| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Biol Chem, Vol. 273, Issue 52, 34683-34686, December 25, 1998
,,,,,§¶
From the DIBIT, Department of Neurosciences, San
Raffaele Scientific Institute and B. Ceccarelli Center, Department of
Pharmacology, University of Milan, 20132 Milan, Italy and the
§ Consiglio Nazionale Ricerche Center of Molecular and
Cellular Pharmacology, University of Milan,
20129 Milan, Italy
 |
ABSTRACT |
|---|
Top
Abstract
Introduction
Procedures
Results
Discussion
References
|
|---|
Neurosecretion competence is intended as the
ability of neurosecretory cells to express dense and clear vesicles
discharged by regulated exocytosis (neurotransmitter release). Such a
property, which so far has never been studied independently, is
investigated here by a heterotypic cell fusion approach, using a clone
of rat pheochromocytoma PC12 cells totally incompetent for
neurosecretion that still largely maintains its typical molecular and
cellular phenotype. When fused with wild-type partners of various
species (rat, human) and specialization (PC12, neuroblastoma SH-SY5Y, HeLa), the defective cells reacquire their competence as revealed by the expression of their secretion-specific proteins. Fused wild-type
cells therefore appear able to complement defective cells by providing
them with factor(s) inducing the reactivation of their secretory
program. The mechanism of action of these factors may consist not in a
coordinate unblocking of transcription but in the prevention of a rapid
post-transcriptional degradation of the mRNAs for
secretion-specific genes.
Regulated exocytosis, a fundamental activity of neurosecretory
cells, is among the processes most intensely investigated. So far,
however, the studies have focussed primarily on the identification of
the proteins involved and on the mechanistic interactions that underly
vesicle discharge (1-4). In contrast, the way by which neurosecretory
cells become competent for neurosecretion has received much less
attention. Competence acquisition requires the appearance of specific
organelles, dense-core (DV)1
and clear or synaptic-like (SLV) vesicles, an event that takes place
during phenotype development. Because of this temporal coincidence, the
first process is widely envisaged as a step of the second (5, 6).
To investigate the relationships between competence and phenotype, we
used clones isolated from a neurosecretory cell model, the rat
pheochromocytoma PC12 line (7, 8). Most PC12 clones, indicated here as
wild-type clones, exhibit DVs from which exocytic release of secretory
proteins and catecholamines occurs after appropriate stimulation.
Moreover, Western blots, immunocytochemistry, and patch clamping
document in these clones the expression and discharge of
acetylcholine-containing SLVs (8-10). An additional clone has been
found to lack en block both types of secretory vesicles,
i.e. to be neurosecretion incompetent, no matter whether analyzed at rest or after a variety of stimulatory treatments, while
maintaining numerous markers typical of PC12 and other neurosecretory cells (7, 8). These results suggest the lack in the defective clone of
one or more factors that impart to the cells their neurosecretion competence, acting independently, at least in part, from the factors governing the rest of the phenotype (8). Here we report direct evidence
supporting this hypothesis, showing that defective PC12 cells reacquire
their competence when fused to various types of wild-type cells,
regardless of their species of origin and of their specialization.
Moreover, molecular results suggest competence to be conferred not at
the transcriptional but at a post-transcriptional level.
Control (clones 15 and 251) and defective (clone 27) rat
pheochromocytoma PC12 clones, all resistant to G418 due to their stable
transfection with an appropriate construct, and the PC12-27 subclone
(c7) resistant also to hygromycin-B and expressing the human isoform of
chromogranin B (CgB) have been described (7, 8). Properties of the
second defective clone (PC12-Trk) are given in Refs. 11 and 12. The
anti-rat CgB antibodies were purified rabbit IgGs (13). The anti-human
CgB was a mouse monoclonal antibody (14). The anti-rat synaptophysin
(Sph) (monoclonal antibody) was purchased from Boehringer Mannheim, the
anti-TrkA was from Santa Cruz, and the anti-synaptotagmin (Syt) and the anti-human Microscopy--
Cell fixation (with mixtures of formaldehyde and
glutaraldehyde), permeabilization, immunolabeling, and confocal
microscopy are specified in Ref. 17. For electron microscopy the cells detached from the dishes and centrifuged were fixed with the aldehyde mixture followed by OsO4. Ultrathin Epon sections were cut
and investigated after uranyl and lead staining (8).
CgB Secretion--
The media of cell preparations exposed to
various treatments, including high K+ with and without
Ca2+, were immunoprecipitated with anti-rat CgB. The washed
precipitates were analyzed on SDS-polyacrylamide gels and processed for
fluorography (8).
