Current Biology, Vol. 12, 1401–1404, August 20, 2002, 2002 Elsevier Science Ltd. All rights reserved. PII S0960-9822(02)01072-2
Neurotrophin-4: A Survival Factor
for Adult Sensory Neurons
although 95% of D-hair receptors are missing in mature
receptors encountered at 2–3 weeks or even later, be-
(Figure 1). Because most NT-3 homozygous knockout
mice die at birth, we used NT-3 ?/? mice for functional
analysis of the role of NT-3 in the survival of D-hair
receptors. Whereas D-hair receptors were still present
in normal proportions in NT-4 ?/? mice until 7 weeks
after birth, in 2–3-week-old NT-3 ?/? mice, D-hair recep-
tors were already reduced by 40% compared to controls
(Figure 1). The magnitude of this loss at 2–3 weeks was
very similar to that previously described in adult
NT-3 ?/? mice .
factor for D-hair receptors before the second post-natal
week and that NT-4 alone becomes necessary for sur-
vival of D-hair receptors beginning around the 7thweek
of life. We excluded the possibility of D-hair receptors
being present but having lost mechanosensitivity by
electrophysiologically isolating fibers with an axon in
the skin (electrical stimulation of the nerve), and we
found that all fibers so isolated in older NT-4 mutants
were nociceptive (our unpublished data). To test when
an anatomical loss of presumptive D-hair receptor neu-
rons takes place, we performed a quantitative histologi-
We found that compared to controls, nerves taken from
NT-4 knockout mice at 5 weeks had slightly fewer axons
(Figure 1B). This reduction in the total number of myelin-
ated axons was apparent even though the number of
D-hair receptors measured physiologically was un-
changed atthis agecompared tocontrols. Thus,in spite
of the normal proportion of functional D-hair receptors
at 7 weeks (Figure 1A), myelinated neurons were none-
theless reduced in number.
The early reduction in myelinated axons may reflect
a survival role for NT-4 earlier in development before
myelinated neurons are specified to become (or differ-
entiate into) functional subtypes. However, two recent
studies indicate that NT-4 does not act as a survival
ber of neurons present in the dorsal root ganglion (DRG)
was found to be unchanged or even increased at birth
[11, 12]. The fact that we observed a small drop in the
number of myelinated axons at 5 weeks might indicate
that NT-4 influences the number of sensory axons that
become myelinated. As reported previously in the ma-
ture NT-4 knockout mouse , all other types of myelin-
ated mechanoreceptors were present in normal num-
bers at 7 weeks of age (Table 1).
A similar small loss of myelinated axons was also
observed in nerves from 2–3-week-old NT-3 ?/? mice
(Figure 1B); some of this axon loss might have included
the previously observed early physiological loss of
D-hair receptors. However, it is also possible that some
of the early reduction is a result of an early role for NT-3
in the embryonic or early postnatal development of the
DRG . There is indeed substantial evidence that the
Cheryl L. Stucky,1,2Jung-Bum Shin,1
and Gary R. Lewin3
Growth Factor and Regeneration Group
Max Delbru ¨ck Institute for Molecular Medicine
The nerve growth factor (NGF) family of neurotrophins
provides a substantial part of the normal trophic sup-
port for sensory neurons during development . Al-
though these neurotrophins, which include Brain-
Derived Neurotrophic Factor (BDNF), Neurotrophin-3
(NT-3), and Neurotrophin-4 (NT-4), continue to be ex-
pressed into adulthood [2, 3], there is little evidence
we have examined the age-dependent neurotrophic
requirements of a specialized type of mechanorecep-
tive neuron, called a D-hair receptor, in the dorsal root
ganglion (DRG). Studies using knockout mice have
demonstrated that the survival of D-hair receptors is
dependent upon both NT-3 and NT-4 [5–7]. Here, we
show that the time period when D-hair receptors re-
quire these two neurotrophins is different. Survival of
D-hair receptors depends on NT-3 early in postnatal
development and NT-4 later in the mature animal. The
age-dependent loss of D-hair neurons in older NT-4
knockout mice was accompanied by a large reduction
(78%) in neurons positive for the NT-4 receptor (trkB)
together with neuronal apoptosis in the DRG. This is
the first evidence that sensory neurons have a physio-
logical requirement for a single neurotrophin for their
continued survival in the adult.
