UNIVERSE* v

lUJNlQlS i j&F
H WWW-CHAMPAICN

LXLOGY

=2

(

FIELDIANA '

Geology

Published by Field Museum of Natural History

Volume 33, No. 24 April 5, 1977

This volume is dedicated to Dr. Rainer Zangerl

Tooth Histology and Ultrastructure

of a Paleozoic Shark,

Edestus heinrichii

Katherine Taylor 1

Committee on Evolutionary Biology
University of Chicago

and

Thomas Adamec 2

University of Chicago
Pritzker School of Medicine

INTRODUCTION

Edestus heinrichii (Newberry and Worthen, 1866), is a Paleozoic
shark known from symphyseal tooth isolates and several articulated
tooth bars. Specimens attributed to this genus have been described
from Russia, Australia, England, and the mid-continental United
States. E. heinrichii is one of 15 species within genus Edestus that
have been distinguished by variations in the dentition size and
morphology. Teeth remain the only anatomic evidence of the genus
thus far described. This paper re-examines the symphyseal den-
tition based on new material from the Pennsylvanian shales of the
Illinois Basin. Aspects of histology, tissue ultrastructure, tooth
ankylosis, gross morphology and embryology of the fossil are exam-
ined. Evidence is provided for the absence of orthodentine in the
symphyseal teeth. This is the first elasmobranch known to have this
condition. The teeth are composed of only two types of dentine:
enameloid and trabecular. The ultrastructure of the denteon in

'Present address: Department of Pathology, University of Chicago.
2 Present address: Department of Pathology, University of North Carolina at
Chapel Hill.

Library of Congress Catalog Card No. : 76-56537

Publication 1253 441 "*"*"»«"!*

JUN 06 1977

University ot hHnois
w* iirHona-r.hamoai&ft

442 FIELDIANA: GEOLOGY, VOLUME 33

trabecular dentine is shown to share a similar fundamental struc-
ture with the osteon of secondary bone. The specimens studied here
are Field Museum of Natural History (FMNH) PF 2848 and PF
2849, of E. heinrichii. They are from the Pennsylvanian shales of
Mecca Quarry in Parke County, Indiana, collected by Dr. Rainer
Zangerl. Recent material for comparison of tissue structure is from
Sphryna tudes and Isurid sharks, from the Field Museum's Depart-
ment of Fishes.

MATERIALS AND METHODS

The Field Museum study collection has at least 27 individual
teeth of E. heinrichii so far identified by X-ray, including five par-
tial tooth whorls of from two to three teeth and one completely
articulated whorl of nine teeth (fig. 1). One of the partial tooth bars
of three articulated teeth with complete crowns and almost com-
plete roots was chosen for sectioning, along with a single isolated
tooth. The fossils remained completely embedded in shale and were
identified by x-ray (pi. 1). In PF 2849, the anterior teeth were cut
serially at 2 mm. intervals into 16 sections and light microscope
slides were hand ground (fig. 2). Serial sections 6, 7, and 12 did not
survive the mounting and grinding process and fragments of them
were used for electron field emission scanning. These fragments
were put through successive 24-hr. periods in propylene oxide until
the embedded epoxide resins were removed, then dehydrated in
absolute alcohol. Some of the specimens at this stage were etched
with hydrochloric acid, then air dried. Dried material was mounted
on aluminium discs and then coated with gold: palladium (40:60) in
an Edward vacuum coating machine. The scans were made by the
senior author and by Dr. John M. Clark of the University of Chicago
Pritzker School of Medicine on the Hitachi HFS II scanning elec-
tron microscope, established by a grant from the Sloan Foundation,
at the Enrico Fermi Institute. The scans were done under PHS
Grant No. 5 T05 GM01939 from the National Institute of General
Medical Science.

CONDITION OF THE FOSSILS AND THEIR PRESERVATION

The hard tissues were almost perfectly preserved in the fossiliza-
tion process. Both tooth specimens were laid down parallel to the
shale's bedding plane, as is the case with the vast majority of the
specimens in the study collection. X-ray photographs of similarly
embedded specimens were made at various angles and checked for

TAYLOR & ADAMEC: PALEOZOIC SHARK 443

angular deformation; none was found. The x-rays (pi. 1) represent
fully sagittal views. Zangerl and Richardson (1963, p. 181) report
that a large cladodontid tooth from the same quarry was embedded
upright and showed no evidence of distortion due to compression.
The shape dimensions are in complete agreement with teeth em-
bedded laterally. Plastic deformation is therefore negligible.

Diagenesis has only slightly modified the morphology and histol-
ogy. The teeth are to some extent decalcified and bituminized. The
burial sediments and diagenetic replacement materials have natu-
rally stained histologic areas uniformly and consistently. Micro-
scopic cavities are neatly stained with iron which is brown to red to
orange in transmitted light. On PF 2848, calcite has filled the basal
canals and made them opaque, and filerite (zinc sulphate) has
formed between the denticles along the borders (pi. 1). The presence
of filerite from decomposition is common in the Mecca fossils ( Zan-
gerl, pers. comm. ).

The depositional environment of the black shales was so acid the
bacterial degradation was not very destructive. This permitted a
slow steady impregnation with hydrocarbons, a condition most
favorable to preservation.

Cracking of the enameloid surface is grossly visible when matrix
is removed from the crown. The cracks occur at regular intervals
remaining fairly equidistant and run from the base of the crown to
the tip (fig. 1). In sagittal view cracks in the trabecular dentine
lining the crown perforate the enameloid and open onto the crown
surface (pi. 4a). The openings are 40-50/u. wide and average .4 to .5
mm. in depth. Electron scans of the enameloid surface (pi. 4b, c)
demonstrate that micro-cracks occur at intervals corresponding to
the channels in the brightfield views. There is no indication from the
examination here that these are anatomic structures. They do not
appear to be in association with the vascular pattern of the trabecu-
lar dentine they penetrate. Zangerl found in gross examination that
the system of macro-cracks is arranged stress-coat fashion and
probably resulted from pressure of the burial mud when it lost its
plasticity (Zangerl and Richardson, 1963, p. 181). The cracks are
presumed to be diagenetic rather than anatomic.

