2331
Development 122, 2331-2337 (1996)
Printed in Great Britain © The Company of Biologists Limited 1996
DEV8264
Identified central neurons convey a mitogenic signal from a peripheral target
to the CNS
Thomas S. Becker*, Gerald Bothe, Alyson J. Berliner† and Eduardo R. Macagno‡
Department of Biological Sciences, Sherman Fairchild Center, Columbia University, New York, NY 10027, USA
*Present address: Center for Cancer Research and Department of Biology, Massachusetts Institute of Technology, Cambridge, MA 02139, USA
†Present address: College of Physicians and Surgeons, Columbia University, New York, NY 10032, USA
‡Author for correspondence at present address: 1011 Fairchild Ctr., Columbia University, New York, NY 10027, USA (e-mail: [email protected])
SUMMARY
Regulation of central neurogenesis by a peripheral target
has been previously demonstrated in the ventral nerve cord
of the leech Hirudo medicinalis (Baptista, C. A., Gershon,
T. R. and Macagno, E. R. (1990). Nature 346, 855-858)
Specifically, innervation of the male genitalia by the fifth
and sixth segmental ganglia (the sex ganglia) was shown to
trigger the birth of several hundred central neurons (PIC
neurons) in these ganglia. As reported here, removal of the
target early during induction shows that PIC neurons can
be independently induced in each side of a ganglion, indicating that the inductive signal is both highly localized and
conveyed to each hemiganglion independently. Further,
since recent observations (Becker, T., Berliner, A. J.,
Nitabach, M. N., Gan, W.-B. and Macagno, E. R. (1995).
Development, 121, 359-369) had indicated that efferent projections are probably involved in this phenomenon, we
individually ablated all possible candidates, which led to
the identification of two central neurons that appear to play
significant roles in conveying the inductive signal to the
CNS. Ablation of a single ML neuron reduced cell proliferation in its own hemiganglion by nearly 50%, on the
average. In contrast, proliferation on the opposite side of
the ganglion increased by about 25%, suggesting the possibility of a compensatory response by the remaining contralateral ML neuron. Simultaneous ablation of both ML
neurons in a sex ganglion caused similar reductions in cell
proliferation in each hemiganglion. Deletion of a single AL
neuron produced a weaker (7%) but nonetheless reproducible reduction. Ablation of the other nine central
neurons that might have been involved in PIC neuron
induction had no detectable effect. Both ML and AL
neurons exhibit ipsilateral peripheral projections, and both
arborize mostly in the hemiganglion where they reside.
Thus, we conclude that peripheral regulation of central
neurogenesis is mediated in the leech by inductive signals
conveyed retrogradely to each hemiganglion by specific
central neurons that innervate this target and the hemiganglion they affect.
INTRODUCTION
about 450 neurons, though this number is later reduced to about
400 through cell death (Macagno, 1980; Stewart et al., 1986;
Baptista and Macagno, 1988a). Two of these ganglia (in
midbody segments 5 and 6; MG5 and MG6) innervate the male
and female sex organs and are therefore termed the ‘sex
ganglia’.
In H. medicinalis, this early round of mitosis occurs in the
sex ganglia before embryonic day 8. Significantly later,
between embryonic days 16 and 25 (E16-25), MG5 and MG6
undergo a second period of proliferation which gives rise to an
additional population per sex ganglion of about 350 small
neurons (Baptista et al., 1990; Becker, 1994). If the male
genitalia are surgically removed, proliferation giving rise to
these neurons is prevented in both sex ganglia (Baptista et al.,
1990). If the nerves connecting MG6 to the male organ (the
sex nerves) are ablated, proliferation is prevented in MG6 only,
while it remains normal in MG5. These and other observations
demonstrate that a nerve-mediated interaction between the
Before the discovery of regulatory neuronal death and targetdependent cell survival (reviewed by Oppenheim, 1991;
Hamburger, 1992), target regulation of neuronal proliferation
was proposed as a retrograde mechanism for adjusting central
neuronal numbers to mesodermal peripheral requirements (e.g.
Detwiler, 1936). Although some of the vertebrate neurotrophic
factors implicated in neuronal survival have also been demonstrated to promote proliferation of neural precursors in vitro
(e.g. DiCicco-Bloom et al., 1993; Tischler et al., 1993), retrograde stimulation of neuronal proliferation has so far eluded
conclusive demonstration in vertebrates. In the leech, however,
target regulation of central neurogenesis has been demonstrated for a specific population, the PIC neurons (Baptista et
al., 1990).
During the first half of embryogenesis in hirudinid leeches,
all 21 midbody ganglia (MG) of the ventral nerve cord generate
Key words: Hirudo, mitogenesis, innervation-dependent
neurogenesis, CNS
2332 T. S. Becker and others
male genitalia and MG5 and MG6 induces the birth of these
small neurons, which are hence termed peripherally induced
central, or ‘PIC’ neurons (Baptista et al., 1990).
