| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,2,2
* Department of Clinical Neuroscience, Karolinska Institute, Stockholm, Sweden;
Department of Neurology and Neuroscience, Weill Medical College of Cornell University, New York, New York, USA;
Department of Cell Biology, University of Oklahoma Health Science Center, Oklahoma City, Oklahoma, USA;
Department of Anatomy and Neurobiology, University of Kentucky College of Medicine, Lexington, Kentucky, USA; and
|| Department of Morphological Brain Science, Kyoto University, Kyoto, Japan
3Correspondence: Experimental Alcohol and Drug Addiction Research Section, CMM L8:01, Department of Clinical Neuroscience, Karolinska Institute, S-171 76 Stockholm, Sweden. E-mail: georgy.bakalkin{at}ki.se
ABSTRACT
The classical view postulates that neuropeptide precursors in neurons are processed into mature neuropeptides in the somatic trans-Golgi network (TGN) and in secretory vesicles during axonal transport. Here we show that prodynorphin (PDYN), precursor to dynorphin opioid peptides, is predominantly located in axon terminals and dendrites in hippocampal and striatal neurons. The molar content of unprocessed PDYN was much greater than that of dynorphin peptides in axon terminals of PDYN-containing neurons projecting to the CA3 region of the hippocampus and in the striatal projections to the ventral tegmental area. Electron microscopy showed coexistence of PDYN and dynorphins in the same axon terminals with occasional codistribution in individual dense core vesicles. Thus, the precursor protein is apparently stored at presynaptic sites. In comparison with the hippocampus and striatum, PDYN and dynorphins were more equally distributed between neuronal somata and processes in the amygdala and cerebral cortex, suggesting regional differences in the regulation of trafficking and processing of the precursor protein. Potassium-induced depolarization activated PDYN processing and secretion of opioid peptides in neuronal cultures and in a model cell line. Regulation of PDYN storage and processing at synapses by neuronal activity or extracellular stimuli may provide a local mechanism for regulation of synaptic transmission. —Yakovleva, T., Bazov, I., Cebers, G., Marinova, Z., Hara, H., Ahmed, A., Vlaskovska, M., Johansson, B., Hochgeschwender, U., Singh, I. N., Bruce-Keller, A. J., Hurd, Y. L., Kaneko, T., Terenius, L., Ekström, T. J., Hauser, K. F., Pickel, V. M., Bakalkin, G. Prodynorphin storage and processing in axon terminals and dendrites
Key Words: neuropeptides • dynorphin • plasticity
NEUROPEPTIDES ARE SYNTHESIZED as large precursor molecules, which translocate into the endoplasmic reticulum (ER) where they fold and undergo posttranslational modifications (1
2
3
4
5)
. Folded and assembled proteins are transported to the Golgi apparatus and then targeted to the regulated secretory pathway (4
, 6
7
8)
. In endocrine cells, proteolytic processing of neuropeptide precursors is initiated by processing convertases (PCs) in the trans-Golgi network (TGN) and continues in the secretory vesicles (3
4
5
, 6
7
8)
. In neurons, processing of neuropeptide precursors is thought to take place during transport from the perikaryon to nerve terminals. This view is based on early studies demonstrating that the conversion of the neurophysin precursor to neurophysin occurs intra-axonally during transport from the hypothalamus to the neurohypophysis (9
, 10)
. More recent in vitro work (11
12
13
14
15)
supports this hypothesis.
The endogenous opioid peptides, dynorphins, enkephalins, and endorphins, mediate or modulate synaptic transmission. The dynorphins, ligands for 
-opioid receptors, have a wide distribution in the brain including the basal ganglia, striatum, hippocampus, cerebral cortex, and spinal cord (16
17
18
19
20)
. Dynorphin A (Dyn A), dynorphin B (Dyn B), and 
-neoendorphin are generated by cleavage of precursor protein PDYN by PC1 and PC2 (21
22
23
24)
. Dynorphins are involved in modulation of reward induced by intake of addictive substances, memory acquisition, and pain processing (25
26
27
28
29)
.
Several cellular processes critical for function of neuropeptides have not yet been characterized. They include regulation of storage and processing of neuropeptide precursors and their targeting to axon terminals and dendrites. Here we focus on the PDYN system as a model to study the biogenesis of neuropeptide precursors in the brain. We characterized biochemical properties of PDYN and identified the subcellular sites in neurons where this protein is stored and processed to mature peptides. In addition, we found that depolarization that induces dynorphin secretion also stimulates PDYN processing using a model cell line and primary neuronal cultures.
