann J. 1962 Studies on mechanisms of Joint and

Hay, E. D.

bone formation in the skeleton rays of fish fins.
Developmental Biology, 5: 1—34.

 

1952 The role of epithelium in amphibian limb
regeneration, studied by haploid and triploid trans—
plants. Am. J. Anat., 15: 4&7-A82.

 

100

1958 The fine structure of blastema cells and "
differentiating cartilage cells in regenerating
limbs of Amblystoma larvae. J. Biophys. Biochem.

Cytol., 5: 583-592.

 

1959 Electron micros00pic observations of muscle
dedifferentiation in regenerating Amblystoma limbs.
Develop. Biol., 1: 555-586.

 

 

Hay, E. D., and Fischman, D. A. 1960 Origin of the regenera-
tion blastema of amputated Triturus viridescens
limbs, studied by autoradiography following injec-
tigns of tritiated thymidine. Anat. Record, 136:

20 .

 

1961 Origin of the blastema in regenerating limbs
of the newt Triturus viridescens. Devel. Biol., 5:
26-590

 

 

Heath, H. D. 1953 Regeneration and growth of chimaeric am-
phibian limbs. J. Exptl. Zool., 122: 339-336.

Hellmich, W. 1930 Untersuchung u. determination des regene-
rative materials bei amphibien. Roux. Arch., 121:
135-204.

 

1931 Histology of regeneration in different species
of adult and larval urodeles. Anat. Rec., 58: 303-
307.

 

Hertwig, G. 1927 Beitrage zum determinations und regenera-
tions problem mittels der transplantation haploid-
kerniger zellen. Roux. Arch. f. Entwmech., 11;:
292-316.

Ide-Rozas, Alberto 1936 Die cytologischen verhaltnisse bei
der regeneration kaulquappenextremitat. Roux. Arch.,

112: 552-607.

Kamrin, A. A., and Singer, M. 1959 The growth influence of
spinal ganglia implanted into the denervated fore-
limb regenerate of the newt, Triturus. J. Morph.,
104: 415.

Kazancev, w. 193A Histologische untersuchunger der regener-
ationsprozesse an amputierten extremitaten beim
axolotl hauptsachlich zwecks klarung der frage nach
der herkunft der zellen des regenerats. Acad. des
Science de l'URSS. Travaux du Lab. de 2001. Exp.
et de Morph. des Animaux, 5: 23-54.

 

 

 

 

Leibman, E. 1949 The leucocytes in regenerating limbs of
Triturus viridescens. Growth, 15: 103-118.

 

lOl

Litwiller, Raymond 1939 Mitotic index and size in regenera-
ting Amphibian limbs. J. Exp. Zool., 82: 280-286.

 

1940 Mitotic indices in regenerating urodele
limbs. II. A study of the diurnal distribution of
cell divisions. Growth, 4: 168-172.

Luther, W. 1948 Regenerations versuche mit holfe von rontgen
strahlen. Verhandl. der Deutschen Zoolgen in Kiel,
1948: 66-68.

L4-Manner, H. w. 1953 The origin of the blastema and of new
‘ " tissues in regenerating forelimbs of adult Triturus
viridescens. J. Exp. Zool., 122: 229-258.

 

 

Messier, B. and LeBlond 1960 Cell proliferation and migra-
tion as revealed by radioautography after injection
of thymidine H3 into male rats and mice. American
Journal of Anatomy, 106: 247-265.

Mettetal, C. 1939 La regeneration des membres chez la sala-
mander et 1e triton. Histologie et determination.
Arch. Anat. Hist. Embryol., 28: 1-214.

 

Michael, M. I. and Faber, J. 1961 The self—differentiation
of the paddle-shaped limb regenerate, transplanted
with normal and reversed proximo-distal orientation
after removal of the digital plate (Ambystoma mexi-
canum). Arch. de. Biol., 12: 301—330.

 

 

Milaire, J. 1956 Contribution a 1'étude morphologique et
cytochimique des bourgeons des membres chez le rat.
Arch. Biol., 51: 297—391.

1957 Contribution a la connaissance morphologique
et cytochimique des bourgeons des membres chez quel-
que reptiles. Arch. Biol., 55: 429-572.

 

Needham, J. 1942 Biochemistry_and Morphogenesis. Cambridge
Univ. Press. London, England.

 

Needham, A. E. 1952 Regeneration and Wound—Healing.
Methuen's Monographs on Biological Subjects. John
Wiley and Sons, Inc., New York.

 

O'Rahilly, R., Gardner, E., and Gray, D. J. 1956 The ecto-
dermal thickening and ridge in the limbs of staged
human embryos. J. Embryol. exp. Morphol., 4: 254-264.

 

Polezajew, and Ginsburg, G. I. 1944 Investigation of ways of
formation of regeneration blastema based on calcula-
tion of mitotic coefficient. Comptes Rend. (Doklady)
de 1'Acad. des Sci. de 1'URSS, 43: 315-317.

 

 

102

Puckett, W. 0. 1936 The effects of x-radiation on limb de-
velopment and regeneration in Amb1ystoma. 1.
Morph., 53: 173-213.