The induction of PIC neurons can be prevented completely
only if the male genitalia are ablated on or before E13; from
E14 to E16, ablation leads to incomplete induction, and after
E16 PIC neuron birth is independent of the target (Baptista et
al., 1990; Becker and Macagno, 1992a,b). These and subsequent observations led to the following conclusions: (a) that
there is a distinct period (E13-E16) during which PIC neuron
induction can take place; (b) that it is the developmental age
of the CNS, not of the peripheral target, that is critical for PIC
neuron induction; and (c) that the most likely candidates for
conveying a mitogenic signal to the sex ganglia are central
neurons that innervate the male genitalia through the sex
nerves (Becker and Macagno, 1992a; Becker et al., 1995). We
therefore set out to determine which central neurons might be
responsible for conveying this mitogenic signal.
It is known that the nerves connecting the male genitalia and
the sex ganglia are established as early as E12 (Jellies and
Kristan, 1988; Baptista and Macagno, 1988b), and through
these sex nerves, prior to the inductive period, a group of 11
bilateral pairs of identified neurons in each sex ganglion make
contact with and innervate the male genitalia (Becker et al.,
1995). Three of the 11 neurons identified in MG6 innervate the
male genitalia through both sex ganglia (RPE, N, MC; Fig.
A
B
MG5
MATERIALS AND METHODS
Animals and culture conditions
Leech embryos were obtained from our breeding colony of Hirudo
medicinalis maintained at 23°C. Embryos were removed from their
cocoons and kept in sterile artificial spring water (0.5 g/l Instant
Ocean, Aquarium Systems). To reduce parasitic infections of the
embryos and adults, quinine hydrochloride (2 mg/l; Sigma) was added
to the water.
C
MG5
MALE ORGAN
1A), and hence remain connected to the target when the sex
nerves of only one sex ganglion are transected. Since disconnecting only MG6 from the target is sufficient to prevent proliferation in MG6, without affecting MG5 (Becker et al., 1995),
we would predict that none of these three neurons is involved
in PIC neuron induction. The remaining eight candidate
neurons fall into two subgroups, four that project ipsilaterally
(AL, ML, Rz, NUT; Fig. 1C) and four that project contralaterally (DONUT, CPL, PC, LPE; Fig. 1B). Since, as shown
here, induction of PIC neurons can occur independently in each
side of the ganglion (see Results; also Becker and Macagno,
1992a), we would predict that one, but not both, of these
subgroups includes the conveyors of the inductive signal to the
CNS. These predictions were tested in a systematic series of
ablations of all of these neurons, the results of which are
reported in this paper.
MG5
MALE ORGAN
MALE ORGAN
MG6
MG6
MG6
RPE
DONUT
AL
N
CPL
ML
MC
PC
Rz
LPE
NUT
Fig. 1. Neurons that innervate
the male genitalia just prior to
the critical period can be
divided into 3 morphological
classes: those that innervate the
target through nerves of both
sex ganglia, MG5 and MG6
(A), those that project
contralaterally to the target on
the opposite side (B), and those
that project ipsilaterally to the
target (C). Among this last
group, the Retzius (Rz) neurons
arborize bilaterally within the
neuropil, while the other three
types (ML, AL and NUT
neurons) have arbors that are
largely confined to the
ipsilateral half of the neuropil
in the embryo. For simplicity,
the projections of the mirrorimage homologues are not
illustrated.
Leech neurons convey mitogenic signals 2333
Embryogenesis lasts about 30 days under these conditions.
Embryos were staged by days of development after egg laying
(Fernandez and Stent, 1982).
Bromodeoxyuridine labeling
Our techniques for the administration and detection of bromodeoxyuridine (Gratzner, 1982), have been described in detail previously (Becker and Macagno, 1992a).
Single cell ablations
To identify those neurons that innervate the male genitalia, the carbocyanine dye DiI (Molecular Probes Inc., USA; 1% in DMSO) was
injected, at E11, into the primordia of the male genitalia using a
micropipette connected to a 20 ml syringe. This usually resulted in the
backfilling of a subset of the previously mapped male organ-innervating neurons (see Fig. 2A) by E12. Despite heavy labeling of the male
organ and lighter labeling of varying subsets of the innervating neurons,
injection of the dye did not interfere with induction of PIC neurons,
which was normal in otherwise untreated animals (data not shown).
At E12, the experimental animals were anesthetized in 8% ethanol,
placed ventral-side-up in a groove cut into a silicone-covered slide and
pressed down with a small strip of coverslip held in place with fine
tungsten pins. With fine iredectomy scissors, a slit was made over
MG6, exposing the ventral aspect of the ganglion. Using epifluorescence optics in an upright compound microscope, single DiI-labeled
neuronal cell bodies were visualized and then impaled with a microelectrode filled with a 1% solution of carboxyfluorescein (Sigma) in
0.2 M NaOH. Application of several negative current pulses (0.5 nA;
duration 0.5 seconds) was sufficient to reveal the branching pattern of
the injected cell, thus allowing confirmation of the filled cell’s identity
(see Fig. 2B). Injection was then continued for 1-2 minutes in the dark
until cell bodies burst. Embryos were subsequently returned to artificial spring water and allowed to develop normally.
