| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Articles |
Isala Klinieken Zwolle, Departments of Pediatrics (E.L.T.v.d.A.), Internal Medicine (A.A.M.F.), Clinical Chemistry (A.P.A., H.E.), 8025 AB Zwolle, The Netherlands; and Rudolf Magnus Institute for Neurosciences, Department of Medical Pharmacology, Utrecht University (M.N., R.Z., D.d.W., J.P.H.B.), 3508 TA Utrecht, The Netherlands
Address all correspondence and requests for reprints to: Dr. J. Peter H. Burbach, Rudolf Magnus Institute for Neurosciences, Utrecht University, P.O. Box 80040, 3508 TA Utrecht, The Netherlands. E-mail: j.p.h.burbach{at}med.uu.nl
Abstract
Familial neurohypophysial diabetes insipidus (FNDI) is an autosomal
dominant trait in which expression of a mutant vasopressin prohormone
reduces vasopressin production. We investigated the NP85 Cys
Gly
mutant vasopressin prohormone in a large kindred in The Netherlands. We
demonstrate that growth retardation is an important early sign in two
children from this kindred, which recuperates by substitution therapy
with 1-desamino-8-D-arginine
vasopressin. To obtain clues about the basis for the
dominant inheritance of FNDI, we analyzed the trafficking and
processing of the mutant vasopressin prohormone in cell lines by
metabolic labeling and immunoprecipitation. The mutant vasopressin
prohormone was retained in the endoplasmic reticulum and thus was not
processed to vasopressin. This defect was not caused by dimerization of
the vasopressin prohormone via its unpaired cysteine residue. High
level expression of the mutant vasopressin prohormone in cell lines
resulted in strong accumulation in the endoplasmic reticulum and an
altered morphology of this organelle. We hypothesize that disturbance
of the endoplasmic reticulum results in dysfunction and ultimately cell
death of the cells expressing the mutant prohormone. Our data support
the hypothesis that FNDI is a progressive neurodegenerative disease
with delayed onset of symptoms. Its treatment requires early detection
of symptoms for which growth parameters are useful.
FAMILIAL NEUROHYPOPHYSIAL diabetes insipidus (FNDI) is characterized by excessive thirst, excessive urine production, and a lack of the day-night rhythm of urine production (1, 2, 3). The disease is caused by a mutation in the vasopressin (VP) prohormone gene, resulting in a defective preprohormone and a deficiency of arginine VP. FNDI reduces not only the plasma levels of VP, but also the amount of VP stored in the posterior pituitary, as evidenced by the absence of the characteristic high signal intensity of the posterior pituitary in T1-weighted magnetic resonance images of most FNDI patients (4, 5, 6, 7, 8 ; see also 9). At present, 36 different mutations have been identified in FNDI families (8, 10, 11, 12, 13, 14, 15, 16, 17). All except 1 (14) are inherited in an autosomal dominant mode. Moreover, a delayed onset of the disease is characteristic (1, 2, 3, 10). The disease is not manifest during the first months or years of life and progressively develops thereafter. The delayed onset and the autosomal dominant inheritance indicate that the mutant VP preprohormone gradually induces the lack of VP secretion despite the presence of the normal allele. It has been speculated that accumulation of the mutant prohormone may block the secretory pathway or can result in degeneration of VP neurons (11, 18). The latter is supported by postmortem studies showing, respectively, a strong reduction in the amount of magnocellular neurons in the supraoptic and paraventricular nuclei of the hypothalamus of FNDI patients (19, 20, 21, 22).
FNDI mutations have been found within the signal sequence, VP, and
neurophysin (NP) moieties of the VP preprohormone (Fig. 1
) (8, 10, 11, 12, 13, 14, 15, 16, 17). The
mutations within the signal sequence reduce signal sequence cleavage
and, correspondingly, endoplasmic reticulum (ER) exit of the mutant
prohormone (18, 23, 24). Most of the FNDI mutations (30 of
36) have been identified within the NP domain of the precursor. NP is
an intracellular binding protein for VP. It has been postulated that
this binding is required for appropriate folding and targeting of the
prohormone to the regulated secretory pathway (25, 26). NP
is tightly folded. Although it is only 93 amino acids long, it contains
7 disulfide bridges. In addition, the central core (amino acids 10–85)
consists of 2 domains of 4 antiparallel ß-strands, separated by an
-helix and a loop (25, 26, 27, 28). Except for 1, all FNDI
mutations within NP are present within this highly folded central core,
suggesting that they affect the folding of the NP domain. The same
hypothesis can be made for the VP2TyrHis mutation within the VP
moiety. Tyrosine at position 2 of VP is essential for binding of VP to
NP, which, in turn, enhances the stability of NP (25, 26, 27, 28).
