|
Advertisement | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
J Biol Chem, Vol. 274, Issue 10, 6536-6545, March 5, 1999
From the Melanin concentrating hormone (MCH)
and neuropeptide EI (NEI) are two peptides produced from the same
precursor in mammals, by cleavage at the
Arg145-Arg146 site and the
Lys129-Arg130 site, respectively. We performed
co-localization studies to reveal simultaneously the expression of MCH
mRNA and proconvertases (PCs) such as PC1/3 or PC2. In the rat
hypothalamus, PC2 was present in all MCH neurons, and PC1/3 was present
in about 15-20% of these cells. PC1/3 or PC2 was not found in
MCH-positive cells in the spleen. In GH4C1
cells co-infected with vaccinia virus (VV):pro-MCH along with VV:furin,
PACE4, PC1/3, PC2, PC5/6A, PC5/6B, or PC7, we observed only efficient
cleavage at the Arg145-Arg146 site to generate
mature MCH. Co-expression of pro-MCH together with PC2 and 7B2 resulted
in very weak processing to NEI. Comparison of pro-MCH processing
patterns in PC1/3- or PC2-transfected PC12 cells showed that PC2 but
not PC1/3 generated NEI. Finally, we analyzed the pattern of pro-MCH
processing in PC2 null mice. In the brain of homozygotic mutants, the
production of mature NEI was dramatically reduced. In contrast, MCH
content was increased in the hypothalamus of PC2 null mice. In the
spleen, a single large MCH-containing peptide was identified in both
wild type and PC2 null mice. Together, our data suggest that pro-MCH is processed differently in the brain and in peripheral organs of mammals.
PC2 is the key enzyme that produces NEI, whereas several PCs may cleave
at the Arg145-Arg146 site to generate MCH in
neuronal cell types.
Prohormones give rise to multiple peptides through cleavage at
specific sites, usually basic residues, by endoproteases (reviewed in
Refs. 1 and 2). Seven prohormone convertases
(PCs)1 have been identified
so far in different mammalian cellular models, and all participate in
selective cleavage of precursors on the C-terminal side of pairs of
basic residues (reviewed in Refs. 3-5). Based on their tissue
distribution and intracellular localization, the mammalian
subtilisin/kexin-like serine proteinases can be subdivided into four
classes: (i) furin (6) and the recently discovered PC7 (7) process
precursors that reach the cell surface via the constitutive secretory
pathway; (ii) PC1/3 and PC2 (reviewed in Refs. 5 and 8) process
precursors whose products are stored in neuroendocrine secretory
granules; (iii) PACE4 (9), the isoforms PC5/6-A (10-12), and PC5/6-B
(13), which are expressed in both endocrine and nonendocrine cells,
conceivably process precursors in both the constitutive and regulated
secretory pathways; (iv) PC4, which is predominantly synthesized in
testicular germ cells (14, 15), appears important for fertilization
(16). Defining the cellular co-localization of each PC with its cognate substrate(s) is necessary in order to ascribe a role for these enzymes
in particular precursor processing events. In addition, dramatic
developmental changes in the tissue expression of the PCs have been
reported (17, 18). These PCs could therefore participate in the
differential processing of neuroendocrine precursors, such as those
encoding multiple bioactive peptides. Indeed, PC1/3 and PC2 cleave
pro-opiomelanocortin at distinct pairs of basic residues in established
cell lines (19) and in a temporal order in the corticotrophs (20).
Specific patterns of cleavage by either PC1/3 or PC2 have also been
demonstrated for proglucagon (21-23), proneurotensin (24),
pro-thyrotropin-releasing hormone (25), proenkephalin (26, 27), and
procholecystokinin (28, 29). The functional relevance of the
proconvertases in a whole animal model was recently established by
disrupting the corresponding genes such as these encoding PC2 (30) and
PC4 (16).
Melanin-concentrating hormone (MCH) and neuropeptide EI (NEI) are two
distinct peptides encoded in the same precursor and co-expressed and
secreted by neurons in the brain (31-34) and different cell types
(macrophages, Sertoli cells, spermatogonia) in the peripheral tissues
of mammals (35-37) (Fig. 1). Both
peptides are synthesized in the perikarya of neurons in the lateral
hypothalamic area (LHA) and zona incerta and serve as
neurotransmitters/neuromodulators in a number of neural functions
(reviewed in Refs. 38 and 39). Recently, several laboratories have
suggested an important role for MCH in the regulation of
satiety/appetite and associated pathologies such as obesity
(40-43).
