Involvement of Cardiotrophin-1 in Cardiac Myocyte-Nonmyocyte Interactions During Hypertrophy of Rat Cardiac Myocytes In Vitro
Background—The mechanism responsible for cardiac hypertrophy is currently conceptualized as having 2 components, mediated by cardiac myocytes and nonmyocytes, respectively. The interaction between myocytes and nonmyocytes via growth factors and/or cytokines plays an important role in the development of cardiac hypertrophy. We found that cardiac myocytes showed hypertrophic changes when cocultured with cardiac nonmyocytes. Cardiotrophin-1 (CT-1), a new member of the interleukin-6 family of cytokines, was identified by its ability to induce hypertrophic response in cardiac myocytes. In this study, we used the in vitro coculture system to examine how CT-1 is involved in the interaction between cardiac myocytes and nonmyocytes during the hypertrophy process.
Methods and Results—RNase protection assay revealed that CT-1 mRNA levels were 3.5 times higher in cultured cardiac nonmyocytes than in cultured cardiac myocytes. We developed anti–CT-1 antibodies and found that they significantly inhibited the increased atrial and brain natriuretic peptide secretion and protein synthesis characteristic of hypertrophic changes of myocytes in the coculture. In addition, non–myocyte-conditioned medium rapidly elicited tyrosine phosphorylation of STAT3 and induced an increase in natriuretic peptide secretion and protein synthesis in cultured cardiac myocytes; these effects were partially suppressed by anti–CT-1 antibodies. Finally, the hypertrophic effects of CT-1 and endothelin-1, which we had previously implicated in the hypertrophic activity in the coculture, were additive in cardiac myocytes.
Conclusions—These results show that CT-1 secreted from cardiac nonmyocytes is significantly involved in the hypertrophic changes of cardiac myocytes in the coculture and suggest that CT-1 is an important local regulator in the process of cardiac hypertrophy.
Cardiac hypertrophy ultimately leads to heart failure, although initially it is an adaptive process that occurs in response to mechanical load or tissue injury. The mechanism responsible for cardiac hypertrophy is complex, and our understanding of its details remains incomplete. It is characterized by an increase in cardiac myocyte (MC) cell size, proliferation of cardiac fibroblasts, and progression of interstitial and perivascular fibrosis. Therefore, the interaction between MCs and the surrounding nonmyocytes (NMCs) is likely to be an important component of the hypertrophy process.1 Consistent with this notion, many growth factors and cytokines, acting as autocrine/paracrine factors, are involved in the hypertrophic response,2 3 4 5 6 7 8 9 10 11 12 13 and “cross talk” between MCs and NMCs via such humoral factors appears to play an important role in the pathophysiology of cardiac hypertrophy. In an earlier study, we used an in vitro, MC/NMC coculture model to show that hypertrophic responses in MCs, including increased cell size and secretion of natriuretic peptides, occur in coculture and that endothelin-1 (ET-1) secreted from NMCs (fibroblasts or myofibroblasts but not endothelial cells) acts as a paracrine factor during the process.14 We also concluded that other, as yet unidentified, humoral factors secreted from NMCs are involved in the hypertrophic response.14
Recently, gp130-dependent signaling pathways were implicated in the heart development and the progression of cardiac hypertrophy.15 16 Cardiotrophin-1 (CT-1) is a novel member of the interleukin-6 (IL-6) family of cytokines identified by expression cloning based on its ability to induce cardiac MC hypertrophy.17 CT-1 binds to the leukemia inhibitory factor (LIF) receptor/gp130 heterodimer and activates both mitogen-activated protein kinase and Janus kinase signal transducers and activators of transcription (STAT) signaling pathways.18 Through these signaling pathways, CT-1 induces MC hypertrophy19 and prolongs the survival of both embryonic and neonatal rat MCs.20 Indeed, the restricted expression of CT-1 mRNA in the primitive heart tube and the continued myocardial expression during embryogenesis suggest that CT-1 may activate gp130-dependent signaling pathways during cardiac development. Because the heart continues to be a prominent site of CT-1 expression throughout adulthood, it may also mediate hypertrophic responses in MCs in vivo through the same gp130-dependent signaling pathways.18 19 20 21 Using an adult animal model of cardiac hypertrophy (stroke-prone spontaneously hypertensive rats/Izm), we previously showed that CT-1 mRNA expression is specifically augmented in the heart.22 This suggests a possible role of CT-1 in in vivo cardiac hypertrophy. Nevertheless, despite the growing evidence concerning the pharmacological actions and the unique expression profile of CT-1, little is known about the function of endogenous CT-1 in the development of cardiac hypertrophy.
