Expression of Nerve Growth Factor (NGF) Isoforms in the Rat Uterus during Pregnancy: Accumulation of Precursor proNGF
http://www.100md.com
内分泌学杂志 2005年第4期
Institute of Anatomy, University of Leipzig, Leipzig D-04103, Germany
Address all correspondence and requests for reprints to: Dr. Edgar Lobos, Institute of Anatomy, University of Leipzig, Liebigstrasse 13, 04103 Leipzig, Germany. E-mail: Edgar.Lobos@medizin.uni-leipzig.de.
Abstract
The mechanisms that promote the transient degenerative changes in the uterus innervation during pregnancy remain incompletely understood. Signaling by the nerve growth factor (NGF)-? is important for maintaining the density of peripheral sympathetic innervation. Here, we analyzed the spatial and temporal expression of NGF isoforms in the rat uterus using RT-PCR, immunoblot analysis, and immunohistochemistry during pregnancy (d 7, 14, and 21), and postpartum (d 1, 8, and 22). Western blot analysis using antibodies to mature NGF-? and to proNGF domain demonstrated a significant decrease in mature NGF-? at gestational d 14 and 21 (term pregnancy) and 1 d postpartum, which paralleled a remarkable accumulation of the 26–28-, 32-, and 60-kDa proNGF forms. There were diminished ratios of mature NGF-? to proNGF independent of uterus growth on the same gestational days. Immunohistochemistry revealed a progressive NGF-? decline throughout pregnancy in the myometrium and a near absence at term pregnancy, which contrasted with increased NGF immunostaining in the intermyometrial connective tissue layers. More importantly, proNGF-specific antibodies identified the increased NGF immunoreactivity in the intermyometrial layers at term pregnancy as proNGF and not mature NGF-?. Alterations in the processing of NGF and accumulation of proNGF in the intermyometrial layers, where axonal degeneration occurs, may contribute significantly to the pregnancy-related uterine denervation and to the control of myometrial activity.
Introduction
IN THE PREGNANT UTERUS, a pronounced transient decrease of autonomous innervation evidenced by a marked axonal degeneration of myometrial and perivascular adrenergic fibers has been reported in animal models (1, 2, 3, 4) and in humans (5, 6). The physiological significance of the uterine denervation during pregnancy is thought to be a reduction of myometrial contractility and the prevention of preterm labor (1). The mechanisms underlying this transient and reversible uterine denervation remain largely unknown. However, several factors have been implicated in the pregnancy-related uterine denervation: 1) the mechanical stretch induced by the growing fetus (7), 2) the high levels of circulating ovarian steroids during pregnancy (1, 8), and 3) a dilution of neurotrophic factors in the pregnant uterus (9, 10). Because the density of sympathetic innervation in effector organs is directly correlated to the expression of neurotrophins especially nerve growth factor (NGF)-? (11, 12), previous studies have analyzed the presence and fluctuations of NGF-? in the pregnant uterus with conflicting results. In the pregnant rat uterus, NGF protein increased as determined by ELISA, but the increased NGF protein (with an assumed increase in neurotrophic activity by NGF-?) was not in accordance with the observed denervation of the uterus (10). In the pregnant guinea pig uterus, a significant impairment of the uterine innervation also occurs, but analysis of the NGF content by ELISA revealed no significant changes of this neurotrophin during pregnancy (13). These authors concluded that alterations in NGF content do not account for the impairment of sympathetic uterine innervation. These results are in contrast to studies that showed that the survival of sympathetic neurons are dependent of NGF-? produced by target tissues (11, 12) and that NGF-? synthesized and secreted by smooth muscle cells is crucial for the development and maintenance of vascular (14) and bladder (15) innervation. Thus, the transient degenerative and regenerative changes in the myometrial innervation in the pregnant rat uterus offer a convenient model for studying the mechanisms involved in the neuronal plasticity of the peripheral nervous system.
Besides the regulation of growth, survival, and differentiation of neurons in both the central and peripheral nervous systems (16), NGF-? contributes to the regulation of functions in nonneuronal cells, but our knowledge of NGF functions in nonneuronal cells is only fragmentary (17). NGF secreted by vascular smooth muscle cells acts in an autocrine manner inducing the migration but not the proliferation of smooth muscle cells that express TrkA receptors (18). NGF is synthesized as a precursor protein (proNGF) that is processed posttranslationally into mature 13.5-kDa NGF-? (for review, see Ref. 19). The processing of proNGF results also in a series of high-molecular-weight intermediate forms whose biological roles are not well understood (19). High-molecular weight glycosylated proNGF forms have been detected in a wide variety of cells in vitro and in vivo (20, 21, 22, 23). ProNGF forms and not mature NGF-? are the predominant forms in human and rat brains and several peripheral tissues (24). ProNGF and its proteolytically processed protein products may differentially activate pro- and antiapoptotic cellular responses through preferential activation of TrkA or p75NT receptors (25). Therefore, it has been suggested that the balance between cell death and survival may be determined by the ratio of secreted proNGF and mature NGF-? (26).
In view of the conflicting results on the NGF-? content in the pregnant uterus and because of the recently recognized importance of NGF processing for its biological activity, the present studies were undertaken to determine in the pregnant and postpartum rat uterus: 1) the temporal and spatial modulation of NGF-? and precursor proNGF and 2) changes in the processing of NGF isoforms and whether these changes correlated with the previously well-described uterine denervation.
Materials and Methods
Animals
All experiments involving animals were carried out in accordance with health standards for the care and use of experimental animals. Virgin female rats (Lewis strain, from the University of Leipzig breeding colony; 60–75 d of age), in estrus, were mated with adult males. The different stages of the cycle in the nonpregnant (NP) rats were determined by examination of vaginal smears. The presence of a vaginal plug after mating was designated as d 1 of pregnancy. The rats were killed at various stages of gestation [d 7 (n = 4), 14 (n = 5), and full-term pregnancy at d 21 (n = 5)] and postpartum d 1 (n = 5), 8 (n = 4), and 22 (n = 5). The uteri were collected and the uterine horns were dissected, freed from the placentas and feti, and immediately frozen and stored at –80 C until use. Uteri from NP rats (n = 5), which were used as controls, and the 22-d postpartum animals were taken at estrus as confirmed by vaginal smears. The mean weight value per uterine horn ± SEM were: NP, 239 ± 8 mg; pregnant d 7, 245 ± 12 mg; d 14, 699 ± 50 mg; d 21, 1398 ± 114 mg; postpartum d 1, 1048 ± 88 mg; postpartum d 8, 150 ± 8 mg; and postpartum d 22, 228 ± 14 mg.
Materials
Rat recombinant NGF-? was purchased from R&D Systems (Minneapolis, MN). Mouse 2.5S NGF was purchased from Sigma-Aldrich (St. Louis, MO). Human postmortem brain tissue cortex was a kind gift from Dr. V. Gundlade (Department of Pathology, University of Leipzig, Leipzig, Germany). Rabbit anti-mature NGF-? antibody (H-20) was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The anti-NGF-? antibody does not cross-react with other neurotrophins such as brain-derived growth factor (BDGF), NT-3, or NT-4. The antibody to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was obtained from Research Diagnostic (Flanders, NJ). Horseradish peroxidase-labeled goat antirabbit and the Vectastain kit were obtained from Vector Laboratories (Burlingame, CA). Rabbit anti-proNGF raised against the proNGF domain (amino acids –60 to –91, code PH 421) was a kind gift from Dr. Max Reinshagen (University of Ulm, Ulm, Germany). Normal rabbit IgG was obtained from Dako Laboratories (Glostrup, Denmark).
Preparation of soluble proteins
Human brain samples were obtained at autopsy 16 h postmortem from a male patient (47 yr old) who was without any neurological disease. The brain tissue was stored at –80 C until use. Uteri from NP rats and of the 22-d postpartum animals were taken at estrus as confirmed by vaginal smears. Frozen uterine horns and human brain were weighed and crushed into a powder with a pestle and mortar under liquid nitrogen. The frozen powder was homogenized in 1 ml ice-cold homogenization buffer (pH 7.4; 20 mM HEPES, 1 mM EDTA, 0.2 M sucrose, 20 μg/ml soybean trypsin inhibitor, 20 μg/ml leupeptin, 5 μg/ml pepstatin A, 5 mM dithiothreitol, 5 μg/ml E-64, 5 μg/ml bestatin, 5 μg/ml aprotinin, 5 μg/ml antipain, and 0.1 mM phenylmethylsulfonyl fluoride) containing 0.5% sodium dodecyl sulfate (SDS). After homogenization, the samples were centrifuged at 80,000 x g for 1 h at 4 C. The supernatants were designated as soluble protein extracts. The protein concentration was determined using the protein assay based on bicinchoninic acid using bovine albumin as the standard (Pierce, Rockford, IL).
Western blot analysis
Protein samples (30 μg/lane) were separated by SDS-PAGE on 12.5% gels for the detection of mature NGF-? and proNGF. Protein samples were diluted in SDS-PAGE sample buffer and boiled for 5 min before being loaded onto the gels. Recombinant rat NGF-? (20 ng/lane from R&D Systems) and human brain cortex (30 μg/lane) were loaded as positive controls to test for the specificity of the anti-mature NGF-? and proNGF antibodies, respectively. To control for antibody specificity, the anti-NGF-? or proNGF antibodies were absorbed to a 10-fold excess of recombinant NGF-?. Proteins were electroblotted using a semidry transblotter. Antibodies were diluted as follows: anti-mature NGF-? and anti-proNGF (1:4,000) and all secondary antibodies (1:5,000). Protein loading was normalized to GAPDH (1:10,000). The proteins were detected using an enhanced chemiluminescence reagent kit (Amersham Biosciences Inc., Piscataway, NJ). The relative density of the bands was densitometrically analyzed with the AlphaImager 2000 (Alpha Innotech Corp., San Leandro, CA). For the densitometric analysis of mature NGF-? and proNGF, the levels for the NP uterus were arbitrarily set as 100%.