In the rat and human brain, exclusively mature peptides,
i.e. cyclic MCH and amidated NEI, were identified (44, 45).
Larger forms were found in the human colon, thymus, or adipose tissues (45) (Fig. 1, A and B). In the rat gut (35),
testis (37), and spleen2 most
of the MCH immunoreactivity also did not correspond to the mature
peptides, but the structure of these pro-MCH derivatives is not yet
established. These tissue-specific differences in MCH-derived peptides
probably result from differential processing of pro-MCH. In the current
study, we examined (i) the cellular co-expression of different
convertases in the rat brain and selected peripheral tissues (spleen,
testis); (ii) the processing of mammalian pro-MCH in
GH4C1 cells expressing various convertases
using vaccinia expression systems; (iii) the processing of rat pro-MCH
after stimulation by 50 ng/ml nerve growth factor (NGF) and 20 mM lithium of PC12 cells engineered to express either PC1/3
or PC2; (iv) the effects of disruption of the PC2 gene on the
processing of pro-MCH in the brain, spleen, and testis of transgenic
mice. Our main finding is that active PC2 is necessary for the
formation of mature NEI in the brain and in cellular models. In
contrast, many convertases (including PC1/3 and PC2) have the capacity
to generate MCH from its precursor both in virally infected
GH4C1 cells and in an animal model.
Co-localization Studies: in Situ
Hybridization/Immunohistochemistry--
Frozen sections (12 µm-thick) of brain, spleen, or testis from adult Wistar rats (180 g
weight) were cut on a cryostat at Vaccinia Virus (VV) Constructs and Infections--
Purified
recombinant VV using the full-length cDNA inserts of the rat MCH
(VV: rMCH), the mouse PC1/3 (VV:PC1/3), PC2 (VV:PC2), PC5/6 A or B
(VV:PC5), PC7 (VV:PC7), human furin (VV:hfurin), human PACE4
(VV:hPACE4), and mouse pro-7B2 (VV:27-kDa 7B2) and active 7B2
(VV:21-kDa 7B2) were constructed as reported previously (19, 22, 46).
The VV:recombinant convertases were found to be expressed at high
levels in GH4C1 cells (Refs. 19, 22, 25, and 26
and data not shown). Conditions for cellular infections were as
described previously (19, 22). Briefly, GH4C1
cells at P15 (50 × 106 cells/plate) were infected
with recombinant VV for 1 h and plated for 18 h, and cell
extracts and 18-h media were collected. In some cases after 18-h
infection, cells were incubated with serum-free medium for 4 h,
and the media were collected. Proteins or peptides were extracted as
described below. Two to four separate sets of experiments were
performed with plates in duplicate for each VV:recombinant convertase
preparation (mean values are shown).
PC1/3 or PC2-PC12 Cell Transfectants--
Wild type (WT) PC12
cells, PC1-expressing PC12 cell lines (L1.2 and E1.2), and
PC2-expressing PC12 cell lines (L2.2 and E2.15) obtained from Dr. P. Kitabgi's laboratory (Institut de Pharmacologie Moléculaire et
Cellulaire, Valbonne, France) were cultured as described previously
(24). The PC12 cell lines were stimulated for 24 h
(T24) or not (control; C) with optimum
concentrations of NGF (50 ng/ml) and LiCl (20 mM) to induce
MCH mRNA synthesis (47). Expression of MCH mRNA and
pro-MCH-derived peptides (MCH, NEI) was determined by RT-PCR (47) and
by RIA coupled to RP-HPLC as described below. MCH mRNA was detected
only in NGF/lithium-treated PC12 cells, whereas PC1/3 or PC2 was
expressed in resting and lithium/NGF-treated cells (Ref. 24 and data
not shown).
PC2 Null Mutant Mice Model--
PC2 null mutant mice were
generated after homologous recombination in introducing the neomycin
gene into the third exon of the mouse PC2 gene (30). This resulted in
the production of an inactive truncated PC2 protein (30). Tissues
(brain, gut, testis) from 12 WT and 12 PC2 null mice (PC2 KO) were
dissected and immediately frozen at RNA Extraction and Semiquantitative RT-PCR
Analysis--
Hypothalami from three wild type and PC2 null mice were
dissected following partial thawing, and total RNA was extracted as described previously (35). The quality of the RNA was monitored by
ethidium bromide staining of 0.8% agarose gel after electrophoresis.