In this report, we examined the involvement of CT-1 in MC/NMC interactions during MC hypertrophy using an in vitro coculture system. We demonstrate dominant expression of CT-1 mRNA in NMCs compared with MCs and show that anti–CT-1 blocking antibodies, developed in our laboratory, suppress hypertrophic responses in MCs. These results suggest that CT-1 functions as a paracrine factor in our in vitro system and implicates it as a possible local regulator during the processes of cardiac hypertrophy and heart failure.
Human ET-1 was purchased from Peptide Institute. Human LIF and rat IL-6 were purchased from Immugenex Corp. BQ123, an ET-A receptor antagonist, was provided by Banyu Pharmaceutical Co Ltd.
Preparation of Recombinant Rat CT-1
The recombinant rat CT-1 was prepared by the GST-fusion system according to the manufacturer’s manual (Pharmacia Biotechnology Inc).
Neonatal rat ventricular MCs and NMCs were prepared on a Percoll gradient and maintained as previously described.14 After MCs were separately collected, the cells were preplated on noncoated dishes for 1 hour to reduce contamination by NMCs. Nonattached cells were then collected; this cell population consisted of >97% MCs, as assessed by immunofluorescence with anti–rat sarcomeric actin antibodies (DAKO Japan Co, Ltd).
NMCs were plated and maintained for 2 or 3 passages as previously reported.14 The absence of endothelial cell contamination of the NMC cultures was confirmed by the lack of 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate acetyl LDL binding (Biochemical Technologies Inc) to the cells.14 NMC-conditioned medium (NMC-CM) was prepared as described previously.14
Cocultures of MCs and NMCs were prepared as previously described14 (the ratio of the cell numbers of NMCs to MCs in the coculture was 30%).
Northern Blot Hybridization Analysis
A 473-bp EcoRI-ApaI fragment of rat CT-1 cDNA was used as a probe.22 Total cellular RNA was extracted from cultured MCs or NMCs with Trizol (Gibco BRL), and Northern blot analysis was performed as previously described.14
RNase Protection Assay
The 473-bp EcoRI-ApaI fragment of rat CT-1 cDNA was ligated with pBluescript II SK(−). Radiolabeled RNA transcripts were then synthesized by use of the DNA template linearized by EcoRI digestion according to the technical manual from Promega Co. The RNase protection assay was performed with the RPA Hybspeed kit (Ambion Inc) as instructed by the manufacturer’s manual. All signals were normalized to an internal control RNA (β-actin) (Ambion Inc).
Preparation of Polyclonal Blocking Antibodies to Rat CT-1
Rabbits were injected with 25 μg SC of recombinant rat CT-1 emulsified in complete Freund’s adjuvant (Difco Laboratories) every 2 weeks. Anti–rat CT-1 antibodies (KCT-1) were purified by protein G–sepharose chromatography (Pharmacia Biotech).
RIAs for Atrial and Brain Natriuretic Peptides
Atrial natriuretic peptide (ANP) and brain natriuretic peptide (BNP) concentrations were measured with specific radioimmunoassays as previously reported.23
Analysis of Protein Synthesis in Cultured Cells
Protein synthesis in cultured cells was evaluated by analysis of [3H]leucine incorporation. Cells were cultured under various experimental conditions for 48 hours. At that time, 3 μCi of [3H]leucine (Amersham Life Science) was added, and the cells were cultured for an additional 24 hours. After being washed twice with ice-cold PBS, the cells were then incubated with 10% trichloroacetic acid for 30 minutes at 4°C. Cell precipitants were then solubilized in 0.2N NaOH for >4 hours. Radioactivity was measured in a liquid scintillation counter.