Cell Fusions--
For fusions between PC12 clones (clones 15 and
27, homotypic and heterotypic), cells were detached with a gentle
stream of medium, resuspended at the concentration of
106/ml, and loaded with either 5 µM Orange
CMTMR or 2 µM Green CMFDA (Molecular Probes) (18). After
three washes in Hanks' balanced salt solution and mixing 50:50 of the
two preparations, the cells were fused with PEG 4000 (Merck) (19).
Sorting of labeled cells was in a FACS Plus (Becton Dickinson).
RNA Extraction, Northern Blotting, and RNase Protection--
RNA
extraction was performed according to the single-step method (20). The
probes for Northern blotting were prepared by reverse
transcriptase-polymerase chain reaction, eluted from agarose gel (Qiaex
II extraction kit, Qiagen) and labeled with [32P]dCTP
(3000 Ci/mmol, Redivue Amersham) by random priming (Megaprime kit,
Amersham) (8). RNAs from control and defective clones were fractionated
on 1.2% agarose, 2.2 mM formaldehyde denaturing gels and
then transferred to nylon membranes (Biodyne A, Pall). Hybridizations
carried out at 65 °C were followed by washes at 65 °C in 0.1×
SSC, 0.1% SDS (8). RNase protection assays were carried out as in Ref.
21. The probes employed included: 5' SgII, positions 2652-2813
(GenBankTM accession number S83173); and 3' SgII, positions
1789-1940 (GenBankTM accession number M93669). The probes
were cloned into pSP73 plasmid (Promega) and transcribed in
vitro (Riboprobe® Systems, Promega).
Run-on--
The assay was as in Ref. 22. PC12-15, PC12-27, and
Fisher rat fibroblasts washed in phosphate-buffered saline were lysed by exposure to a nuclear buffer containing 0.3 M sucrose,
60 mM KCl, 15 mM NaCl, 20 mM
Tris-HCl, pH 8, 0.5 mM spermidine, 2 mM EGTA,
0.5 mM EDTA, 0.5 mM The diversity of defective PC12-27 with respect to control clones
(such as PC12-15 and PC12-251, Fig.
1A) should first be discussed.
In terms of microscopy (Fig. 1, A and B) and
neurosecretory markers (7, 8), the differences do not go beyond general interclone variability (7). PC12-27 cells, however, lack not only DVs
and SLVs (7) but also their cargo and membrane proteins (including CgB,
Fig. 1B; SgII; VAMP-2; Syt; Sph) (7, 8), whereas in
wild-type clones 15 these proteins are abundant and exhibit prominent
punctate patterns (Fig. 1A). The defect of PC12-27 is not
restricted to the secretory organelles but includes many (if not all)
of the components participating in the regulated exocytosis/endocytosis
cycle, such as the G protein, rab3a, and the tSNAREs present in the
plasmalemma, syntaxin-1, and SNAP-25 (7, 8). From here on the proteins
mentioned so far will be referred to as secretion-specific. In
conclusion, PC12-27 do not appear as undifferentiated cells. Rather,
they are PC12 that, although possessing properties typical of the
neurosecretory phenotype, are totally incompetent for regulated
neurosecretion.
INTRODUCTION
Top
Abstract
Introduction
Procedures
Results
Discussion
References
EXPERIMENTAL PROCEDURES
Top
Abstract
Introduction
Procedures
Results
Discussion
References
1-integrin subunit were purified rabbit IgGs
(15, 16).
-mercaptoethanol, 0.2%
Nonidet P-40. After 5 min on ice, the nuclei were pelleted, washed, and
resuspended (final concentration, 108 nuclei/ml) in a
buffer containing 50 mM Hepes buffer, 5 mM
MgCl2, 0.5 mM dithiothreitol, 1 µg/ml bovine
serum albumin, and 25% glycerol. The isolated nuclei exhibited a well
preserved appearance. Labeling of nascent transcripts was performed for
20 min with 50 µl of the nuclear suspension together with
[32P]UTP (800 Ci/mmol). RNA was extracted and hybridized
to DNA probes immobilized in duplicate (1 and 2 µg) on nylon
membranes under the same conditions as for Northern blots, with the
last wash conducted for 20 min at room temperature in the presence of
RNase A (1 µg/ml). The employed cDNA probes were: SgII, full
length as well as the 5' and 3' regions; CgB exon 4; VAMP-2, full
length; together with 18 S and 
phage as the positive and negative controls.
RESULTS
Top
Abstract
Introduction
Procedures
Results
Discussion
References
View larger version (130K):
[in a new window]
Fig. 1.