This method of cell ablation was employed instead of the more conventional photoablation method because we had earlier found that
irradiation of the sex ganglia with light of short wavelength for over
3-5 minutes sometimes interfered with subsequent PIC neurogenesis
(data not shown). In addition, this method does not appear to have
any deleterious effects on neighboring, non-injected neurons. To
examine this issue, we ablated one of a pair of Retzius (Rz) neurons,
which are easily identified by their large size and central position in
each ganglion, and are serotonergic (Lent, 1973). A day after one Rz
cell was injected with dye as described above and presumably killed,
staining with antibodies against serotonin confirmed, in ten out of ten
cases, the ablation of the injected cell. The other Rz neuron, which
was directly adjacent to and electrically coupled to its contralateral
homologue, appeared to be unaffected (data not shown).
Cell counts and statistics
In these experiments, cell ablations were carried out only in MG6
because it was technically easier to visualize dye-filled cells in MG6
in live embryos. BrdU-positive cell nuclei in MG5 and MG6 were
generally counted separately in each hemiganglion using a manual
counter and compound microscope. Due to the innately high animalto-animal variation in the numbers of induced mitoses, even among
unoperated animals (Baptista et al., 1990; Becker and Macagno,
1992a; Becker, 1994), we compared operated and unoperated hemiganglia in MG6 to control hemiganglia in MG5 in the same animal,
or total labeled cell counts in MG6 to those in MG5. Statistical significance of differences between means was evaluated using Student’s
t-test. All values presented are means ± standard deviations.
RESULTS
PIC neurons are induced in hemiganglionic domains
Induction of PIC neurons is normally localized to the two sex
ganglia, but can occur in just one sex ganglion under experimental conditions in which a single ganglion remains
connected to the male organ (Baptista et al., 1990, Becker et
al., 1995). Induction in only one sex ganglion is also observed
in many cases if the target is ablated during the inductive
period (Becker and Macagno, 1992a). For example, after
ablation of the male genitalia at E14, at the beginning of the
inductive period, induction occurred in a single sex ganglion
in 21 out of the 26 animals tested (9 cases of MG5 alone and
12 cases of MG6 alone; Becker and Macagno, 1992a).
Further examination of these 26 experimental animals then
revealed that in several of them induction of PIC neurons was
restricted to domains smaller than a single ganglion (Fig. 2C).
Cell proliferation was confined almost exclusively to a hemiganglion in seven cases (in 5 animals in MG6 alone, in one
animal in both MG5 and MG6). The numbers of dividing cells
in the 7 induced hemiganglia ranged from 39 to 129 cells; in
each case, there were 12 or fewer dividing cells in the other
half of the ganglion, a level of proliferation similar to background values measured in non-sex ganglia.
Given our previous finding that central neurons were likely
to be involved in conveying the peripheral mitogenic signal
into the CNS, one interpretation of this result is that the
mitogenic signal is transmitted within a ganglion by a central
neuron whose arbor is confined largely to one side of the
Fig. 2. (A,B) Micrographs of a live E12 embryo showing some DiI
backfilled neuronal somata in (A) and a single carboxyfluoresceinfilled N neuron (A and B). A was photographed through rhodamine
and fluorescein filters, B through fluorescein filters. (C) Proliferation
of PIC neurons is sometimes observed in hemiganglionic domains
after incomplete induction. Micrograph of a midbody ganglion
(MG6) from an E21 specimen after ablation of the male genitalia at
E14 (the middle of the inductive period). (D) Killing of an ML
neuron (the arrowhead indicates the debris of the soma) at E12
resulted in a much decreased level of cell proliferation in the
hemiganglion on the same side, and an increase on the other side. In
all panels, anterior is up. Bars, 50 µm.
2334 T. S. Becker and others
neuropil. To explore this possibility, we systematically ablated,
individually and in pairs, all central neurons that innervate the
male genitalia during the critical period for PIC neuron
induction.
Table 1. Single ML ablations in MG6
Numbers of cells labeled with BrdU
Ganglion MG5
Ganglion MG6
Specimen
L
R
R−L
R+L
O
N
N−O
Ablating a single ML neuron significantly reduces
PIC neuron induction in its own hemiganglion, but
increases it in the opposite hemiganglion
Single, identified neurons were ablated at E12 (Fig 2A,B; see
Materials and Methods). This stage precedes the beginning of
the critical period of interaction between the sex ganglia and
the male organ (Becker and Macagno, 1992a), but is after projections of most of these central neurons have reached their
target (Becker et al., 1995). Effects of the ablations on induced
cell proliferation were assayed by counting the number of ganglionic nuclei that incorporated BrdU (Gratzner, 1982) at E20,
a time in embryogenesis when PIC neurons are normally being
generated at a relatively high rate (Baptista et al., 1990; Becker,
1994). Using this assay, unoperated embryos averaged 91±28
labeled cells in MG5 and 89±22 in MG6, or an average of 45
labeled cells per hemiganglion (n=8; Becker, 1994). In unoperated animals, there is no statistically significant difference
between the two sex ganglia in the number of labeled cells.