Western Blot Analysis for STAT3
After stimulation, whole-cell protein extracts were obtained and applied to a 10% SDS-PAGE; the resolved proteins were electrophoretically transferred onto polyvinylidene difluoride membranes (Bio-Rad Laboratories). The membranes were then blocked with 5% skim milk (Difco Laboratories) and probed with anti-STAT3 antibodies (Santa Cruz Biotechnology, Inc) or phosphospecific STAT3 antibodies that recognize only activated STAT3 (New England Biolabs, Inc). The antibodies were applied at a 1:1000 dilution for 1 hour to detect STAT3 or tyrosine-phosphorylated STAT3, respectively. Immune complexes were visualized on Konica Medical film (Konica Co) with an enhanced chemiluminescence system according to the manufacturer’s instructions (NEN Life Science Products).
Western Blot Analysis for CT-1 Purified From NMC-CM
We coupled anti–CT-1 antibodies to a HiTrap NHS-activated column (Pharmacia Biotech) as instructed by the manufacturer’s manual. Ten milliliters of the NMC-CM was applied to the column. The eluate was lyophilized and applied to an SDS-PAGE. Western blot analysis was performed as indicated above using anti–CT-1 antibodies raised by the synthetic 21 peptides corresponding to rat CT-1 (183 to 203) (OCT-6).
Data are presented as mean±SEM. Unpaired Student’s t tests were used to determine significant differences between 2 groups, and ANOVA with subsequent Fisher’s test was used to determine significant differences among 3 or 4 groups. A value of P<0.05 was considered significant.
CT-1 mRNA Expression in MCs and NMCs
To determine which cell types synthesize CT-1 in the heart, we analyzed CT-1 mRNA expression in cultured MCs and NMCs by use of Northern blots. The 1.4-kb CT-1 mRNA signal was clearly detected in NMCs; on the contrary, little signal was detected in MCs (Figure 1A⇓). We also evaluated CT-1 mRNA expression using RNase protection assay, which is a much more sensitive procedure. CT-1 mRNA protected bands were detected in both MCs and NMCs (Figure 1B⇓). However, CT-1 mRNA levels in NMCs were 3.5 times higher than those in MCs (Figure 1C⇓).
Effects of Anti–CT-1 Antibodies in MC/NMC Interactions
To investigate the function of CT-1, we developed anti–CT-1 antibodies. Initially, the capacity of purified anti–CT-1 antibodies to block the effects of CT-1 was tested. MCs were stimulated with 10−9 mol/L CT-1, which is a dose for maximal hypertrophic activity,17 in the presence of 0.1 mg/mL of anti–CT-1 antibodies. As shown in Figure 2⇓, whereas CT-1 significantly increased ANP and BNP secretion (Figure 2A⇓) and [3H]leucine incorporation (Figure 2B⇓) in MCs, the augmentation was completely suppressed in the presence of anti–CT-1 antibodies. To elucidate the specificity of the antibodies for CT-1, we also examined whether the antibodies attenuate the effects of other IL-6 related cytokines, IL-6 and LIF, or a known trophic factor, ET-1. The antibodies did not affect ANP and BNP secretion (Figure 2A⇓) and [3H]leucine incorporation (Figure 2B⇓) in MCs induced by LIF or ET-1. IL-6 had little effect on MCs, as previously reported by others17 (Figure 2⇓, A and B). In addition, the antibodies completely blocked both morphological changes and STAT3 tyrosine phosphorylation induced by CT-1 (Figure 3⇓, A and B).
Next we examined the effects of the blocking antibodies in MC/NMC cocultures. As we have previously reported,14 MCs showed hypertrophic responses when cocultured with NMCs. The presence of 0.1 mg/mL of purified anti–CT-1 blocking antibodies partially, but significantly, inhibited the increased secretion of natriuretic peptides and [3H]leucine incorporation normally seen in MC/NMC cocultures (Figure 4⇓, A, B, and C). By contrast, when the antibodies were added to pure MC cultures, they had no effect at all (Figure 4⇓, A, B, and C). These results indicate that CT-1 contributes significantly to the hypertrophic activity seen in MC/NMC cocultures.
