/ . Embryol. exp. Morph. Vol. 16, 1, pp. 91-104, August 1966
Printed in Great Britain
91
Studies on pattern regulation in hydra
I. Regional differences in time required for
hypostome determination
By GERALD WEBSTER 1 & LEWIS WOLPERT 2
From the Zoology Department, King's College, University of London
INTRODUCTION
Many systems having a linear pattern of spatial differentiation are capable of
regulating this pattern during development or regeneration. This is very evident
in the early development of the sea-urchin embryo, the development of cellular
slime moulds, and in the regeneration of hydroids, planarians and vertebrate
limbs. Hydra is a classic example of such a system in that removal or isolation of
almost any region leads to the reconstitution of the original pattern. As a
result of earlier studies on such systems attempts were made to provide a general
explanation in terms of axial gradients, apical dominance, and polarity (Huxley
& de Beer, 1934; Child, 1941). Work since this time has not invalidated in any
striking manner these concepts, but has rather tended to extend them (e.g.
Rose, 1952, 1957). In some ways the concepts have tended to remain descriptive
rather than explanatory (Webster & Wolpert, 1966).
Spiegelman (1945) has attempted to specify some of the minimum requirements for regulation in such systems in terms of these concepts. He emphasizes
that the crux of the problem of regulation lies in the fact that while the potentialities for forming a particular region exist throughout a large part of the
system, the region is normally formed from only a small part. This means that
the potentialities of most of the system are not realized. He also points out that
this fact cannot be accounted for by a simple gradient in rate of differentiation,
since a difference between two regions in rate of differentiation does not of itself
imply a restriction of differentiation to the region of greater rate. What is
needed in such a system is a' principle of limited realization', that is, a mechanism
for suppressing potentialities. Gradients, or differences, are, however, a necessary component of such systems, since the fact that two regions are capable of
a particular differentiation, and yet only one succeeds in differentiating, means
that there must be some difference between them. If such differences vary in a
1
Author's address: Strangeways Research Laboratory, Wort's Causeway, Cambridge,
England.
2
Author's address: Department of Biology as applied to Medicine, Middlesex Hospital
Medical School, London, W. 1, England.
92
G. WEBSTER & L. WOLPERT
continuous fashion, then they are gradients. These ideas seem fundamental to
any consideration of pattern formation but, as will be discussed elsewhere, may
not be applicable without modification to regions other than the dominant
region (Webster & Wolpert, 1966).
The properties of a dominant region have been characterized by Huxley &
de Beer (1934): (1) it is the first region to be formed; (2) its formation, once
initiated, is an autonomous process, independent of the formation of other
regions; (3) once formed, it exerts an organizing influence on other regions;
(4) it inhibits the formation of further dominant regions.
Hypostome
-— Tentacles
— Digestive zone
Budding zone
— Peduncle
Basal disc
B
Fig. 1. A, Diagram of hydra showing the principal regions of the axis. B, Regions
used in measurement of time for hypostome determination in isolated pieces and in
other transplantation experiments: a, Subhypostomal region; b, proximal digestive zone; c, distal peduncle; d, proximal peduncle.
There is a considerable amount of evidence which suggests that the distal
hypostome (Fig. 1A) is the dominant region in hydra. The organizing ability
of the hypostome was revealed by Browne (1909), who showed that a small
fragment of hypostome was capable of eliciting the formation of tentacles and a
short axis when grafted to a host animal. All other regions of the animal (with
the exception of the basal disc) were absorbed. These results were confirmed by
Yao (1945). Browne also observed that during distal regeneration the properties
of the hypostome were acquired by the distal region some time before the
tentacles were produced.
Rand, Bovard & Minnich (1926) shed further light on the organizing properties of the hypostome when they demonstrated that a grafted hypostome and
tentacles could suppress distal regeneration in the host animal. No other region
possessed this property. These workers drew two important conclusions with
Pattern regulation in hydra. I
93
regard to the role of the hypostome: (1) it 'initiates or controls the development of structures appropriate in relation to itself; (2) it 'inhibits the operation
of a developmental mechanism where such operation would result in the
formation of structures inconsistent with the attainment of normal form'.
The organizing properties of the hypostome—one of the characteristics of the
dominant region of a regulative system—are thus well established.