Semiquantitative RT-PCR analysis was performed using competition
between reverse-transcribed RNA from tissues and a short template
(MCH
Preliminary experiments established that primers amplify target and
internal MCH Peptide Extraction--
Virus-infected
GH4C1 cells were sonicated in 1 ml of 0.06 N HCl, 0.1% (v/v) trifluoroacetic acid solution. Extracts
were passed through a C18 silica cartridge (C18 Sep Pak, Waters
Associates, Milford, MA), and absorbed peptides were eluted with 55%
(v/v) CH3CN and lyophilized.
Peptides were extracted from PC-transfected PC12 cells with 0.1 N HCl. Extracts were centrifuged 10 min at 3,500 × g. Supernatants were kept at 95 °C for 10 min and stored
at
After homogenization by a Polytron homogenizer, tissues were boiled for
10 min in 0.1 M HCl and centrifuged twice at 10000 × g at room temperature for 20 min. The supernatants were
neutralized with NaOH, and the peptide solutions were lyophilized. The
dried peptide extracts were resuspended in water and used either for a
direct RIA or an RP-HPLC purification.
RP-HPLC Condition--
Reverse-phase HPLC (RP-HPLC) was
performed with a C18 Merck Lithosorb column 100-RP (5 µm) or a C18
PLRP-S column 300A (8 µm) (Polymer Laboratories, Amherst, MA). The
mobile phase was 0.1% trifluoroacetic acid solution. The peptides were
eluted with an acetonitrile/trifluoroacetic acid 0.1% gradient from 20 to 60% in 40 min with a 1-ml/min flow rate after an isocratic step at
20% for 10 min. The retention times of the synthetic rat/human MCH and
NEI were, respectively, 31 or 33 min and 8-9 or 13 min, depending on
the column used. Synthetic MCH and NEI were gifts from Drs. Carl M. Hoeger and Jean Rivier (Salk Institute, La Jolla, CA). Synthetic
NEI-GRR (43) was custom-synthesized by Neosystem (Strasbourg, France).
MCH and NEI Radioimmunoassays--
RIAs were performed as
described (36, 45). Briefly, 50 µl of rat MCH or NEI antiserum (final
dilution 1:300,000), 50 µl of the standard peptide or biological
samples, 50 µl of the iodinated peptide (105 cpm/ml), and
100 µl of phosphate buffer were mixed. After 24 or 48 h of
incubation at 4 °C, bound radioactivity was precipitated by adding
50 µl of bovine serum, 50 µl of 0.6% Tween 20, and 350 µl of
25% polyethylene glycol 6000 for 10-20 min. Tubes were then centrifuged for 10 min at 4000 rpm at 4 °C, and the supernatant was
discarded. The pellets were counted in a Crystal Packard for automatic
calculation of the bound/total ratio. Blank values (i.e. radioactivity precipitated without antibodies) were always <6%. The
routine sensitivity of the assays was 250 and 50 fmol/sample for the
MCH and NEI RIA, respectively. The NEI antiserum recognized equally
well the amidated NEI peptide and NEI-GR, but NEI-GRR cross-reacted
<0.1% with this antiserum (43, 45).
Co-localization of MCH mRNA and Proconvertases in the
LHA--
Using 33P-labeled RMCH2 oligoprobe on the one
hand and different proconvertase antisera on the other, we
simultaneously examined the expression of MCH mRNA and the amounts
of PC1/3 and PC2 protein in the rat LHA. As shown in Fig.
2, A and B, both
PC1/3 and PC2 immunoreactivities were found within or close to MCH
mRNA-expressing neurons. At higher magnification, strict
co-localization was demonstrated for PC2 and MCH mRNA in all
perikarya, whereas only part of the MCH mRNA-expressing neurons
co-localized with PC1/3 (Fig. 2, C and E, and
Fig. 2, D and F, respectively). We estimated that
about 15-20% of MCH neurons contained both PC1 and PC2 (27 ± 7 of 150 individual neurons examined per rat; n = 2 rats).