To test whether CT-1 secreted into the culture medium by NMCs mediates the hypertrophic effects, MCs were incubated with NMC-CM containing anti–CT-1 blocking antibodies. It was observed that the antibodies significantly suppressed the increase in secretion of natriuretic peptides and protein synthesis induced by NMC-CM (Figure 4⇑, D, E, and F). Furthermore, NMC-CM induced tyrosine phosphorylation of STAT3 in MCs (Figure 5A⇓), and the time course of tyrosine phosphorylation of STAT3 was similar to that of CT-1–induced phosphorylation (Figure 5B⇓), although the time of peak phosphorylation of STAT3 was not completely identical to that of CT-1. The anti–CT-1 blocking antibodies partially inhibited the tyrosine phosphorylation of STAT3 elicited by NMC-CM (Figure 5C⇓). Thus, the tyrosine phosphorylation of STAT3 induced by NMC-CM appears to be mediated, at least in part, by CT-1. The partial inhibitory effect of anti–CT-1 antibodies suggests that other humoral factors present in NMC-CM also stimulate tyrosine phosphorylation of STAT3.
We also performed Western blotting analysis on samples of NMC-CM using anti–CT-1 antibodies (OCT-6). As shown in Figure 5D⇑, the band responsible for CT-1 was detected in the NMC-CM. We estimated that the NMC-CM contains ≈0.5 μg/L (3×10−11 to 2×10−11 mol/L) of CT-1 by comparing the intensity of the band with that obtained with different quantities of recombinant rat CT-1, as indicated. The concentration of CT-1 is high enough to induce hypertrophic responses, because we and others17 have confirmed that CT-1 induced MC hypertrophy at concentrations as low as 10−11 mol/L.
Additive Hypertrophic Effect of CT-1 and ET-1 in MCs
In an earlier report, we showed that ET-1 plays a role in the MC hypertrophic response in MC/NMC cocultures and that NMC-CM contains ET-1.14 Then we investigated whether the hypertrophic effects of CT-1 and ET-1 were additive. Simultaneous administration of CT-1 and ET-1 augmented secretion of ANP and BNP (Figure 6A⇓) and [3H]leucine incorporation into cells (Figure 6B⇓) to a significantly greater degree than either CT-1 or ET-1 alone. Addition of anti–CT-1 blocking antibodies suppressed the responses to the levels induced by ET-1 alone (Figure 6⇓). These additive hypertrophic effects of CT-1 and ET-1 were observed in both maximal and submaximal conditions (Figure 6C⇓). Moreover, the increases in MC size induced by CT-1 and ET-1 appear to be additive in nature (Figure 7⇓). These results clearly show a summation of the hypertrophic effects of CT-1 and ET-1 in MCs, and they emphasize that MC hypertrophy is a manifestation of the effects of multiple factors.
The simultaneous administration of 0.1 mg/mL CT-1 blocking antibodies and the endothelin type A receptor antagonist BQ123 10−5 mol/L in MC/NMC coculture suppressed ANP secretion more profoundly than those in coculture with either the antibodies or BQ123 alone (P<0.05 versus coculture with CT-1 blocking antibodies alone and P<0.1 versus coculture with BQ123 alone) (Figure 8A⇓). Moreover, the coadministration of the anti–CT-1 antibodies and BQ123 inhibited the NMC-CM–induced increase in MC size to a greater degree than either the antibodies or BQ123 alone (P<0.05 versus NMC-CM with the antibodies alone and P<0.05 versus NMC-CM with BQ123 alone) (Figure 8⇓, B through G). These results indicate that CT-1 and ET-1 secreted by NMCs elicit the hypertrophic response in MCs in an additive fashion.
The mechanism responsible for cardiac remodeling is currently considered to have 2 components mediated by MCs and NMCs, respectively. To investigate the interaction between MCs and NMCs, we established an in vitro MC/NMC coculture system. Using this system, we found that ET-1 secreted from NMCs is implicated in the hypertrophic response observed in MC/NMC coculture.14 Because ET-1 blockade did not inhibit the response completely, however, it was concluded that other as yet unidentified factors might also be involved. Thus, we hypothesized that CT-1 secreted from NMCs also contributes to the development of MC hypertrophy.
To examine this hypothesis, we first studied the expression profile of CT-l mRNA in cultured cardiocytes. Northern blot analysis and RNase protection assays revealed that CT-1 mRNA was expressed to a greater degree in NMCs than in MCs. To determine whether CT-1 acts as an NMC-derived trophic factor, we next studied the effect of anti–CT-1 blocking antibodies in pure MC cultures and in MC/NMC cocultures. Addition of the antibodies to the pure MC cultures had no affect on either secretion of ANP and BNP or protein synthesis assessed by [3H]leucine incorporation. However, when added to the cocultures, CT-1 blocking antibodies significantly suppressed the hypertrophic response. NMC-CM induced rapid (within 5 minutes) tyrosine phosphorylation of STAT3 in MCs, as is the case for exogenously administered CT-1. Anti–CT-1 blocking antibodies partially inhibited NMC-CM–induced STAT3 tyrosine phosphorylation and significantly suppressed the NMC-CM–induced hypertrophic response in MCs. These lines of evidence all indicate that, like ET-1, CT-1 acts as a local trophic factor and contributes to the process of MC hypertrophy in this in vitro model of cardiac hypertrophy.