 

Riddiford, L. M. 1960 Autoradiographic studies of triti-
ated thymidine infused into the blastema of the
early regenerate in the adult newt, Triturus.

J. Exp. Zool., 144: 25-32. *

 

\Rose, 3. Meryl, 1948 Epidermal dedifferentiation during
blastema formation in regenerating limbs of Tri-
turus viridescens. J. Exp. Zool., 108: 337—362.

 

Rose, R. C., Quastler, H., and Rose, S. M. 1953 Regenera—
tion of x-rayed limbs of adult Triturus provided
with unirradiated epidermis. Anat. Rec., 117: 619.

 

1955 Regeneration of x-rayed salamander limbs
provided with normal epidermis. Science, 122:
1018—1019.

Salpeter, M., and Singer, M. 1960 Differentiation of the
submicroscopic adepidermal membrane during limb
regeneration in adult Triturus including a note on
the use of the term basement membrane. Anat.
Record, 155: 27-32.

Saunders, J. W. 1948 The proximo-distal sequence of origin
of the parts of the chick wing and the role of the
ectoderm. J. Exp. Zool., 108: 363-404.

 

Saunders, J. W., Gasseling, M. T., and Gfeller, M. D. 1958
Interactions of ectoderm and mesoderm in the ori-
gin of axial relationships in the wing of the fowl.
J. Exp. Zool., 151: 39.

 

Scheuing, M. R., and Singer, M.. 1957 The effects of micro-
‘ quantities of beryllium ion on the regenerating
forelimb of the adult newt, Triturus. J. Exp.
Zool., 155: 301.

Singer, M., and Salpeter, M. 1961 The role of the wound
epithelium in vertegrate regeneration. Intern.
Symposium on Growth, Purdue Univ., Lafayette,
Indiana, 1960.

Singer, M., Ray, Eva K. and Peadon, Annie M. 1964 Regional
growth differences in the early regenerate of the
adult newt, Triturus viridescens, correlated with
the position of the larger nerves, Folia Biologica,
15: 347-362.

 

 

Taban, C.

Thornton,

Thornton,

Thornton,

103

1955 Quelques problemes de regeneration chez les
urodeles. Rev. Suisse Zool., 5g: 387.

 

C. S. 1938 The histogenesis of muscle in the re-
generating forelimb of larval Amb1ystoma punctatum.

J. MOrph., 5g: 17-47.

1938 The histogenesis of the regenerating fore-
limb of larval Amblystoma after exarticulation of
the humerus. J. Morph., 5g: 219-241.

 

 

1942 Studies on the origin of the regeneration
blastema in Triturus viridescens. J. Exp. Zool.,
Q9: 375-385.

 

 

1954 The relation of epidermal innervation to
limb regeneration in Amblystoma larvae. J. Exp.
Zool., 127: 577-602.

 

1956 Epidermal modifications in regenerating
and non-regenerating limbs of anuran larvae. J.
Exp. Zool., 131: 373.

 

1957 The effect of apical cap removal on limb
regeneration in Amb1ystoma larvae. J. Exp. Zool.,
134: 357.

1958 The inhibition of limb regeneration in uro-
dele larvae by localized irradiation with ultra-
violet light. J. Exp. Zool., 137: 153.

 

 

1960 Influence of an eccentric epidermal cap on
limb regeneration in Amblystoma larvae. Develop.
Biol., 5: 551-569.

 

1962 Influence of head skin on limb regeneration
in urodele amphibians. J. Exp. Zool., 150: 5-16.

 

1965 Influence of the wound skin on blastemal
cell aggregation. Reprinted from Regeneration in
Animals and Related Problems, edited by V. Kiortsis
and H. A. L. Trampusch. 333-340.

 

C. S., and Steen, T. P. 1962 Eccentric blastema

formation in aneurogenic limbs of Ambystoma larvae
following epidermal cap deviation. Devel. Biol.,

5: 328—343.

 

 

C. S., and Thornton, Mary T. 1965 The regeneration
of accessory limb parts following epidermal cap
transplantation in Urodeles. Experientia, g1: 146.

 

104

Tschumi, P. A. 1955 Versuche uber die wachstumsweise von
hinterbeinknospen von Xenppus laevis daud, und die
bgdeutgng der epidermis. Rev. Suisse Zool., 62:
2 1-28 . _—

 

1956 Die bedeutung der epidermislests fur die
entwicklung der beine von Xenopus laevis. Rev.
Suisse Zool., 55: 707.

1957 The growth of the hindlimb bud of Xenopus
laevis and its dependence upon the epidermis. 1.

Van Stone, J. M. 1955 The relationship between innervation
and regenerative capacity in hind limbs of Rana
sylvatica. J. Morphol., 51: 345-392.

Weiss, P. 1939 Principles of Development. lknumr Holt &
Co., New York.

Zwilling, E. 1956 Interaction between limb bud ectoderm
and mesoderm in the chick embryo. J. Exp. Zool.,

11g: 173.

   

”71117171777[11171711777317