Any part of hydra can become any other part following transplantation
(Browne, 1909), and the work of many early investigators showed that most
regions (exceptions are the tentacles and basal disc) are capable of complete
regulation and therefore of forming a hypostome.
Regeneration in hydra is always polarized, i.e. distal structures (hypostome
and tentacles) are formed from distal ends and proximal structures (peduncle
and basal disc) from proximal ends. Polarity is rigorously maintained in
isolated pieces (Morgan, 1901; Tardent, 1960) but can sometimes be altered
in graft combinations (Peebles, 1900; King, 1901; Browne, 1909; Goetsch,
1929); in such cases the new polarity is usually determined in relation to the
position of a hypostome.
This series of papers will consider the factors controlling the formation and
localization of the dominant region—that is, the hypostome—in hydra. Effectively we will be concerned with regulation of a simple linear pattern consisting
of two regions, one of which is the hypostome. It should be noted that this whole
formulation of the problem is quite different from that of Burnett (1961, 1962),
who considers regeneration of hydra as being due to the growth of a new
hypostome and tentacles from a growth zone. He is concerned primarily with
the factors controlling growth within this zone. Apart from any theoretical
objections that can be raised against such a formulation of the problem,
Campbell (1965) has shown that growth is not localized in such a zone.
In this paper the experiments were designed to investigate regional differences in rate of hypostome formation. Earlier work on hydra has shown that
tentacle regeneration in hydra occurs more rapidly from distal than from
proximal regions (Peebles, 1897; Browne, 1909; Weimer, 1928; Rulon & Child,
1937; Spangenberg & Eakin, 1961). However, although tentacle morphogenesis
is dependent on the presence of a hypostome, the time required for this process
does not necessarily reflect the time required for hypostome differentiation. A
more direct measurement of the rate of hypostome differentiation is required.
Browne's observations suggest a test which will indicate the presence of a
hypostomal organizer at an early stage of regeneration. A piece from a regenerating animal when transplanted to a host will elicit the formation of a secondary
axis if a hypostome is present and will be absorbed if it is not. Although this
test is ultimately dependent upon formation of tentacles, the primary factor
involved is some change in properties which confers upon a region the power to
resist absorption when transplanted. By analogy with similar changes occurring
during embryonic development (Spemann, 1938; Weiss, 1939), when regions
94
G. WEBSTER & L. WOLPERT
acquire independence of, or resistance to, environmental influences, it seems
permissible to refer to this change as hypostome determination.
MATERIALS AND METHODS
The majority of experimental work was carried out using Hydra littoralis
grown from a clone supplied by Dr H. M. Lenhoff.
Animals were cultured by the methods of Loomis & Lenhoff (1956) in a
medium ('M' solution) containing 0-001 M tris (hydroxy) methylaminomethane,
0-001 M-CaCl2,0-001 M-NaHCO3,0-0001 M-KCI and 0-0001 M-MgCl2 in deionized
water; the pH was adjusted to 7-5-7-8 with N-HC1 (Muscatine, 1961). Hydra
were grown at room temperature (20-22 °C) and were fed daily with washed
brine shrimp nauplii and washed after feeding. Under these conditions, and with
weekly subculturing, they grew logarithmically with a doubling-time of about
2\ days.
In the majority of experiments actively growing 'adult' animals—possessing
one or two buds—were used 15-20 h after feeding.
Operative procedures
Simple cutting and more elaborate transplantation experiments were performed using techniques similar to those employed by earlier workers (Rand,
1899; Peebles, 1900; King, 1901; Browne, 1909). Operations were carried out
on a plasticene surface under ' M ' solution. Strict cleanliness was maintained
but sterile conditions were not necessary and no antibiotics were used. Animals
were kept in an incubator at 26 °C throughout the course of an experiment and
were examined every 24 h.
Pieces to be transplanted from intact animals were obtained by isolating rings
from selected body levels (Fig. IB), and splitting these up into smaller fragments ; small pieces which had been isolated for some time and allowed to heal
became spherical and were split open or cut in half before use as grafts. Transplantation was accomplished by inserting the graft into small incisions made in a
host animal, care being taken that the endoderm of graft and host was in close
contact to facilitate adhesion (Papenfuss, 1934). With persistence and practice
about 90 % of the transplantation experiments were successful.