RT-PCR
Total RNA was extracted from frozen homogenized uterine tissue using the Qiagen RNA extraction kit (Rneasy, QIAGEN GmbH, Hilden, Germany). First-strand cDNA was synthesized using Moloney murine leukemia virus reverse transcriptase according to the instructions of Amersham Biosciences Inc. (cDNA synthesis kit) in a 15-μl vol reaction containing 2 μg deoxyribonuclease (DNase)-treated total RNA. The housekeeping gene GAPDH was used as an internal control to determine the relative concentration of the NGF-?. PCR was carried out using the primer pairs derived from the published rat GAPDH sequence (GenInfo no. 204248; forward, 5'-CTCCGCCCCTTCCGCTGATG-3'; and reverse, 5'-GTCCACCACCCTGTTGCTGTAG-3') and primers for NGF-? (GenInfo no. 205691; forward, 5'-CTTCAGCATTCCCTTGACAC-3'; and reverse, 5'-TGAGCACACACACGCAGGC-3'). Amplified PCR products (NGF-?, 592 bp; GAPDH, 615 bp) were size-fractionated by electrophoresis through a 2% agarose gel and stained with ethidium bromide (0.2 μg/ml). The DNA bands were visualized and quantified by densitometry and expressed as the ratio of the densitometric analysis for each target gene normalized by the GAPDH control. cDNA sequencing was performed on PCR-amplified products to confirm their identity.
Immunohistochemical localization of NGF-? and precursor proNGF
Rat uteri after isolation were immediately fixed in 4% paraformaldehyde in PBS for 24 h. After fixation, samples were dehydrated through a graded series of alcohol and cleared in methylbenzoate and xylol before paraffin embedding. Sections (7 μm) were mounted on paper glue-coated glass slides and deparaffinated. Uteri from NP and the 22-d postpartum animals were taken at estrus as confirmed by vaginal smears. To differentiate between connective tissue and smooth muscle cells, the uteri were stained with hematoxylin-eosin and van Gieson’s standard reagents. Endogenous peroxidase activity was quenched using a solution of 3% H2O2 in 10% methanol PBS for 3 min. Sections were washed in PBS and Tris-buffered saline, and nonspecific binding sites were blocked by incubating with 1.5% normal goat serum for 30 min at room temperature. After rapid washing with PBS, sections were incubated with rabbit antiserum to NGF-? (1:400) diluted in PBS containing 0.25% BSA and 0.1% Triton X-100 and incubated at 4 C overnight. Sections were extensively washed and incubated with biotinylated goat-antirabbit IgG (1:400). The bound primary antibodies were visualized by avidin-biotin-peroxidase detection using the Vectastain kit (Vector Laboratories) and 3-amino-9-ethylcarbazole as the color-developing reagent prepared according to the manufacturer’s instructions. For the immunohistochemical localization of proNGF signals, cryostat sections were incubated first with proNGF-specific antibody (1:400) at 4 C overnight. After washing, the sections were incubated with horseradish peroxidase-labeled goat antirabbit IgG (1:400) for 30 min at room temperature. After further washing, sections were incubated for 10 min at room temperature with the Tyramide Signal Amplification System (TSA Biotin System; PerkinElmer, Boston, MA). Bound antibodies were visualized with the Vectastain kit as described above. The sections were embedded in water-soluble glycerine gelatin. To control for antibody specificity, slides were incubated with anti-NGF-? antibody absorbed to 10-fold excess of recombinant NGF-? or with preimmune rabbit IgG serum. In both cases, no immunostaining was detected.
Statistical analysis
Data presented are the mean values from four to five animals ± SEM. Data are representative of three independent experiments. Statistical analysis of the data was performed using a Student’s t test with a Bonferroni correction for analysis of multiple comparisons. The differences were considered significant at P < 0.01.
Results
Specificity of the NGF-? and proNGF antibodies
To analyze the specificity of the anti-mature NGF-? and proNGF antibodies, we used purified mature NGF-? and postmortem human parietal cortex as a positive controls for the mature and proNGF isoforms, respectively. As shown in Fig. 1A, the anti-mature NGF-? antibody recognized both proNGF isoforms with molecular masses of 28–85 kDa in the normal human brain cortex (lane 1) and mature NGF-? with a molecular mass of 13 kDa (lane 2). Mature NGF-? was not detected in the human brain protein extract. The reactivity to the 13-kDa mature NGF-? and to the proNGF bands with molecular masses of 28–85 kDa was eliminated or dramatically reduced after incubation with the murine 2.5S NGF form (Fig. 1B). The anti-proNGF antibody recognized proteins with molecular masses of 28–75 kDa in the human brain extract, and, more importantly, the proNGF antibody did not cross-react with mature NGF-? (Fig. 1C). Thus, this confirms proNGF antibody specificity toward the proNGF isoforms only. Incubation of the proNGF-specific antibodies with purified mouse NGF-? had no effect on the immunoreactivities toward the 32- to 75-kDa bands (Fig. 1D). Incubation with preimmune rabbit serum revealed no specific immunostaining, thus revealing the specificity of the staining (data not shown).
FIG. 1. Specificity of anti-mature NGF-? and proNGF antibodies. A, Anti-mature NGF-? antibodies recognize proNGF isoforms (28–85 kDa) in the human brain cortex (lane 1) and mature NGF-? (lane 2). B, NGF-? and proNGF immunoreactive bands were neutralized after overnight incubation of the mature NGF-? antibody with a 10-fold excess (by weight) of murine 2.5S NGF. C, Anti-proNGF antibodies recognize proNGF bands (28–75 kDa) in the human brain cortex and exhibit no cross-reactivity with mature NGF-? (lane 2). D, ProNGF immunoreactivity toward the 28–75 bands was unchanged after absorption of the proNGF antibody with 2.5S murine NGF.
Expression of mature NGF-? and precursor proNGF in the uterus during pregnancy and postpartum
Using the antiserum against mature NGF-?, we investigated by Western blot analysis whether the expression of mature NGF-? was altered in the soluble protein fraction of uterine extracts during pregnancy and postpartum compared with the NP uterus. A remarkable decrease in the 13-kDa band intensity of NGF-? that comigrated with the recombinant rat NGF-? band (rNGF-?) occurred in the uterus during the course of pregnancy (Fig. 2A, lower arrow). Incubation with the secondary antibody or with control rabbit IgG resulted in no detectable staining of any 13-kDa band (data not shown), demonstrating its specificity. The densitometric analysis of band intensities after protein normalization against GAPDH expression is shown in Fig. 2B. In early pregnancy (d 7), no significant differences in band immunoreactivity were observed compared with the NP uterus (P > 0.01). The NGF-? immunoreactivity declined significantly at d 14 by 35.8 ± 10% and at d 21 by 70 ± 6% (P < 0.001). One day postpartum, a 57 ± 8% decrease in NGF-? was still evident (P < 0.001). Uterine NGF-? immunoreactivity returned to levels comparable with the NP uterus at postpartum d 8 (P > 0.01). Surprisingly, and in contrast to the reduced NGF-? signal, our analysis revealed a striking increase in a 60-kDa band during pregnancy (Fig. 2A, upper arrow), probably representing the precursor proNGF recently described in dorsal root ganglia and sympathetic neurons from rats (23, 27). The antibodies also immunostained a 75-kDa band that did not show any significant variation during middle and late pregnancy.
FIG. 2. Western blot analysis of NGF-? and proNGF in the NP, pregnant, and postpartum rat uterus. Uterine proteins (30 μg/lane) and 30 ng recombinant NGF-? (rNGF-?) were analyzed as described in Materials and Methods. A, Representative Western blot analysis of mature NGF-? (lower arrow) in the NP, pregnant (d 7, 14, and 21), and postpartum uterus (postpartum d 1, 8, and 22). Uteri from the NP and 22-d postpartum animals were taken at the 22-d postpartum animals were taken at estrus as confirmed by vaginal smears. B, Densitometric analysis of mature NGF-? protein expressed as a percentage of the NP uterus (100%) after GAPDH normalization. C, Representative Western blot analysis using proNGF-? antibody demonstrating the accumulation of the 60- and 26- to 32-kDa proNGF forms. D, Densitometrical analysis of proNGF after normalization to GAPDH. Each bar represents the mean ± SEM of each group of rats (n = 4–5 rats/group). Shaded bars, Pregnant; black bars, postpartum. *, P < 0.01; **, P < 0.001; significant differences in NGF-? or proNGF levels from the NP uterus.
Increased proNGF levels in the pregnant uterus
To assess the identity of the main 60-kDa band, we used an antiserum specific to the proNGF domain corresponding to amino acids –60 to –91 of the prosequence position. The reactivity of this proNGF antibody has been previously extensively characterized (24). The prodomain-specific antibody identified the 60-kDa protein band as the precursor proNGF (Fig. 2C). When the proNGF bands were normalized to the corresponding GAPDH intensities (Fig. 2D), a significant 137 ± 10% increase in proNGF in the uterus was observed at d 14, which increased further to 167 ± 3% at term pregnancy (P < 0.001, respectively). Postpartum d 1 proNGF remained increased by 153 ± 7% (P < 0.001). The intensity of the 60-kDa proNGF immunoreactive band returned to levels comparable with the NP uterus at postpartum d 8 (P > 0.01) (Fig. 2D). In addition, the proNGF antibody also identified 26–28 kDa and 32-kDa immunoreactive bands in the pregnant uterus at gestational d 14 and 21 and at postpartum d 1. These 26- to 32-kDa immunoreactive bands were not detected in the NP uterus before gestational d 14 nor after postpartum d 8 when the NGF-? band immunostaining returned to levels comparable or higher than the NP uterus. These results show disturbances in the posttranslational processing of NGF during pregnancy and that these could be responsible for the progressive proNGF accumulation in the uterus.
Ratio of mature NGF-?/proNGF in pregnant and postpartum uterus
Previous Western blot studies have shown that this particular antibody recognizes mature NGF-? and proNGF with similar affinities (24). Because purified proNGF protein is not commercially available to perform a quantitative analysis, we calculated pregnancy-related changes in the ratio of the mature NGF ? to proNGF band immunoreactivities after normalization to GAPDH expression. After compensating for the increase in horn weight during pregnancy, taking into account the mean weight of the uterus on the defined gestational days, the NGF-? to proNGF ratios (Fig. 3) showed a 59% decline at d 14 (ratio 0.82 ± 0.1 vs. 2.01 ± 0.25 in NP; P < 0.01), followed by a 83% decline at d 21 (ratio 0.332 ± 0.14; P < 0.001). The ratio was still significantly decreased by 80% at postpartum d 1 (ratio 0.39 ± 0.13; P < 0.001), returning to values not significantly different from the NP uterus after postpartum d 8 (Fig. 2). Thus, the marked decrease in the ratio of mature NGF-? to proNGF during pregnancy was independent of any increase in uterus weight during gestation.
FIG. 3. Alterations in the ratios of mature NGF-? to proNGF per horn (mgr) throughout pregnancy. The ratios were calculated by dividing the normalized single ratios of NGF-? by the ratios of proNGF, and then the ratios were calculated per uterus horn (mgr). White bars, NP; shaded bars, pregnant; dark bars, postpartum. Bars represent the mean of the respective individual ratios ± SEM. *, P < 0.01; **, P < 0.001; significantly different ratio from the NP uterus.