MATERIALS AND METHODS
Plasmids and recombinant PDYN
The 970bp KpnI–XhoI fragment of pRcCMV-rat-PDYN plasmid with full-length PDYN cDNA was cloned into PCB6+-vector, giving rise to pCMV-r-PDYN. To produce recombinant rat PDYN, DNA fragment coding rat PDYN23–248 was cloned into BamHI- and PstI(pol)-sites of pQE30 (Qiagen, Germany), and the His-tagged PDYN was expressed in E. coli expression system and purified over a Ni–nitrolotriacetic acid-agarose column (Qiagen, Germany). Sequences were verified by nucleotide sequencing.
Experimental animals/brain dissection
Ventral striatum (nucleus accumbens and adjacent caudate-putamen), ventral tegmental area (VTA)/substantia nigra, dentate gyrus, and CA3 regions of hippocampus were dissected at –10°C from frozen sections of brains of male Sprague-Dawley rats (500–700 g) using either a small scalpel blade or punching with a pipette tip. The border of the brain structure to be dissected was outlined by visually comparing each brain section under strong fluorescent illumination (Osram Dulux S 11W 21) with the relevant plate of the Swanson rat brain atlas (30)
. Kainic acid (10 mg/kg ip) was administered to male Sprague-Dawley rats (250–275 g) euthanized by CO2 overdose 30 and 90 min after injection.
Cell/tissue extracts
Buffers were supplemented with 5x protease (Complete, Boehringer Mannheim, Mannheim, Germany) and proteasome (5 µM MG132; Calbiochem, Darmstadt, Germany) inhibitors, and procedures were carried out on ice. MIN6 cells were exposed for 1 h on ice in the dark to 100 mM iodoacetamide in PBS before harvesting to prevent disulfide interchange. Iodoacetamide solution was freshly prepared and kept shortly in dark tubes for each experiment. Tissues were extracted with buffer C (31)
(20 mM HEPES, pH 7.9; 0.42 M NaCl, 25% glycerol, 1.5 mM MgCl2, 0.4 mM EDTA, 0.5 mM DTT, and 0.2% Nonidet P-40), 1% Triton X-100–10 mM Tris-HCl buffer, pH 8.0 (TB), 1 M acetic acid, or SDS-buffer (0.45 M Tris-HCl, pH 8.5, 2.5% glycerol, and 4% SDS). Tissue homogenates in buffer C were mixed with iodoacetamide (10 mM final) and centrifuged at 20,000 g for 10 min, and the supernatant was kept at –80°C. Homogenates prepared in four volumes of TB were incubated for 15 min, insoluble material collected by centrifugation at 20,000 g for 10 min was solubilized in SDS buffer and boiled for 5 min, and DNA was sheared by ultrasonic irradiation. For acetic acid extractions, tissues were boiled for 5 min in 1 M acetic acid supplied with 0.5 mM DTT, kept on ice for 5 min, ultrasonicated for 2 min, mixed with iodoacetamide (10 mM final), and centrifuged at 20,000 g for 15 min, and the supernatant was loaded onto a Sep-Pac C18 reverse-phase columns (Waters Co., Milford, MA). For SDS extraction, tissue samples combined with 4 volumes (w/v) of SDS buffer, containing 50 mM DTT and protease and proteasome inhibitors and heated up to 95°C, were boiled for 5 min, homogenized, and ultrasonicated for 2 min. Protein concentration was determined with the Micro bicinchoninic acid (BCA; Pierce, Rockford, IL; SDS-containing solutions) or Bio-Rad (Bio-Rad Laboratories, Hercules, CA) protein assay.
Anti-PDYN antibodies
Rabbit polyclonal antisera were generated against five rat PDYN peptides (PDYN31–45, PDYN140–159, PDYN CTF235–248, Dyn A1–17, and Dyn B1–13) all conjugated with keyhole limpet hemocyanin via Cys added to their N terminus, as well as against recombinant rat PDYN (Suppl. Fig. 1). IgG fractions were purified with Protein A-Sepharose, and antibodies were characterized by Western blot (WB) and immunocytochemistry (IC; Suppl. Fig. 1; see also ref 32
and Results). Affinity-purified anti-rat PDYN C-terminal fragment 235–248 antibodies demonstrated high selectivity and were used for WB and IC experiments, and anti-recombinant PDYN23–248 antibodies were used for IC. For PDYN and Dyn A dual-labeling, a guinea pig anti-rat-PDYN antiserum extensively characterized in dot-blot and IC by Lee et al. (33)
and rabbit anti-Dyn A antibodies (IgG fraction) were used for immunoperoxidase and immunogold labeling, respectively. Anti-Dyn A antibodies (84+) did not react with PDYN in WB and IC with HeLa cells transected with pCMV-r-PDYN plasmid and with MIN6 cells expressing this protein used as a positive control.