After transplantation, animals were left undisturbed in the operating dish
for 1 h to allow grafts to heal in place and were then transferred to individual
solid watch-glasses or Petri dishes containing * M ' solution, and placed in the
incubator.
Due to the time required for transplantation, most experiments were done on
batches of 5-10 animals and the results presented in the tables are usually the
total of at least two experiments. There was rarely any marked discrepancy
between the results of such repeated experiments, but in cases where there was,
the two batches are tabulated separately.
Pattern regulation in hydra. I
95
Nature and form of the secondary axis
The distal structures (tentacles and axis) formed as a result of transplanting
a fragment of hypostome are derived primarily from the tissues of the host,
the graft usually remaining as the hypostome of the new axis (Browne, 1909;
Yao, 1945). Yao observed that in a few cases 'self-differentiation' occurred and
the graft transformed partially or entirely into parts of the new axis in addition
to becoming the new hypostome. Since in the present experiments grafts were
not marked, no distinction could be made between induced and self-differentiated axes, but for the purpose of this investigation this was of little consequence since the primary consideration was the presence or absence of a
hypostome and the subsequent formation of distal structures, irrespective of
how or from what material they were formed. Therefore, throughout these
papers the formation of secondary axes as a result of grafting will be referred
to for convenience as induction.
3
4
Fig. 2. Different forms of induced secondary axes.
The induced structures obtained in these experiments as a result of grafting
were of four distinct types, and comparable to those described by Browne and
Yao (Fig. 2). Type 1: a short axis of diameter comparable to that of the distal
regions of the host, a conical hypostome with mouth and two or more tentacles
around the circumference of the hypostome—in other words, a more or less
typical distal end. Type 2: a short axis of small diameter, a small hypostome and
two or more tentacles—the whole structure being much smaller and of slightly
different proportions to the host distal end. Type 3: a short axis comparable
with type 1 but with only one tentacle at its apex and no normal hypostome.
Type 4: a single tentacle (rarely two) with no axis, obtained rarely.
96
G. WEBSTER & L. WOLPERT
All these types with the exception of type 4 were classified as positive inductions. Type 1 is of the typical form described by earlier workers (Browne and
Yao), produced by grafting a hypostome fragment into distal regions. Type 2 is
of the type described by Browne as being obtained when a hypostome is
implanted in proximal regions; in the present experiments it was obtained under
similar conditions. Type 3 is typical of the regeneration which is obtained at a
distal end in an animal which has been treated with low concentrations of an
inhibitory but not lethal substance, e.g. mercaptoethanol (unpublished observations). These induced axes are stable over a period of several days at least and,
after the host animals have been fed, develop accessory tentacles and become
identical with type 1. Type 4, although clearly a definite response on the part
of the host, and therefore presumably indicating some degree of hypostomal
influence, was not considered as a case of positive induction since in the majority
of cases these tentacles disappeared (were absorbed) within 48 h, and were
always absorbed after feeding. They are apparently rather unstable structures and
are excluded from the results for this reason, bearing in mind that a primary
criterion for hypostome determination is resistance to absorption. The comparative rarity of type 4 inductions means that their exclusion has a negligible effect
on the results.
Table 1. Formation of hypostome from the subhypostomal region and
from the distal peduncle
Source of graft
Subhypostome
Time
after
cutting (h)
No. of
successful
grafts
No. of
positive
inductions
0-j
19
18
18
18
9
20
20
0
8(44%)
14 (78 %)
0
0
1 (5 %)
13 (65 %)
4
6
Distal peduncle
0-4
6
12
18
EXPERIMENTS AND RESULTS
Experiment 1. Formation of hypostome from the subhypostomal
region and from the distal peduncle
Animals were cut at the subhypostomal level (just proximal to the ring of
tentacles) or at the distal end of the peduncle (just proximal to the youngest
bud) to remove distal regions (see Fig. 1). At various times after cutting the
distal tip of the regenerating piece was removed and transplanted to the middigestive zone of an intact host hydra. The transplanted piece was approximately the same size as a normal adult hypostome. Results are shown in Table 1.
The results show a distinct difference between the two regions in the time at
Pattern regulation in hydra. I
97
which a determined hypostome is formed. Pieces from the subhypostomal region
induced 4 h after isolation while those from the distal peduncle produced only
one positive induction after 12 h and required 18 h before a substantial number
of pieces were determined. All the positive cases were type 1 or 3 inductions.