Results and Discussion
Molecular markers that unambiguously identify distinct
functional classes of sensory neurons are unavailable
for nearly all physiologically characterized sensory neu-
rons [8, 9]. Thus, the existence of one functional class
of sensory neuron, the D-hair mechanoreceptor, can
only be shown by electrophysiological methods .
on access to both NT-3 and NT-4 for survival. In order to
determine the time frame during which D-hair afferents
depend on these two neurotrophins, we used the in
vitro skin nerve preparation from mouse to obtain a
quantitative estimate of the number of D-hair receptors
(2–3 weeks) or mature (5–24 weeks) mice deficient in
either NT-3 or NT-4. To our surprise, we found that
1These authors contributed equally to this work.
2Present Address: Department of Cell Biology, Neurobiology and
Anatomy, Medical College of Wisconsin, 8701 Watertown Plank
Road, Milwaukee, Wisconsin 53226-0509.
Table 1. Proportions of Physiologically Characterized A? Fibers
in NT-3 and NT-4 Mutants
Percent SAMPercent RAM
WT 2–3 weeks
NT-3 ?/? 2–3 weeks
NT-4 ?/? 2–3 weeks
NT-4 ?/? 5–7 weeks
48% (n ? 11/23)
45% (n ? 10/22)
50% (n ? 10/20)
52% (n ? 12/23)
55% (n ? 12/22)
50% (n ? 10/20)
Among A? fibers, the proportion of slowly and rapidly adapting
mechanoreceptors (SAM and RAM) was also measured. The table
summarizes this data from all the groups measured. No significant
differences were observed in the proportion of RAM and SAM in
any of the genotypes examined at the various ages. Interestingly,
this data indicates that a substantial loss of SAM had not yet taken
place in young (2–3-week-old) NT-3 ?/? mice, as has been de-
scribed in the adult . This was perhaps not so surprising because
the morphological equivalent of SAM, the Merkel cell neurite com-
when the mice were 2 weeks of age .
Figure 1B). We believe this later axon loss reflects the
selective death of D-hair receptor neurons through pro-
grammed cell death.
During this time we directly demonstrated the pres-
ence of TUNEL-positive neurons in DRGs taken from
8-week-old NT-4 knockout mice (Figure 2E). The total
number of TUNEL-positive cells found was small (5 cells
found in 770 sections taken from 4 mice). However,
the low number of neurons undergoing apoptosis was
expected. Thus, the axons lost between 7 and 10 weeks
(approximately 50 axons in total) should represent a
minimum estimate of the total number of D-hair recep-
tors normally present in the saphenous nerve. Approxi-
mately 30% of DRG neurons have myelinated axons
; thus, a 10% loss in myelinated axons from the
saphenous nerve would in the DRG be observed as at
most a 3% loss in the total DRG neuron number over a
prolonged period of time in the absence of NT-4 (3–4
To confirm that the loss of functionally defined recep-
tors and their axons and cell bodies reflects a true loss
of DRG neurons, we took a different approach. We per-
formed nonradioactive in situ hybridization experiments
with a probe that recognizes mRNA coding for the trkB
tyrosine kinase receptor, the cognate receptor for NT-4
. Since NT-4-dependent D-hair receptors must ex-
press this receptor ,we hypothesized that substantial
age-dependent loss of trkB-positive neurons must take
place during the time that D-hair mechanoreceptors are
lost. Indeed, when we compared the proportion of trkB-
positive neurons in DRG sections of mature wild-type
mice to those in sections of NT-4 knockout mice, we
observed a substantial reduction (78%) in the number
of trkB-positive neurons. More interestingly, when we
looked at the number of trkB-positive neurons in 5-
week-old NT-4 knockout mice and compared it to that
from mature wild-type mice, there was no discernable
difference (Figure 2; but see ). About 7% of the wild-
type DRG neurons were clearly trkB positive, in agree-
ment with previous work . Our data would therefore
suggest that a substantial proportion of trkB-positive
neurons are D-hair receptors and that they disappear
Figure 1. D-Hair Receptors Depend on NT-3 and NT-4 during Post-
(A) Graphs show the proportion of A? fibers (conduction velocity ?