Western blotting
Aliquots of cell/tissue extracts were resolved by SDS-PAGE on 10% Tricine 16 cm long gels under reducing condition; samples were heated for 5 min at 95°C in the presence of 50 mM DTT and then alkylated with 200 mM iodoacetamide at room temperature. Extracts of MIN6 cells and pituitary tissue were also analyzed under nonreducing condition in which DTT was omitted. Proteins were transferred at 4°C onto a nitrocellulose membrane (Schleicher and Schuell, Dassel, Germany), which then was blocked in 5% nonfat dry milk, incubated with primary antibodies to PDYN anti-rCTF (1/600) or BiP (1/350; BD Transduction Laboratories, Lexington, KY), and developed with the enhanced chemiluminescence (ECL) detection system (Amersham, Little Chalfont, UK).
RIA
The procedure and antidynorphin antibodies were described elsewhere (34
, 35)
. Briefly, cells/tissue extracts in 1 M acetic acid and cell culture medium mixed with acetic acid were run through a SP-Sephadex ion exchange C-25 column, and peptides were eluted and analyzed by RIA. Anti-Dyn A antibodies demonstrated 100% molar crossreactivity with Dyn A (9
10
11
12
13
14
15
16
17)
and <0.1% molar crossreactivity with Dyn B, Dyn A(1–8), -neoendorphin, Leu-enkephalin, and big dynorphin. Anti-Dyn B antiserum showed 100% molar crossreactivity with big dynorphin, 0.8% crossreactivity with Leu-morphine (29 aa C-terminally extended Dyn B), and <0.1% crossreactivity with Dyn A(1–17), Dyn A(1–8), -neoendorphin, and Leu-enkephalin. Dyn A, Dyn B, and Leu-enkephalin-Arg RIA readily detected these peptides in brain tissues of wild-type (WT) mice (36)
and rats (34)
, whereas it did not detect any peptide immunoreactivity in the striatum, hippocampus, and frontal cerebral cortex of PDYN knockout mice used as negative controls (for details, see ref 35
), demonstrating high specificity of this method and the absence of interference with other peptides present in tissue extracts.
Measurement of molar PDYN content
To prepare the PDYN standard solution, SDS extracts of rat pituitary tissue were resolved on 10% Tricin-SDS gels, gel area (26–30 kDa) with PDYN band was excised, and proteins were extracted with 100 mM NaCl buffer. Molar content (Fig. 1
) of PDYN in the preparation was determined using the procedure adapted from Lewis et al. (37)
and Nyberg et al. (38)
and based on trypsin digestion followed by Leu-enkephalin-Arg RIA. Aliquots of the PDYN standard solution were incubated with immobilized TPCK-trypsin (Pierce) in 0.1 M NH4HCO3 buffer, pH 8.0 for 16 h at 37°C to liberate Leu-enkephalin-Arg, a marker of PDYN (38)
. Reaction was stopped by addition of acetic acid (1 M final) and incubation at 95°C for 5 min. After centrifugation, the mixture was subjected to SP-Sephadex C-25 chromatography (34)
and Leu-enkephalin-Arg was measured by Leu-enkephalin-Arg RIA. Efficacy of PDYN digestion was 100% as determined in WB experiments using anti-rCTF PDYN antibodies. Recovery of synthetic Leu-enkephalin-Arg from the reaction mixture with either PDYN or trypsin was 86%, and this value was used for calculations of molar PDYN content in the standard solution. Trypsin or PDYN alone did not produce any signal in Leu-enkephalin-Arg RIA. PDYN content was calculated taking into consideration that a PDYN molecule contains three Leu-enkephalin-Arg sequences. SDS extracts of rat tissues and an aliquot of the PDYN standard solution were analyzed using 10% Tricin-SDS gel. Loading of tissue extracts was controlled using MemCod Reversible Protein Stain Kit (Pierce). Intensities of PDYN bands were measured, and the content of PDYN in brain strictures was calculated in fmol/mg protein.
|
Light microscopic immunocytochemistry
Immunocytochemistry on brain sections was performed with anti-rCTF (1:3000), anti-rec PDYN (1:5000), or anti-dynorphin A (1:3000) antibodies using standard peroxidase-based method (EnVision, DAKO, Sweden). Controls included prebleed IgG fractions instead of primary antibody (Ab) and anti-rCTF antibodies blocked with 10 nM of antigenic peptide.