Semiquantitative RT-PCR analysis of NGF-? mRNA
We analyzed whether the abundance of NGF mRNA was altered in the uterus throughout pregnancy and postpartum compared with the NP uterus. The linearity of the PCR amplification reactions was assessed for each primer pair previously allowing the setting up of conditions under which PCR amplification was in the logarithmic phase (data not shown). No PCR products were obtained when samples were amplified without RT, indicating the absence of genomic DNA. The primers specific for NGF mRNA amplified a PCR product of 592 bp as shown in the NP, pregnant, and postpartum uterus (Fig 4A). The densitometric analysis of the uterine NGF mRNA after normalization to GAPDH mRNA (Fig. 4B) revealed no significant differences in mRNA expression throughout gestation, postpartum vs. the NP uterus (P > 0.1). Similar results were obtained when the RT-PCR was normalized to the mRNA expression of calponin (data not shown).
FIG. 4. Semiquantitative RT-PCR analysis of uterine NGF mRNA during gestation. A, Representative agarose gel of RT-PCR amplification of NGF mRNA from NP, different days of pregnancy and postpartum rat uterus. The primers used to amplify NGF are described in Material and Methods. First-strand cDNAs were made from 2 μg uterine RNA. The numbers on the right side of the gels indicate the sizes of the amplified products. B, Graphic compilation of the PCR data generated for NGF mRNA normalized to the expression of GAPDH. Circles, mean ± SEM (n = 3–4 rats/group). P > 0.01, no significant differences in NGF mRNA from the NP uterus.
Immunohistochemical localization of NGF-? and proNGF
To analyze the temporal and spatial changes in the expression of NGF in the uterus, we performed immunohistochemical analysis using first an antibody against mature NGF-?. In the NP rats, NGF-? immunoreactivity was detected predominantly in the inner and outer myometrial layers and in the smooth muscle cells of the uterine blood vessels (Fig. 5A). In addition, expression of NGF-? was detected in the epithelium of uterine glands as seen in the inset in Fig. 5A. As pregnancy progressed, a major decrease in NGF-? immunostaining occurred in the myometrial layers, the vascular smooth muscle cells, and the glandular epithelium, which was more pronounced at gestational d 21 (Fig. 5B). This coincided with increased NGF immunostaining in some intermyometrial areas and areas around the outer myometrium (arrows) (Fig. 5B). Additional histochemical analysis clearly established that the areas with increased NGF immunoreactivity are connective tissue areas that stained red with the van Gieson stain, in contrast to the yellow-brown stain of the smooth muscle cells (Fig. 5C). After parturition, uterine NGF-? immunostaining remained low, but restoration to an intensity comparable with the NP uterus was evident at postpartum d 22 (Fig. 5D). No immunoreaction was observed in the NP uterus when antiserum was preabsorbed with an excess of NGF-?, demonstrating the specificity of the immunostaining (Fig. 5E).
FIG. 5. Immunohistochemical localization of NGF-? and proNGF in the NP, pregnant, and postpartum uterus using polyclonal anti-mature NGF-? and proNGF antibodies. A, Uterus from NP rats, inner (im), and outer myometrial layers (om), vascular smooth muscle cells (v); uterine glandulae (gl) are NGF-? immunoreactive as shown in the inset in A. B, At d 21, NGF-? immunostaining of the myometrium was nearly abolished in the myometrium but was increased in some intermyometrial areas and in connective tissue areas around the outer myometrium (arrows). C, van Gieson staining of the uterus at d 21 showing connective tissue areas red (arrows), clearly distinguishable from the yellow-brown smooth muscle cells from the inner and outer myometrium. D, NGF-? immunostaining of the postpartum rat uterus at d 22 is comparable with the NP uterus. E, NP uterus incubated with a NGF-? antibody preabsorbed overnight with a 10-fold excess of murine 2.5S NGF showing the specificity of the staining. F, Absence of proNGF immunoreactivity signal in the NP rat uterus after incubation with proNGF-specific antibodies. G, Uterus at term pregnancy (d 21) showing markedly increased proNGF immunoreactivity in some intermyometrial areas and in connective tissue areas around the outer myometrium (arrows). H, Absence of specific staining in the intermyometrial connective tissue areas was observed using rabbit preimmune serum demonstrating the specificity of the staining. I, At postpartum d 22, no specific proNGF immunoreactivity was observed. Scale bars, 100 μm.
Because the anti-mature NGF-? antibody cannot differentiate between the mature and the proNGF isoforms, we sought to differentiate their immunolocalization using the proNGF antibody, which does not recognize mature NGF-?. Immunohistochemical analysis using an enhanced signal amplification system localized proNGF to the intermyometrial and connective tissue areas surrounding the outer myometrium. ProNGF immunoreactivity increased from d 14 onwards (data not shown) and was maximal at gestational d 21 (Fig. 5G; arrows). Furthermore, control experiments using preimmune serum revealed no specific signal in any intermyometrial areas, demonstrating the specificity of the binding and simultaneously demonstrating the absence of nonspecific endogenous peroxidase activity in this area (Fig. 5H). In contrast, proNGF immunostaining was not detectable in the NP uterus (Fig. 5F) or at postpartum d 22 (Fig. 5I). Thus, the increase in NGF immunoreactivity in the intermyometrial tissue area is due to accumulation of proNGF during pregnancy.
Discussion
In this study, we report that in the rat uterus, the expression of mature NGF-? is significantly reduced during middle and late pregnancy and 1 d postpartum compared with the NP uterus. Antibodies to mature NGF-? recognized both mature and proNGF forms in contrast to the proNGF antibody, which recognized only the proNGF isoform. Interestingly, mature NGF-? was not present in the human brain cortex, in agreement with previous studies showing the presence of only proNGF (24). At term, uterine NGF-? protein fell by approximately 75% as demonstrated by Western blot analysis. Immunohistochemistry analysis revealed that mature NGF-? was expressed also in the uterine glands, and its presence might also contribute to this finding. However, because the endometrium is not innervated, endometrial changes in the expression of the NGF isoforms will have little bearing on changes in myometrial innervation.
Our findings strongly suggest that NGF-? produced in the myometrium is a significant regulator of myometrial innervation because the decrease in NGF-? parallels the well-documented axonal degeneration of the myometrial sympathetic nerves (1, 9, 28). Our results contrast to previous reports on the increase in NGF-? protein content in the pregnant uterus based on ELISA that used antisera with undefined reactivities toward the NGF isoforms (10, 13). These studies assumed incorrectly to measure mature NGF-? only, which is not the case as recently demonstrated by the discovery that proNGF instead of mature NGF is the main form in the mammalian brain and in other tissues (24, 27). We demonstrate here that only the unprocessed proNGF forms increased at gestational d 14 to term and 1 d postpartum coinciding, according to the literature, with the axonal degeneration of small sensory fibers and sympathetic axons of the myometrium (1, 9, 28).
Thus, reduced supply of prosurvival mature neurotrophin in the pregnant uterus is linked to its transient neurodegenerative changes. This finding not only agrees with the growing evidence that reduced neurotrophin support leads to neurodegenerative processes of the central nervous system (29) but also extends it to neurodegenerative processes in peripheral organs. However, NGF is produced and secreted by smooth muscle cells, which also express TrkA (18); thus, its decline during pregnancy may result also in alterations of the nonneuronal effects of NGF on smooth muscle cells, which are not completely understood. An autocrine action of NGF on smooth muscle cells includes the induction of chemotaxis without induction of proliferation (18). In animal models of hypertension, hypertrophy and hyperplasia of smooth muscle cells are linked in a time-dependent way to increased NGF secretion measured by ELISA (30, 31). However, in these studies, the reactivity of the antiserum toward the NGF isoforms was not defined. During pregnancy, the myometrium undergoes both hyperplasic and hypertrophic changes, which parallel with an accumulation of proNGF forms as shown here, suggesting that prolonged muscle stretch might influence the processing of NGF in association with altered muscle function.
Using a proNGF-specific antibody that does not recognize mature NGF-?, we demonstrated a marked increase of the 60- and 26- to 32-kDa proNGF forms at d 14 of gestation until term and at 1 d postpartum. A 60-kDa band, representing the unprocessed glycosylated proNGF, and 26- to 32-kDa bands corresponding to the nonglycosylated translation products of rat proNGF were previously detected by this antibody in rat dorsal root ganglia, colon, and spinal cord tissues (23). A 32-kDa precursor proNGF was previously detected in the human brain, which was the main NGF isoform in both normal brain and Alzheimer’s disease brain (24). Furthermore, the proNGF antibody identified a 13-kDa band in the rat uteri with a similar molecular weight to mature NGF-? which arises because rat proNGF contains two dibasic cleavage sites necessary for its processing as previously reported (23, 32).
The accumulation of the proNGF forms in the pregnant uterus might result from either impaired processing of NGF by furin or the proconvertases enzymes (25, 33) or from alterations in the glycosylation/deglycosylation of proNGF. High-molecular weight forms of proNGF are synthesized and secreted by sympathetic neurons, but these proNGF forms do not support neuronal survival (27). In contrast, it was recently reported that proNGF exhibits neurotrophic activity but shows at least a 5-fold less activity than the mature NGF-? (34). Accumulation of proNGF in the brain of Alzheimer’s patients was recently demonstrated and postulated to be due to disturbances in the posttranslational processing of NGF (24) because no difference in NGF mRNA expression between Alzheimer’s patients and control individuals was found (35, 36). In cultured neurons, proNGF binds with high affinity and preferentially to p75NTR and induces p75NTR-dependent apoptosis (25). The secreted 26- to 28-kDa proNGF forms were recently shown to be a pathophysiological death-inducing ligand after brain injury (37). On this basis, it has been suggested that the balance between cell death and survival in neurons may be determined by the ratio of proNGF and mature NGF (26). The data presented here support this idea because the decline in the ratio of mature NGF-? to proNGF during gestation paralleled with the reported degeneration of myometrial and perivascular nerves (1, 9, 28). The recovery of the ratio at d 8 postpartum coincided with the restoration of uterine sympathetic innervation reported previously (1, 9, 28). The increase in proNGF protein in the uterus is probably not related to enhanced transcription because RT-PCR during gestation revealed no increased NGF mRNA, in contrast to the significantly elevated proNGF protein demonstrated at this time. Factors that modulate the expression of mRNA NGF and protein include ovarian hormones (38, 39) and cytokines (40). Long-term estrogen treatment of infantile/prepubertal rats results in a complete loss of the intrauterine sympathetic nerves (41). It is conceivable that the high levels of hormones during pregnancy might induce alterations in the processing of neurotrophins in the uterus.