In agreement with this, the only mutation causing recessive FNDI
(VP7ProLeu within VP) changes a residue within VP that is not
required for NP binding (25, 26, 27).
|
Subjects and Methods
Subjects
A family with FNDI was studied. We recently described the
genetic screening of 1 branch of the family (17). The
pedigree of this branch (branch A) is shown in Fig. 2A. In this manuscript 3 more branches of
this family were studied. The family includes 56 individuals
symptomatic with polyuria and polydipsia from 4 generations (22 in
branch A, 14 in branch B, 15 in branch C, and 5 in branch D). All
family members reported an onset in early infancy. The pedigree shows
an autosomal dominant inheritance pattern of clinical overt diabetes
insipidus.
|
Subject B-IV-9 was diagnosed with FNDI at the age of 7 yr. The presence
of the nucleotide (nt) 2110 TG mutation in one of the VP prohormone
alleles of this subject was confirmed by PCR and restriction analysis
with Sau96I (Fig. 3). The
details of this method were described previously (17). The
same method was employed to analyze the presence of the nt 2110 TG
mutation in the VP prohormone genes of the other members of branches B,
C, and D of the kindred.
|
The human VP gene (GenBank accession no. M11166) was a gift from Dr. Jim Battey (Bethesda, MD) (29). The gene was cloned into the hygromycin B resistance-conferring expression plasmid pRSVhyg (Dr. Van Tol, University of Toronto, Toronto, Canada) as previously described (11). The resulting plasmid was named phygHVP.
A PCR fragment encompassing the last part of the second intron and
first part of the third exon of the NP85CG mutant VP prohormone gene
was obtained by PCR of patient material with the following primers:
NAR1 M13FW, 5'-GTTGTAAAACGACGGCCAGCCGGCAGGGAGGGTGTGGG-3';
and AVP3-M13BK,
5'-GAAACAGCTATGACCATGCCTCTCTCCCCTTCCCTCTT-3'
(17). These primers amplify a 409-bp PCR fragment
extending from nt 2008 in intron 2 to nt 2379 in intron 3 of the VP
prohormone gene. Due to a polymorphism in the second intron, the
5'-primer is specific for the mutant gene (17), resulting
in amplification of the mutant gene fragment only. The
underlined primer sequences are M13 sequences, which were
introduced because the primers were previously employed for primer-dye
cycle sequencing with M13 primers (17). The PCR fragment
was digested with NarI, and the resulting 160-kb
NarI fragment (nt 2059–2219) containing the nt 2110 TG
mutation, was substituted for the wild-type NarI fragment in
the phygHVP plasmid via a series of three cloning steps. The cloned
NarI fragment was completely sequenced to verify the
presence of the nt 2110 TG mutation and the absence of mutations
induced by the PCR.
Culture and transfection of cells
The rat adrenal pheochromocytoma cell line PC12 transfected with mouse prohormone convertase 2 (PC12/PC2) was a gift from Dr. Sharon Tooze (30) and was cultured in poly-L-lysine-coated flasks in DMEM (Life Technologies, Inc., Gaithersburg, MD) supplemented with 4 mmol/L glutamine, nonessential amino acids (Life Technologies, Inc.), 200 IU/mL penicillin, 200 µg/mL streptomycin, 5% FCS, and 10% horse serum (Life Technologies, Inc.) in an atmosphere of 10% CO2.
The mouse pituitary tumor cell line AtT-20 (ATCC CCL89, American Type Culture Collection, Manassas, VA) and the mouse neuroblastoma cell line Neuro-2A (ATCC CCL 131) were cultured in DMEM (Life Technologies, Inc.) supplemented with 4 mmol/L glutamine, nonessential amino acids (Life Technologies, Inc.), 200 IU/mL penicillin, 200 µg/mL streptomycin, and 10% FCS (Life Technologies, Inc.) in an atmosphere of 5% CO2.