CT-1 is reported to share its receptors with LIF,19 and LIF also can be a candidate for an NMC-derived hypertrophic factor.24 In our in vitro model, LIF gene expression in serum-starved NMCs was not detected by Northern blot analysis, although CT-1 gene expression was detected in this condition. In addition, in vivo, in both humans and rats, the LIF gene expression was lower than that of CT-1 in the heart (LIF gene expression cannot be detected by Northern blot and can only be found by reverse transcription–polymerase chain reaction; unpublished data, 1998). Furthermore, CT-1 gene expression appears to be increased both in the ventricles of a rat model of cardiac hypertrophy (stroke-prone spontaneously hypertensive rats/Izm versus Wistar-Kyoto/Izm)22 and in the ventricles of patients with ischemic cardiomyopathy, in contrast to no significant increase in LIF gene expression (unpublished data, 1998). These lines of evidence suggest that mainly CT-1, not LIF, acts as a natural ligand to activate gp130-dependent signals in those situations.
In addition to its hypertrophic effect, CT-1 is also reported to inhibit MC apoptosis.21 Ciliary neurotrophic factor and LIF, 2 other members of the IL-6 superfamily of cytokines, share the receptor components LIF receptor/gp130 with CT-1 and have a survival effect on motor neurons in vitro. Interestingly, ciliary neurotrophic factor is expressed in Schwann cells in the adult rat sciatic nerve, and LIF is expressed in nonneuronal cells, including fibroblasts and Schwann cells, in the rat superior cervical ganglia.25 26 Thus, under certain circumstances, interaction between neurons and nonneuronal cells via cytokines expressed in the surrounding nonneuronal cells may contribute to the survival of differentiated neurons. That CT-1 secreted from surrounding NMCs is involved in the maintenance of differentiated MCs in healthy or injured hearts is an attractive hypothesis. However, whether CT-1 promotes survival of MCs in vivo remains to be determined.
In conclusion, we show that NMCs are the principal source of CT-1 in culture and that when secreted, CT-1 contributes significantly to the MC hypertrophic response in vitro. These results suggest a possible role of CT-1 as a local regulator during the processes of cardiac hypertrophy and heart failure in vivo.
This work was supported in part by research grants from the Japanese Ministry of Education, Science, and Culture; the Japanese Ministry of Health and Welfare; the Japanese Society for the Promotion of Science “Research for the Future” program (JSPS-RFTF96I00204 and JSPS-RFTF98L00801); the Smoking Research Foundation; and the Yamanouchi Foundation for Research on Metabolic Disorders. We thank M. Okumura and M. Yamashita for their excellent secretarial work.
- Received October 28, 1998.
- Revision received April 27, 1999.
- Accepted April 30, 1999.
- Copyright © 1999 by American Heart Association
Weber KT, Brilla CG. Pathological hypertrophy and cardiac interstitium: fibrosis and renin-angiotensin-aldosterone system. Circulation. 1991;83:1849–1865.
Long CS, Ordahl CP, Simpson PC. α1-Adrenergic receptor stimulation of sarcomeric action isogene transcription in hypertrophy of cultured rat heart muscle cells. J Clin Invest. 1989;83:1078–1082.
Knowlton KU, Michel MC, Itani M, Shubeita HE, Ishihara K, Brown JH, Chien KR. The alpha 1A-adrenergic receptor subtype mediates biochemical, molecular, and morphologic features of cultured myocardial cell hypertrophy. J Biol Chem. 1993;268:15374–15380.
Shubeita HE, McDonough PM, Harris AN, Knowlton KU, Glembotski CC, Brown JH, Chien KR. Endothelin induction of inositol phospholipid hydrolysis, sarcomere assembly, and cardiac gene expression in ventricular myocytes: a paracrine mechanism for myocardial cell hypertrophy. J Biol Chem. 1990;265:20555–20562.