Immunohistochemical studies identified NGF-? to be predominantly located in the myometrium and in the smooth muscle cells of uterine vessels in agreement with recent studies (38, 41). The decrease in NGF-? immunoreactivity in the uterine blood vessels of the pregnant rat parallels the reported decline of NGF-? in the human uterine arteries of pregnant women (42). The decline in mature prosurvival NGF-? immunostaining was concomitant with an increased NGF-immunoreactivity in intermyometrial connective tissue areas. Using the proNGF-specific antibody, we demonstrated that the increased intermyometrial NGF-immunoreactivity at term pregnancy is due to the accumulation of proNGF and not mature NGF-?. No proNGF immunohistochemistry signal was detected in the NP uterus or in the postpartum uterus (d 22). Epitopes in denatured antigens recognized in Western blot analysis may be different from epitopes on nondenatured antigens recognized in immunohistochemistry; therefore, one cannot expect the same level of sensitivity detection. This appears to be the case because the proNGF immunoreactivity in the pregnant myometrium was visible only after the use of an enhanced detection system.
In summary, we demonstrate for the first time alterations in NGF isoforms during pregnancy, accumulation of proNGF, and decreased ratios of mature NGF-? to proNGF in the pregnant rat uterus coinciding with the transient uterine axonal degeneration of the pregnant myometrium reported by previous studies. Our observations strongly support the idea that under certain circumstances when the balance between a proneurotrophin and mature neurotrophin favors the proneurotrophins, it may cause degeneration of neurons. Additional studies on the biological effects of NGF-?, proNGF, and its intermediates on neuronal and nonneuronal cells will provide insights into the multiple mechanisms involved in the maintenance of relative myometrial quiescence in pregnancy.
Acknowledgments
We thank Dr. John S. Hothersall (University College London, London, UK) for critical reading of the manuscript.
References
Owman C 1981 Pregnancy induces degenerative and regenerative changes in the autonomic innervation of the female reproductive tract. Ciba Found Symp 83:252–279
Sporrong B, Alm P, Owman C, Sjoberg NO, Thorbert G 1981 Pregnancy is associated with extensive adrenergic nerve degeneration in the uterus. An electronmicroscopic study in the guinea-pig. Neuroscience 6:1119–1126
Alm P, Lundberg LM, Wharton J, Polak JM 1988 Organization of the guinea-pig uterine innervation. Distribution of immunoreactivities for different neuronal markers. Effects of chemical- and pregnancy-induced sympathectomy. Histochem J 20:290–300
Moustafa FA 1988 Changes in cholinergic and noradrenergic nerves in the pregnant and postpartum uterus of the albino rat and guinea pig. Acta Anat (Basel) 132:310–316
Wikland M, Lindblom B, Dahlstrom A, Haglid KG 1984 Structural and functional evidence for the denervation of human myometrium during pregnancy. Obstet Gynecol 64:503–509
Morizaki N, Morizaki J, Hayashi RH, Garfield RE 1989 A functional and structural study of the innervation of the human uterus. Am J Obstet Gynecol 160:218–228
Lundberg LM, Alm P, Thorbert G 1989 Local mechanical effects and humoral factors evoke degeneration of guinea pig uterine innervation. Acta Obstet Gynecol Scand 68:487–496
Marshall JM 1981 Effects of ovarian steroids and pregnancy on adrenergic nerves of uterus and oviduct. Am J Physiol 240:C165–C174
Haase EB, Buchman J, Tietz AE, Schramm LP 1997 Pregnancy-induced uterine neuronal degeneration in the rat. Cell Tissue Res 288:293–306
Varol FG, Duchemin AM, Neff NH, Hadjiconstantinou M 2000 Nerve growth factor (NGF) and NGF mRNA change in rat uterus during pregnancy. Neurosci Lett 294:58–62
Korsching S, Thoenen H 1983 Nerve growth factor in sympathetic ganglia and corresponding target organs of the rat: correlation with density of sympathetic innervation. Proc Natl Acad Sci USA 80:3513–3516
Shelton DL, Reichardt LF 1984 Expression of the ?-nerve growth factor gene correlates with the density of sympathetic innervation in effector organs. Proc Natl Acad Sci USA 81:7951–7955
Brauer MM, Shockley KP, Chavez R, Richeri A, Cowen T, Crutcher KA 2000 The role of NGF in pregnancy-induced degeneration and regeneration of sympathetic nerves in the guinea pig uterus. J Auton Nerv Syst 79:19–27
Head RJ 1991 Hypernoradrenergic innervation and vascular smooth muscle hyperplastic change. Blood Vessels 28:173–178
Steers WD, Kolbeck S, Creedon D, Tuttle JB 1991 Nerve growth factor in the urinary bladder of the adult regulates neuronal form and function. J Clin Invest 88:1709–1715
Bibel M, Barde, YA 2000 Neurotrophins: key regulators of cell fate and cell shape in the vertebrate nervous system. Genes Dev 14:2919–2937
Tessarollo L 1998 Pleiotropic functions of neurotrophins in development. Cytokine Growth Factor Rev 9:125–137
Donovan MJ, Miranda RC, Kraemer R, McCaffrey TA, Tessarollo L, Mahadeo D, Sharif S, Kaplan DR, Tsoulfas P, Parada L, Toran-Allerand CD, Hajjar DP, Hempstead B 1995 Neurotrophin and neurotrophin receptors in vascular smooth muscle cells. Regulation of expression in response to injury. Am J Pathol 147:309–324
Fahnestock M, Yu G, Coughlin MD 2004 ProNGF: a neurotrophic or an apoptotic molecule? Prog Brain Res 146:107–110
Saboori AM, Young M 1986 Nerve growth factor: biosynthetic products of the mouse salivary glands. Characterization of stable high molecular weight and 32,000-dalton nerve growth factors. Biochemistry 25:5565–5571
Edwards RH, Selby MJ, Garcia PD, Rutter WJ 1988 Processing of the native nerve growth factor precursor to form biologically active nerve growth factor. J Biol Chem 263:6810–6815
Lakshmanan J, Burns C, Smith RA 1988 Molecular forms of nerve growth factor in mouse submaxillary glands. Biochem Biophys Res Commun 152:1008–1014
Reinshagen M, Geerling I, Eysselein VE, Adler G, Huff KR, Moore GP, Lakshmanan J 2000 Commercial recombinant human beta-nerve growth factor and adult rat dorsal root ganglia contain an identical molecular species of nerve growth factor prohormone. J Neurochem 74:2127–2133
Fahnestock M, Michalski B, Xu B, Coughlin MD 2001 The precursor pro-nerve growth factor is the predominant form of nerve growth factor in brain and is increased in Alzheimer’s disease. Mol Cell Neurosci 18:210–220
Lee R, Kermani P, Teng KK, Hempstead BL 2001 Regulation of cell survival by secreted proneurotrophins. Science 294:1945–1948
Chao MV, Bothwell M 2002 Neurotrophins: to cleave or not to cleave. Neuron 33:9–12
Hasan W, Pedchenko T, Krizsan-Agbas D, Baum L, Smith PG 2003 Sympathetic neurons synthesize and secrete pro-nerve growth factor protein. J Neurobiol 57:38–53
Klukovits A, Gaspar R, Santha P, Jancso G, Falkay G 2002 Functional and histochemical characterization of a uterine adrenergic denervation process in pregnant rats. Biol Reprod 67:1013–1017
Dawbarn D, Allen SJ 2003 Neurotrophins and neurodegeneration. Neuropathol Appl Neurobiol 29:211–230
Clemow DB, Spitsbergen JM, McCarty R, Steers WD, Tuttle JB 1998 Arterial nerve growth factor (NGF) mRNA, protein, and vascular smooth muscle cell NGF secretion in hypertensive and hyperactive rats. Exp Cell Res 244:196–205
Clemow DB, Steers WD, Tuttle JB 2000 Stretch-activated signaling of nerve growth factor secretion in bladder and vascular smooth muscle cells from hypertensive and hyperactive rats. J Cell Physiol 183:289–300
Dicou E, Lee J, Brachet P 1986 Synthesis of nerve growth factor mRNA and precursor protein in the thyroid and parathyroid glands of the rat. Proc Natl Acad Sci USA 83:7084–7088
Seidah NG, Benjannet S, Pareek S, Savaria D, Hamelin J, Goulet B, Laliberte J, Lazure C, Chretien M, Murphy RA 1996 Cellular processing of the nerve growth factor precursor by the mammalian pro-protein convertases. Biochem J 314:951–960
Fahnestock M, Yu G, Michalski B, Mathew S, Colquhoun A, Ross GM, Coughlin MD 2004 The nerve growth factor precursor proNGF exhibits neurotrophic activity but is less active than mature nerve growth factor. J Neurochem 89:581–592
Goedert M, Fine A, Hunt SP, Ullrich A 1986 Nerve growth factor mRNA in peripheral and central rat tissues and in the human central nervous system: lesion effects in the rat brain and levels in Alzheimer’s disease. Brain Res 387:85–92
Jette N, Cole MS, Fahnestock M 1994 NGF mRNA is not decreased in frontal cortex from Alzheimer’s disease patients. Brain Res Mol Brain Res 25:242–250
Harrington AW, Leiner B, Blechschmitt C, Arevalo JC, Lee R, Morl K, Meyer M, Hempstead BL, Yoon SO, Giehl KM 2004 Secreted proNGF is a pathophysiological death-inducing ligand after adult CNS injury. Proc Natl Acad Sci USA 101:6226–6230
Bjorling DE, Beckman M, Clayton MK, Wang ZY 2002 Modulation of nerve growth factor in peripheral organs by estrogen and progesterone. Neuroscience 110:155–167
Chavez-Genaro R, Crutcher K, Viettro L, Richeri A, Coirolo N, Burnstock G, Cowen T, Brauer MM 2002 Differential effects of oestrogen on developing and mature uterine sympathetic nerves. Cell Tissue Res 308:61–73
Sherer TB, Neff PS, Hankins GR, Tuttle JB 1998 Mechanisms of increased NGF production in vascular smooth muscle of the spontaneously hypertensive rat. Exp Cell Res 241:186–193
Chalar C, Richeri A, Viettro L, Chavez-Genaro R, Bianchimano P, Marmol NM, Crutcher K, Burnstock G, Cowen T, Brauer MM 2003 Plasticity in developing rat uterine sensory nerves: the role of NGF and TrkA. Cell Tissue Res 314:191–205
Naves FJ, Vazquez MT, Jose IS, Martinez-Almagro A, Vega JA 1998 Pregnancy-induced denervation of the human uterine artery correlates with local decrease of NGF and TrkA. Ital J Anat Embryol 103:279–290(Edgar Lobos, Claudia Gebh)
Address all correspondence and requests for reprints to: Dr. Edgar Lobos, Institute of Anatomy, University of Leipzig, Liebigstrasse 13, 04103 Leipzig, Germany. E-mail: Edgar.Lobos@medizin.uni-leipzig.de.