GROSS MORPHOLOGY, VASCULARIZATION,
AND ANKYLOSIS

The tooth base presumably grows continually in a longitudinal
direction from the time the crown comes into place functionally

444

FIELDIANA: GEOLOGY, VOLUME 33

Fig. 1. Edestus heinrichii, UF 30 (FMNH), showing a complete symphyseal tooth
bar of nine successive teeth. Enameloid flanges can be seen extending posteriorly;
the stress-coat-like cracks in the enameloid are approximated.

until the whole tooth including its base is shed from the anterior-
most position on the whorl. The crown is full-sized when it comes
into place in the posterior-most position. The replacement-shedding
process proceeds at a constant rate so that seven to nine teeth are
maintained on each bar. The tooth crown is defined by the area
covered with enameloid. The cusp is non-equilateral; the anterior
edge rises at a sharp angle to the root; the posterior edge slopes at a
wider angle. There are up to 1 1 denticles on the anterior crown bor-
der of the adult tooth and 13 along the posterior border. Crenula-

post

Fig. 2. Edestus heinrichii, PF 2849, showing position of coronal sections (A) seen
in Plate 2, and the position of the sagittal section (B). The basal sinus seen in the
serial sections is approximated by dotted lines.

B. PF 284

Plate 1. a, Sagittal X-ray of Edestus heinrichii, PF 2849. Three adult symphyseal
teeth are seen in anatomic articulation. The arrows along the posterior border of
the third tooth indicate a radio-opaque area of pyrite. Filerite, a decomposition phe-
nomenon, has formed between the denticles, b, Sagittal x-ray of E. heinrichii, PF
2848, is an isolated tooth that was shed anteriorly from tooth bar.

445

446 FIELDIANA: GEOLOGY, VOLUME 33

tions on the denticles are not apparent on x-ray but can be seen
under magnification on exposed specimens of E. heinrichii. The
denticles along at least the anterior border are crenulated. The
crowns are so closely spaced that the adjacent borders overlap all in
the same direction (fig. 1). Flanges of enameloid extend out 1.5 cm.
behind the crown on the top of the tooth base troughs (fig. 1) and
occur symmetrically on each side of the bar. These flanges also
occur in Edestus minor, although considerably reduced.

The teeth are well vascularized. The pattern is characterized by
major arterial branching and venous anastomosing throughout the
tooth base, with substantially smaller arterioles supplying the
central crown region and a finer nutrient network going to the apical
lining and terminating at the enameloid junction. The vessels run
along the longitudinal axis of the root from back to front, diminish
in size from frequent branching, and slant upward into the crown.
The vessels do not converge toward the crown's apex but remain at
right angles to the posterior border as can be seen in a sagittally sec-
tioned tooth (pi. 3b). The largest canals are centrally placed in the
root. In the central vascular network there is clearly a single channel
that is the major arterial and venous supply for each tooth. Karpin-
sky ( 1899, pp. 404-421) described the presence of similar large single
channels in Helicoprion without discussing their function. The
channels' successive branching is clearly demonstrated on the serial
enlargements (pi. 2). The central canal slants upward toward the
crown and runs in this specimen just to one side of the midline. The
canal may conduct arteries, veins, and nerves as is the typical verte-
brate circulatory and innervation pattern.

The trabeculation of the tooth bases is more rugged on the ex-
posed outer surface of the tooth bar. This is particularly noticeable
on Plate 3a of the serial sections. The external and internal root
surfaces facing into the troughs have much smaller trabecles indi-
cating less stress between teeth than between the whorl and the
jaws. These internal areas of ankylosis have uniform surfaces and
emissary foramina (pi. 3a).

The nature of the ankylosis of the teeth to one another has not
been fully detailed before. In his schematic drawing of what he
called "Protopirata heinrichii" C. R. Eastman (1902) reproduced
the presence of a basal sinus which he does not name or discuss. It
has otherwise been assumed that the tooth bases were fully in con-
tact with each other (Newberry, 1889; Hay, 1910). This was not
found to be the case here. The trough of a tooth base and the base of

crown II
: .'"'tiM 'r\\ •■' end 0!

IO:/.'#;K/*v**7. tooth
r..r#»Ct«CoV.\-*basei

Plate 2. Coronal serial section of slides 1, 3, 5, 8, 11, 16. Slides 1, 3, 5, and 8 show
two articulated teeth. The slides show the course of the central canal vascular
supply, demonstrate the basal sinus between articulated teeth, and show the pos-
terior extension of the enameloid flange on the crown.

447

Pf28 . 48 t i 500, /<

A. PF 2849. slide 16

Plate 3. a, PF 2849, coronal section of tooth showing distinction between crown
and base. Crown is covered by thin enameloid (see wide arrows), and is composed of
Types 1 and 2 trabecular dentine. Interdenteonal hard tissue characterizing Type 2
is shown by thin arrows. Type 3 trabecular dentine is restricted to the outer milli-
meter of the base and is an open spongiosum lacking denteons. Emissary foramina
in Type 3 are associated with rough ligamentous attachment (the trabecles), and
with the vascular supply (the foramina), b, PF 2848, the vascular pattern in this
sagittally cut fossil tooth shows that the vasculature within the denteon lumen run
perpendicular to the surface in Type 1 trabecular dentine, and at right angles to the
tooth surface in Type 2. c, PF 2849, at higher magnification the absence of ortho-
dentine is demonstrated. Type 1 trabecular dentine is subjacent to the enameloid.
Here again the regular stress-coat cracking in the enameloid can be seen.

448

Plate A A

fYPe -2 fcj

Ju*&**i

Plate 4 B

Plate 4 C i ■ 10 ^

Plate 4. PF 2849. a, Enlargement of the crown tip showing cracks from the
enameloid perforating the adjacent trabecular dentine, b, c, Electron scans of the
cracks do not show them to be in association with the vascular pattern. There is a
distinct difference in fracture pattern between hypermineralized enameloid and the
more fibrous trabecular dentine. The juncture between the tissues shows up clearly.