Electron microscopy
The brain tissue of male Sprague-Dawley rats was fixed by aortic arch perfusion with 3.8% acrolein (Polysciences, Warrington, PA) in a solution of 2% paraformaldehyde in 0.1 M phosphate buffer, sectioned at Vibratome at 40–50 µm thickness, and processed for immunoperoxidase labeling using an established protocol with the avidin-biotin complex (ABC) method (39)
. For this, the sections were incubated for 48 h at 4°C with anti-rCTF antibodies (1:500), followed by secondary immunoreagents and final processing for electron microscopy as described by Leranth and Pickel (40)
and Chan et al. (41)
. The hippocampal regions and the accumbens shell were identified using the rat brain atlas of Paxinos and Watson (42)
, and the immunolabeled profiles were classified according to the criteria from Peters et al. (43)
.
For PDYN and Dyn A dual-labeling, a guinea pig anti-rat-PDYN antiserum (33)
and rabbit anti-Dyn A antibodies (IgG fraction) were used for immunoperoxidase and immunogold labeling, respectively. Sections were blocked in 0.8% BSA and 0.1% gelatin in PBS for 10 min and subsequently incubated for 2 h in a 1:50 dilution of anti-rabbit (for anti-Dyn A) IgG conjugated with 1 nm colloidal gold (Amersham, Little Chalfont, UK). The sections were placed in 2% glutaraldehyde for 10 min to ensure the adherence of the bound colloidal gold. The gold particles were silver-enhanced using the IntenS-EM kit (Amersham) for 7 min at room temperature. The sections were fixed in 2% osmium tetroxide in 0.1 M PB for 60 min, followed by dehydration with a graded series of ethanols and propylene oxide. These sections then were incubated overnight in 1:1 mixture of propylene oxide and Epon (Electron Microscopy Science, Fort Washington, PA). The sections were transferred to 100% Epon for 2–3 h and then flat-embedded between sheets of Aclar plastic.
PDYN mRNA studies
Brains of male Sprague-Dawley rats were frozen and cut into coronal sections stored at –30°C. In situ hybridization was carried out under established conditions for studying PDYN mRNA expression levels in the rat and human brain (44
, 45)
.
Cell culture, transfection, and secretion assay
Cells cultured in Iscove’s medium with 10% FBS were transfected with DNA-lipofectamin mixture in OPTI-MEM medium. MIN6 cells (46)
were cultured in Dulbecco’s modified Eagle’s medium (DMEM), 10% heat-inactivated FBS, 0.1% ß-mercaptoethanol, and 20 mM glucose (Glc). To measure secretion, MIN6 cells were incubated in a modified Krebs-Ringer solution (125 mM NaCl, 5.9 mM KCl, 1.28 mM CaCl2, 1.2 mM MgCl2, 25 mM HEPES, 0.1% BSA, pH 7.4, and 3 mM Glc; basal conditions) or in the same buffer with 25 mM KCl (stimulated conditions) for 20 min (47)
. Acetic acid was added to cells or culture medium to a 1 M final concentration, and peptides were analyzed by RIA.
Cerebral cortical cultures were prepared from Sprague-Dawley rats (17–19 day gestation), and secretion assay was performed as described earlier (48)
. Briefly, cultures cultivated for 14–20 days were incubated for 0.5 or 2 h with Locke 5 mM K+, containing (mM) 149 NaCl, 5 KCl, 1 MgCl2, 4 NaHCO3, 10 HEPES, 2.3 CaCl2, 5 Glc, pH 7.5, or Locke 90 mM K+. Culture medium mixed with acetic acid (1 M final) and cell extracts in 1 M acetic acid were run through SP-Sephadex ion-exchange C-25 column and analyzed by RIA. To measure the levels of PDYN and PDYN processing intermediates, Leu-enkephalin-Arg was analyzed in cell extracts, prepared with 1 M acetic acid, lyophilized, and treated with trypsin (34
, 38)
.