Pro-MCH Processing in GH4C1
Cells--
Pro-MCH processing was determined in
GH4C1 cells after co-infection of recombinant
vaccinia virus engineered to express rat pro-MCH or either of the
prohormone convertases listed in Tables I
and II. Media were collected during an
18-h period of infection (18-h media) and following a further 4 h
of infection in a fresh serum-free media (4-h media). Cell extracts
were isolated after the 22 h of infection. Immunoreactive peptide
content was measured with specific RIAs in total cellular extracts or
collected media and in fractions following RP-HPLC analysis. Both MCH
and NEI RIA have previously been shown to be highly specific for their respective peptides (36, 45). However, it is worth noting that MCH
antisera recognized equally well the mature peptide, processing
intermediates and the full-length precursor (unpublished data),2 whereas the NEI antisera can identify amidated NEI,
NEI-G, and NEI-GR forms but cross-react poorly with other pro-MCH
derivatives and the precursor itself (36, 45). Therefore, the MCH-IR
concentrations in whole cell extracts or collected media reflected the
total amounts of intracellular pro-MCH derivatives (processed plus
unprocessed forms), and the NEI-IR concentrations indicated the amounts
of processed NEI peptide.
No MCH-IR or NEI-IR was detected in WT GH4C1
cells, whereas infection with VV:MCH resulted solely in the production
of pro-MCH (Tables I and II). We assessed first the processing
capability of one ubiquitous PC (furin) and the two PCs (PC1/3 and PC2)
found co-localized with MCH and NEI in pro-MCH-expressing perikarya by
coinfecting GH4C1 cells with VV:MCH and either
of the PC. As shown in Table I, MCH-IR was found in both media and cell
extracts after furin, PC1/3, or PC2:VV infection, whereas NEI-IR was
absent. The ratio of MCH peptide was very similar in cell extracts,
18-h media, and 4-h media in all of the PC:VV co-infection, consistent with the constitutive secretion of prohormone derivatives in the GH4C1 cells. The cleavage efficiency at the
Arg145-Arg146 site was quite similar between
PC1/3 and PC2 and tends to be higher in cell extracts after the
furin:VV infection.
We extended this analysis to the other PCs and determined the
intracellular content of pro-MCH peptides in
GH4C1 cells infected with either of PC:VV. As
shown in Table II and Fig. 3A,
all tested prohormone convertases were able to cleave the pro-MCH at
the pair of basic residues Arg145-Arg146 to
yield mature MCH with an order of efficiency that was PC7 > furin > PACE4 > PC1/3
The PC2-specific binding protein 7B2 has been shown to be essential for
the expression of active PC2 (50-52). VV:PC2 plus 27-kDa 7B2 produced
a large amount of mature MCH (63% cleavage efficiency; not shown). In
contrast, simultaneous infection of GH4C1 cells with VV:MCH, the 7B2 precursor (VV:27-kDa 7B2), or the processed 7B2
(VV:21-kDa 7B2) and VV:PC2 resulted in a weak cleavage at the
Lys129-Arg130 site (0.6-0.4% cleavage
efficiency) as further demonstrated by RP-HPLC analysis (Fig.
3B). In addition, only mature NEI was revealed (Fig.
3B), and putative processing intermediates such as the
pro-MCH-(130-165) (44-min peak, Fig. 7) were lacking (data not shown).
These results indicated that in the vaccinia expression system, any of
the prohormone convertases may potentially liberate MCH without being
able to produce an equivalent amount of NEI. Coinfection of
GH4C1 cells with VV:PC7 or VV:PC2 plus 7B2
(precursor or active form) resulted in weak processing to mature
NEI.
PC2 but Not PC1/3 May Generate NEI in Stably PC-transfected PC12
Cells--
We hypothesized that PC1/3 and PC2 may be responsible for
cleavage at pairs of basic residues found in pro-MCH to yield mature NEI and MCH but only under restricted physiological conditions not
reached in the vaccinia expression system. Accordingly, PC1/3- or
PC2-stably transfected PC12 cells from P. Kitabgi's laboratory (Valbonne, France) were used in an additional study. It has been demonstrated that wild type PC12 cells lack both of these PCs and that
the stably transfected cell lines express large amounts of active
enzyme capable of processing endogenous prohormones such as
proneurotensin (24). Furthermore, we have previously shown that MCH
mRNA is absent in resting PC12 cells but can be induced by the
addition of NGF and lithium for 24 h, resulting in MCH mRNA
production (47). The levels of MCH-IR and NEI-IR were determined using
RIA on RP-HPLC fractions of cellular extracts from control (Fig. 4,
C) or NGF/lithium-treated (Fig. 4, T24) PC12
cells. The MCH RIA revealed a broad peak of immunoreactivity in all
PC12 cell lines tested under resting or stimulated conditions (data not
shown). Further Western blot analysis demonstrated cross-reactivity of
MCH antiserum to several protein species in PC12 cellular extracts (not
shown). Therefore, it was not possible to ascertain whether PC1/3 or
PC2 cleave at the Arg145-Arg146 site in this
cellular model. However, a single peak, corresponding to mature NEI,
was identified only in the PC2-transfected cells (L2.2) treated with
NGF and lithium (Fig. 4). The level of
NEI was estimated to about 0.3 pmol/106 cells,
i.e. at least 10-fold the amount of peptide found in
VV:MCH/PC2 plus 7B2-infected GH4C1 cells (see
Table II and Fig. 3B). No NEI immunoreactivity was found in
WT and PC1-transfected cells or in unstimulated L2.2 cells (Fig. 4).