449

450 FIELDIANA: GEOLOGY, VOLUME 33

the successive tooth it holds are not completely ankylosed forming a
basal sinus (fig. 2). The sinus is patent only between adjacent tooth
bases and is not a continuous channel throughout the intermandibu-
lar whorl. The successive basal sinuses are not artifacts of this
particular fossil nor a result of the specimens having partially rotted
apart. Tracings from blow-ups were cut out and a reassembly at-
tempted that would close off the basal sinuses. Such a realignment
was not structurally possible. The basal sinus is a real anatomical
feature. There may have been more mobility between teeth than had
been supposed with the basal sinus tissues cushioning compressive
and shearing stresses, a condition also more conducive for anterior
tooth shedding.

HISTOLOGY

Remarkable conservatism in the retention of tooth types is a sub-
class character of elasmobranchs. This conservatism over a long
stratigraphic sequence seems to be the case for the 110-million-year
span of Edestus, from the Mississippian through the early Triassic.
Edestus was a successful form. Only two types of dentine — tra-
becular dentine and enameloid — occurred in its symphyseal teeth.
It is the first shark for which the lack of orthodentine has been docu-
mented (pi. 4a).

In the early literature terms for different dentine types proliferate
that were often defined differently by individual researchers. 0rvig's
(1951, 1967a, c) consolidation and reordering of terms for the hard
tissues of elasmobranchs is followed here with one exception. Tra-
becular dentine is used here for what would ordinarily be called
osteodentine. We have not been able to identify the interstitial acel-
lular banding between the denteons as bone. Osteoblasts may in
certain instances transform into odontoblasts ( Pflugfelder, 1930),
but invoking such a process without evidence is unwarranted here.

The histology of edestid teeth has been examined previously
(Hay, 1910; Nielson, 1932, 1952; Zangerl, 1966). Hay made sagittal
and coronal sections of only the tooth base of E. heinrichii, therefore
not observing the absence of orthodentine in the crown. His speci-
men came from the same general area, western Indiana, as those
examined here. The two correspond exactly in tooth base structure.
Hay refers to the trabecular dentine of the base as "vasodentine," a
tissue containing capillary canals instead of dentineal tubules ; and
from gross rather than histologic examination reports that the
tooth crown covering "is probably true enamel" (Hay, 1912, p. 50).

TAYLOR & ADAMEC: PALEOZOIC SHARK 451

Zangerl ( 1966) described the histology of the closely related edes-
tid Ornithoprion hertwigi. The outermost layer, "which probably
constituted the orthodentine with its vitrodentine surface" (Zang-
erl, 1966, p. 31), was missing. In light of its absence inE. heinrichii
it was probably originally absent in 0. hertwigi also ( Zangerl, pers.
comm.). A section through the trabecular dentine of a large O. hert-
wigi tooth shows trabecular dentine corresponding exactly to the
type 1 (see pi. 3a, b; 4a) crown lining found inE. heinrichii. The clear
interstitial banding of acellular calcified tissue is absent, as it is in
E. heinrichii, and the dentine tubules do not define the denteon
margins.

Dentine is homologous among all vertebrates, the matrix being
secreted by mesodermally derived odontoblasts. The odontoblasts
retreat along the front of the matrix accumulation, leaving hair-like
cell processes, called Tomes' fibers, behind (pi. 5b). Orthodentine is
the same histologically in fish, reptiles, and mammals. Its absence
in this species and probably the whole family is a feature for which
there is no ready explanation. Peyer (1968, p. 65) emphasizes that
in all known elasmobranchs, both fossil and extant, the outermost
coat of compact dentine is orthodentine. It is undoubtedly lacking
in E. heinrichii. Some elasmobranch teeth consist almost entirely of
orthodentine and there is a transition to teeth of very largely trabec-
ular dentine with orthodentine forming a very thin coating. E. hein-
richii is interpreted here as an evolutionary form in which the ten-
dency toward reduction of orthodentine has culminated in its com-
plete absence. Holocephalians characteristically lack orthodentine
also; this is not to suggest that Edestus is more closely related to
them than to elasmobranchs, but that this is a feature of convergent
evolution.

TRABECULAR DENTINE HISTOLOGY AND
ULTRASTRUCTURE

Three morphological types of trabecular dentine were found at the
light-microscope level. The tissue is one of the most widely distrib-
uted hard tissues in early elasmobranch teeth with the same histo-
logic and ultrastructural characteristics abundantly represented in
modern sharks. All three morphological types are seen in a single
tooth organ. Type 1 trabecular dentine is a dense packing of den-
teons enclosing a fine capillary system lining the tooth crown (pi.
3a, 4a). There is diagnostically no interstitial tissue between the
denteons in Type 1 and the calcified peritubular lining is much re-

C *>

•st-

o. 3

6 o

CO < * -

c 2

i- c

S "3

s c

.3 CD

JS In

-^ CO

CO

cs .g

be £

c **

^ i «

E 3 ^

CD o -u

■s I e

m x) o

T3 c x;

i- a o,

B CO CO

5 .2

^ -3

a
>.

bx
O

I i

2 « u

^ "3

e -°

5 °

S c

O o

>

£

i:

2

CJ

CD

u

So

cr,

01

h

CO

u

_2

_c

CO

u

Mm

JS

c

~J

k

be a

CO

3

CO

'C

PQ

-4->

X

9

C

02

CN

CO

_c

X
CM

Ol,

'■3

cc

M

i-

CL,

A

JO

a

CO

eg 3

452

TAYLOR & ADAMEC: PALEOZOIC SHARK

A.

453

Plate 6. PF 2849, a, Fiber-mineral bundles within the denteon wall are arranged
in circular fashion, b, Scan at 2,000 magnifications of the denteon wall shows
branching fiber bundles, c, The fracture pattern of the interdenteonal tissue is that
of a woven-fibered hard tissue. The interstitium between denteons seen here is
characteristic of Type 2 trabecular dentine.

duced and frequently absent. Type 2 is immediately subjacent to
Type 1 and constitutes the central crown region and most of the
tooth base. The denteons are separated by an acellular interden-
teonal hard tissue (pi. 3a). The type of interstitium found here has
been referred to by Radinsky ( 1961) as interosteonal hard substance
and Peyer (1968) as cell-free interosteonal hard substance. The

Plate 7. a, Fossil denteon in Type 2 trabecular dentine in brightfield shows
Tome's fibers quite clearly. The black dots are radio-opaque pyrite. b, In polarizing
light the denteon is seen to be composed of an inner dark ring of different refraction
and therefore different fiber-crystal orientation than the bright outer ring, c, Recent
Type 2 trabecular dentine from the hammerhead shark, shows a consistent inner
ring of transversely oriented fibers, and an outer ring of more longitudinally ori-
ented fibers, d, e, In modern tissue as in the fossil, denteons with lamellae of com-
mon fiber orientation are interspersed with lamellae of alternating pitch. The inter-
denteonal hard tissue is the frothy material between the denteons. f, A natural
growth surface of a denteon from a modern Isurid shark shows the rope-like sub-
structure of the denteon wall.