RESULTS
Biochemical characterization of PDYN in rat brain
PDYN, similarly to other neuropeptide precursor proteins, is synthesized at low levels in the central nervous system (CNS) and other tissues and has not yet been characterized. To identify PDYN, rabbit polyclonal antibodies against six epitopes of this protein were characterized by WB and IC (Figs. 2
and 3
; Supplemental Fig. 1; see also ref 32
). Affinity-purified anti-rat PDYN C-terminal fragment 235–248 antibodies (anti-rCTF) with high specificity were selected for analysis of PDYN by WB (Fig. 2)
and IC (Figs. 3
and 4
). Anti-recombinant rat PDYN23–248 antibodies (anti-rec-PDYN) with epitope residing in PDYN111–158 labeled well PDYN in cells transfected with PDYN-expression plasmid (Supplementary Fig. 1C) and were used for IC (Fig. 3)
. PDYN epitopes recognized by these two antibodies do not overlap with sequences of mature opioid peptides.
|
|
|
PDYN was extracted from cell and tissue samples by 1) the high ionic strength buffer C with nonionic detergent Nonidet P-40 (Fig. 2A
), 2) 1 M acetic acid to prevent protein degradation (Fig. 2B and C
), and 3
) 2% SDS buffer for complete extraction (Fig. 2D
). Acetic acid extracts were concentrated on Sep-Pac C18 reverse-phase columns, and PDYN was eluted with 60% acetonitrile (Fig. 2B
, lane 2). A dominant band with an apparent molecular size of 28 kDa identical to the calculated PDYN molecular size was detected in buffer C extracts of HeLa cells transfected with pCMV-PDYN plasmid, mouse pancreatic ß-cell line MIN6 cells, and rat pituitary gland extracted with buffer C (Fig. 2A
) and also in the pituitary gland, hippocampal dentate gyrus, and striatum after acidic extraction and chromatography (Fig. 2B and C
). Both MIN6 cells and rat pituitary express PDYN at high levels (22
, 49)
. The 28 kDa band was absent in the pituitary of PDYN knockout mice (Suppl. Fig. 2A). In addition to 28 kDa PDYN, a lower 10 kDa band was observed in the rat pituitary gland. Both bands were blocked by preincubation of antibodies with rCTF but not with the structurally similar CTF of human PDYN. The 10 kDa band apparently corresponds to the abundant 10 kDa C-terminal PDYN processing intermediate previously identified in the pituitary gland (22)
. The ratio of 28 to 10 kDa bands varied from 10/1 to 0.5/1 between extract preparations. SDS extraction yielded twice as much PDYN compared with other extraction methods (Suppl. Fig. 2B and C) and demonstrated high PDYN levels in the rat pituitary gland; intermediate levels in the striatum, hippocampus, cortex, and spinal cord; and absence of PDYN in the cerebellum (Fig. 2D
). This pattern is consistent with the distribution of PDYN mRNA and peptides in the brain (16
17
18
19
20)
. C-terminal PDYN processing intermediates were not evident in brain structures, suggesting that PDYN is either rapidly processed to opioid peptides or is stored in unprocessed form in neurons.
Higher PDYN recovery in SDS buffer suggested low solubility of this protein. Indeed, after four consecutive extractions of striatal tissue with 1% TB, a substantial amount of PDYN was found in the insoluble fraction while the molecular chaperon BiP located in the ER was successfully extracted (Fig. 2E
). The 28 kDa insoluble PDYN was absent in the pituitary gland of PDYN knockout mice (Suppl. Fig. 2D; ref 50
). Six Cys residues remain at the PDYN N terminus after cleavage of a 21 aa-long signal peptide as predicted by the method of von Heijne (www.cbs.dtu.dk/services/SignalP; ref 51
). These cysteines may form intermolecular disulphide bonds between PDYN molecules producing PDYN oligomers. Indeed, high molecular PDYN complexes were identified in the insoluble fractions of MIN6 cells and also in the rat pituitary gland analyzed by PAGE under nonreducing conditions (Fig. 2F
, lanes 1 and 3). These complexes dissociated to PDYN monomers in the presence of the reducing agent DTT (lanes 2 and 4). Thus, a substantial fraction of insoluble PDYN may consist of oligomers stabilized by disulphide bonds.