A direct, if crude, comparison of the rates of hypostome determination in the
two regions is possible if an estimate is made of the time for 50 % of the pieces
to become determined (T50). For the subhypostomal region T50 = 4-5 h; for
the distal peduncle T50 = 16-5 h.
In both pieces a determined hypostome was detected a considerable time
before tentacle regeneration occurred. Animals cut subhypostomally developed
tentacle buds about 18-20 h after cutting as compared with peduncles where
tentacles did not appear until about 30 h after cutting. In both regions therefore
the delay between hypostome determination and tentacle morphogenesis was
about 13-15 h.
The most obvious difficulty in evaluating the above results is that in the case
of the subhypostomal region the formation of a hypostome is taking place in a
piece of considerably greater size than in the case of the distal peduncle region.
An animal from which the hypostome and tentacles only have been removed
has about 4 times the mass (protein content) and 3-5 times the number of cells
(DNA content) as an isolated peduncle (Webster, 1964). There is thus a close
correlation between the relative sizes of the two pieces and the relative times for
hypostome formation, and the possibility that the size of the regenerating piece
rather than the position of the hypostome-forming region on the linear axis is
determining the time in this experiment must be considered. The next experiment was designed to investigate this possibility.
Experiment 2. Formation of hypostome by isolated
pieces of similar size
Pieces of approximately the same size (volume) were isolated from four
positions on the linear axis: the subhypostomal region (a), the proximal digestive
zone (b), the distal peduncle (c), and the proximal peduncle (d) (Fig. 1B). The
average volume of the isolated pieces—calculated from ocular micrometer
measurements—was 1-6 x 10~2 mm3, ranging from 1-0 x 10~2 to 2-8 x 10~2. At
various times after isolation the pieces were transplanted to the mid-digestive
zones of intact host animals as in Exp. 1. None of the pieces at the time of transplantation showed any signs of tentacle formation. Results are shown in
Table 2.
The first point to emerge from an inspection of Table 2 is that isolation of the
subhypostomal region from more proximal regions has no measurable effect on
the time for hypostome determination, thus proving that the results obtained
in Exp. 1 were not due to the size differences between the two pieces. The Tm
of 4-5 h can therefore be taken as the value for the time for hypostome determination in this region.
7
J EEM 16
98
G. WEBSTER & L. WOLPERT
In the case of the distal peduncle, however, isolation of this region from more
proximal regions has a considerable effect on the time for hypostome determination. The r 50 calculated from Table 2 is 28 h, an increase of 75 % in the
time compared with that measured in Exp. 1 (T50 = 16-5 h).
The r 50 for the isolated proximal digestive zone is 22 h. The rate in situ for
this region was not measured since it is known (see Burnett, 1961) that regeneration of tentacles from this region is interfered with by developing buds in the
budding zone, just proximal to the distal cut surface.
Table 2. Formation of hypostome by isolated pieces of similar size
Source of graft
Sub-hypostome (a)
Proximal digestive zone (6)
Distal peduncle (c)
Proximal peduncle (d)
Time
after
isolation (h)
No. of
successful
grafts
No. of
positive
inductions
6
7
12
18
22
26
24
30
30
40
50
19
11
14(74%)
0
1 (10 %)
10
15
14
18
20
20
10
10
14
6(40%)
7(50%)
12(67%)
4 (20 %)
14(70%)
0
1(10%)
2 (14 %)
No T50 can be obtained for the isolated proximal peduncle since even after
50 h only a very small number of transplanted pieces induced. Fragments from
this region do not survive well after isolation, tending to disintegrate after 24 h
or so. The endoderm is very sparse when the pieces are isolated, and seems to
disappear entirely in many pieces which have been isolated some time, thus
making transplantation very difficult—hence the small number of successful
grafts. This region, unlike the other three, was never observed to undergo total
reconstitution but the positive inductions indicate that, in some animals at least,
this region is capable of forming a hypostome, albeit very slowly. It is felt that
more significance should be attached to the small number of positive cases than
to the larger number of negative cases in this sort of experiment when considering the possible properties of the region in situ, since a new factor—survival
in isolation—is introduced.