Ito H, Hirata Y, Adachi S, Tanaka M, Tsujino M, Koike A, Nogami A, Murumo F, Hiroe M. Endothelin-1 is an autocrine/paracrine factor in the mechanism of angiotensin II-induced hypertrophy in cultured rat cardiomyocytes. J Clin Invest. 1993;92:398–403.
Gardner DG, Newman ED, Nakamura KK, Nguyen KP. Endothelin increases the synthesis and secretion of atrial natriuretic peptide in neonatal rat cardiocytes. Am J Physiol. 1991;261:E177–E182.
Parker TG, Packer SE, Schneider MD. Peptide growth factors can provoke “fetal” contractile protein gene expression in rat cardiac myocytes. J Clin Invest. 1990;85:507–514.
Ito H, Hiroe M, Hirata Y, Tsujino M, Adachi S, Shichiri M, Koike A, Nogami A, Marumo F. Insulin-like growth factor-I induces hypertrophy with enhanced expression of muscle specific genes in cultured rat cardiomyocytes. Circulation. 1993;87:1715–1721.
Thaik CM, Calderone A, Takahashi N, Colucci WS. Interleukin-1 beta modulates the growth and phenotype of neonatal rat cardiac myocytes. J Clin Invest. 1995;96:1093–1099.
Harada M, Itoh H, Nakagawa O, Ogawa Y, Miyamoto Y, Kuwahara K, Ogawa E, Igaki T, Yamashita J, Masuda I, Yoshimasa T, Tanaka I, Saito Y, Nakao K. Significance of ventricular myocytes and nonmyocytes interaction during cardiac hypertrophy. Circulation. 1997;96:3737–3744.
Hirota H, Yoshida K, Kishimoto T, Taga T. Continuous activation of gp130, a signal-transducing receptor component for interleukin 6-related cytokines, causes myocardial hypertrophy in mice. Proc Natl Acad Sci U S A. 1995;92:4862–4866.
Yoshida K, Taga T, Saito M, Suematsu S, Kumanogoh A, Tanaka T, Fujiwara H, Hirata M, Yamagami T, Nakahata T, Hirabayashi T, Yoneda Y, Tanaka K, Wang WZ, Mori C, Shiota K, Yoshida N, Kishimoto T. Targeted disruption of gp130, a common signal transducer for the interleukin 6 family of cytokines, leads to myocardial and hematological disorders. Proc Natl Acad Sci U S A. 1996;93:407–411.
Pennica D, King KL, Shaw KJ, Luis E, Rullamas J, Luoh SM, Darbonne WC, Knutzon DS, Yen R, Chien KR, Baker JB, Wood WI. Expression cloning of cardiotrophin 1, a cytokine that induces cardiac myocyte hypertrophy. Proc Natl Acad Sci U S A. 1995;92:1142–1146.
Wollert KC, Taga T, Saito M, Narazaki M, Kishimoto T, Glembotski CC, Vernallis AB, Heath JK, Pennica D, Wood WI, Chien KR. Cardiotrophin-1 activates a distinct form of cardiac muscle cell hypertrophy: assembly of sarcomeric units in series VIA gp130/leukemia inhibitory factor receptor-dependent pathways. J Biol Chem. 1996;271:9535–9545.
Sheng Z, Pennica D, Wood WI, Chien KR. Cardiotrophin-1 displays early expression in the murine heart tube and promotes cardiac myocyte survival. Development. 1996;122:419–428.
Sheng Z, Knowlton K, Chen J, Hoshijima M, Brown JH, Chien KR. Cardiotrophin 1 (CT-1) inhibition of cardiac myocyte apoptosis via a mitogen-activated protein kinase-dependent pathway: divergence from downstream CT-1 signals for myocardial cell hypertrophy. J Biol Chem. 1997;272:5783–5791.
King KL, Lai J, Winer J, Luis E, Yen R, Hooley J, Williams PM, Mather JP. Cardiac fibroblasts produce leukemia inhibitory factor and endothelin, which combine to induce cardiac myocyte hypertrophy in vitro. Endocrine. 1996;5:85–93.
Sendtner M, Stockli KA, Thoenen H. Synthesis and localization of ciliary neurotrophic factor in the sciatic nerve of the adult rat after lesion and during regeneration. J Cell Biol. 1992;118:139–148.