Regional variation in PDYN distribution patterns
Light microscopic examination demonstrated identical overall distribution patterns of PDYN immunoreactivity (PDYN-IR) produced with anti-rCTR and anti-recPDYN antibodies in the rat brain (Fig. 3)
. No labeling was evident with prebleed IgG fractions of either anti-PDYN antibodies or with anti-rCTF antibodies preincubated with 10 nM of antigenic peptide. PDYN distribution was generally consistent with previous analyses of dynorphins with high levels in the amygdala, hippocampus, and striatum and lower amounts in the cerebral cortex (16
17
18
19
20)
. The PDYN-IR distribution patterns showed differences between these four structures (Fig. 3B
) and may be roughly classified into four general categories: 1) IR limited to the cell body, 2) IR present in the cell body and in the adjacent processes (presumably dendrites), 3) IR limited to axonal fibers, and 4) diffuse stippled labeling that resembles axon terminal-like puncta (16)
. Neurons in the cerebral cortex displayed PDYN-IR mostly in cell bodies with little labeling in the fibers or axonal puncta that prevailed in distinct regions of hippocampus. Neurons in the central nucleus of amygdala displayed equally strong cell body and fiber labeling. Hippocampus displayed PDYN-IR in hilus and stratum lacunolum-moleculare of the CA3 region, with no light microscopic detection of IR in the CA1 region. Occasionally, PDYN-labeled cell bodies were detected in the granule cell layer in the ventral hippocampus. A mixed pattern of light microscopic PDYN-IR was seen in the nucleus accumbens core and shell regions. The PDYN-IR ventral striatal fibers might represent axons of neurons projecting from the nucleus accumbens to the VTA and substantia nigra (20)
. The dorsal striatum (Fig. 3A
) had a pattern of PDYN-IR similar to the nucleus accumbens. The PDYN-IR pattern was nearly identical to that revealed for Dyn A (Fig. 3B
). Anti-Dyn A (1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17)
antibodies used in this analysis did not cross-react with PDYN, as was evident from the control cell transfection experiments (data not shown). The presence of PDYN-IR in brain structures such as the CA3 region in the hippocampus and the VTA, which lack PDYN-containing neuronal cell bodies (16
17
18
19
20
, 52)
, suggests that PDYN is transported to and stored in axon terminals.
Ultrastructural localization of PDYN and dynorphins
Electron microscopic immunocytochemistry showed PDYN-IR in many somatodendritic profiles in rat hippocampal dentate gyrus and CA3 areas and in the nucleus accumbens shell. The hippocampal PDYN-IR somata were identified as granule cells based on their confinement to the granule cell layer, their small size (10–12 µm), round shape, and one visible apical dendrite extending into the molecular layer. Within these cells, PDYN-IR was primarily associated with vesicular organelles resembling dense core vesicles (Fig. 4A
). Occasionally, the PDYN-labeled granules were affiliated with the Golgi apparatus, as was also seen in PDYN-IR medium spiny neurons of the ventral striatum (data not shown). In both these regions, PDYN-IR also was localized to dense cytoplasmic granules or showed a diffuse cytoplasmic distribution in dendritic profiles including spines. The distribution of PDYN-IR in larger dendrites and spines of the dentate molecular layer is seen in Fig. 4B,C
.
PDYN-IR was exclusively seen in small axons and axon terminals in the hilus and supragranular regions of the dentate gyrus and in stratum lucidum of CA3. These terminals contained one to five dense core vesicles and numerous small synaptic vesicles. Within the hilus and CA3, the majority of PDYN-labeled axonal profiles were mossy fiber terminals (19)
, since they were large (>1.0 µm in diameter) and formed synapses with multiple dendritic spines as well as dendritic shafts (Fig. 4D
). Since PDYN was present in VTA in both unprocessed and processed forms (see the next sections), PDYN localization was further analyzed in the nucleus accumbens. Axonal profiles also were frequently labeled for PDYN in the nucleus accumbens shell, where their distribution was similar to that of the somata containing PDYN-IR. The terminals were substantially smaller than those in the hippocampal formation but still contained a mixed population of large dense core and small clear vesicles. In dual labeling experiments, the PDYN was shown to coexist with dynorphin A in the same axon terminal, where there was occasional codistribution in individual dense core vesicles (Fig. 4E
).
PDYN mRNA-PDYN mismatch in the CA3 subfield of the hippocampus and VTA
In the rat hippocampus, in situ hybridization histochemistry detected intense expression of the PDYN gene in the dentate gyrus and much lower expression in the CA region (Fig. 5
A). In the rat striatum, the hybridization signal was intense with the highest levels in the nucleus accumbens (Fig. 5B,C
). In the VTA, which receives projections from striatal dynorphin-containing neurons, PDYN mRNA was not detectable. The patterns identified are consistent with previous publications (18
, 52)
. The negligible levels of PDYN mRNA in the CA3 subfield of the hippocampus and VTA largely exclude the possibility of a local synthesis of the precursor at these sites. The PDYN mRNA-PDYN mismatch can underlie previously reported differences in localization of precursor mRNA and mature dynorphins (18)
.