10 ms?1or, in 2–3-week-old mice, ? 8 ms?1) classified as
A-mechanonociceptors (white) or D-hair receptors (black) in wild-
type, NT-3, or NT-4-deficient mice over time. In wild-type mice,
about 35% of the A? fibers recorded in the skin nerve preparation
are consistently found to be D-hair receptors [5, 6, 7, 10, 17]. Com-
paredto controlwild-typelittermates, youngNT-3?/? micealready
had a reduced proportion of D-hair receptors by 2–3 weeks. The
experimenter was always blind to the genotype of the recorded
mouse. For comparison, data reported in an earlier study of adult
NT-3 ?/? mice are also plotted . In contrast to results with NT-
3-deficient mice, no loss of D-hair receptors was observed in 2–3-
week-old or 5–7-week-old NT-4 ?/? mice. Again, the data from
adult mice (?7 weeks old) are plotted for comparison . Data in
all groups were obtained from at least six mice, and n refers to the
total number of single neurons recorded.
(B) Graphsshow thequantification ofthe totalnumber ofmyelinated
axons in the saphenous nerve from NT-3 and NT-4-deficient mice.
Semi-thin sections of the saphenous nerve were prepared as pre-
viously described, and the total number of myelinated axons was
counted [5–7]. The number of myelinated axons found in nerves
from wild-type 2–3-week-old littermate mice was equivalent to that
found in adult wild-type mice [5–7]. A significant reduction (p ? 0.01
in an unpaired Student’s t test) in axon number was found in nerves
taken from 2–3-week-old NT-3 ?/? and 3–5-week-old NT-4 ?/?
mice compared to controls. In addition, a further loss of axons was
found in nerves taken from adult NT-4 ?/? mice compared to those
of NT-4 ?/? mice of 3–5 weeks of age.
phenotypic fate of sensory neurons can be modified by
neurotrophin availability in the first two weeks of life [13,
14]. However, electrophysiological experiments can be
used to reliably classify sensory neurons only after the
second postnatal week, in part because the myelination
of sensory axons is still in progress until 2 weeks of
age. In the time period when D-hair receptors become
exclusively dependent on NT-4(?7 weeks), a significant
additional axonal loss of 10% was observed in the sa-
phenous nerve (i.e., the difference in the number of ax-
ons in NT-4 mutants at 3–5 weeks compared to adults;
life. Neurotrophin-4 is exceptional in that, unlike other
neurotrophins [1, 20], it is not critical for the embryonic
development of the majority of sensory neurons. This
attribute has allowed us to demonstrate the functional
importance of NT-4 for a physiologically characterized
mature neuronal type: D-hair receptors. The switch in
neurotrophin dependence exhibited by D-hair receptors
is distinct from other examples in which neurotrophins
change their neurotrophin dependence very early in de-
velopment before they are terminally differentiated .
This report provides the first evidence that fully differen-
tiated sensory neurons in the adult still may require
access to neurotrophic factors for long-term survival.
neurons themselves may synthesize NT-4 ( and our
unpublished data). Future experiments that utilize time-
specific or tissue-specific inactivation of other trophic-
factor genes may reveal that the long-term survival of
many classes of peripheral neurons in adult animals
depends on the continued presence of single neuro-
We used mice heterozygous for deletion of the NT-3 gene and
homozygous for deletion of the NT-4 gene, respectively. NT-4 and
NT-3 mutant mice were obtained from the Jackson laboratories
(Bar Harbor, Maine); the generation of these mice was originally
described in [23, 24]. Mice with a null mutation of the NT-4 gene
are viable and fertile, and we therefore used an appropriate control
strain obtained from the Jackson laboratories (129S3, catalog #
from genomic DNA isolated from tail biopsies, and genotyping was
performed after experiments were completed and analyzed so that
experimenters were blind to the genotype of each preparation.