The pieces from the subhypostomal region and the proximal digestive zone
were about a fifth by volume of the whole digestive zone; the pieces from the
distal and proximal ends of the peduncle were about a third of the volume of the
whole peduncle. Using the values for regional protein and DNA content
(Webster, 1964), and assuming that these values are uniform throughout a
region, i.e. the same at distal and proximal ends (almost certainly true for the
digestive zone at least), it may be concluded that the size of the four regions in
Pattern regulation in hydra. I
99
terms of mass of living material or number of cells is similar. It is probable
therefore that the differences between the four regions in time taken for hypostome determination are not due to size differences.
This experiment shows therefore that an important factor governing the time
for hypostome determination in an isolated region is the original position of the
region on the linear axis of the animal. The time is short in regions from distal
levels and long in those from proximal levels.
The times, rates and relative rates of hypostome determination for the four
isolated regions are summarized in Table 3.
Table 3. Regional differences in time for hypostome determination
Proximal
Subdigestive
zone
hypostome
(b)
(a)
Time (h) (7-BO)
Rate (l/r 60 )
Relative rate =
regional rate
sub-hypo, rate
Distal
peduncle
(c)
. Proximal
peduncle
4-5
0-22
22
005
28
004
> 50
< 002
100
22
16
< 9
DISCUSSION
The technique for assessing hypostome determination requires some comment.
The formation of a secondary axis as a result of transplantation is not only
dependent upon the acquisition of inductive ability by the region, and upon a
morphogenetic response by the host to this inductive influence, but also upon
acquisition by the graft of resistance to absorption. Providing that the piece is
always transplanted to the same region of the host, the morphogenetic response
of the host can be ruled out as an important variable in this test. This indeed
confers upon the method a distinct advantage over that involving simple
observation of tentacle formation in regenerating pieces, where the responding
region varies with the level of the axis. It is, however, conceivable that resistance
to absorption and inductive ability are independent properties and that under
certain conditions one may be present without the other. There is no simple
means of resolving this difficulty and it can only be pointed out that, in the
entire course of the experimental work, a transplanted piece either induced a
new axis or was absorbed within 24-48 h. In some cases the piece transformed
into a peduncle and basal disc, but was never observed to persist unchanged.
In view of this difficulty, it is evident that the method is really a test of both
acquisition of resistance to absorption and of ability to induce. When hypostome determination is referred to, these are the changes which will be implied.
The main conclusions to be drawn from the experiments are that an important factor governing the time for hypostome determination in an isolated
7-2
100
G. WEBSTER & L. WOLPERT
region is the original position of the region on the linear axis of the animal. It
seems reasonable to postulate the existence of an axial gradient in properties
which is expressed when regions are isolated as a difference in time for hypostome determination. The time required for hypostome determination by the
subhypostomal region is independent of any more proximal regions being
present. In contrast the time required by the distal peduncle region is dependent
on the presence of more proximal regions. Finally, the hypostome is determined
some time before any visible signs of regeneration, such as tentacle formation,
are evident.
There is a considerable amount of experimental evidence that tentacle
regeneration in hydra occurs more rapidly from distal than from proximal levels
of the axis (Peebles, 1897; Browne, 1909; Weimer, 1928; Rulon & Child, 1937;
Spangenberg & Eakin, 1961) and this Was also noted in Exp. 1. The results of
Exp. 2 suggest that the difference in time required for tentacle regeneration is a
consequence of the difference in the time for hypostome determination and that
the formation of the hypostome is the primary, if not the only, time-determining,
step in distal regeneration. This conclusion is borne out by the observation in
Exp. 1 that the interval between the formation of a hypostome and the appearance of tentacles is practically identical (13-15 h) in animals regenerating from
two levels of the axis; if regional differences in time for tentacle formation do
exist then they must be very small.
The time of about 4 h for hypostome determination by a subhypostomal
region fits in well with the studies of Ham & Eakin (1950), who concluded
from studies on the effect of inhibitors that an important change in response to
chemical agents occurs after about 4 h regeneration.
It has been known for many years that differences exist between distal and
proximal levels of marine hydroids in the rate at which a hydranth is reconstituted (reviews by Child, 1941; Tardent, 1960, 1963). The most detailed
investigation is that of Barth (1938), who showed that the 'inherent' rate of
hydranth reconstitution in isolated pieces of the stem of Tubularia decreased
rapidly as the pieces came from successively more proximal levels. Evidently
there is present in the stem an axial gradient in properties which is expressed as
graded differences in the rate at which hydranth reconstitution can occur.