454

Plate 8. a, The denteon wall appears to be composed of continuous and discon-
tinuous super-bundles in left-handed coils, PF 2848. b, Lamellae arranged circularly
around the lumen are composed of spiralling left-handed super-bundles.

455

o

£

I

Q

ad S

tJ< CO

00 — '

5*
i ■

s •«

£ o

2 S
*~ c

i

8 I

fe Eh

fl

b 3

00 >-

X a
eg w

a> a)

<l

456

457

458 FIELDIANA: GEOLOGY, VOLUME 33

tissue is here called interdenteonal because it is the interstitium of
denteons and is laid down by odontoblasts. A very distinct fracture
pattern was observed on untreated natural surfaces of this tissue,
as shown in Plate 6c. The fracture pattern of this interdenteonal
tissue is that of a woven-fibered hard tissue. The denteons in Type 2
characteristically have a thick peritubular lining (pi. 5a, b, c). The
substance is quite distinct from the interdenteonal tissue under
polarizing light; it has a different birefringence and does not have
the fracture pattern of a woven fibered tissue (pi. 7b). The margin
between the peritubular lining and the innermost lamella is regular
and the absence of odontoblast spaces suggests an erosion surface
instead of an initial growth border (pi. 5a, b). The substance is prob-
ably a lime salt, phosphate accumulation — a mineral storage to
which the circulatory system has ready access. The interdenteonal
tissue is avascular and therefore cannot serve as a ready access
mineral store in the adult tooth as can the peritubular lining. Type 3
trabecular dentine is found in the outer portion of the tooth base,
the external millimeter of all of the non-crown dentine (pi. 3a). It is a
spongy system of open trabecles lacking denteon organization. The
spongiosum grades into Type 2 in the internal portion of the root.

It is the ultrastructure of Types 1 and 2 that scanning has signifi-
cantly added to — that is the nature of the denteon and its matrix.
Scanning evidence emphasizes the similarities in extracellular
events between dentine and bone. It is well known that the denteon
analog, the osteon in secondary bone, grows by the internal apposi-
tion of successive lamellar rings, but the ultrastructure has not been
previously demonstrated for dentine. Annular concentric growth
within a lamella of fossil trabecular dentine appears to be accom-
plished by left-handed coils of super-bundles (pi. 8a, b). This growth
arrangement is confirmed in modern material. A scanning micro-
graph of a growing denteon in an immature Isurid tooth shows
clearly the rope-like spirals within the lamella which when mature
will enclose the denteon lumen (pi. 7f ). The constituents of the rope-
like bundles are organized at the scanning level like a structural
protein complexed with a mineral and a bituminized carbohydrate
phase. The fibrous bundles appear to have an exoskeleton probably
of fluor-apatite, which coalesces each to adjacent bundles in a fash-
ion exactly like that described for bone (pi. 9a-d). Denteons, because
they do not remodel, retain their original structure, a structure in
which there is evidence of continuous and discontinuous spirals of
super-bundles. The spatial relationship between apatite crystal dis-
tribution and structural protein is similar to that found in bone.

TAYLOR & ADAMEC: PALEOZOIC SHARK 459

There is alternation of pitch in the lamellar spirals of the denteon
wall. In polarizing light the fossil denteon is shown to have bands of
different refractive indices indicating a difference in fiber-crystal
orientation. The bright ring seen in Plate 7a is a lamella of longitudi-
nally oriented fibers and crystals. The dark inner band is of trans-
versely oriented constituents. Denteons with alternate pitch lamel-
lae are equally distributed in Types 1 and 2 trabecular dentine. The
alternate pitch and common fiber orientation is seen quite clearly in
modern trabecular dentine from a hammerhead shark (pi. 7c, d, e).
The lamellar organization of trabecular dentine seen here agrees
with that found in bone.

Enameloid

The thin outer crown covering of enameloid at the light micro-
scope level is best seen in Plates 2 and 3. Enameloid is considered a
hyper-mineralized type of dentine formed inside the basement mem-
brane of the enamel epithelium and therefore not the homolog of
enamel in higher vertebrates, which is of ectodermal origin. Peyer
(1968) distinguishes the enameloid of elasmobranchs from the true
enamel of reptiles and mammals on the basis of its ( 1 ) not originat-
ing by the mineralization of cell processes of ameloblasts, (2) by
lacking the birefringence of true enamel, and (3) the direction of
growth being not centrifugal but centripetal. Enameloid is, in fact,
formed inductively by ectodermal and mesodermal elements (Shellis
and Miles, 1974). The enameloid seen here is essentially identical to
that found in other elasmobranch teeth.

Discussion

Trabecular dentine, on the whole, has attained certain physiologi-
cal features that are significant in the evolution of hard tissues in
general. Paleozoic trabecular dentine behaves in every way like the
modern tissue at the histologic and ultrastructural levels. Bone
Haversian systems share several homologous properties with the
denteon in trabecular dentine. The denteon has acquired the cen-
tripetal growth pattern that proceeds by the successive apposition
of lamellae to surround an arteriole and venule. The presumed col-
lagen fibrils in the denteons are roughly parallel as they are in
Haversian bone. The inotropic calcification system is the same in
both. Angiogenesis and microcirculation determine the structure of
Haversian bone and this is the inference made here for trabecular
dentine: that it is formed as a result of the vascular invasion of the
anlagen by capillaries. There is a spatial relationship between fluor-

460 FIELDIANA: GEOLOGY, VOLUME 33

apatite crystal distribution and collagen periodicity as there is in
bone.