Abstract
The mechanisms that promote the transient degenerative changes in the uterus innervation during pregnancy remain incompletely understood. Signaling by the nerve growth factor (NGF)-? is important for maintaining the density of peripheral sympathetic innervation. Here, we analyzed the spatial and temporal expression of NGF isoforms in the rat uterus using RT-PCR, immunoblot analysis, and immunohistochemistry during pregnancy (d 7, 14, and 21), and postpartum (d 1, 8, and 22). Western blot analysis using antibodies to mature NGF-? and to proNGF domain demonstrated a significant decrease in mature NGF-? at gestational d 14 and 21 (term pregnancy) and 1 d postpartum, which paralleled a remarkable accumulation of the 26–28-, 32-, and 60-kDa proNGF forms. There were diminished ratios of mature NGF-? to proNGF independent of uterus growth on the same gestational days. Immunohistochemistry revealed a progressive NGF-? decline throughout pregnancy in the myometrium and a near absence at term pregnancy, which contrasted with increased NGF immunostaining in the intermyometrial connective tissue layers. More importantly, proNGF-specific antibodies identified the increased NGF immunoreactivity in the intermyometrial layers at term pregnancy as proNGF and not mature NGF-?. Alterations in the processing of NGF and accumulation of proNGF in the intermyometrial layers, where axonal degeneration occurs, may contribute significantly to the pregnancy-related uterine denervation and to the control of myometrial activity.
Introduction
IN THE PREGNANT UTERUS, a pronounced transient decrease of autonomous innervation evidenced by a marked axonal degeneration of myometrial and perivascular adrenergic fibers has been reported in animal models (1, 2, 3, 4) and in humans (5, 6). The physiological significance of the uterine denervation during pregnancy is thought to be a reduction of myometrial contractility and the prevention of preterm labor (1). The mechanisms underlying this transient and reversible uterine denervation remain largely unknown. However, several factors have been implicated in the pregnancy-related uterine denervation: 1) the mechanical stretch induced by the growing fetus (7), 2) the high levels of circulating ovarian steroids during pregnancy (1, 8), and 3) a dilution of neurotrophic factors in the pregnant uterus (9, 10). Because the density of sympathetic innervation in effector organs is directly correlated to the expression of neurotrophins especially nerve growth factor (NGF)-? (11, 12), previous studies have analyzed the presence and fluctuations of NGF-? in the pregnant uterus with conflicting results. In the pregnant rat uterus, NGF protein increased as determined by ELISA, but the increased NGF protein (with an assumed increase in neurotrophic activity by NGF-?) was not in accordance with the observed denervation of the uterus (10). In the pregnant guinea pig uterus, a significant impairment of the uterine innervation also occurs, but analysis of the NGF content by ELISA revealed no significant changes of this neurotrophin during pregnancy (13). These authors concluded that alterations in NGF content do not account for the impairment of sympathetic uterine innervation. These results are in contrast to studies that showed that the survival of sympathetic neurons are dependent of NGF-? produced by target tissues (11, 12) and that NGF-? synthesized and secreted by smooth muscle cells is crucial for the development and maintenance of vascular (14) and bladder (15) innervation. Thus, the transient degenerative and regenerative changes in the myometrial innervation in the pregnant rat uterus offer a convenient model for studying the mechanisms involved in the neuronal plasticity of the peripheral nervous system.
Besides the regulation of growth, survival, and differentiation of neurons in both the central and peripheral nervous systems (16), NGF-? contributes to the regulation of functions in nonneuronal cells, but our knowledge of NGF functions in nonneuronal cells is only fragmentary (17). NGF secreted by vascular smooth muscle cells acts in an autocrine manner inducing the migration but not the proliferation of smooth muscle cells that express TrkA receptors (18). NGF is synthesized as a precursor protein (proNGF) that is processed posttranslationally into mature 13.5-kDa NGF-? (for review, see Ref. 19). The processing of proNGF results also in a series of high-molecular-weight intermediate forms whose biological roles are not well understood (19). High-molecular weight glycosylated proNGF forms have been detected in a wide variety of cells in vitro and in vivo (20, 21, 22, 23). ProNGF forms and not mature NGF-? are the predominant forms in human and rat brains and several peripheral tissues (24). ProNGF and its proteolytically processed protein products may differentially activate pro- and antiapoptotic cellular responses through preferential activation of TrkA or p75NT receptors (25). Therefore, it has been suggested that the balance between cell death and survival may be determined by the ratio of secreted proNGF and mature NGF-? (26).
In view of the conflicting results on the NGF-? content in the pregnant uterus and because of the recently recognized importance of NGF processing for its biological activity, the present studies were undertaken to determine in the pregnant and postpartum rat uterus: 1) the temporal and spatial modulation of NGF-? and precursor proNGF and 2) changes in the processing of NGF isoforms and whether these changes correlated with the previously well-described uterine denervation.
Materials and Methods
Animals
All experiments involving animals were carried out in accordance with health standards for the care and use of experimental animals. Virgin female rats (Lewis strain, from the University of Leipzig breeding colony; 60–75 d of age), in estrus, were mated with adult males. The different stages of the cycle in the nonpregnant (NP) rats were determined by examination of vaginal smears. The presence of a vaginal plug after mating was designated as d 1 of pregnancy. The rats were killed at various stages of gestation [d 7 (n = 4), 14 (n = 5), and full-term pregnancy at d 21 (n = 5)] and postpartum d 1 (n = 5), 8 (n = 4), and 22 (n = 5). The uteri were collected and the uterine horns were dissected, freed from the placentas and feti, and immediately frozen and stored at –80 C until use. Uteri from NP rats (n = 5), which were used as controls, and the 22-d postpartum animals were taken at estrus as confirmed by vaginal smears. The mean weight value per uterine horn ± SEM were: NP, 239 ± 8 mg; pregnant d 7, 245 ± 12 mg; d 14, 699 ± 50 mg; d 21, 1398 ± 114 mg; postpartum d 1, 1048 ± 88 mg; postpartum d 8, 150 ± 8 mg; and postpartum d 22, 228 ± 14 mg.
Materials
Rat recombinant NGF-? was purchased from R&D Systems (Minneapolis, MN). Mouse 2.5S NGF was purchased from Sigma-Aldrich (St. Louis, MO). Human postmortem brain tissue cortex was a kind gift from Dr. V. Gundlade (Department of Pathology, University of Leipzig, Leipzig, Germany). Rabbit anti-mature NGF-? antibody (H-20) was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The anti-NGF-? antibody does not cross-react with other neurotrophins such as brain-derived growth factor (BDGF), NT-3, or NT-4. The antibody to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was obtained from Research Diagnostic (Flanders, NJ). Horseradish peroxidase-labeled goat antirabbit and the Vectastain kit were obtained from Vector Laboratories (Burlingame, CA). Rabbit anti-proNGF raised against the proNGF domain (amino acids –60 to –91, code PH 421) was a kind gift from Dr. Max Reinshagen (University of Ulm, Ulm, Germany). Normal rabbit IgG was obtained from Dako Laboratories (Glostrup, Denmark).
Preparation of soluble proteins
Human brain samples were obtained at autopsy 16 h postmortem from a male patient (47 yr old) who was without any neurological disease. The brain tissue was stored at –80 C until use. Uteri from NP rats and of the 22-d postpartum animals were taken at estrus as confirmed by vaginal smears. Frozen uterine horns and human brain were weighed and crushed into a powder with a pestle and mortar under liquid nitrogen. The frozen powder was homogenized in 1 ml ice-cold homogenization buffer (pH 7.4; 20 mM HEPES, 1 mM EDTA, 0.2 M sucrose, 20 μg/ml soybean trypsin inhibitor, 20 μg/ml leupeptin, 5 μg/ml pepstatin A, 5 mM dithiothreitol, 5 μg/ml E-64, 5 μg/ml bestatin, 5 μg/ml aprotinin, 5 μg/ml antipain, and 0.1 mM phenylmethylsulfonyl fluoride) containing 0.5% sodium dodecyl sulfate (SDS). After homogenization, the samples were centrifuged at 80,000 x g for 1 h at 4 C. The supernatants were designated as soluble protein extracts. The protein concentration was determined using the protein assay based on bicinchoninic acid using bovine albumin as the standard (Pierce, Rockford, IL).
Western blot analysis
Protein samples (30 μg/lane) were separated by SDS-PAGE on 12.5% gels for the detection of mature NGF-? and proNGF. Protein samples were diluted in SDS-PAGE sample buffer and boiled for 5 min before being loaded onto the gels. Recombinant rat NGF-? (20 ng/lane from R&D Systems) and human brain cortex (30 μg/lane) were loaded as positive controls to test for the specificity of the anti-mature NGF-? and proNGF antibodies, respectively. To control for antibody specificity, the anti-NGF-? or proNGF antibodies were absorbed to a 10-fold excess of recombinant NGF-?. Proteins were electroblotted using a semidry transblotter. Antibodies were diluted as follows: anti-mature NGF-? and anti-proNGF (1:4,000) and all secondary antibodies (1:5,000). Protein loading was normalized to GAPDH (1:10,000). The proteins were detected using an enhanced chemiluminescence reagent kit (Amersham Biosciences Inc., Piscataway, NJ). The relative density of the bands was densitometrically analyzed with the AlphaImager 2000 (Alpha Innotech Corp., San Leandro, CA). For the densitometric analysis of mature NGF-? and proNGF, the levels for the NP uterus were arbitrarily set as 100%.
RT-PCR
Total RNA was extracted from frozen homogenized uterine tissue using the Qiagen RNA extraction kit (Rneasy, QIAGEN GmbH, Hilden, Germany). First-strand cDNA was synthesized using Moloney murine leukemia virus reverse transcriptase according to the instructions of Amersham Biosciences Inc. (cDNA synthesis kit) in a 15-μl vol reaction containing 2 μg deoxyribonuclease (DNase)-treated total RNA. The housekeeping gene GAPDH was used as an internal control to determine the relative concentration of the NGF-?. PCR was carried out using the primer pairs derived from the published rat GAPDH sequence (GenInfo no. 204248; forward, 5'-CTCCGCCCCTTCCGCTGATG-3'; and reverse, 5'-GTCCACCACCCTGTTGCTGTAG-3') and primers for NGF-? (GenInfo no. 205691; forward, 5'-CTTCAGCATTCCCTTGACAC-3'; and reverse, 5'-TGAGCACACACACGCAGGC-3'). Amplified PCR products (NGF-?, 592 bp; GAPDH, 615 bp) were size-fractionated by electrophoresis through a 2% agarose gel and stained with ethidium bromide (0.2 μg/ml). The DNA bands were visualized and quantified by densitometry and expressed as the ratio of the densitometric analysis for each target gene normalized by the GAPDH control. cDNA sequencing was performed on PCR-amplified products to confirm their identity.