Ablation of a single ML neuron (see diagram, Fig. 1, right)
consistently reduced cell proliferation in the ganglion containing the ablated cell. In the 21 cases in which a single ML
neuron was successfully ablated in MG6 (Table 1), we detected
78±24 (mean ± s.d.) labeled cells in this ganglion, a significant
(P<0.01; one-tailed t-test) decrease of about 13% relative to
unoperated controls (see above). However, the distribution of
labeled cells in MG6 was distinctly asymmetric: there were
24±9 labeled cells in the experimental (operated) hemiganglion
as opposed to 54±18 in the (non-operated) hemiganglion on
the opposite side (Table 1). Compared to unoperated controls,
the operated side averaged 20 fewer labeled cells, equivalent
to a 45% reduction (P<0.0001). In contrast, the non-operated
hemiganglion averaged 10 more labeled cells than controls, or
a 23% increase (P<0.002).
The other sex ganglion, MG5, was not affected by the
ablation of a single ML neuron in MG6 (see Table 1). The
average number of labeled cells, 87±27 per ganglion, was not
significantly different from unoperated control values.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
56
86
35
35
38
41
39
30
28
35
29
35
37
45
28
42
35
50
75
53
43
70
71
32
32
52
50
47
24
30
38
37
46
42
46
33
44
33
42
69
54
47
14
−15
−3
−3
14
9
8
−6
2
3
8
11
5
1
5
2
−2
−8
−6
1
4
126
157
67
67
90
91
86
54
58
73
66
81
79
91
61
86
68
92
144
107
90
38
23
19
34
20
25
16
20
4
19
26
21
27
19
17
28
27
12
46
31
28
84
91
55
55
43
45
36
35
25
50
43
50
66
47
32
54
53
79
81
60
44
46
68
36
21
23
20
20
15
21
31
17
29
39
28
15
26
26
67
35
29
16
Average
s.d.
42.6
14.8
44.7
13.2
23.8
9.1
53.7
17.8
29.9
15.0
Ablating both ML neurons in a ganglion
substantially decreases, but does not abolish,
induced cell proliferation
Although ablating an ML neuron reduces but does not abolish
induced cell proliferation in its own hemiganglion, it is
nonetheless possible that ML neurons are the sole conveyors
of the mitogenic signal if the remaining homologue compensates for the ablation. In fact, the increase in the number of
labeled cells above control values on the non-operated side of
MG6 could be due to a compensatory response by the
remaining ML neuron. It is indeed the case that leech neurons
have the capacity to increase their arbors in response to
homologue ablation, even spreading substantively to the
opposite hemiganglion (see Wolszon, 1995, for a recent
review). Ablating both ML neurons in the same ganglion
would resolve this question.
In a series of double ablation experiments, we were successful in identifying and deleting both ML neurons in MG6
in four embryos (Table 2). In each of these four embryos, large
bilateral decreases in labeled cells were observed in MG6
(Table 2), but induced cell proliferation was not abolished. The
average number of labeled cells in the operated MG6 was
46±9. Relative to unoperated controls, this is a decrease of
nearly 48%. Approximately normal levels of cell proliferation
(within one standard deviation) were measured in MG5 (Table
2).
For comparison, in addition to ablating one ML neuron, we
determined the background level of cell proliferation present
in MG6 by cutting both of the sex nerves connecting this
ganglion to the male organ before the period of induction.
Measured also at E20, the average number of labeled cells in
MG6 was 12±4 (n=5; Table 3), a much smaller number than
was measured after the ablation of both ML neurons. It is
apparent that the induction of PIC neurons by the male
genitalia does not depend upon the ML neurons alone, and that
other cells must be involved as well.
2.1
7.4
87.3
27.1
N+O
122
114
74
89
63
70
52
55
29
69
69
71
93
66
49
82
80
91
127
91
72
77.5
24.0
Statistics: one-tailed t-test, MG5L:MG5R, 0.105; MG6N:MG6O, 6.76E09; MG5:MG6, 0.007.
L, left; R, right; O, operated; N, non-operated; R−L and N−O, difference
between hemiganglia; R+L, N+O, total number of labeled cells per ganglion.
Table 2. Double ML ablations in MG6
Numbers of cells labeled with BrdU
Specimen
MG5
MG6
1
2
3
4
120
113
68
112
56
48
34
46
MG5−MG6
64
65
34
66
Average
s.d.
103.3
23.8
46.0
9.1
57.3
15.5
Statistics: one-tailed t-test, MG5:MG6, 0.0026.