Expression of secretion-specific proteins by
rat PC12 cells from control and defective clones, before and after
fusion. The 15 and 27 clones are illustrated in A and
B (rat CgB immunolabeling) and A' and
B' (phase contrast microscopy), respectively.
C-E, rat CgB immunolabeling of a homotypic
27-27 fusion (C) and of 27-15 hybrids, analyzed 48 h
(D) and 2 weeks (E) after fusion, respectively.
F, hybrids of the same preparation as those shown in
E, but they were immunolabeled for Sph. The abundance of DVs
in the same cells is shown at the ultrastructural level in G
and H (bars, 0.1 µm). I shows the
FACS pattern obtained with cells loaded with the orange and green dyes
before (left) and after (right) PEG 4000 fusion.
The cells in C-H were from populations gated in the area
delimited by the continuous line.
Fusions between PC12 Cells-- To establish whether neurosecretion competence operates under positive or negative control, parallel cell aliquots of the defective 27 and the control 15 PC12 clones were loaded with either orange or green fluorescent vital dyes (18), fused with PEG 4000 (19) and sorted by FACS. The cells exhibiting double fluorescence (Fig. 1I), were processed by immunolabeling (confocal microscopy) and for electron microscopy. Of these fused cells, the homotypic 27/27 never showed appearance of DVs and secretion-specific proteins (Fig. 1C and not shown), whereas the 15/15 never showed any loss of them (not shown). In contrast, over 90% of the 27/15 hybrids resembled controls in their punctate immunolabeling for DV and SLV proteins as seen already 48 h after fusion until weeks later (Fig. 1, D and E for CgB and Fig. 1F for Sph). At the electron microscope, numerous DVs were seen in the hybrids, localized preferentially below the plasmalemma (Fig. 1, G and H) as in control PC12 (23). Moreover, when a hybrid population was pulse-labeled with [35S]sulfate, release of [35S]-CgB was found to take place by typical regulated exocytosis. Such a release, in fact, required a period of chase after the pulse, necessary for the transport of labeled protein to DVs, and depolarization (with high K+) in a Ca2+-dependent manner (Fig. 2) (8). We conclude that in the hybrids the defect of PC12-27 cells had been fully complemented. This suggests neurosecretion competence to be controlled positively and not negatively.
|
Further information was obtained by additional inter-PC12 fusions, carried out with two defective PC12 clones, both incompetent for neurosecretion, PC12-27 (8) and PC12-Trk (11, 12). For convenience, the PC12-27 cells employed in these experiments were from a subclone (c7) stably transfected with the human CgB (localized primarily in the Golgi complex) and resistant not only to G418 but also to hygromycin-B (8). The Trk clone overexpressing the nerve growth factor receptor, TrkA, isolated from another parental population, differs from clone 27 in general morphology, responsiveness to nerve growth factor, and rate of growth (11, 12). The 27/Trk hybrids, resistant to hygromycin-B, were identified by double immunostaining for TrkA and human CgB. Their neurosecretion competence was investigated by an antibody specific for the rat isoform of CgB, with consistently negative results. In fact, neither expression of the rat secretory protein nor changes in the distribution of the human CgB from Golgi to punctate were observed (not shown), suggesting that no complementation had taken place. The results therefore appear compatible with the possibility that the same regulatory factor(s) lacking in PC12-27 is (are) responsible also for the defect in the Trk clone.
Fusions with Human Cells-- An important point that could not be established with the inter-PC12 fusion experiments is whether the competent phenotype of the hybrids resulting from the PC12-27 and PC12-15 fusions was due to the contribution of control cells, whose competence was unaffected by the fusion with defective cells, or also to competence reacquisition by the latter. In fact, no way was available to distinguish the products coded by the genes of the two partners.
To investigate the question, the approach was extended to human cells,
and CgB expressed by the hybrids was identified with specific anti-rat
and anti-human antibodies. Two human lines were employed: neuroblastoma
SH-SY5Y (Fig. 3, a,
A, and A'), whose cells are able to acquire some
neurosecretion competence upon differentiation (24), and epithelial
HeLa (not shown). After fusion with PC12-27, the preparations were
exposed to G418 to eliminate the unfused human cells, which are
sensitive to the antibiotic. The rat/human hybrids were identified by
specific immunostaining for various human antigens, including
1-integrin (15).