One of the major points of difference of trabecular dentine from
bone is that there is no clastic resorption. It would serve no func-
tional advantage in a system of continually replaced teeth. The only
remodelling phenomenon seen is the perivascular erosion and infill-
ing, for which there is only indirect evidence. An additional differ-
ence is that the denteons in trabecular dentine do not incarcerate
blast cells into lacunae as bone does.

This is not to suggest that trabecular dentine is the precursor of
bone. The advent of bone, however, was by transformation within
an existing developmental system and trabecular dentine demon-
strates features of a system fundamental to many hard tissues. It is
not clear what parameters of growth and function evolved in the
precursor of these hard tissues and which elements constitute the
final modification.

EMBRYOLOGY

Introduction

It is now a relatively secure developmental fact that structures
originating from epithelia require a mesenchymal association for
both development and differentiation into adult forms (Fleisch-
majer and Billingham, 1968). This fact is well illustrated in the
studies of the kidney (Grobstein, 1955), the integument and its
appendages (Wessells, 1967, 1970; Kollar and Baird, 1970a, b), the
lungs (Spooner and Wessells, 1970; Wessells, 1970), pancreas (Die-
telein-Lievre, 1970), and some portions of the thymic immune sys-
tem (Harrison et al., 1970), and many others. Involvement in such a
diversity of systems speaks toward the epithelial-mesenchymal
interaction as being a fundamental developmental process in the
ontogeny of present day organisms. It is not unreasonable, there-
fore, to attempt to explain certain observations of developmental
events in fossil forms by invoking similar processes.

Edestus heinrichii, unrelated to the modern radiation of sharks,
represents an evolutionary endpoint. To date there has been no
satisfactory explanation of the ontogeny of the diverse types of
hard tissue found in this organism, especially in the case of trabecu-
lar dentine.

Current hypotheses (0rvig, 1967c) relating to the origins of type 2
trabecular dentine center around the idea that the presumptive

TAYLOR & ADAMEC: PALEOZOIC SHARK 461

odontoblasts differentiate into competent dentine-secreting cells
retain this competency at the external borders of the developing
tooth to produce type 1 trabecular dentine, but somehow lose this
competency in the interior portions, dedifferentiate, and then pro-
ceed along an alternative developmental program to become osteo-
blasts involved in the deposition of the interstitial bands (if, in fact,
they are bone), then redifferentiate again into odontoblasts which
secrete the trabecular denteons. This hypothesis is untenable from
the standpoint of a single cell population undergoing three develop-
mental sequences. An explanation based on observations of the
embryology of modern organisms is more economic. Indeed, Herold
( 1971) has shown that one cell type undergoing maturation, but not
re-differentiation, produces the pattern of osteodentine seen in
certain teleost teeth.

Hypothetical embryo gene sis of trabecular dentine in E. heinrichii

The participation of the mesenchyme in the induction of tooth
ontogeny was clarified by Wild (1955a, b) with his observation that
the mesenchymal component of urodele teeth migrated from their
origin in the neural crest. Since that time there has been consider-
able controversy over the major determinant of tooth development,
i.e., whether the mesenchyme or the oral epithelium determines the
ultimate type of dentition at a particular locus. Pourtois ( 1964) and
Miller (1969) have shown that in the stages preceding epithelial
invagination, the isolated epithelial components can develop into
the correct dental type (incisor or molar). However, Kollar and
Baird (1970a, b) demonstrated the inductive predominance of the
dental papilla in later stages of tooth morphogenesis. These con-
trary results can be easily reconciled if one were to propose that
initially the epithelium is provisionally determined and exerts an
inductive effect on the presumptive dental mesenchyme which, as
tooth development proceeds, assumes complete control over the
final morphogenetic outcome. The implication here, then, is that the
cells of the dental papilla become autonomous in their ability to dif-
ferentiate after their initial induction by the epithelial component.
Huggins et al. ( 1934) demonstrated this fact in developing dog teeth
by implanting isolated mesenchymal components into abdominal
muscles, where calcified dentine was formed by the odontoblasts in
the absence of direct epithelial contact with the enamel organ. The
nature of the dentine formed was varied, from orthodentine grading
through trabecular forms to material indistinguishable from true
bone. This latter finding is of utmost interest in the discussions of

462 FIELDIANA: GEOLOGY, VOLUME 33

the origins of the trabecular dentine found in the teeth of fossil
sharks and indeed of the modern elasmobranchs with similar hard
tissue types. The major consideration here is that the trabecular
pattern arose when the odontoblastic layer was in the presence of,
but not in contact with, the internal enamel epithelium. This sug-
gests that the resultant interaction was incomplete in the sense that
there was no physical epithelial-mesenchymal interface to order the
secretory events. It is also significant that in these experiments
there seemed to be a time dependency for the pattern of dentine
seen, with the trabecular dentine formed in the implants continued
beyond 24 days. Combining these two points, one has the basis for
an interesting speculation for the genesis of a trabecular dentine
pattern: trabecular dentine is the natural maturation pattern of
secretion by induced odontoblasts in the absence of an internal
enamel epithelial interface, but in the presence of a distant, indirect
or primitive epithelial influence.

Against the above background it is possible to construct a sound
hypothesis regarding the generation of trabecular dentine during
the morphogenesis of elasmobranch teeth. Peyer (1968) describes
the histiogenesis of the teeth of two modern sharks, Squalus acan-
thias and Scylorhinus canicula, and shows that, in spite of an evag-
inating surface development of the teeth, all embryological cell
types characteristic of higher vertebrate forms are present. Here we
shall term the analog of the internal enamel epithelium, the internal
odontogenic epithelium, to avoid assumptions about the develop-
mental future of this tissue.