There is some evidence which suggests that the observed axial differences in
rate of reconstitution do not reflect the rate at which all the processes involved
in hydranth formation occur. Lund (1923), working with Obelia, was able to
show that the axial rate differences arose as a result of differences in the time at
which visible regeneration (out-growth of the coenosarc) commenced at different levels; once the process had started it proceeded at similar rates in all
pieces, irrespective of their original axial position. Similar results were obtained
by Steinberg (1954,1955) with Tubularia. The process of hydranth reconstitution
was found to be divisible into two phases, the first consisting of the formation
of a distal thickening as a result of distal migration of the coenosarc, the second
Pattern regulation in hydra. I
101
of actual hydranth differentiation. Steinberg observed that the axial difference
in rate of reconstitution arose as a result of a difference in the time at which the
distal thickening was formed, the process of hydranth differentiation occurring
at the same rate in all pieces. These results suggest that in other hydroids, as in
hydra, the primary time-determining step occurs at an early stage in the process
of reconstitution.
During hydranth reconstitution, differentiation or organization seems to
proceed in a disto-proximal direction (Child, 1941; Rose, 1955) and it has been
demonstrated in Tubularia that distal parts of the hydranth become determined
before proximal parts when tested by isolation (Davidson & Berrill, 1948). It is
known that the most distal region of some hydroids possesses organizing ability,
comparable with that of the hypostome of hydra, and will induce a secondary
axis when transplanted (Corymorpha, Child, 1941; Cordylophora, Beadle &
Booth, 1938; Moore, 1952).
These facts suggest that the time-determining step which occurs at an early
stage of reconstitution is, in all hydroids, the formation of a distal organizer or
dominant region.
The term ' rate of hypostome formation' has been carefully avoided in the
description of our results since what has been measured here is the time required
for hypostome determination. Even the concept of rate of hypostome formation
is an obscure one, implying as it does a continuous transformation of nonhypostomal region into hypostome from the time when the original hypostome
is removed. There is no reason to believe that this is the case. Barth (1938), in
his work on the reconstitution of the hydranth in Tubularia, uses rate of regeneration to mean the amount of stem transformed into hydranth per unit time.
However, as already pointed out above, hydranth reconstitution is not a unitary
process, and it is not meaningful to talk about a rate for such a complex process.
There is no evidence to suggest that either hypostome formation or subsequent
hydranth morphogenesis can be treated as continuous processes with respect to
time, particularly if switch mechanisms are involved, as in the case of hypostome
determination which seems to be an 'all or none' phenomenon.
This raises the whole problem of what determines the time required for hypostome determination. It seems that if the formation of a hypostome is viewed as
a 'developmental pathway' then regional time-differences could arise in three
main ways: (1) all regions start from the same point on the pathway and at the
same time, but differ in their rate of development; (2) all regions start from the
same point, develop at the same rate, but differ in the time at which development is initiated; (3) all regions start at the same time, develop at the same
rate, but differ in the point on the pathway from which they start. It is of great
importance to establish which of these is in fact operative and it should be noted
that only case (1) requires a gradient in rate of development. As pointed out
above, for hydranth reconstitution the difference in time required for morphogenesis seems to result from the operation of the second mechanism.
102
G. WEBSTER & L. WOLPERT
Some consideration must be given to the difference in time for hypostome
determination from the distal peduncle in isolated pieces as compared with
intact peduncles. The increase in time required resulting from isolation could be
the result either of a reduction in the total size of the piece, or of the removal of
more proximal regions, or of the presence of a second cut surface. No definite
decision can be made between these alternatives with the evidence at present
available, but the view favoured at the moment is that the result is a size effect.
The experiments of Barth (1938) and Steinberg (1954) on Tubularia led these
workers to conclude that the size of the piece (volume or number of cells) had a
definite influence on the rate of hydranth reconstitution though a more specific
effect due to the absence of proximal regions was not considered. The role of the
second (exposed) cut surface was excluded in Barth's work where ligatures were
employed, and the cut surface has been minimized as an important factor in the
reconstitution of hydra by the few workers who have considered the problem
(see, for example, Morgan, 1901). The view of Child (1941) that cutting somehow
'activates' a region and initiates the formation of a new dominant region is
probably no longer tenable (Rose & Rose, 1941).