Immunohistochemical localization of NGF-? and precursor proNGF
Rat uteri after isolation were immediately fixed in 4% paraformaldehyde in PBS for 24 h. After fixation, samples were dehydrated through a graded series of alcohol and cleared in methylbenzoate and xylol before paraffin embedding. Sections (7 μm) were mounted on paper glue-coated glass slides and deparaffinated. Uteri from NP and the 22-d postpartum animals were taken at estrus as confirmed by vaginal smears. To differentiate between connective tissue and smooth muscle cells, the uteri were stained with hematoxylin-eosin and van Gieson’s standard reagents. Endogenous peroxidase activity was quenched using a solution of 3% H2O2 in 10% methanol PBS for 3 min. Sections were washed in PBS and Tris-buffered saline, and nonspecific binding sites were blocked by incubating with 1.5% normal goat serum for 30 min at room temperature. After rapid washing with PBS, sections were incubated with rabbit antiserum to NGF-? (1:400) diluted in PBS containing 0.25% BSA and 0.1% Triton X-100 and incubated at 4 C overnight. Sections were extensively washed and incubated with biotinylated goat-antirabbit IgG (1:400). The bound primary antibodies were visualized by avidin-biotin-peroxidase detection using the Vectastain kit (Vector Laboratories) and 3-amino-9-ethylcarbazole as the color-developing reagent prepared according to the manufacturer’s instructions. For the immunohistochemical localization of proNGF signals, cryostat sections were incubated first with proNGF-specific antibody (1:400) at 4 C overnight. After washing, the sections were incubated with horseradish peroxidase-labeled goat antirabbit IgG (1:400) for 30 min at room temperature. After further washing, sections were incubated for 10 min at room temperature with the Tyramide Signal Amplification System (TSA Biotin System; PerkinElmer, Boston, MA). Bound antibodies were visualized with the Vectastain kit as described above. The sections were embedded in water-soluble glycerine gelatin. To control for antibody specificity, slides were incubated with anti-NGF-? antibody absorbed to 10-fold excess of recombinant NGF-? or with preimmune rabbit IgG serum. In both cases, no immunostaining was detected.
Statistical analysis
Data presented are the mean values from four to five animals ± SEM. Data are representative of three independent experiments. Statistical analysis of the data was performed using a Student’s t test with a Bonferroni correction for analysis of multiple comparisons. The differences were considered significant at P < 0.01.
Results
Specificity of the NGF-? and proNGF antibodies
To analyze the specificity of the anti-mature NGF-? and proNGF antibodies, we used purified mature NGF-? and postmortem human parietal cortex as a positive controls for the mature and proNGF isoforms, respectively. As shown in Fig. 1A, the anti-mature NGF-? antibody recognized both proNGF isoforms with molecular masses of 28–85 kDa in the normal human brain cortex (lane 1) and mature NGF-? with a molecular mass of 13 kDa (lane 2). Mature NGF-? was not detected in the human brain protein extract. The reactivity to the 13-kDa mature NGF-? and to the proNGF bands with molecular masses of 28–85 kDa was eliminated or dramatically reduced after incubation with the murine 2.5S NGF form (Fig. 1B). The anti-proNGF antibody recognized proteins with molecular masses of 28–75 kDa in the human brain extract, and, more importantly, the proNGF antibody did not cross-react with mature NGF-? (Fig. 1C). Thus, this confirms proNGF antibody specificity toward the proNGF isoforms only. Incubation of the proNGF-specific antibodies with purified mouse NGF-? had no effect on the immunoreactivities toward the 32- to 75-kDa bands (Fig. 1D). Incubation with preimmune rabbit serum revealed no specific immunostaining, thus revealing the specificity of the staining (data not shown).
FIG. 1. Specificity of anti-mature NGF-? and proNGF antibodies. A, Anti-mature NGF-? antibodies recognize proNGF isoforms (28–85 kDa) in the human brain cortex (lane 1) and mature NGF-? (lane 2). B, NGF-? and proNGF immunoreactive bands were neutralized after overnight incubation of the mature NGF-? antibody with a 10-fold excess (by weight) of murine 2.5S NGF. C, Anti-proNGF antibodies recognize proNGF bands (28–75 kDa) in the human brain cortex and exhibit no cross-reactivity with mature NGF-? (lane 2). D, ProNGF immunoreactivity toward the 28–75 bands was unchanged after absorption of the proNGF antibody with 2.5S murine NGF.
Expression of mature NGF-? and precursor proNGF in the uterus during pregnancy and postpartum
Using the antiserum against mature NGF-?, we investigated by Western blot analysis whether the expression of mature NGF-? was altered in the soluble protein fraction of uterine extracts during pregnancy and postpartum compared with the NP uterus. A remarkable decrease in the 13-kDa band intensity of NGF-? that comigrated with the recombinant rat NGF-? band (rNGF-?) occurred in the uterus during the course of pregnancy (Fig. 2A, lower arrow). Incubation with the secondary antibody or with control rabbit IgG resulted in no detectable staining of any 13-kDa band (data not shown), demonstrating its specificity. The densitometric analysis of band intensities after protein normalization against GAPDH expression is shown in Fig. 2B. In early pregnancy (d 7), no significant differences in band immunoreactivity were observed compared with the NP uterus (P > 0.01). The NGF-? immunoreactivity declined significantly at d 14 by 35.8 ± 10% and at d 21 by 70 ± 6% (P < 0.001). One day postpartum, a 57 ± 8% decrease in NGF-? was still evident (P < 0.001). Uterine NGF-? immunoreactivity returned to levels comparable with the NP uterus at postpartum d 8 (P > 0.01). Surprisingly, and in contrast to the reduced NGF-? signal, our analysis revealed a striking increase in a 60-kDa band during pregnancy (Fig. 2A, upper arrow), probably representing the precursor proNGF recently described in dorsal root ganglia and sympathetic neurons from rats (23, 27). The antibodies also immunostained a 75-kDa band that did not show any significant variation during middle and late pregnancy.
FIG. 2. Western blot analysis of NGF-? and proNGF in the NP, pregnant, and postpartum rat uterus. Uterine proteins (30 μg/lane) and 30 ng recombinant NGF-? (rNGF-?) were analyzed as described in Materials and Methods. A, Representative Western blot analysis of mature NGF-? (lower arrow) in the NP, pregnant (d 7, 14, and 21), and postpartum uterus (postpartum d 1, 8, and 22). Uteri from the NP and 22-d postpartum animals were taken at the 22-d postpartum animals were taken at estrus as confirmed by vaginal smears. B, Densitometric analysis of mature NGF-? protein expressed as a percentage of the NP uterus (100%) after GAPDH normalization. C, Representative Western blot analysis using proNGF-? antibody demonstrating the accumulation of the 60- and 26- to 32-kDa proNGF forms. D, Densitometrical analysis of proNGF after normalization to GAPDH. Each bar represents the mean ± SEM of each group of rats (n = 4–5 rats/group). Shaded bars, Pregnant; black bars, postpartum. *, P < 0.01; **, P < 0.001; significant differences in NGF-? or proNGF levels from the NP uterus.
Increased proNGF levels in the pregnant uterus
To assess the identity of the main 60-kDa band, we used an antiserum specific to the proNGF domain corresponding to amino acids –60 to –91 of the prosequence position. The reactivity of this proNGF antibody has been previously extensively characterized (24). The prodomain-specific antibody identified the 60-kDa protein band as the precursor proNGF (Fig. 2C). When the proNGF bands were normalized to the corresponding GAPDH intensities (Fig. 2D), a significant 137 ± 10% increase in proNGF in the uterus was observed at d 14, which increased further to 167 ± 3% at term pregnancy (P < 0.001, respectively). Postpartum d 1 proNGF remained increased by 153 ± 7% (P < 0.001). The intensity of the 60-kDa proNGF immunoreactive band returned to levels comparable with the NP uterus at postpartum d 8 (P > 0.01) (Fig. 2D). In addition, the proNGF antibody also identified 26–28 kDa and 32-kDa immunoreactive bands in the pregnant uterus at gestational d 14 and 21 and at postpartum d 1. These 26- to 32-kDa immunoreactive bands were not detected in the NP uterus before gestational d 14 nor after postpartum d 8 when the NGF-? band immunostaining returned to levels comparable or higher than the NP uterus. These results show disturbances in the posttranslational processing of NGF during pregnancy and that these could be responsible for the progressive proNGF accumulation in the uterus.
Ratio of mature NGF-?/proNGF in pregnant and postpartum uterus
Previous Western blot studies have shown that this particular antibody recognizes mature NGF-? and proNGF with similar affinities (24). Because purified proNGF protein is not commercially available to perform a quantitative analysis, we calculated pregnancy-related changes in the ratio of the mature NGF ? to proNGF band immunoreactivities after normalization to GAPDH expression. After compensating for the increase in horn weight during pregnancy, taking into account the mean weight of the uterus on the defined gestational days, the NGF-? to proNGF ratios (Fig. 3) showed a 59% decline at d 14 (ratio 0.82 ± 0.1 vs. 2.01 ± 0.25 in NP; P < 0.01), followed by a 83% decline at d 21 (ratio 0.332 ± 0.14; P < 0.001). The ratio was still significantly decreased by 80% at postpartum d 1 (ratio 0.39 ± 0.13; P < 0.001), returning to values not significantly different from the NP uterus after postpartum d 8 (Fig. 2). Thus, the marked decrease in the ratio of mature NGF-? to proNGF during pregnancy was independent of any increase in uterus weight during gestation.
FIG. 3. Alterations in the ratios of mature NGF-? to proNGF per horn (mgr) throughout pregnancy. The ratios were calculated by dividing the normalized single ratios of NGF-? by the ratios of proNGF, and then the ratios were calculated per uterus horn (mgr). White bars, NP; shaded bars, pregnant; dark bars, postpartum. Bars represent the mean of the respective individual ratios ± SEM. *, P < 0.01; **, P < 0.001; significantly different ratio from the NP uterus.