Peyer's microscopic sections show that during early morphogene-
sis there are presumptive odontoblasts subjacent to the internal
odontogenic epithelium and thus subjected to the initial inductive
influence. As maturation of the tooth anlagen continues, many of
the initially induced cells become crowded into the center of the
dental papilla, creating a population of potentially dentine-secreting
cells in the deep interior of the tooth. It is reasonable to propose, as
the observations of Huggins et al. (1934, 1970) suggest, that this
cell population, deprived of an epithelial-mesenchymal interface
with which to orient its secretory activity, will nevertheless retain
its secretory ability long after exposure to the epithelial inductive
influence and deposit dentine circumferentially to produce the tra-
becular pattern. Those odontoblasts remaining juxtaposed to the
internal odontogenic epithelium align their secretion product along
the epithelial-mesenchymal interface. This product is reminiscent of

TAYLOR & ADAMEC: PALEOZOIC SHARK 463

orthodentine but in E. heinrichii would actually be enameloid, pro-
duced under the influence of the ontogenetically primitive epithe-
lium. The epithelial contribution could be either a purely inductive
effect on the subjacent mesenchyme or a product manufactured by
the epithelium and incorporated into the enameloid with the mesen-
chymal component. It is noteworthy in this regard that Shellis and
Miles (1974) have demonstrated a matrix component in the enamel-
oid of certain teleost fishes that is secreted by the internal dental
epithelium. Indeed, it is well known that matrix materials play an
essential role in the differentiative program of tissues invoking epi-
thelial-mesenchymal inter-actions (Dodson, 1963; Fell and Grob-
stein, 1968; Wessell and Evans, 1968; Bernfield, 1970; Goetnick
and Sekellick, 1972; Vracko and Benditt, 1972). Thus, while the
exact timing of developmental stages, and the functional life span of
the cells involved in trabecular dentine deposition are not known, it
is tempting to look toward such matrix influences as an organizer
for the succession of developmental events.

The more centrally located odontoblasts could be subject to an
inductive gradient from the internal odontogenic epithelium to
produce the types 1 and 2 trabecular dentines seen in E. heinrichii.
It is germane to note that in several modern radiations of sharks,
the appearance of a type 2 trabecular dentine occurs concomitantly
with the disappearance of the internal odontogenic epithelium
(Peyer, 1968), at which time the autonomy of the most centrally
placed odontoblasts might be assured.

While the trabecular morphology seen in E. heinrichii might be
explained by invoking the above embryological processes, the pres-
ence of the interstitial banding pattern seen is thus far not treated
by these hypotheses. However, the ideas presented so far can easily
include such a phenomenon. Considering the histology of the tooth
primordium before hard tissue deposition, one should recall that the
mesenchyme of the dental papilla is composed of two populations of
cells — the neural crest cells, which are the presumptive odonto-
blasts, and a population of uncommitted pluripotential mesenchy-
mal cells which will eventually differentiate into such diverse tis-
sues as fibroblastic connective tissue, vascular elements, and other
mesodermally derived tissue. If, in the course of deposition of the
dentine by the odontoblasts located in the central portions of the
tooth, some of the pluripotential mesenchymal cells are trapped
between the expanding trabecles of dentine, one could easily theo-
rize the establishment of an interdenteonal fibroblast population.

464 FIELDIANA: GEOLOGY, VOLUME 33

The findings of Huggins et al. ( 1934, 1970) showed that implants of
both odontoblast and matrix components can induce bone and
cementum formation from surrounding fibroblasts, one can envision
a similar process occurring with these trapped fibroblasts. That is
to say, the expanding dentine network might exert an inductive
effect on the interdenteonal fibroblasts to produce hard tissue meta-
plasia manifested as the interstitial bands. The absence of banding
in the type 1 dentine seen in E. heinrichii could be accounted for by
the exclusion of fibroblasts from the region immediately subjacent
to the internal odontogenic epithelium, this area being exclusively
populated by the neural crest presumptive odontoblasts.

Evidence for such a developmental program is present in both the
light and the scanning electron micrographs discussed in the first
section of this paper. From the light micrographs it is evident that
the interstitial bands are acellular. However, the scanning electron
micrographs show that the matrix of the interstitial bands is com-
posed of woven fibers, probably originally collagenous, oriented per-
pendicular to the axis of the dentinal matrix. This suggests that two
different cell populations separately secreted their respective matri-
ces, and that the interstitial population oriented its product along
the lines of stress between adjacent denteons. The interstitial hard
substance may be related to cementum, judging from the loosely
woven texture of the fibrous matrix, the lack of incarcerated cells
usually associated with true bone, and the interrelationship with
the forming dentine. It is noteworthy that the banding seen in E.
heinrichii is not the type discussed by 0rvig ( 1967c) under the head-
ing of osteodentine or osteo-semi-dentine, where a definite bony
superstructure precedes the deposition of dentine. The evidence in
the sections of E. heinrichii indicates that the dentine was the first
of the hard tissues to be laid down, since the interstitial bands are
discontinuous and narrow, suggesting that indeed they are the
result of isolation of fibroblastic strands at the matrix border of the
enlarging trabecles. This last point must be stressed, since primary
bone formation with secondary dentine deposition should be mani-
fested by well-defined, continuous interstitial bands. The existence
of the hard tissue class known as mesodentine is attributed to the
presence of a mesenchymal cell population intermediate between
osteoblasts and odontoblasts (0rvig, 1967c). This point is not ac-
ceptable when one considers the mass of data showing that in a
broad spectrum of present day organisms possessing denticles, the
odontoblasts do not originate from the same precursor population
as do the osteoblasts ( Horstadius, 1950; Wild, 1955a; Fowler, 1972;

TAYLOR & ADAMEC: PALEOZOIC SHARK 465

Kelly and Bluemink, 1974). Odontoblasts derive from the neural
crest and are thus mesectodermal, whereas osteoblasts derive from
a purely mesodermal cell population. Mesodentine can be more logi-
cally considered within the above concept, originating via trapping
of fibroblastic elements within the expanding dentinal matrix,
rather than as a result of programmed loss of odontoblast function
in the course of the secretory process.