|
PDYN/dynorphins ratio in axon terminals
We next evaluated whether PDYN is present in the CA3 area and VTA in unprocessed or processed forms and assessed the proportion of unprocessed PDYN compared with mature peptides in these structures. PDYN and dynorphin content was measured using the combination of RIA and WB with trypsin digestion (Fig. 1)
. PDYN was partially purified from rat pituitary tissue by PAGE and extracted from the gel, and protein aliquots were used as a standard in WB experiments. Molar concentration of PDYN in a preparation was determined by protein digestion with trypsin that liberates Leu-enkephalin-Arg, a PDYN marker (38)
, followed by SP-Sephadex C-25 chromatography and detection of Leu-enkephalin-Arg by RIA. SDS extracts of brain structures along with an aliquot of PDYN standard solution were analyzed by WB, and molar content of this protein was calculated (Fig. 6
A,B). Dynorphins were extracted from aliquots of the same tissue preparations that were used to analyze PDYN, subjected to SP-Sephadex C-25 chromatography, and analyzed by RIA.
|
The molar content of PDYN substantially exceeded levels of Dyn A, Dyn B, and Leu-enkephalin-Arg by 2- to 10-fold in structures where PDYN is synthesized (the dentate gyrus and striatum) and in areas receiving projections of PDYN-containing neurons (the CA3 area and VTA; Fig. 6 A,B
). Only in the dentate gyrus content of PDYN was similar to that of Dyn A. No PDYN C-terminal processing intermediates were evident in the dentate gyrus, CA3 area, and striatum, while a 10 kDa band with the intensity similar to that of the 28 kDa PDYN band was observed in the VTA. Both 28 and 10 kDa PDYN bands were blocked by preincubation of antibodies with rCTF but not with hCTF (Fig. 6A
).
Regulation of PDYN processing
PDYN processing and dynorphin secretion may be regulated in concert or differentially in axon terminals. Short-lasting, 90 min stimulation of neuronal activity by kainic acid (10 mg/kg ip) decreased the levels of Dyn A, Dyn B, and Leu-enkephalin-Arg approximately by half in the CA3 area, whereas PDYN content remained unchanged (Fig. 6C
). No changes were observed 30 min after kainic acid injection (data not shown). Neuronal degeneration was first seen in the hippocampus at 4 h but not earlier after kainic acid injection (10 mg/kg ip; ref 53
) and, therefore, is not relevant for the observed changes. The peptide decrease may occur due to stimulation of release and degradation of opioid peptides or inhibition of PDYN processing by kainic acid. Regardless of the mechanism, these data suggest that PDYN processing and dynorphin secretion are differentially regulated in axon terminals.
PDYN processing at the sites of its location may be regulated by cell depolarization that also induces the release of mature neuropeptides. Incubation of model cell line MIN6 cells that produce PDYN with a depolarizing potassium concentration for 20 min stimulated secretion of Dyn A (Fig. 7
) and Dyn B (data not shown). The total levels of Dyn A in cells and medium were increased, demonstrating that PDYN processing was activated by depolarization. No significant changes in PDYN levels in cells were induced (data not shown), suggesting that a small PDYN fraction is processed to dynorphins on stimulation. Cultures of embryonic cortical neurons producing dynorphins have been previously established as a model system to study dynorphin secretion particularly induced by neuronal depolarization (48)
. Incubation of cortical neurons for 0.5 or 2 h in the presence of depolarizing potassium concentrations stimulated release into the medium of Dyn B, Leu-enkephalin-Arg (by 3.5- to 6-fold after 2 h incubation; Table 1
) and Dyn A (by 1.8-fold; data not shown), whereas peptide levels in cells were not affected. The total levels of PDYN-derived peptides in cells + medium were increased by 3-fold demonstrating that PDYN processing was activated by K+ depolarization (Table 1)
. Leu-enkephalin-Arg was the dominant mature peptide in neuronal culture, suggesting that its levels were elevated due to processing of PDYN or PDYN long fragments but not due to conversion from dynorphins. The cellular levels of trypsin-liberated Leu-enkephalin-Arg were 22–53-fold greater than those of free Leu-enkephalin-Arg (Table 1)
, demonstrating that PDYN products are present in neurons predominantly in unprocessed or partially processed forms.
|
|
DISCUSSION
Neuropeptide precursors have long been thought to be processed into short functional fragments in secretory granules during transport from neuronal soma to axon terminals (1
, 9
, 10)
. In the present study, we provide pivotal evidence that PDYN along with dynorphin opioid peptides is present in axon terminals in rat brain. We also show that in the CA3 and VTA regions containing PDYN terminals largely arising from neurons in the dentate gyrus and ventral striatum, the molar content of PDYN is much greater than that of dynorphins. We hypothesize that PDYN storage and processing at synapses by neuronal activity or extracellular stimuli may provide a local mechanism for regulation of synaptic transmission (Fig. 8
). Consistently, throughout the brain, dense networks of axon terminals are immunopositive for the processing enzymes PC1 and PC2 (54)
.