|
In the unfused PC12-27, SH-SY5Y, and HeLa cells, as well as in all homotypic fusions, no signal for DV and SLV proteins could ever be appreciated (Fig. 3A' and not shown). In contrast, clear positivity was recorded in many but not all hybrids of both the PC12-27/SH-SY5Y (Fig. 3, B' and C') and the PC12-27/HeLa (Fig. 3, D' and E') type. Even more important, the CgB expressed by the hybrids was recognized specifically by the anti-rat antibody (Fig. 3, B' and D'), for which human cells are negative (Fig. 3A'), with a distribution coinciding with that of the other secretion-specific proteins, including Syt (Fig. 3, C' and E'). In contrast with the anti-human CgB antibody, no immunostaining was appreciated. In terms of intensity and distribution, the signal for the rat CgB and the other vesicle proteins in PC12-27/SH-SY5Y hybrids resembled that of control PC12, whereas in PC12-27/HeLa it was less prominent and more clustered in areas that might correspond to the Golgi complex (Fig. 3, D' and E'). In the latter, therefore, the intracellular traffic of secretion-specific proteins was possibly delayed. Still, a punctate distribution of the label was also visible (cf. Fig. 3, B'-E' to Fig. 1, A and D-F), and typical DVs were recognized by electron microscopy (not shown). We conclude, therefore, that the neurosecretion competence established in the hybrids, no matter whether by fusion with SH-SY5Y or HeLa cells, was sustained by the expression of secretion-specific proteins encoded by genes of PC12-27 cells.
Molecular Mechanism(s) Underlying the Competence Defect-- In previous studies, Northern blot assays of PC12-27 cells had revealed the mRNAs levels for several missing cargo and membrane vesicle proteins to be much lower (0-8%) than in controls, with no major changes of their long term turnover (8). These results suggested that the defect of clone 27 consists of a blockade of transcription of the secretion-specific genes. When, however, run-on experiments were carried out (Fig. 4, left panel), the results appeared incompatible with the above conclusion, because of the three transcripts investigated, the PC12-27 signal was decreased (moderately) for only one (the vesicle membrane vSNARE, VAMP-2) and unchanged or even increased by 2-fold for the others (the secretory proteins, CgB and SgII) (Fig. 4, left panel). These major discrepancies between Northern blots and run-on results appeared specific to PC12-27 cells. When in fact a nonsecretory cell type (Fisher rat fibroblasts) was analyzed, a good match between the very low or negative Northern blots for the three mRNAs and the hardly appreciable (SgII) or completely negative (CgB and VAMP-2) run-on signals was observed (not shown).
|
To explain the PC12-27 transcription results, a variety of control
experiments were carried out. The possibility that the excess SgII
run-on signal was due to transcriptional attenuation (25) could be
excluded because (a) the results obtained with the whole
cDNA were duplicated using probes containing either the 5' or the
3' sequences (Fig. 4, left panel) and (b) no
additional mRNA fragments (including unspliced or partially spliced
transcripts) but only the processed bands (accounting for ~5% of
controls) were revealed by RNase protection (Fig. 4, right
panel). Moreover, when the latter assay was carried out in PC12-27
by using a sense RNA probe, no signal appeared, thus excluding
transcription of the noncoding strand (26) in the SgII gene (Fig. 4,
right panel). Also unlikely appears a blockade of mRNA
transport across the nuclear envelope, with ensuing retention into the
nucleus. In fact, when the nuclear and cytoplasmic SgII and VAMP-2
mRNAs were assayed separately by Northern blot, their ratios were
found to be the same in PC12-27 and PC12-15 clones (not shown), despite the huge differences of the absolute values.
| |
DISCUSSION |
|---|
|
Top
Abstract
Introduction
Procedures
Results
Discussion
References
|
|---|
Taken as a whole, our fusion results provide information about neurosecretion competence. Only when the fusion of PC12-27 cells was homotypic or with the cells of another similarly defective PC12 clone, Trk, did the resulting cells remain incompetent; otherwise they were mostly competent for neurosecretion. Moreover, the results with rat/human hybrids (fusions of PC12-27 with either SH-SY5Y or HeLa) demonstrate that the competence was sustained by the reactivation of the PC12 program with expression of gene products that, before fusion, were not expressed in the defective clone. The simplest hypothesis that emerges from our data is that one or more regulatory factors are complemented by the partner to the defective cell and that this is enough to re-establish the competence of the latter. At the moment, we know neither the nature nor the timing of expression of these factors. They may be stably expressed by cells and preexist therefore fusions. Alternatively, expression could start only in the hybrid, in response to signals contributed by the fused PC12.