These results then support the hypothesis of preferential cleavage by
PC2 at the Lys129-Arg130 site in neuroendocrine
cells. The same pattern of pro-MCH processing was observed in another
PC2-transfected PC12 cell line (E2.15; Ref. 24) (data not shown).
PC2 Is Necessary for the Formation of NEI but Not MCH in the
Hypothalamus--
To test directly whether PC2 was actually involved
in producing MCH and NEI, we determined whether pro-MCH processing
could be affected by the lack of active PC2 in a transgenic mice model (30). Peptides were extracted from the whole brain, hypothalamus, brain
without hypothalamus, spleen, and testis of WT or KO mice, and MCH-IR
and NEI-IR contents were measured by RIA in crude extracts and after
RP-HPLC separation.
The results of representative experiments concerning MCH and NEI assays
in the brain, spleen, and testis of WT and KO mice are summarized in
Table III. As expected, MCH-IR and NEI-IR
contents were quite similar in WT mice. Surprisingly, the KO mice
showed higher amounts of MCH-immunoreactive materials by comparison
with the WT mice, solely in the hypothalamus. On the other hand, the NEI-IR content was drastically diminished in both whole brain and
hypothalamus of KO mice compared with WT (
The apparent increase in MCH-IR in the hypothalamus of PC2 null mice
could result either from enhanced synthesis, possibly reflected at the
level of MCH mRNA, or regulation at post-translational levels, such
as an increase in storage of MCH-containing secretory granules. The
first possibility was examined by using a semiquantitative RT-PCR assay
(48). As shown in Fig. 6, the levels of
MCH mRNA were similar in WT and KO mice, indicating that changes in
MCH-IR contents were not accompanied by modifications in MCH gene
expression. Further attempts to reveal differences in the pattern of
pro-MCH derivatives using either Western blot (not shown) or RP-HPLC
analysis (Fig. 5, left panels) with MCH antiserum
demonstrated the presence of a single peak corresponding to mature MCH
in both WT and KO mice. These results are therefore consistent with the
hypothesis of an increase of storage of MCH in the soma located within
the LHA of the KO mice.
A Large Form of MCH Peptide Is Produced in Peripheral Tissues of
Mice--
In peripheral organs such as spleen and testis, RIA of cell
extracts for MCH-IR indicated no significant differences between the WT
and KO mice, suggesting that PC2 was not involved in the processing of
pro-MCH in these tissues (Table III). No NEI-IR could be detected in
spleen or testis of both types of mice (Table III). Further RP-HPLC
analysis with spleen extracts revealed a single peak of MCH-IR at 44 min, and no MCH-IR or NEI-IR was found co-eluting at the position of
the corresponding synthetic peptide (Fig.
7). The material corresponding to the
44-min peak displayed the characteristics of the intermediate cleavage
form, pro-MCH-(130-165), already identified in several human
peripheral tissues (45). The apparent increase in MCH-IR material in KO
mice versus WT mice noted in Fig. 7 was not reproducibly
found with other spleen peptide samples (not shown).
It is now well established that PC1/3 and PC2 are the major
enzymes involved in the processing of precursors directed to the regulatory secretion pathway within endocrine and neuroendocrine cells.
On the other hand, furin, PACE4, PC5/6, and PC7 exhibit a widespread
distribution and are capable of processing proproteins within
constitutively secreting cells (1-5, 7, 11, 53). Information regarding
the role of these enzymes in the processing of the precursor encoding
MCH and NEI was so far lacking. We suggest here that PC2 is necessary
for the generation of mature NEI in neuronal cells, whereas all of the
known PCs are potentially capable of producing MCH from its precursor
in mammals.