Semiquantitative RT-PCR analysis of NGF-? mRNA
We analyzed whether the abundance of NGF mRNA was altered in the uterus throughout pregnancy and postpartum compared with the NP uterus. The linearity of the PCR amplification reactions was assessed for each primer pair previously allowing the setting up of conditions under which PCR amplification was in the logarithmic phase (data not shown). No PCR products were obtained when samples were amplified without RT, indicating the absence of genomic DNA. The primers specific for NGF mRNA amplified a PCR product of 592 bp as shown in the NP, pregnant, and postpartum uterus (Fig 4A). The densitometric analysis of the uterine NGF mRNA after normalization to GAPDH mRNA (Fig. 4B) revealed no significant differences in mRNA expression throughout gestation, postpartum vs. the NP uterus (P > 0.1). Similar results were obtained when the RT-PCR was normalized to the mRNA expression of calponin (data not shown).
FIG. 4. Semiquantitative RT-PCR analysis of uterine NGF mRNA during gestation. A, Representative agarose gel of RT-PCR amplification of NGF mRNA from NP, different days of pregnancy and postpartum rat uterus. The primers used to amplify NGF are described in Material and Methods. First-strand cDNAs were made from 2 μg uterine RNA. The numbers on the right side of the gels indicate the sizes of the amplified products. B, Graphic compilation of the PCR data generated for NGF mRNA normalized to the expression of GAPDH. Circles, mean ± SEM (n = 3–4 rats/group). P > 0.01, no significant differences in NGF mRNA from the NP uterus.
Immunohistochemical localization of NGF-? and proNGF
To analyze the temporal and spatial changes in the expression of NGF in the uterus, we performed immunohistochemical analysis using first an antibody against mature NGF-?. In the NP rats, NGF-? immunoreactivity was detected predominantly in the inner and outer myometrial layers and in the smooth muscle cells of the uterine blood vessels (Fig. 5A). In addition, expression of NGF-? was detected in the epithelium of uterine glands as seen in the inset in Fig. 5A. As pregnancy progressed, a major decrease in NGF-? immunostaining occurred in the myometrial layers, the vascular smooth muscle cells, and the glandular epithelium, which was more pronounced at gestational d 21 (Fig. 5B). This coincided with increased NGF immunostaining in some intermyometrial areas and areas around the outer myometrium (arrows) (Fig. 5B). Additional histochemical analysis clearly established that the areas with increased NGF immunoreactivity are connective tissue areas that stained red with the van Gieson stain, in contrast to the yellow-brown stain of the smooth muscle cells (Fig. 5C). After parturition, uterine NGF-? immunostaining remained low, but restoration to an intensity comparable with the NP uterus was evident at postpartum d 22 (Fig. 5D). No immunoreaction was observed in the NP uterus when antiserum was preabsorbed with an excess of NGF-?, demonstrating the specificity of the immunostaining (Fig. 5E).
FIG. 5. Immunohistochemical localization of NGF-? and proNGF in the NP, pregnant, and postpartum uterus using polyclonal anti-mature NGF-? and proNGF antibodies. A, Uterus from NP rats, inner (im), and outer myometrial layers (om), vascular smooth muscle cells (v); uterine glandulae (gl) are NGF-? immunoreactive as shown in the inset in A. B, At d 21, NGF-? immunostaining of the myometrium was nearly abolished in the myometrium but was increased in some intermyometrial areas and in connective tissue areas around the outer myometrium (arrows). C, van Gieson staining of the uterus at d 21 showing connective tissue areas red (arrows), clearly distinguishable from the yellow-brown smooth muscle cells from the inner and outer myometrium. D, NGF-? immunostaining of the postpartum rat uterus at d 22 is comparable with the NP uterus. E, NP uterus incubated with a NGF-? antibody preabsorbed overnight with a 10-fold excess of murine 2.5S NGF showing the specificity of the staining. F, Absence of proNGF immunoreactivity signal in the NP rat uterus after incubation with proNGF-specific antibodies. G, Uterus at term pregnancy (d 21) showing markedly increased proNGF immunoreactivity in some intermyometrial areas and in connective tissue areas around the outer myometrium (arrows). H, Absence of specific staining in the intermyometrial connective tissue areas was observed using rabbit preimmune serum demonstrating the specificity of the staining. I, At postpartum d 22, no specific proNGF immunoreactivity was observed. Scale bars, 100 μm.
Because the anti-mature NGF-? antibody cannot differentiate between the mature and the proNGF isoforms, we sought to differentiate their immunolocalization using the proNGF antibody, which does not recognize mature NGF-?. Immunohistochemical analysis using an enhanced signal amplification system localized proNGF to the intermyometrial and connective tissue areas surrounding the outer myometrium. ProNGF immunoreactivity increased from d 14 onwards (data not shown) and was maximal at gestational d 21 (Fig. 5G; arrows). Furthermore, control experiments using preimmune serum revealed no specific signal in any intermyometrial areas, demonstrating the specificity of the binding and simultaneously demonstrating the absence of nonspecific endogenous peroxidase activity in this area (Fig. 5H). In contrast, proNGF immunostaining was not detectable in the NP uterus (Fig. 5F) or at postpartum d 22 (Fig. 5I). Thus, the increase in NGF immunoreactivity in the intermyometrial tissue area is due to accumulation of proNGF during pregnancy.
Discussion
In this study, we report that in the rat uterus, the expression of mature NGF-? is significantly reduced during middle and late pregnancy and 1 d postpartum compared with the NP uterus. Antibodies to mature NGF-? recognized both mature and proNGF forms in contrast to the proNGF antibody, which recognized only the proNGF isoform. Interestingly, mature NGF-? was not present in the human brain cortex, in agreement with previous studies showing the presence of only proNGF (24). At term, uterine NGF-? protein fell by approximately 75% as demonstrated by Western blot analysis. Immunohistochemistry analysis revealed that mature NGF-? was expressed also in the uterine glands, and its presence might also contribute to this finding. However, because the endometrium is not innervated, endometrial changes in the expression of the NGF isoforms will have little bearing on changes in myometrial innervation.
Our findings strongly suggest that NGF-? produced in the myometrium is a significant regulator of myometrial innervation because the decrease in NGF-? parallels the well-documented axonal degeneration of the myometrial sympathetic nerves (1, 9, 28). Our results contrast to previous reports on the increase in NGF-? protein content in the pregnant uterus based on ELISA that used antisera with undefined reactivities toward the NGF isoforms (10, 13). These studies assumed incorrectly to measure mature NGF-? only, which is not the case as recently demonstrated by the discovery that proNGF instead of mature NGF is the main form in the mammalian brain and in other tissues (24, 27). We demonstrate here that only the unprocessed proNGF forms increased at gestational d 14 to term and 1 d postpartum coinciding, according to the literature, with the axonal degeneration of small sensory fibers and sympathetic axons of the myometrium (1, 9, 28).
Thus, reduced supply of prosurvival mature neurotrophin in the pregnant uterus is linked to its transient neurodegenerative changes. This finding not only agrees with the growing evidence that reduced neurotrophin support leads to neurodegenerative processes of the central nervous system (29) but also extends it to neurodegenerative processes in peripheral organs. However, NGF is produced and secreted by smooth muscle cells, which also express TrkA (18); thus, its decline during pregnancy may result also in alterations of the nonneuronal effects of NGF on smooth muscle cells, which are not completely understood. An autocrine action of NGF on smooth muscle cells includes the induction of chemotaxis without induction of proliferation (18). In animal models of hypertension, hypertrophy and hyperplasia of smooth muscle cells are linked in a time-dependent way to increased NGF secretion measured by ELISA (30, 31). However, in these studies, the reactivity of the antiserum toward the NGF isoforms was not defined. During pregnancy, the myometrium undergoes both hyperplasic and hypertrophic changes, which parallel with an accumulation of proNGF forms as shown here, suggesting that prolonged muscle stretch might influence the processing of NGF in association with altered muscle function.
Using a proNGF-specific antibody that does not recognize mature NGF-?, we demonstrated a marked increase of the 60- and 26- to 32-kDa proNGF forms at d 14 of gestation until term and at 1 d postpartum. A 60-kDa band, representing the unprocessed glycosylated proNGF, and 26- to 32-kDa bands corresponding to the nonglycosylated translation products of rat proNGF were previously detected by this antibody in rat dorsal root ganglia, colon, and spinal cord tissues (23). A 32-kDa precursor proNGF was previously detected in the human brain, which was the main NGF isoform in both normal brain and Alzheimer’s disease brain (24). Furthermore, the proNGF antibody identified a 13-kDa band in the rat uteri with a similar molecular weight to mature NGF-? which arises because rat proNGF contains two dibasic cleavage sites necessary for its processing as previously reported (23, 32).
The accumulation of the proNGF forms in the pregnant uterus might result from either impaired processing of NGF by furin or the proconvertases enzymes (25, 33) or from alterations in the glycosylation/deglycosylation of proNGF. High-molecular weight forms of proNGF are synthesized and secreted by sympathetic neurons, but these proNGF forms do not support neuronal survival (27). In contrast, it was recently reported that proNGF exhibits neurotrophic activity but shows at least a 5-fold less activity than the mature NGF-? (34). Accumulation of proNGF in the brain of Alzheimer’s patients was recently demonstrated and postulated to be due to disturbances in the posttranslational processing of NGF (24) because no difference in NGF mRNA expression between Alzheimer’s patients and control individuals was found (35, 36). In cultured neurons, proNGF binds with high affinity and preferentially to p75NTR and induces p75NTR-dependent apoptosis (25). The secreted 26- to 28-kDa proNGF forms were recently shown to be a pathophysiological death-inducing ligand after brain injury (37). On this basis, it has been suggested that the balance between cell death and survival in neurons may be determined by the ratio of proNGF and mature NGF (26). The data presented here support this idea because the decline in the ratio of mature NGF-? to proNGF during gestation paralleled with the reported degeneration of myometrial and perivascular nerves (1, 9, 28). The recovery of the ratio at d 8 postpartum coincided with the restoration of uterine sympathetic innervation reported previously (1, 9, 28). The increase in proNGF protein in the uterus is probably not related to enhanced transcription because RT-PCR during gestation revealed no increased NGF mRNA, in contrast to the significantly elevated proNGF protein demonstrated at this time. Factors that modulate the expression of mRNA NGF and protein include ovarian hormones (38, 39) and cytokines (40). Long-term estrogen treatment of infantile/prepubertal rats results in a complete loss of the intrauterine sympathetic nerves (41). It is conceivable that the high levels of hormones during pregnancy might induce alterations in the processing of neurotrophins in the uterus.