In further pursuit of the question of the presence of bone in E.
heinrichii, one must consider whether the trabecles themselves
represent a stage in the evolution of bone. There are several develop-
mental dissimilarities that preclude making such an association.
First is the difference in the cellular origins already discussed at
length above. Second, previous discussions propose that the den-
tinal trabecles of E. heinrichii are formed in a centripetal direction
away from the presumptive regions of the interstitial bands, and
toward elements of the concurrently developing vascular supply,
i.e., toward the nutritional supply of the odontoblasts ahead of the
dentine and predentine deposition which is modulated by an ex-
panding vascular network. The smaller dentine pattern seen in the
type 1 dentine could well be the result of such a process occurring
around abundant terminal small vessels near the periphery of the
tooth. Third, there is no evidence in any of the material from E.
heinrichii studied of hard tissue clasis along the free edge of the
denteons, as shown by the absence of irregular erosions, thus elimi-
nating an important developmental aspect of most bone, since an
elementary feature of more rapid bone growth and development is
the role of osteoclastic resorption in the modelling process ( Grune-
berg, 1937). However, it is generally agreed that both intramem-
branous and endochondral bone ontogeny in higher vertebrates pro-
ceeds with osteoblastic activity preceding the osteoclastic contribu-
tion to remodelling.

The presence of trabecular dentine in edestids might thus repre-
sent a level in evolutionary advance to modern bone with complete
dependence on blastic activity for this process. Clastic activity
might then be interpreted as an evolutionary adaptation to adjust
final form more adequately to ultimate function. However, the rela-
tion between bone and edestid trabecular dentine cannot be estab-
lished on this basis alone.

Finally, the Tome's processes within the dentinal matrix function
differently from bone canaliculi, since their primary purpose is the
mineralization of the matrix, not the nutritional supply of the em-
bedded cell population, as is the case for the canaliculi. Herold

466 FIELDIANA: GEOLOGY, VOLUME 33

(1971) concludes that teleost osteodentine bears no relationship to
bone in that there are critical differences in the formation process,
matrix structure, and in cellular components. However, unlike the
teleost system studied, the edestid trabecular dentine does show
evidence of appositional growth within the denteons, a feature used
by Herold ( 1971 ) to disassociate osteodentine from bone.

In conclusion, what has been presented in this section is an at-
tempt to explain the presence of the various hard tissues seen in E.
heinrichii on the basis of what is known about the histiogenesis of
similar tissues in modern vertebrates. It was not intended to place
this edestid in an evolutionary perspective, nor to relate it to other
forms with similar or dissimilar histology. Rather, the purpose here
was to explore an approach to the interpretation of paleohistology
with the suggestion that application of such a perspective to similar
material might lead to a more rational construction of evolutionary
history.

REFERENCES

Bendix-Almgreen, S. E.

1962. The anatomy of the Permian edestids from East Greenland. Medd. Dansk.
Geol. Fore., 15(1), pp. 152-153.

1968. Affinities of bradyodonts. In Orvig, T., ed., Current problems of lower ver-
tebrate phylogeny, Nobel Symp. #4, Interscience Publ., pp. 153-170.

Bernfield, M.
1970. Collagen synthesis during epithelial-mesenchymal interactions. Dev. Biol,
22(2), pp. 213-231.

DlETELIN-LlEVRE, F.

1970. Tissus exocrine et endocrine du pancreas chez l'embryon du poulet; origine
et interactions tissulaires dans la differentiation. Dev. Biol, 22( 1), pp. 138-156.

Dodson, J. W.

1963. On the nature of tissue interactions in embryonic skin. Exp. Cell. Res., 31( 1),
pp. 233-235.

Eastman, C. R.

1902. Notice of interesting new forms of Carboniferous fish remains. Amer. Nat.,
36(431), pp. 849-854.

1903. On the nature of Edestus and related forms. Mark Anniversary Volume
(Harvard Univ.), pp. 279-289.

Fell, P. E. and C. Grobstein
1968. The influence of extra-epithelial factors of the growth of embryonic pancre-
atic epithelium. Exp. Cell Res., 53( 1), pp. 301-304.

TAYLOR & ADAMEC: PALEOZOIC SHARK 467

Fleischmajer, R. and R. E. Billingham, eds.
1968. Epithelial-mesenchymal interaction. 18th Hahnemann Symp., Williams &
Wilkins.

Fowler, M. J.
1972. Cellular heterogeneity in the chick wing mesoblast. 1. Analysis by equilib-
rium density gradient centrifugation. Dev. Biol., 28(4), pp. 583-602.

Goetnick, P. F. and M. J. Sekellick
1972. Observations of collagen synthesis, lattice formation, and morphology of
scaleless and normal embryonic skin. Dev. Biol., 28, pp. 636-648.

Grobstein, C.
1955. Inductive interactions in the development of the mouse metanephros. J.
Exp. Zool, 130, pp. 319-340.

Gruneberg, H.
1937. The relations of endogenous and exogenous factors in bone and tooth
development. The teeth of the Gray-Lethal mouse. J. Anat., 71(2), pp. 236-244.

Harrison, B. H., C. I. Patt, and J. T. Albright

1970. The epithelial origin of lymphocites in the palatine tonsil of rabbits. Tissue
Cell, 2(1), pp. 33-45.

Hay, O. P.

1910. On the nature of Edestus and related genera, with description of a new
genus and three new species. Proc. U.S. Nat. Mus., 42, pp. 43-61.

1912. On an important species of Edestus: with description of a new species -Edes-
tus mints. Proc. U.S. Nat. Mus., 42, pp. 31-38.

Herold, R. C.

1971. Osteodentinogenesis. An ultrastructural study of tooth formation in the
Pike, Esox lucius. Zellforsch., 112, pp. 1-14.

Hitchcock, E.
1856. Account of the discovery of the fossil jaw of an extinct family of sharks from
the coal formation. Proc. Amer. Assn. Adv. Sci., pp. 229-230.

HORSTADIUS, S.

1950. The neural crest: its properties and derivatives in the light of experimental
research. Oxford Univ. Press.

Huggins, C. B., H. R. McCarroll, and A. A. Dahlberg
1934. Transplantation of tooth germ elements and the experimental heterotropic
formation of dentin and enamel. J. Exp. Med., 60(2), pp. 199-210.

Huggins, C. B., S. Wiseman, and A. A. Reddi
1970. Transformation of fibroblasts by allogenic and xenogenic transplants of
demineralized tooth and bone. J. Exp. Med., 132(6), p. 1250.

468 FIELDIANA: GEOLOGY, VOLUME 33

Karpinskii, A. P.
1899. Ueber die Reste von Edestiden un die neue Gattung H elicoprion. Vses. Min.