In contrast to the distal peduncle, isolation of the subhypostomal region has
no apparent effect on the time for hypostome determination, even though the
total size of the regenerating piece is reduced by a very much greater amount.
A 75 % increase in the time required for hypostome determination would
change the r 50 for this region from 4-5 to about 8 h; it is possible that such an
increase does occur but that the relative crudity of the experimental technique,
coupled with the variability of the material, makes a change of this magnitude
undetectable.
We have shown in this paper that there is an axial gradient in time for hypostome determination and this is probably the time-determining step in distal
regeneration. As pointed out in the introduction, graded differences in properties
are a necessary but not sufficient factor for regulation to occur. In addition a
'principle of limited realization' is required. This will be investigated in the
following paper.
SUMMARY
1. The regulation of hydra into hypostomal and non-hypostomal regions is
considered as a model for the regulation of a linear pattern into two compartments. Regulation of such a system requires graded differences and a 'principle
of limited realization'.
2. The hypostome is the dominant region in hydra and the time for its
determination in isolated pieces depends upon the original position of the piece
along the linear axis. The time increases in a disto-proximal direction from 4 h
to more than 20 h.
3. The determination of the hypostome is the first detectable event in distal
regeneration and is probably the time-determining step in this process. Work on
other hydroids is reinterpreted in these terms.
Pattern regulation in hydra. I
103
4. The time for hypostome determination in proximal regions is significantly
affected by the presence of more proximal regions, while in distal regions the
time appears independent of the presence of more proximal regions.
RESUME
Etudes sur la regulation chez Vhydre. I. Differences regionales
dans le temps requis pour la determination de Vhypostome
1. La regulation de l'hydre en regions hypostomale et non-hypostomale est
consideree comme un modele de regulation d'un type lineaire en deux compartiments. La regulation d'un tel systeme requiert des differences graduees et
un 'principe de realisation limitee'.
2. L'hypostome est la region dominante de l'hydre et le temps requis pour sa
determination en fragments isoles depend de la localisation primitive du fragment le long de l'axe lineaire. La duree augmente en direction disto-proximale,
de 4 heures a plus de 50 heures.
3. La determination de l'hypostome est le premier evenement decelable dans
la regeneration distale et est probablement l'etape determinante de ce processus.
Le travail fait sur d'autres hydroides est re-interprete dans ce sens.
4. Le temps necessaire a la determination de l'hypostome dans les regions
proximales est significativement affecte par la presence de regions plus proximales, tandis que dans les regions distales ce temps apparait independant de la
presence de regions plus proximales.
One of us (G. W.) is indebted to the Agricultural Research Council for a Postgraduate
Research Studentship.
REFERENCES
L. G. (1938). Quantitative studies of the factors governing the rate of regeneration
in Tubularia. Biol. Bull. mar. biol. Lab., Woods Hole 74, 155-77.
BEADLE, L. C. & BOOTH, F. A. (1938). Reorganisation of tissue masses in Cordylophora.
J. exp. Biol. 15, 303-26.
BROWNE, E. (1909). The production of new hydranths in Hydra by the insertion of a small
graft. /. exp. Zool. 7, 1-23.
BURNETT, A. L. (1961). The growth process in Hydra. /. exp. Zool. 146, 21-84.
BURNETT, A. L. (1962). In Regeneration, pp. 27-52. Ed. D. Rudnick. New York: Ronald
Press Co.
CAMPBELL, R. D. (1965). Cell proliferation in Hydra: an autoradiographic approach. Science,
N.Y. 148, 1231-2.
CHILD, C. M. (1941). Patterns and Problems of Development. University of Chicago Press.
DAvrosoN, M. E. & BERRILL, N. J. (1948). Regeneration of primordia and developing
hydranths of Tubularia. J. exp. Zool. 107, 465-78.
GOETSCH, W. (1929). Das Regenerationsmaterial und seine experimentelle Beeinflussung.
Arch. EntwMech. Org. Ill, 211-311.
HAM, R. G. & EAKIN, R. E. (1958). Time sequence of certain physiological events during
regeneration in Hydra. /. exp. Zool. 139, 33-54.
HUXLEY, J. S. & DE BEER. G. R. (1934). The Elements of Experimental Embryology. Cambridge University Press.