Leech neurons convey mitogenic signals 2335
Table 3. MG6 sex nerve transections
Numbers of cells labeled with BrdU
Ganglion MG5
Specimen
L
R
1
2
3
4
5
42
61
43
49
54
47
51
50
63
53
Average
s.d.
49.8
7.9
52.8
6.1
R−L
5
−10
7
14
−1
3.0
9.0
Ganglion MG6
R+L
O
N
89
112
93
112
107
8
5
4
4
6
11
5
6
5
5
102.6
10.9
5.4
1.7
6.4
2.6
N−O
3
0
2
1
−1
1.0
1.6
N+O
19
10
10
9
11
11.8
4.1
Statistics: one-tailed t-test, MG5L:MG5R, 0.2494; MG6N:MG6O, 0.1151;
MG5:MG6, 0.0001.
Nomenclature: same as Table 1.
Table 4. Single AL ablations in MG6
Numbers of cells labeled with BrdU
Ganglion MG5
Specimen
L
R
1
2
3
4
5
56
33
27
26
19
56
32
47
37
22
Average
s.d.
32.2
14.2
38.8
13.2
Ganglion MG6
R−L
R+L
0
−1
20
11
3
112
65
74
63
41
6.6
8.8
71.0
25.9
O
N
49
28
22
25
18
61
37
28
30
22
28.4
12.1
35.6
15.2
N−O
N+O
12
9
6
5
4
110
65
50
55
40
7.2
3.3
64.0
27.2
Statistics: one-tailed t-test, MG5L:MG5R, 0.0853; MG6N:MG6O, 0.0040;
MG5:MG6, 0.0963.
Nomenclature: same as Table 1.
A small but significant effect follows the ablation of
an AL neuron, but not the ablation of the other
neurons that innervate the male genitalia
A significant but small reduction in induced cell proliferation
also followed the killing of a single AL neuron (see diagram,
Fig. 1). In five successful ablations of an AL neuron, we
measured a reduction of only about 7% (P<0.004) in cell proliferation in the operated hemiganglion relative to the unoperated side (Table 4). No ablations of both AL neurons in the
same ganglion were accomplished; since these neurons
innervate the target later than the others (Becker et al., 1995),
we were almost never able to identify both in the same
ganglion.
For the other nine types of neurons that innervate the male
genitalia (Fig. 1), neither single ablations nor simultaneous
ablations of both homologues in the same ganglion had statistically significant effects on cell proliferation (data not shown).
DISCUSSION
The principal finding reported here is the identification of two
leech central neurons that, apparently through their innervation
of the male genitalia, are responsible to a very significant
degree, if not entirely (see discussion of compensatory effects
below), for the induction of a population of central neurons
specific to the sex ganglia. While the regulation of the size of
central neuronal populations by peripheral factors has been
previously proposed (see Introduction), our observations are
the first to demonstrate that signals leading to the birth of
neurons can be conveyed retrogradely along axonal processes,
and are also the first to show conclusively that single identified neurons, rather than groups or classes of neurons, can
convey specific developmental signals in a retrograde fashion
from their target organ to other cells in the CNS, signals that
in this case locally control the generation of neurons in the area
of the CNS where these cells reside.
Are the ML and AL neurons the only ones
responsible for PIC neuron induction?
We measured a significant reduction of induced cell proliferation in a hemiganglion as a result of the ablation of the local
ML neuron, and also, but to a much lesser degree, following
the ablation of an AL neuron. Since none of the single or even
paired ablations of the other identified neurons yielded reliable
results, it seems reasonable to conclude that only the ML and
AL neurons have pivotal roles in this phenomenon among
those that innervate the male genitalia during the critical period
for PIC neuron induction. However, the reduction in cell proliferation was never complete when either of these neurons was
ablated individually. Nor is a simple linear summation of their
individual effects in their own hemiganglia (reductions of 45%
and 7% for ML and AL neurons, respectively) sufficient to
account for all the induction. A possible explanation for these
apparently contradictory observations is that the ML and AL
neurons act together to regulate the induction of the PIC
neurons, and can compensate for each other’s absence.
However, it is also possible that incomplete abolition of PIC
proliferation stemmed from the induction being underway
when we ablated the ML neuron; we relied on dye travelling
from the target back to the CNS to identify these neurons and
hence they must have been connected to the target already at
the time of ablation.
Our results from double cell kills do support the idea that
there is an interaction between the pair of ML neurons that
leads to a compensatory response by the non-ablated cell. It is
well documented that when a leech neuron is ablated in the
early embryo, its homologues in the same and/or other
segments expand their central and peripheral arbors to take
over the vacated territory and targets (e.g., Macagno et al.,
1990; Wolszon, 1995). We would therefore predict that the
remaining ML neuron in MG6 would expand its central or
peripheral arbor in response to the ablation of its homologue,
and as a result partially compensate for its absence. The fact
that we do observe an increase in induced cell proliferation in
the non-operated hemiganglion may be a consequence of the
hypertrophy of the local ML neuron, as proposed above in the
Results. For instance, the remaining ML neuron may expand
its field in the target and thus have access to more of the
putative inducing activity. We intend to test these ideas in
future studies by initially comparing the morphology of this
neuron under normal and experimental conditions.