The first criterion to define the physiological importance of
endoproteases involved in processing of a specific precursor is
obviously to find one or several of these enzymes in the cells expressing the prohormone of interest. In this study, we demonstrate strict co-localization of PC2 with MCH mRNA in neurons of the LHA,
the main site of synthesis of pro-MCH in the rat brain (32). These
results support and extend previous in situ hybridization data showing that most of MCH mRNA-expressing magnocellular cells of the LHA contained also PC2 mRNA (54). These results indicate that PC2 represents a strong candidate for the processing of pro-MCH in
these neurons. In addition, PC1/3 was also found in some but not all
MCH mRNA-expressing perikarya. In a semiquantitative analysis, we
estimated that about 15-20% of MCH-positive cells in the LHA co-express also PC1/3. On the other hand, MCH-expressing cells in the
gut (macrophages) or testis (spermatogonia, Sertoli cells) lacked PC1/3
or PC2 (data not shown).
The cleavage susceptibility of pro-MCH to various prohormone
convertases was analyzed first by vaccinia expression systems that
generated high levels of pro-MCH and PCs, thereby facilitating the
processing analysis. GH4C1 cells were
co-infected with recombinant vaccinia virus engineered to express
pro-MCH and different prohormone convertases, namely furin, PACE4,
PC1/3, PC2, PC2 with various forms of 7B2, PC5/6-A, PC5/6-B, and PC7.
Thereafter, peptides produced and MCH precursor were identified by RIA
and RP-HPLC analysis. All seven processing enzymes are capable of
cleaving on the C-terminal side of the
Arg145-Arg146 site to produce high amounts of
MCH, the most efficient being PC7 and PC2 in the presence of 7B2 (see
Tables I and II and Fig. 3). In great contrast, large amounts of NEI-IR
were not found with any of the above combinations of pro-MCH and PCs,
although minor cleavage at the Lys129-Arg130
site was observed to produce mature NEI with PC7 or PC2 expressed with
either 27-kDa 7B2 or 21-kDa 7B2 (see Fig. 3, A and
B). Several interpretations may be proposed to explain these
data: (i) none of the tested enzymes may actually cleave at this site
of pro-MCH; (ii) all or some of these PCs process pro-MCH to produce an
NEI form undetectable by our RIA, for instance NEI-GRR (45); (iii) the
enzyme(s) involved in cleavage at this site required defined conditions
that may not be reproduced in this vaccinia expression model. We cannot
rule out, on one hand, the likelihood that a novel uncharacterized
convertase is required for this cleavage or, on the other hand, that
trimming by carboxypeptidases was reduced in these cells, preventing
exposure of epitopes recognized by NEI antiserum (43, 45). However, we
favor another possibility that is discussed below.
Because PC1/3 and PC2 co-localized with MCH in the LHA, it was of
obvious interest to investigate the role of these processing enzymes in
cellular models that expressed PC(s) and pro-MCH in the appropriate
intracellular compartment. We have examined therefore the processing of
pro-MCH in PC12 cell lines stably transfected with PC1/3 or PC2 (24).
It has been shown that the same transfectants expressed active PC1/3 or
PC2, resulting in different processing patterns of proneurotensin close
to those found in the gut (PC1/3 transfectants) or the brain (PC2
transfectants) (24). We had previously revealed that MCH mRNA
production may be induced in PC12 cells by adding NGF and lithium salt
to the media for 24 h (47). Interpretation of results using our
MCH RIA was impeded by the cross-reactivity of the MCH antiserum with
unknown proteins of high molecular weights expressed in large amounts
in all of the PC12 cell lines tested so far (data not shown). However,
the data obtained with the NEI-RIA support the conclusion that PC2 but
not PC1/3 may cleave efficiently at the
Lys129-Arg130 site to generate NEI when PC12
cells are committed to differentiate into neuron-like cells in the
presence of NGF.