BARTH,
104
G. WEBSTER & L. W O L P E R T
KING, H. D. (1901). Observations and experiments on regeneration in Hydra viridis. Arch.
EntwMech. Org. 13, 135-78.
LOOMIS, W. F. & LENHOFF, H. M. (1956). Growth and sexual differentiation of hydra in
mass culture. /. exp. Zool. 132, 555-74.
LUND, E. J. (1923). Experimental control of organic polarity by the electric current. III.
Normal and experimental delay in the initiation of polyp formation in Obelia internodes.
/. exp. Zool. 37, 69-87.
MOORE, J. (1952). The induction of regeneration in the hydroid Cordylophora lacustris. J. exp.
Biol. 29, 72-93.
MORGAN, T. H. (1901). Regeneration. New York: Macmillan Co.
MUSCATINE, L. (1961). Symbiosis in marine and freshwater coelanterates. In The Biology of
Hydra. Ed. W. F. Loomis and H. M. Lenhoff. Coral Gables, Florida: University of Miami
Press.
PAPENFUSS, E. F. (1934). Reunition of pieces in Hydra with special reference to the role of the
three layers and to the fate of differentiated parts. Biol. Bull. mar. biol. Lab., Woods Hole
67, 223^3.
PEEBLES, F. (1897). Experimental studies on Hydra. Arch. EntwMech. Org. 5, 794-819.
PEEBLES, F. (1900). Experiments in regeneration and in grafting of Hydrozoa. Arch. EntwMech.
Org. 10, 435-88.
RAND, H. W. (1899). The regulation of graft abnormalities in Hydra. Arch. EntwMech. Org.
9, 161-214.
RAND, H. W., BOVARD, J. F. & MINNICH, D. E. (1926). Localization of formative agencies in
Hydra. Proc. natn. Acad. Sci. U.S.A. 12, 565-70.
ROSE, S. M. (1952). A hierarchy of self-limiting reactions as the basis of cellular differentiation
and growth control. Am. Nat. 86, 337-54.
ROSE, S. M. (1955). Specific inhibition during differentiation. Ann. N.Y. Acad. Sci. 60,
1136-59.
ROSE, S. M. (1957). Cellular interaction during differentiation. Biol. Rev. 32, 351-82.
ROSE, S. M. & ROSE, F. C. (1941). The role of a cut surface in Tubularia regeneration.
Physiol. Zool. 14, 328-43.
RULON, O. & CHILD, C. M. (1937). Observations and experiments on developmental pattern
in Pelmatohydra oligactis. Physiol. Zool. 10, 1-13.
SPANGENBERG, D. B. & EAKIN, R. E. (1961). A study of the variation in the regeneration
capacity of Hydra. J. exp. Zool. 147, 259-70.
SPEMANN, H. (1938). Embryonic Development and Induction. New Haven, Conn.: Yale
University Press.
SPIEGELMAN, S. (1945). Physiological competition as a regulatory mechanism in morphogenesis. Q. Rev. Biol. 20, 121-46.
STEINBERG, M. S. (1954). Studies on the mechanism of physiological dominance in Tubularia.
J. exp. Zool. Ill, 1-26.
STEINBERG, M. S. (1955). Cell movement, rate of regeneration and the axial gradient in
Tubularia. Biol. Bull. mar. biol. Lab., Woods Hole 108, 219-34.
TARDENT, P. (1960). In Developing Cell Systems and Their Control, pp. 21-43. Ed.
D. Rudnick. New York: Ronald Press Co.
TARDENT, P. (1963). Regeneration in the hydrozoa. Biol. Rev. 38, 292-333.
WEBSTER, G. (1964). A study of pattern regulation in hydra. Ph.D. Thesis, University of
London.
WEBSTER, G. & WOLPERT, L. (1966). In preparation.
WEIMER, B. R. (1928). The physiological gradients of hydra. I. Reconstitution and budding
in relation to length of piece and body level in Pelmatohydra oligactis. Physiol. Zool. 1,
183-230.
WEISS, P. (1939). Principles of Development. New York: Henry Holt and Co.
YAO, T. (1945). Studies on the organiser problem in Pelmatohydra oligactis. I. The induction
potency of the implant and the nature of the induced hydranth. /. exp. Biol. 21, 147-50.
{Manuscript received 28 January 1966)