The AL neurons are also excellent candidates for this compensatory response, as they normally have a small role in this
process anyway. The best way to test whether the ML and AL
neurons together are sufficient for PIC neuron induction would
be to ablate all four of these neurons in an individual ganglion.
However, this was not feasible for practical reasons. The axon
of the AL neuron reaches the male genitalia later than that of
2336 T. S. Becker and others
the ML neuron. It is however, occasionally backfilled by DiI
injection into the target, even at the beginning of the critical
period, which is why it could be ablated in a few of the experiments we performed. The small chance of backfilling AL and
ML neurons simultaneously made it impossible to ablate both
cell types in the same ganglion.
Whether the AL neuron is the only neuron that can compensate for the ablation of the ML neuron and mediate
induction of PIC neurogenesis is not clear, but it is noteworthy that the ML and AL neurons have similar branching and
projection patterns that are consistent with predictions based
on the pattern of proliferation. As shown in a previous paper
(Becker et al., 1995), the ML neuron branches in the ipsilateral neuropil and projects ipsilaterally to the male genitalia.
The AL neuron, first described by Zipser (1980), also branches
and projects ipsilaterally. If additional neurons were involved,
one might therefore expect them to have projection patterns
similar to these two neurons. In backfills performed on adult
leeches, there are additional ipsilateral neurons innervating the
male genitalia (Becker et al., 1995), but these neurons do not
reach the target before E15, close to the end of the critical
period, and are therefore not likely to be important for PIC
neuron induction. However, until we have means other than
dye backfilling from the target to identify possible candidates,
the possibility that these late arrivals can also partially take
over the functions of the ML neurons will remain untested.
What might be the mechanisms of PIC neuron
induction?
Although our observations delimit the range of possible mechanisms for the induction of PIC neurogenesis, they leave some
fundamental questions about the induction of PIC neurons
unanswered. For example, what component of the target is
required to generate the inductive signal? What is the inductive
signal? How is it conveyed to the sex ganglia? Is the signal
itself mitogenic or does it affect cell division indirectly? How
does this signal trigger the proliferation of the precursors of the
PIC neurons?
Two models consistent with our observations can be readily
proposed. One model would have the inductive signal itself as
a mitogen released by specific cells in the target. The axon
terminals of the ML and AL neurons would then express a
receptor for this mitogen on their surfaces and, upon reaching
and innervating the target, bind the mitogen, internalize it and
transport it retrogradely to the sex ganglia. There the ML and
AL neurons would either re-expose the mitogen on their
surfaces or release it near the susceptible precursors of the PIC
cells.
A number of biomolecules, shown to have mitogenic effects
on specific populations of vertebrate cells, could play such a
role in the leech. Among well-known growth factors, for
example, basic fibroblast growth factor (bFGF) has been
shown to be a mitogen for embryonic rat spinal cord neuroblasts (Ray and Gage, 1994), while both bFGF and epidermal
growth factor (EGF) have been found to be mitogenic for
neural cell lines derived from mouse cerebellum (Kitchens et
al., 1994; see also Kilpatrick et al., 1995 for review). Nerve
growth factor (NGF) appears to be a potent inducer of proliferation of both immature (Lillien and Claude, 1985) and adult
(Tischler et al., 1993) rat chromaffin cells, as well as some
subclones of PC12 cells (Burstein and Greene, 1982). Hepato-
cyte growth factor (HGF) has been reported to be a mitogen
for Schwann cells (Krasnoselsky et al., 1994).
Among mitogenic growth factors, NGF and insulin are
thought to be internalized once they are bound to appropriate
membrane receptors on axonal processes. In the case of NGF,
the NGF-receptor complex is known to be transported to other
parts of the cell before being cycled back to the cell surface.
The leech homologue of NGF, or a similar molecule, would fit
with the model we proposed above. Receptors for some of
these important growth factors are beginning to be characterized in the leech (Nitabach and Macagno, 1995), and their
possible involvement in PIC cell induction will be tested in the
future.
A second model for PIC neuron induction would have the
inductive signal not itself a mitogen, but rather a signal that
affects the fate of the developing ML and AL neurons. In at
least one case of a central neuron in the leech, the Retzius cell,
cell fate is determined by contact with the male genitalia
(reviewed by French and Kristan, 1995). A molecular factor
detected by the axon terminals at the target, for example, might
cause a cytoplasmic messenger to travel retrogradely to the
nucleus and there cause novel gene expression, such as the
expression of a mitogen that affects the proliferation of PIC
cell precursors. Such a mechanism for affecting gene
expression retrogradely has been reported, for example, in
Aplysia neurons (Ambron et al., 1992; Schmied et al., 1993).