To further assess the physiological role of PC2 in the pro-MCH
processing in vivo, we studied the cleavage of this
precursor in a strain of mice lacking functional PC2 (30). Using this model, we showed here that cleavage at the
Lys129-Arg130 site was dramatically impaired,
whereas cleavage at Arg145-Arg146 was not
altered in the hypothalamus. As an internal control, no difference in
processing of pro-MCH was observed in cells expressing MCH-related
peptides but lacking PC2, such as immune cells located in the spleen
(Table III). PC2 corresponds therefore to the key endoprotease
responsible for mature NEI production by virtue of cleavage at the
Lys129-Arg130 site in neuronal cells, whereas
this enzyme may also be involved but is not the only enzyme that can
cleave pro-MCH at the Arg145-Arg146 site to
release MCH (Fig. 8).
Strikingly, as regards NEI production, the results obtained using the stably transfected PC12 cell lines or PC2-KO mice do not agree with the vaccinia expression data mentioned above. Similar discrepancies have been noted previously between data obtained from vaccinia virus-infected models and other cellular models, such as cells stably transfected with PC-encoding vector, when secretogranin II (55), proglucagon (21, 22), and proneurotensin3 processing was analyzed. These differences could be due to a possible alteration in the formation and storage of fully functional secretory granules in the vaccinia virus-infected cells. More specifically, it is well established that optimal PC2 enzyme activity requires acidic pH and high Ca2+ levels (56, 57). Furthermore, PC2 cleavage of precursors operates late along the biosynthetic pathway after the dissociation of PC2 from the C-terminal CT peptide of 7B2. These conditions may not be ideally achieved in the secretory pathway of vaccinia virus-infected cells, leading to conflicting results when compared with other cellular or animal models (55). Based on the in vitro data, numerous prohormone convertases may be candidates for processing at the Arg145-Arg146 site to produce mature MCH (see Table II and Fig. 8). PC2 and PC1/3 have been found co-expressed with MCH in the neurons of LHA, and they represent therefore the strongest candidates. However, PC2 is not required for efficient processing of pro-MCH to produce mature MCH in the PC2 null mice model, and PC1/3 was found only in 15-20% of MCH-expressing neurons in the rat brain. Therefore, other yet uncharacterized proconvertase enzymes may possibly be involved in this cleavage. Among them, the PC5/6-A exhibits some overlapping of expression with MCH in the LHA4 and may potentially cleave at the Arg145-Arg146 site to liberate MCH (see Table II). PC7 is widely expressed in the hypothalamus, and it could also be present in MCH neurons, but co-localization studies are presently lacking. An intriguing result concerns the 2-3-fold increase in mature MCH content observed selectively in the hypothalamus of PC2 null mice in comparison with controls (see Table III). The elevated MCH concentrations were not paralleled by enhanced gene transcription and/or MCH mRNA stability, as illustrated by our semiquantitative RT-PCR data. One could envisage first that the targeting of MCH-containing granules to axonal projections is somehow altered. Transport of secretory vesicles involves sequential steps and numerous protein factors (58). Impaired processing of one of these regulatory factors in the PC2 null mice may contribute to a partial decrease in granular transit. Alternatively, chronic elevation of hypothalamic MCH in PC2 null mice may be directly dependent upon the lack of NEI production or other peptidic secretagogues involved in the regulation of pro-MCH synthesis and release. In this context, it is worth noting that similar increase in MCH content was observed in fat/fat mice, which produced the NEI-GRR form but very low amounts of fully processed amidated NEI (43). The physiological significance of MCH content enhancement in PC2 null mice remains at this stage a matter for speculation. Production of pro-MCH derivatives was not modified in spleen and testis of mutant PC2 mice, in agreement with co-localization studies that demonstrated that macrophage-like cells in the spleen and Sertoli cells or spermatogonia in the testis do not express detectable levels of PC2. The levels of MCH-IR contents in mice were close to those found in rat (35) or human (45) peripheral tissues. Interestingly, no peak corresponding to either mature MCH or NEI was found in mouse spleen. The MCH immunoreactivity was exclusively associated with a peptidic form corresponding to a 44-min peak. The same product was previously identified as an intermediate cleavage form of pro-MCH, bearing NEI and MCH, in human peripheral tissues (45). This large MCH-IR form may represent a novel peptide with hitherto