Ovo. Zapiski, ser. 1, 36, pp. 361-476.
1924. Zamechaniya o zubnykh segmentakh Edestidae i ob ikh orientirovke (re-
marks on the tooth segments of Edestidae and their position). Izv. Imp. Akad.
Nauk,(N.S.) 3, pp. 379-388.

Kelly, R. O. and J. G. Blueminck
1974. An ultrastructural analysis of cell and matrix differentiation during early
limb development in Xenopus laervis. Dev. Biol., 37, pp. 1-17.

Kollar, E. J.

1970. The induction of hair follicles by dermal papillae. J. Invest. Derm., 55(6),
pp. 374-378.

Kollar, E. J., and G. R. Baird
1970a. Tissue interactions in embryonic mouse tooth germs. I. Reorganization of

the dental epithelium during tooth germ reconstruction. J. Embryol. Exp.

Morphol.,24(l),pp. 159-171.
1970b. Tissue interactions in embryonic mouse tooth germs. II. The inductive role

of the dental papilla. J. Embryol. Exp. Morphol, 24( 1), pp. 173-186.

Miller, W. A.
1969. Inductive changes in early tooth development: I. A study of mouse tooth
development on the chick chorio-allantois. J. Dent. Res., suppl. to no. 5, pp.
719-725.

Newberry, J. S.

1889. The Paleozoic fishes of North America. Monogr. U.S. Geol. Surv., 16.

Newberry, J. S. and A. H. Worthen
1866. Descriptions of new species of vertebrates, mainly from the sub-Carbonifer-
ous limestones and coal measures of Illinois. 111. Geol. Surv., 2, pp. 9-134.
1870. Descriptions of fossil vertebrates. 111. Geol. Surv., 4, pp. 347-374.

NlELSON.E.

1932. Permo-Carboniferous fishes from East Greenland. Medd. Grdnland, 86(3),
pp. 1-63.

1952. On a new species of little known Edestidae from the Permian and Triassic of
East Greenland. Medd. Grdnland, 144(5), pp. 1-55.

Obruchev, D. V.

1953. Izuchenie Edestus i raboty A. P. Karpinskogo, (The study of Edestus and
the works of A. P. Karpinskii). Tr. Paleontol. Inst. Akad. Nauk SSSR., 45, pp.
1-85.

0RVIG, T.

1951. Histological studies of Placoderms and fossil elasmobranchs. Arkiv. Zool.,
2, pp. 321-454.

1965. Paleohistological notes. 2. Certain comments on the phyletic significance of
acellular bone tissue in early lower vertebrates. Arkiv. Zool., 2( 16), pp. 551-556.

TAYLOR & ADAMEC: PALEOZOIC SHARK 469

1967a. Histological studies of Ostracoderms, Placoderms, and fossil elasmo-
branchs. Arkiv. Zool, 19(1), pp. 1-39.

1967b. The dermal skeleton: general considerations. In 0rvig, T., ed., Current

problems of lower vertebrate phylogeny, Nobel Symp. #4, Interscience Publ.,

pp. 373-397.
1967c. Phylogeny of tooth tissues: evolution of some calcified tissues. In Miles,

A. E. W. and R. C. Grenlach, eds., Structural and chemical organization of

teeth. Academic Press, pp. 45-105.

Peyer.B.
1968. R. Zangerl, translator. Comparative odontology. Univ. of Chicago Press.

Pflugfelder, c.

1930. Das Gebiss der Gymnodonten. Ein. Beitr. Histogenese Dentins Z. Anat.
Entw.-Gesch., 93, pp. 543-566.

POURTOIS, M.

1964. Comportment en culture in vitro des ebauches dentaires de rongeurs pre-
levees aux stades de predifferenciation. J. Embryol. Exp. Morphol., 12, pp.
381-405.

Radinsky, L.
1961. Tooth histology as a taxonomic criterion for cartilaginous fishes. J.
Morphol., 109, pp. 73-81.

Shellis, R. P. and A. E. W. Miles
1974. Autoradiographic study of the formation of enameloid and dentine matrices
in Teleost fishes using tritiated amino acids. Proc. Roy. Soc. London, B, 185,
pp. 51-72.

Spooner, B. S. and N. K. Wessells

1970. Mammalian lung development: interactions in primordium formation and
bronchial morphogenesis. J. Exp. Zool., 175(4), pp. 445-454.

Vracko, R. and E. D. Benditt
1972. Basal lamina: the scaffold for orderly cell replacement. Observations on
regeneration of injured skeletal muscle fibers and capillaries. J. Cell Biol., 55(2),
pp. 406-419.

Wessells, N. K.

1967. Differentiation of epidermis and epidermal derivatives. New Eng. J. Med.,
277(1), pp. 21-32.

1970. Mammalian lung development: interactions in the formation and morpho-
genesis of trachael buds. J. Exp. Zool., 175(4), pp. 455-466.

Wessells, N. K. and J. Evans

1968. The ultrastructure of oriented cells and extracellular materials between
developing feathers. Dev. Biol., 18( 1), pp. 42-61.

Wild, C. E.
1955a. The Urodele neuro-epithelium: I. The differentiation of the cranial neural
crest. J. Exp. Zool., 130(3), pp. 573-595.

470 FIELDIANA: GEOLOGY, VOLUME 33

1955b. The Urodele neuro-epithelium: II. The relationship between phenylalanine
metabolism and differentiation of neural crest cells. J. M- .^7(2), PP-

313-344. m

Zangerl, R. H •

1966. A new shark of the family Edestidae, Ornithoprion hertwigi. Fieldiana:

Geol., 16(1), pp. 1-43.
1973. Interrelationships of early Chondrichthyans. In Greenwood, P. H., R. S.

Miles, and C. Patterson, eds., Interrelationships of fishes, J. Linn. Soc, 53, pp.

1-14.

Zangerl, R. and E. S. Richardson, Jr.

1963. The paleoecological history of two Pennsylvanian black shales. Fieldiana:
Geol., Mem., 4.

ZWILLING.E.

1955. Ectoderm-mesoderm relationship in the development of the chick embryo
limb bud. J. Exp. Zool., 128, pp. 423-441.

:*>£S2V** J

t<«".

7»-W