While in the examples we have discussed above the effect
is thought to be directly mitogenic to the PIC cell precursors,
it is worth noting that some factors have been shown to
enhance cell proliferation indirectly. The neurotrophin NT-3,
for example, stimulates embryonic rat SCG neuroblast proliferation by enhancing precursor survival rather than by
affecting mitosis directly (DiCicco-Bloom et al., 1993). Since
NT-3 has also been reported to be a direct mitogen for cultured
neural crest cells (Kalcheim et al., 1992), it is clear that factors
can play different roles for different cell types (see Chao, 1992,
for a recent review of this subject). Therefore, whatever
molecular signals are involved in the induction of PIC cells in
the leech, their actual role will have to be determined experimentally.
The molecules discussed above are all soluble diffusible
factors, which presumably can act at some distance from their
release sites. Their range of action, however, depends upon the
distribution of components of the extracellular matrix that can
immobilize them. A different type of mitogenic factor, especially worth mentioning in light of our results, is the
membrane-bound neuronal mitogen for Schwann cells and
oligodendrocytes (Wood and Bunge, 1975; Salzer et al.,
1980a,b; Mason et al., 1990). A 50×103 Mr membrane-bound
protein with this activity has recently been purified from fetal
bovine brains (Nordlund et al., 1992) but has yet to be cloned.
Such a mitogen requires cell-cell contact in order to exert its
effects. The fact that PIC neuron induction is so highly
localized supports the idea that the mitogenic signal might be
anchored on or close to the ML and AL neurons.
In conclusion, we have shown that developmental signals
can be conveyed retrogradely by identified neurons. In this
case, the developmental signal results in the selective birth of
segment-specific neurons termed PIC cells. This strategy for
target regulation of central neuron number is very different
from the well described mechanism of selective cell survival
Leech neurons convey mitogenic signals 2337
in response to target availability found in vertebrates and may
have evolved as a parsimonious alternative to increasing
central neuronal numbers in response to growing demands of
the periphery.
We thank Kavitha S. Becker for the artwork, and Nicholas Necles
for photographic work. We also thank Laura Wolszon and Darcy
Kelley for their critical readings of the manuscript. T. S. B. would like
to thank Ludger Tüschen and Mark Bertelmann for help and encouragement during the course of these studies. G. B. was supported by a
postdoctoral fellowship of the Deutsche Forschungsgemeinschaft.
Supported by an NIH grant to E. R. M.
REFERENCES
Ambron, R. T., Schmied, R., Huang, C. C., and Smedman, M. (1992). A
signal sequence mediates the retrograde transport of proteins from the axon
periphery to the cell body and then into the nucleus. J. Neurosci. 12, 28132818.
Baptista, C. A. and Macagno, E. R. (1988a). The role of the sexual organs in
the generation of postembryonic neurons in the leech Hirudo medicinalis. J.
Neurobiol. 19, 707-726.
Baptista, C. A. and Macagno, E. R. (1988b). Modulation of the pattern of
axonal projections of a leech motor neuron by ablation or transplantation of
its target. Neuron 1, 949-962.
Baptista, C. A., Gershon, T. R. and Macagno, E. R. (1990). Peripheral organs
control central neurogenesis in the leech. Nature 346, 855-858
Becker, T. and Macagno, E. R. (1992a). CNS control of a critical period for
peripheral induction of central neurons in the leech. Development 116, 427434.
Becker, T. and Macagno, E. R. (1992b). Central neurogenesis in the leech. In:
Determinants of Neuronal Identity. (ed. M Shankland and E. R. Macagno),
pp. 79-95. New York: Academic Press.
Becker, T. (1994). Induction of segment-specific neurons in the leech, Hirudo
medicinalis. PhD thesis, Columbia University, New York: New York.
Becker, T., Berliner, A. J., Nitabach, M. N., Gan, W.-B and Macagno, E. R.
(1995). Target-induced neurogenesis in the leech CNS involves efferent
projections to the target. Development 121, 359-369.
Burstein, D. E. and Greene, L. A. (1982). Nerve growth factor has both
mitogenic and anti-mitogenic activity. Dev. Biol. 94, 477-482.
Chao, M. V. (1992). Neurotrophin receptors: a window into neuronal
differentiation. Neuron 9, 583-593.
Detwiler, S. (1936). Neuroembryology New York: Macmillan.
DiCicco-Bloom, E., Friedman, W. J. and Black, I. B. (1993). NT-3
stimulates sympathetic neuroblast proliferation by promoting precursor
survival. Neuron 11, 1101-1111.
Fernandez, J. and Stent, G. S. (1982). Embryonic development of the
hirudinid leech Hirudo medicinalis: Structure, development and
segmentation of the germinal plate. J. Embryol. Exp. Morph. 72, 71-96.