In the study of the mechanisms that sustain competence expression, a complete dissociation was observed between RNase protection and Northern blot on the one side and run-on on the other side. The significance of these tests is different. The first two reveal the steady state levels of specific transcripts distinguishing in addition (RNase protection) unspliced and partially spliced fragments as well as other artifacts, which were not found in PC12-27 cells. Run-on, on the other hand, reveals the elongation of the transcripts taking place in isolated nuclei incubated in vitro (22). In view of these differences, we cannot exclude that an inhibition of transcription, effective in intact cells, was relieved during isolation of the nuclei. It should be noted, however, that (a) to our knowledge a process of this kind has never been reported and (b) in our hands, it did not occur in a true non-neurosecretory cell type, Fisher rat fibroblasts. In addition, the PC12 results cannot be explained by the degradation of rapidly turning-over protein(s) specifically regulating the breakdown of secretion-specific mRNAs because preincubation with cycloheximide (7 µM for up to 5 h) did not induce appreciable changes of the SgII mRNA levels either in control or in defective PC12 (Northern blots, not shown). The hypothesis that we favor is therefore that in PC12-27 transcription of secretion-specific genes is followed by an extensive product degradation, not present in controls, occurring immediately after or even during the course of RNA synthesis. Direct evidence for the latter process is still missing in the present experimental model. However, RNA binding proteins regulating initial steps of gene product degradation have been described in the brain (25, 26).
In conclusion, our results identify one of the fundamental properties
of neurosecretory cells, their competence for neurosecretion, as a
regulated process that cells can lose en block by arresting the
expression of their secretion-specific proteins without fully compromising their typical phenotype. Reappearance of rat
secretion-specific proteins after fusion of defective PC12 to partners
strongly suggests the existence of governing factor(s) that PC12-27
cells are unable to express. Our findings may ultimately open a new
chapter in the field of neurosecretory cell differentiation. In
addition, the discovery of new mechanisms could have impact also in
applicative fields, such as cell engineering and cell transplantation
where the role of neurosecretory cells has remained so far only marginal.
| |
ACKNOWLEDGEMENTS |
|---|
We are grateful to Flavia Valtorta for the
generous gift of the PC12-Trk defective clone, to Antonio Malgaroli and
Mara Zocchi for the anti-Syt and anti-1-integrin
antibodies, to W. B. Huttner and H. H. Gerdes for the SgII
and CgB genomic DNAs, and to Silvia Heltai for the FACS sorting experiments.
| |
FOOTNOTES |
|---|
* This work was supported by Grant ERB B104 CT960058 from the European Union Biotechnology Program, Grant RG 520/95 from the Human Frontier Science Program, Grant 97.01144.PF49 from the Target Project on Biotechnology of the Consiglio Nazionale Ricerche, and Grant 199/bi from the Armenise-Harvard Foundation and Telethon Foundation (to J. M.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
¶ To whom correspondence should be addressed: DIBIT, Scientific Institute San Raffaele, Via Olgettina 58, 20132 Milano, Italy. Tel.: 39-2-26432770; Fax: 39-2-26434813; E mail: meldolesi.jacopo{at}hsr.it.
The abbreviations used are: DV, dense-core vesicle(s); SLV, clear or synaptic-like vesicle(s); CgB, chromogranin B; Sph, synaptophysin; Syt, synaptotagmin; FACS, fluorescence-activated cell sorter; SgII, secretogranin II.
| |
REFERENCES |
|---|
|
Top
Abstract
Introduction
Procedures
Results
Discussion
References
|
|---|
This article has been cited by other articles:
|
|
 |
|
  H. Murata, N. Tajima, Y. Nagashima, M. Yao, M. Baba, M. Goto, S. Kawamoto, I. Yamamoto, K. Okuda, and H. Kanno Von Hippel-Lindau Tumor Suppressor Protein Transforms Human Neuroblastoma Cells into Functional Neuron-like Cells Cancer Res., December 1, 2002; 62(23): 7004 - 7011. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Grundschober, M. L. Malosio, L. Astolfi, T. Giordano, P. Nef, and J. Meldolesi Neurosecretion Competence. A COMPREHENSIVE GENE EXPRESSION PROGRAM IDENTIFIED IN PC12 CELLS J. Biol. Chem., September 20, 2002; 277(39): 36715 - 36724. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. D. Blagoveshchenskaya and D. F. Cutler Sorting to Synaptic-like Microvesicles from Early and Late Endosomes Requires Overlapping but Not Identical Targeting Signals Mol. Biol. Cell, May 1, 2000; 11(5): 1801 - 1814. [Abstract] [Full Text]
|
|
|
|
|
|
C. Leoni, A. Menegon, F. Benfenati, D. Toniolo, M. Pennuto, and F. Valtorta Neurite Extension Occurs in the Absence of Regulated Exocytosis in PC12 Subclones Mol. Biol. Cell, September 1, 1999; 10(9): 2919 - 2931. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |
K+ = 3.5 mM; +K+ = 53.5 mM;