Figure 2. Loss of trkB-Positive Neurons and Neuronal Apoptosis in
old NT-4 ?/? Mice
Representative photomicrographs of nonradioactive in situ hybrid-
ization experiments on DRG sections in which a DIG-labeled probe
specific for the cytoplasmic tail of mouse trkB (nt 1165–1489) was
used (see reference  for methodological details). (A) In sections
from wild-type mice, a small number of neurons were strongly la-
beled with the trkB probe, and (B) the number of trkB-positive neu-
rons in DRG sections from younger (5-week-old) NT-4 ?/? mice
was similar. (C) However, the number of positive cells in sections
from adult NT-4 ?/? DRGs was dramatically reduced. (D) Quantifi-
cation of the percentage of trkB-positive cells over many different
sections from two independent experiments is also shown. Note a
?/? mice compared to controls. In a separate set of experiments,
TUNEL-positive cells were detected in DRG sections of 8-week-old
control and NT-4 ?/? mice. (F) As expected, no TUNEL-positive
cells were observed in any of the control sections because neuronal
apoptosis is not thought to occur normally in the DRGs of adult
mice. (E) However, occasional TUNEL-positive cells were observed
in DRG sections taken from 8-week-old NT-4 ?/? mice. The exam-
morphology. Sense controls were always clear of staining, and sec-
tions from NT-4 ?/? and controls were always processed on the
same slides for the trkB and TUNEL analyses.
We used an in vitro skin-nerve preparation to record from function-
ally single primary afferents in micro-dissected filaments of the sa-
phenous nerve as described previously [5, 6, 7, 10]. Once the re-
ceptive field of a single sensory axon had been identified with a
mechanical search stimulus, the receptive properties of the unit
were systematically examined with a range of quantitative stimuli.
We used a probe fixed to a micrometer to apply sustained force to
the receptive field for 10 s in order to classify those neurons with
a rapidly or slowly adapting response to mechanical stimuli. D-hair
receptor neurons are characteristically activated by extremely low
(?10 m/s) and a large receptive field allows them to be easily identi-
ities between 1.2 and 10 m/s were always found to be A? fiber
trophic support in the adult NT-4 knockout mouse. It is
trkB receptors [17, 18]. Because SAMs are unaffected
in NT-4 mutants , we presume that the remaining
Several studies have indicated that the percentage of
trkB-positive neurons is higher than we report here;
however, many neurons express a low density of trkB
receptors and therefore may not have been scored as
positive in our experiments .
becomesdependent onaspecificneurotrophic factorin
require another neurotrophin (NT-3) in early postnatal
Histology, TUNEL Assay, and In Situ Hybridization
Histological procedures to quantify the number of axons in the sa-
phenous nerve were carried out as previously described [5, 6, 7,
17]. In situ hybridization experiments to quantify the number of DRG
cell bodies that express trkB mRNA were carried out according to
protocols described . For the detection of apoptosis, paraffin
sections (5 ?M) of all the DRGs from 5 mice per genotype were
dewaxed,permeatedwith 0.1%TritonX-100,and washedwithPBS.
TUNEL (TdT-mediated dUTP nick end labeling) staining was per-
formed according to the manufacturer’s instructions (in situ cell
death detection kit, fluorescein, Roche).
This work was supported by grants from the Deutsche Forschungs-
gemeinschaft (SPP-1025 and SPP-1026), and C.L.S. is supported
Current Biology Download full-text
by National Institutes of Health grant NS40538. We are grateful to
Anke Kanehl for excellent technical assistance. We would also like
to thank Annette Hammes for help with the TUNEL assays and
Andreas Eilers for helpful comments on the manuscript.
18. LeMaster,A.M.,Krimm, R.F.,Davis,B.M.,Noel, T.,Forbes,M.E.,
Johnson, J.E.,and Albers,K.M. (1999).Overexpression ofbrain-
derived neurotrophic factor enhances sensory innervation and
selectively increases neuron number. J. Neurosci. 19, 5919–
19. Karchewski, L.A., Kim, F.A., Johnston, J., McKnight, R.M., and
Verge, V.M. (1999). Anatomical evidence supporting the poten-
tial for modulation by multiple neurotrophins in the majority of
adult lumbar sensory neurons. J. Comp. Neurol. 413, 327–341.
20. Ibanez, C.F. (1996). Neurotrophin-4: the odd one out in the neu-
rotrophin family. Neurochem. Res. 21, 787–793.
21. Davies, A.M. (1997). Neurotrophin switching: where does it
stand? Curr. Opin. Neurobiol. 7, 110–118.
22. Heppenstall, P.A., and Lewin, G.R. (2001). BDNF but not NT-4
is requiredfor normalflexion reflexplasticity andfunction. Proc.
Natl. Acad. Sci. USA 98, 8107–8112.
23. Ernfors, P., Lee, K.F., Kucera, J., and Jaenisch, R. (1994). Lack
system and loss of limb proprioceptive afferents. Cell 20,
24. Liu, X., Ernfors, P., Wu, H., and Jaenisch, R. (1995). Sensory
but not motor neuron deficits in mice lacking NT4 and BDNF.