French, K. A. and Kristan, W. B. Jr. (1995). Cell-cell-interactions that
modulate neuronal development in the leech. J. Neurobiol. 25, 640-651.
Gratzner, H. G. (1982). Monoclonal antibody to 5-bromo and 5iododeoxyuridine: a new reagent for detection of DNA replication. Science
218, 474-475.
Hamburger, V. (1992). History of the discovery of neuronal death in embryos.
J. Neurobiol. 23, 1116-1123.
Jellies, J. and Kristan, W. B. Jr. (1988). An identified cell is required for the
formation of a major nerve during embryogenesis in the leech. J. Neurobiol.
19, 153-165.
Kalcheim, C., Carmeli, C. and Rosenthal, A. (1992). Neurotrophin 3 is a
mitogen for cultured neural crest cells. Proc. Natl. Acad. Sci. USA 89, 16611665.
Kilpatrick, T. J., Richards, L. J. and Bartlett, P. F. (1995). The regulation of
neural precursor cells within the mammalian brain. Molec. Cell. Neursci. 6,
2-15.
Kitchens, D. L., Snyder, E. Y. and Gottlieb, D. I. (1994). FGF and EGF are
mitogens for immortalized neural progenitors. J. Neurobiol. 25, 797-807.
Krasnoselsky, A., Massay, M. J., DeFrances, M. C., Michalopoulos, G.,
Zarnegar, R. and Ratner, N. (1994). Hepatocyte growth factor is a mitogen
for Schwann cells and is present in neurofibromas. J. Neurosci. 14, 72847290.
Lent, C. M. (1973). Retzius cells: Neuronal effectors controlling mucus release
by the leech. Science 179, 693-696.
Lillien, L. E. and Claude, P. (1985). Nerve growth factor is a mitogen for
cultured chromaffin cells. Nature 317, 632-634.
Macagno, E. R. (1980). Number and distribution of neurons in leech segmental
ganglia. J. comp. Neurol. 190, 283-302.
Macagno, E. R., Gao, W.-Q., Baptista, C. A. and Passani, M. B. (1990).
Competition or inhibition? Developmental strategies in the establishment of
peripheral projections by leech neurons. J. Neurobiol., 21, 107-119.
Mason, P. W., Chen, S. J. and De Vries, G. H. (1990). Evidence for the
colocalization of the axonal mitogen for Schwann cells and
oligodendrocytes. J. Neurosci. Res. 26, 296-300.
Nordlund, M., Hong, D., Fei, X. and Ratner, N. (1992). Schwann cells and
cells in the oligodendrocyte lineage proliferate in response to a 50,000 dalton
membrane-associated mitogen present in developing brain. Glia 5,182-192.
Nitabach, M. N. and Macagno, E. R. (1995). Cell and tissue specific
expression of putative protein kinase mRNAs in the embryonic leech, Hirudo
medicinalis. Cell Tissue Res., 280, 479-489.
Oppenheim, R. W. (1991). Cell death during the development of the nervous
system. Ann. Rev. Neurosci. 14, 543-501
Ray, J. and Gage, F. H. (1994). Spinal cord neuroblasts proliferate in response
to basic fibroblast growth factor. J. Neurosci. 14, 3548-3564.
Salzer, J. L.,Williams, A. K., Glaser, L. and Bunge, R. P. (1980a). Studies of
Schwann cell proliferation. II. Characterization of the stimulation and
specificity of the responses to a neurite membrane fraction. J. Cell Biol. 84,
753-766.
Salzer, J. L., Bunge, R. P. and Glaser, L. (1980b). Studies of Schwann cell
proliferation. III. Evidence for the surface localization of of the neuritic
mitogen. J. Cell Biol. 84, 767-778.
Schmied, R., Huang, C. C., Zhang, X. P., Ambron, D. A. and Ambron, R. T.
(1993). Endogenous axoplasmic proteins and proteins containing nuclear
localization signal sequences use the retrograde axonal transport/nuclear
import pathway in Aplysia neurons. J. Neurosci. 13, 4064-4071.
Stewart, R. R., Spergel, D. and Macagno, E. R. (1986). Segmental
differentiation in the leech nervous system: The genesis of cell number in the
segmental ganglia of Haemopis marmorata. J. comp. Neurol. 253, 253-259.
Tischler, A. S., Riseberg, J. C., Hardenbrook, M. A. and Cherington, V.
(1993). Nerve growth factor is a potent inducer of proliferation and neuronal
differentiation for adult rat chromaffin cells in vitro. J. Neurosci. 13, 15331542.
Wolszon, L. R. (1995). Cell-cell interactions define the innervation patterns of
central leech neurons during development. J. Neurobiol. 27, 335-352.
Wood, P. M. and Bunge, R. P. (1975). Evidence that sensory axons are
mitogenic for Schwann cells. Nature 256, 662-664.
Zipser, B. (1980). Horseradish peroxidase nerve backfilling in the leech. Brain
Res. 182, 441-445.
(Accepted 2 May 1996)