Immunohistochemical studies identified NGF-? to be predominantly located in the myometrium and in the smooth muscle cells of uterine vessels in agreement with recent studies (38, 41). The decrease in NGF-? immunoreactivity in the uterine blood vessels of the pregnant rat parallels the reported decline of NGF-? in the human uterine arteries of pregnant women (42). The decline in mature prosurvival NGF-? immunostaining was concomitant with an increased NGF-immunoreactivity in intermyometrial connective tissue areas. Using the proNGF-specific antibody, we demonstrated that the increased intermyometrial NGF-immunoreactivity at term pregnancy is due to the accumulation of proNGF and not mature NGF-?. No proNGF immunohistochemistry signal was detected in the NP uterus or in the postpartum uterus (d 22). Epitopes in denatured antigens recognized in Western blot analysis may be different from epitopes on nondenatured antigens recognized in immunohistochemistry; therefore, one cannot expect the same level of sensitivity detection. This appears to be the case because the proNGF immunoreactivity in the pregnant myometrium was visible only after the use of an enhanced detection system.
In summary, we demonstrate for the first time alterations in NGF isoforms during pregnancy, accumulation of proNGF, and decreased ratios of mature NGF-? to proNGF in the pregnant rat uterus coinciding with the transient uterine axonal degeneration of the pregnant myometrium reported by previous studies. Our observations strongly support the idea that under certain circumstances when the balance between a proneurotrophin and mature neurotrophin favors the proneurotrophins, it may cause degeneration of neurons. Additional studies on the biological effects of NGF-?, proNGF, and its intermediates on neuronal and nonneuronal cells will provide insights into the multiple mechanisms involved in the maintenance of relative myometrial quiescence in pregnancy.
Acknowledgments
We thank Dr. John S. Hothersall (University College London, London, UK) for critical reading of the manuscript.
References
Owman C 1981 Pregnancy induces degenerative and regenerative changes in the autonomic innervation of the female reproductive tract. Ciba Found Symp 83:252–279
Sporrong B, Alm P, Owman C, Sjoberg NO, Thorbert G 1981 Pregnancy is associated with extensive adrenergic nerve degeneration in the uterus. An electronmicroscopic study in the guinea-pig. Neuroscience 6:1119–1126
Alm P, Lundberg LM, Wharton J, Polak JM 1988 Organization of the guinea-pig uterine innervation. Distribution of immunoreactivities for different neuronal markers. Effects of chemical- and pregnancy-induced sympathectomy. Histochem J 20:290–300
Moustafa FA 1988 Changes in cholinergic and noradrenergic nerves in the pregnant and postpartum uterus of the albino rat and guinea pig. Acta Anat (Basel) 132:310–316
Wikland M, Lindblom B, Dahlstrom A, Haglid KG 1984 Structural and functional evidence for the denervation of human myometrium during pregnancy. Obstet Gynecol 64:503–509
Morizaki N, Morizaki J, Hayashi RH, Garfield RE 1989 A functional and structural study of the innervation of the human uterus. Am J Obstet Gynecol 160:218–228
Lundberg LM, Alm P, Thorbert G 1989 Local mechanical effects and humoral factors evoke degeneration of guinea pig uterine innervation. Acta Obstet Gynecol Scand 68:487–496
Marshall JM 1981 Effects of ovarian steroids and pregnancy on adrenergic nerves of uterus and oviduct. Am J Physiol 240:C165–C174
Haase EB, Buchman J, Tietz AE, Schramm LP 1997 Pregnancy-induced uterine neuronal degeneration in the rat. Cell Tissue Res 288:293–306
Varol FG, Duchemin AM, Neff NH, Hadjiconstantinou M 2000 Nerve growth factor (NGF) and NGF mRNA change in rat uterus during pregnancy. Neurosci Lett 294:58–62
Korsching S, Thoenen H 1983 Nerve growth factor in sympathetic ganglia and corresponding target organs of the rat: correlation with density of sympathetic innervation. Proc Natl Acad Sci USA 80:3513–3516
Shelton DL, Reichardt LF 1984 Expression of the ?-nerve growth factor gene correlates with the density of sympathetic innervation in effector organs. Proc Natl Acad Sci USA 81:7951–7955
Brauer MM, Shockley KP, Chavez R, Richeri A, Cowen T, Crutcher KA 2000 The role of NGF in pregnancy-induced degeneration and regeneration of sympathetic nerves in the guinea pig uterus. J Auton Nerv Syst 79:19–27
Head RJ 1991 Hypernoradrenergic innervation and vascular smooth muscle hyperplastic change. Blood Vessels 28:173–178
Steers WD, Kolbeck S, Creedon D, Tuttle JB 1991 Nerve growth factor in the urinary bladder of the adult regulates neuronal form and function. J Clin Invest 88:1709–1715
Bibel M, Barde, YA 2000 Neurotrophins: key regulators of cell fate and cell shape in the vertebrate nervous system. Genes Dev 14:2919–2937
Tessarollo L 1998 Pleiotropic functions of neurotrophins in development. Cytokine Growth Factor Rev 9:125–137
Donovan MJ, Miranda RC, Kraemer R, McCaffrey TA, Tessarollo L, Mahadeo D, Sharif S, Kaplan DR, Tsoulfas P, Parada L, Toran-Allerand CD, Hajjar DP, Hempstead B 1995 Neurotrophin and neurotrophin receptors in vascular smooth muscle cells. Regulation of expression in response to injury. Am J Pathol 147:309–324
Fahnestock M, Yu G, Coughlin MD 2004 ProNGF: a neurotrophic or an apoptotic molecule? Prog Brain Res 146:107–110
Saboori AM, Young M 1986 Nerve growth factor: biosynthetic products of the mouse salivary glands. Characterization of stable high molecular weight and 32,000-dalton nerve growth factors. Biochemistry 25:5565–5571
Edwards RH, Selby MJ, Garcia PD, Rutter WJ 1988 Processing of the native nerve growth factor precursor to form biologically active nerve growth factor. J Biol Chem 263:6810–6815
Lakshmanan J, Burns C, Smith RA 1988 Molecular forms of nerve growth factor in mouse submaxillary glands. Biochem Biophys Res Commun 152:1008–1014
Reinshagen M, Geerling I, Eysselein VE, Adler G, Huff KR, Moore GP, Lakshmanan J 2000 Commercial recombinant human beta-nerve growth factor and adult rat dorsal root ganglia contain an identical molecular species of nerve growth factor prohormone. J Neurochem 74:2127–2133
Fahnestock M, Michalski B, Xu B, Coughlin MD 2001 The precursor pro-nerve growth factor is the predominant form of nerve growth factor in brain and is increased in Alzheimer’s disease. Mol Cell Neurosci 18:210–220
Lee R, Kermani P, Teng KK, Hempstead BL 2001 Regulation of cell survival by secreted proneurotrophins. Science 294:1945–1948
Chao MV, Bothwell M 2002 Neurotrophins: to cleave or not to cleave. Neuron 33:9–12
Hasan W, Pedchenko T, Krizsan-Agbas D, Baum L, Smith PG 2003 Sympathetic neurons synthesize and secrete pro-nerve growth factor protein. J Neurobiol 57:38–53
Klukovits A, Gaspar R, Santha P, Jancso G, Falkay G 2002 Functional and histochemical characterization of a uterine adrenergic denervation process in pregnant rats. Biol Reprod 67:1013–1017
Dawbarn D, Allen SJ 2003 Neurotrophins and neurodegeneration. Neuropathol Appl Neurobiol 29:211–230
Clemow DB, Spitsbergen JM, McCarty R, Steers WD, Tuttle JB 1998 Arterial nerve growth factor (NGF) mRNA, protein, and vascular smooth muscle cell NGF secretion in hypertensive and hyperactive rats. Exp Cell Res 244:196–205
Clemow DB, Steers WD, Tuttle JB 2000 Stretch-activated signaling of nerve growth factor secretion in bladder and vascular smooth muscle cells from hypertensive and hyperactive rats. J Cell Physiol 183:289–300
Dicou E, Lee J, Brachet P 1986 Synthesis of nerve growth factor mRNA and precursor protein in the thyroid and parathyroid glands of the rat. Proc Natl Acad Sci USA 83:7084–7088
Seidah NG, Benjannet S, Pareek S, Savaria D, Hamelin J, Goulet B, Laliberte J, Lazure C, Chretien M, Murphy RA 1996 Cellular processing of the nerve growth factor precursor by the mammalian pro-protein convertases. Biochem J 314:951–960
Fahnestock M, Yu G, Michalski B, Mathew S, Colquhoun A, Ross GM, Coughlin MD 2004 The nerve growth factor precursor proNGF exhibits neurotrophic activity but is less active than mature nerve growth factor. J Neurochem 89:581–592
Goedert M, Fine A, Hunt SP, Ullrich A 1986 Nerve growth factor mRNA in peripheral and central rat tissues and in the human central nervous system: lesion effects in the rat brain and levels in Alzheimer’s disease. Brain Res 387:85–92
Jette N, Cole MS, Fahnestock M 1994 NGF mRNA is not decreased in frontal cortex from Alzheimer’s disease patients. Brain Res Mol Brain Res 25:242–250
Harrington AW, Leiner B, Blechschmitt C, Arevalo JC, Lee R, Morl K, Meyer M, Hempstead BL, Yoon SO, Giehl KM 2004 Secreted proNGF is a pathophysiological death-inducing ligand after adult CNS injury. Proc Natl Acad Sci USA 101:6226–6230
Bjorling DE, Beckman M, Clayton MK, Wang ZY 2002 Modulation of nerve growth factor in peripheral organs by estrogen and progesterone. Neuroscience 110:155–167
Chavez-Genaro R, Crutcher K, Viettro L, Richeri A, Coirolo N, Burnstock G, Cowen T, Brauer MM 2002 Differential effects of oestrogen on developing and mature uterine sympathetic nerves. Cell Tissue Res 308:61–73
Sherer TB, Neff PS, Hankins GR, Tuttle JB 1998 Mechanisms of increased NGF production in vascular smooth muscle of the spontaneously hypertensive rat. Exp Cell Res 241:186–193
Chalar C, Richeri A, Viettro L, Chavez-Genaro R, Bianchimano P, Marmol NM, Crutcher K, Burnstock G, Cowen T, Brauer MM 2003 Plasticity in developing rat uterine sensory nerves: the role of NGF and TrkA. Cell Tissue Res 314:191–205
Naves FJ, Vazquez MT, Jose IS, Martinez-Almagro A, Vega JA 1998 Pregnancy-induced denervation of the human uterine artery correlates with local decrease of NGF and TrkA. Ital J Anat Embryol 103:279–290(Edgar Lobos, Claudia Gebh)