|
PDYN-IR was also observed in dense granular aggregates in dendrites as well as somata of dentate granule cells and striatal neurons, where ultrastructural dual labeling showed coexpression with dynorphin peptides (the present and companion report ref 32
). These results are consistent with prior evidence that neuropeptides are present in the dendritic tree of many neurons (55
56
57)
. Dendritic secretion of dynorphins, oxytocin, and vasopressin appears to be an important mechanism in the local regulation of synaptic transmission (55
, 58
, 59)
. These data along with the presence of PDYN in dendrites in the hippocampus, striatum, and amygdala suggest that targeting of neuropeptide precursors to dendrites and their storage and processing in this compartment represent a general mechanism within the brain.
A fraction of PDYN is apparently present in neurons in insoluble form that may consist of oligomers stabilized by disulfide bridges. The disulfide bonds are essential for the assembly and oligomerization of secretory proteins and are a prerequisite for their intracellular transport (4
, 6
, 8
, 60
, 61)
. PDYN oligomers may also represent the storage form of the protein. The concentration of condensed proteins in dense core vesicles was estimated to be 
42 mM (8)
and is much higher than estimates for dynorphins, which are 1 mM (62)
. Thus, >10-fold amplification of the signal is possible if PDYN is stored in a condensed form and is processed to opioid peptides when required (Fig. 8)
.
Several PDYN processing intermediates have been identified in the rat pituitary gland (22)
. One of them, 10 kDa PDYN C-terminal fragment, was also found in the VTA in the present study. WB analysis with anti-Dyn antibodies demonstrated that levels of PDYN processing intermediates were low in brain structures (not shown); further analysis of these fragments was hindered by insufficient sensitivity of mass-spectrometry. The mere presence of such intermediates would substantiate the conclusion on the localization and storage of unprocessed or partially processed PDYN forms in axon terminals and dendrites.
Depolarization-induced activation of PDYN processing was observed in embryonic neuronal cultures and a model cell line. This may occur due to activation of either synthesis or activity of PC2 or decondensation of the granule content in response to depolarization. With respect to the association of processing and secretion, PDYN may be similar to proatrial natriuretic factor, the cleavage and secretion of which are concomitantly activated by extracellular stimuli (63)
. Storage of PDYN in axon terminals and dendrites and stimulation of its processing by neuronal activity or extracellular signals may represent a novel mechanism for the local regulation of synaptic transmission. It is important to elucidate whether the accumulation of neuropeptide precursors in nerve terminals and dendrites is a general phenomenon and to establish mechanisms of regulation of neuropeptide processing by synaptic activity.
ACKNOWLEDGMENTS
We thank Drs. T. A. Milner and E. Colago for participation in the electron microscopy studies and valuable discussion; Drs. N. Pasikova, I. Usynin, A. Nikoshkov, I. Gileva, and A. Matskevitch for help in immunostaining/RIA experiments and generation and characterization of antibodies; Dr. L. Devi for rat PDYN cDNA; Dr. J. Miyazaki for MIN6 cells; and Drs. N. E. Tabori, B. Berthelsson; and A. Stiene-Martin for help with manuscript preparation and technical assistance. We are also grateful to Dr. Stanley J. Watson, the Molecular and Behavioral Neuroscience Institute, University of Michigan, for the encouraging discussion. This work was supported by grants from the Swedish Science Council and AFA foundation to G. Bakalkin, National Institutes of Health Grants to K. F. Hauser (DA-13278), Y. L. Hurd (DA-313312), and V. M. Pickel (DA-004600).
FOOTNOTES
1 These authors contributed equally to this study. 
2 These authors contributed equally to this study.
Received for publication March 31, 2006. Accepted for publication May 15, 2006.
REFERENCES
1 receptor mRNA distribution in the rat CNS: Comparison to receptor binding and prodynorphin mRNA. Mol. Cell Neurosci. 5,124-144[CrossRef][Medline]This article has been cited by other articles:
|
|
 |
|
  M. H. Ossipov, I. Bazov, L. R. Gardell, J. Kowal, T. Yakovleva, I. Usynin, T. J. Ekstrom, F. Porreca, and G. Bakalkin Control of Chronic Pain by the Ubiquitin Proteasome System in the Spinal Cord J. Neurosci., August 1, 2007; 27(31): 8226 - 8237. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||