Nature 375, 238–241.
25. Mannsfeldt, A.G., Carroll, P., Stucky, C.L., and Lewin, G.R.
lian sensory neurons. Mol. Cell. Neurosci. 13, 391–404.
Received: September 27, 2001
Revised: May 30, 2002
Accepted: July 1, 2002
Published: August 20, 2002
1. Huang, E.J., and Reichardt, L.F. (2001). Neurotrophins: roles in
neuronal development and function. Annu. Rev. Neurosci. 24,
2. Maisonpierre, P.C., Belluscio, L., Friedman, B., Alderson, R.F.,
Wiegand, S.J., Furth, M.E., Lindsay, R.M., and Yancopoulos,
G.D. (1990). NT-3, BDNF, and NGF in the developing rat nervous
system: parallel as well as reciprocal patterns of expression.
Neuron 5, 501–509.
3. Bergman, E., Ulfhake, B., and Fundin, B.T. (2000). Regulation
of NGF-family ligands and receptors in adulthood and senes-
cence: correlation to degenerative and regenerative changes in
cutaneous innervation. Eur. J. Neurosci. 12, 2694–2706.
4. Lindsay, R.M. (1988). Nerve growth factors (NGF, BDNF) en-
hance axonal regeneration but are not required for survival of
adult sensory neurons. J. Neurosci. 7, 2394–2405.
C., Wolf, E., Brem, G., Toyka, K.V., Thoenen, H., and Meyer,
M. (1996). Specific subtypes of cutaneous mechanoreceptors
require neurotrophin-3 following peripheral target innervation.
Neuron 16, 287–295.
6. Stucky, C.L., DeChiara, T., Lindsay, R.M., Yancopoulos, G.D.,
and Koltzenburg, M. (1998). Neurotrophin 4 is required for the
survival of a subclass of hair follicle receptors. J. Neurosci. 18,
7. Minichiello, L., Casagranda, F., Tatche, R.S., Stucky, C.L., Post-
igo, A., Lewin, G.R., Davies, A.M., and Klein, R. (1998). Point
mutationintrkB causeslossofNT4-dependent neuronswithout
major effects on diverse BDNF responses. Neuron 21, 335–345.
8. Lewin, G.R. (1996). Neurotrophins and the specification of neu-
ronal phenotype. Philos. Trans. R. Soc. Lond. B Biol. Sci. 351,
9. Caterina, M.J., and Julius, D. (1999). Sense and specificity: a
molecular identity for nociceptors. Curr. Opin. Neurobiol. 9,
properties of mouse sensory neurons innervating hairy skin. J.
Neurophysiol. 78, 1841–1850.
L.F. (2000). Loss of brain-derived neurotrophic factor-depen-
dent neural crest-derived sensory neurons in neurotrophin-4
mutant mice. Proc. Natl. Acad. Sci. USA 97, 2297–2302.
12. Agerman, K., Baudet, C., Fundin, B., Willson, C., and Ernfors,
P. (2000). Attenuation of a caspase-3 dependent cell death in
Cell. Neurosci. 16, 258–268.
13. Ritter,A.M.,Lewin, G.R.,Kremer,N.E.,and Mendell,L.M.(1991).
Requirement for nerve growth factor in the development of my-
elinated nociceptors in vivo. Nature 350, 500–502.
14. Lewin, G.R., Ritter, A.M., and Mendell, L.M. (1992). On the role
of nerve growth factor in the development of myelinated noci-
ceptors. J. Neurosci. 12, 1896–1905.
15. Lawson, S.N. (1992). Morphological and biochemical cell types
of sensory neurons. In Sensory Neurons, Diversity, Develop-
ment, and Plasticity. Edited by S.A. Scott. (London: Oxford Uni-
versity Press), pp 27–59.
16. Mu, X., Silos-Santiago, I., Carroll, S.L., and Snider, W.D. (1993).
Neurotrophin receptor genes are expressed in distinct patterns
in developing dorsal root ganglia. J. Neurosci. 13, 4029–4041.
17. Carroll, P., Lewin, G.R., Koltzenburg, M., Toyka, K.V., and
Thoenen, H. (1998). A role for BDNF in mechanosensation. Nat.
Neurosci. 1, 42–46.