unknown biological activity. Further experiments are required to establish the structure of this material in mice, but it is attractive to suggest that a similar kind of pro-MCH processing operates in peripheral tissues of human and mouse and leads to the release of an unprocessed NEI-MCH product. Interestingly, fully processed MCH and NEI peptides, albeit in low amounts, were identified in rat gut and testis (35, 37), raising the possibility of species-specific regulation of MCH precursor processing in peripheral organs. As illustrated in Fig. 8, the enzymes that participate in the cleavage at the Lys129-Arg130 pair of pro-MCH in peripheral tissues remain unknown. Based on the cell-specific expression of PC7 in testis and spleen (7) and data from the vaccinia expression systems, we proposed that PC7 could be a putative candidate to cleave at this site. However, the ubiquitous enzyme furin or PACE4 may also be involved in the processing of MCH precursor in cells lacking the regulated secretory pathway. In conclusion, this study reveals the preeminent role of PC2 in the
production of NEI and provides evidence that PC2, PC1/3, and/or P5/6A
are potential candidates involved in the release of MCH from its
precursor in neurons of LHA. In addition, PC7 or other ubiquitous
convertases of the constitutive secretory pathway may possibly cleave
at the Lys129-Arg130 site to generate a large
product, encompassing the NEI and MCH domains of the precursor, found
predominantly in peripheral tissues of mice and humans. As for the
other prohormones studied so far, the differential processing of
pro-MCH reflects the tissue-specific expression of the various
prohormone convertases potentially involved in the cleavage at basic
sites (reviewed in Refs. 3 and 4). The structural basis for the
Lys129-Arg130 site preference of PC2 in the
brain or of the putative ubiquitous processing enzyme in peripheral
tissues remains to be elucidated. The temporal as well as the
tissue-specific regulation of pro-MCH processing should be further
examined in cells (other than PC12 cell lines) expressing stably both
pro-MCH and the proconvertases of interest and in the future PC1/3 or
PC5/6A null mice models.
We thank Dr. P. Kitabgi (Institut de
Pharmacologie Moléculaire et Cellulaire) for helpful advice and
critical reading of the manuscript. We appreciate the expert technical
assistance of Dianne Savaria in the GH4C1 cell
infections and in the isolation of the various vaccinia virus
recombinants. We also thank Josée Hamelin and Maya Mammarbachi
for help in the construction and DNA sequencing of the vaccinia
transfer vector PMJ:rMCH, Dr. Carole Rovère (Institut de
Pharmacologie Moléculaire et Cellulaire) for help in producing
the PC-transfected PC12 cell lines, Dr. D. Fellmann (CNRS ESA 6025, Besançon, France) for providing the deleted MCH competitor
(MCH
* This work was supported by a group grant from the Medical Research Council of Canada (MRC GR11474), by the Association pour la Recherche sur le Cancer (Grant ARC 6987), the Ministère de l'Education Nationale (Grant 95G0099), the Association Française contre les Myopathies (AFM; ASI 1995-1997), and the "Institut Danone" (Appel d'Offres 1998).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.
§ Recipient of an "Association pour le Developpement de l'Enseignement et des Recherches-Provence Alpes Cote d'Azur" (Contrat d'Allocataire de Recherche 9312/2679) studentship (1994-1996) and an AFM fellowship (1997). Supported by the "Institut de Recherche Servier". Present address: Laboratory of Molecular Genetics, The Rockefeller University, 1230 York Ave., New York, NY 10021-6399.
¶ Present address: Smithkline & Beecham Pharmaceuticals, Dept. of Neuroscience, Neuroimaging Research, New Frontiers Science Park, North, Third Ave., Harlow, Essex CM19 5AW, UK.
¶¶ To whom correspondence should be addressed: Institut de Pharmacologie Moléculaire et Cellulaire, CNRS UPR 411, 660 route des Lucioles, Sophia-Antipolis, 06560 Valbonne, France. Tel.: 33 4 93 95 77 53; Fax: 33 4 93 95 77 08; E-mail: Nahonjl{at}ipmc.cnrs.fr.
2 G. Hervieu and J.-L. Nahon, unpublished data.
3 P. Kitabgi, personal communication.
4 P. Villeneuve and A. Beaudet, personal communication.
The abbreviations used are: PC, proconvertase; MCH, melanin-concentrating hormone; NEI, neuropeptide EI; LHA, lateral hypothalamic area; NGF, nerve growth factor; rMCH, rat MCH; WT, wild type; PCR, polymerase chain reaction; RT-PCR, reverse transcription PCR; RIA, radioimmunoassay; HPLC, high pressure liquid chromatography; RP-HPLC, reverse phase HPLC; VV, vaccinia virus.
Copyright © 1999 by The American Society for Biochemistry and Molecular Biology, Inc. This article has been cited by other articles:
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Advertisement | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||