PC12/PC2 and AtT-20 cells were transfected with gene constructs cloned into the pRSVhyg vector by incubating cells plated in 6-cm dishes overnight in 2 mL serum-free medium with 15 µg plasmid DNA and 25 µL Lipofectamine (Life Technologies, Inc.) as previously described (30). Stable PC12/PC2 clones were selected in medium supplemented with 100 µg/mL hygromycin B (Roche Molecular Biochemicals, Indianapolis, IN), and maintained in medium with 50 µg/mL hygromycin B. Stable AtT-20 clones were selected on and maintained in medium with 200 µg/mL hygromycin B. After selection, hygromycin-resistant clones were pooled and employed for the experiments described here. Analysis of (pooled) clones from independent PC12/PC2 transfections gave similar results. In addition, similar results were obtained for clones expressing different levels of wild-type or mutant prohormone.
Neuro-2A cells were transiently transfected with gene constructs cloned into the pRSVhyg vector with the calcium phosphate method (31).
Antisera
Rabbit antiserum D7 raised against swine NP and cross-reacting with human NP was a gift from Dr. Iain Robinson (32). Rabbit antiserum HenryK was raised against rat NP and displays cross-reactivity with human NP (33). Monoclonal antibody 1D3 was raised against the C-terminus of the ER protein protein-disulfide-isomerase and recognizes both protein-disulfide-isomerase and calreticulin (34).
Labeling of cells, immunoprecipitation, and gel electrophoresis
Labeling of cells with 25 µCi [35S]cysteine (ICN Biomedicals, Inc., Costa Mesa, CA) for the times indicated and immunoprecipitation with the D7 antiserum were performed as previously described (11), except for the supplementation of medium with 10% FCS instead of 5% FCS and 10% horse serum in the case of AtT-20 cells. Where indicated, endoglycosidase H digestions of the immunoprecipitates were performed with 4 U endoglycosidase H (Roche Molecular Biochemicals) for 32 h at 37 C in digestion buffer (50 mmol/L sodium citrate, pH 5.5, and 0.2% SDS). Immunoprecipitated proteins were analyzed by 10% Tricine-SDS-PAGE (35). Depending on the experiment, nonreducing or reducing gel electrophoresis was employed. Nonreducing gel electrophoresis gave sharper bands, thus improving the separation of proteins. However, because the separation between the VP prohormone and the slightly slower migrating background band from the medium was better on reducing gels, we used reducing SDS-PAGE when we wanted to focus at the VP prohormone (in the case of endoglycosidase H digestions). The MW-SDS-17S marker (Sigma, St. Louis, MO) was used as a molecular weight marker. Protein gels were dried, and radioactive bands were detected by analysis on a BAS1000 phosphorimager (Fujix, Fuji, Tokyo, Japan).
Immunofluorescence
Cells were grown on coverslips and transiently transfected for 48 h. Fixation and immunofluorescence were performed as previously described (36, 37). Immunofluorescence was performed with a 1:1000 dilution of the HenryK antiserum followed by a 1:300 dilution of fluorescein-conjugated donkey antirabbit Igs (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA). In the case of double labeling, the second labeling was performed with a 1:50 dilution of the 1D3 monoclonal antibody (hybridoma supernatant) followed by a 1:2400 dilution of Cy3-conjugated donkey antimouse Igs (Jackson ImmunoResearch Laboratories, Inc.). The coverslips were mounted with DABCO/Mowiol and examined with a x63 Planapo objective on a Leitz DMIRB fluorescence microscope (Leica Corp., Voorburg, The Netherlands) interfaced with a Leica Corp. TCS4D confocal scanning laser microscope (Leica Corp., Heidelberg, Germany).
Results
Clinical characterization of individuals affected with FNDI
Recently, a novel diabetes insipidus mutation was identified in a
Dutch kindred (Fig. 2A
) (17). Subjects A-IV-21 and
A-III-18 were the first patients in this family who visited the
out-patient clinics. A-IV-21 was a 2-yr-old boy, who presented with
failure to thrive. In this subject, formal assessment of the diagnosis
of diabetes insipidus was carried out in the hospital. The patient was
subjected to a standard water deprivation test with hourly measurement
of plasma and urinary osmolality (38, 39). The test was
concluded when plasma osmolality rose to 315 mosmol/kg while urinary
osmolality as well as the plasma VP concentration remained relatively
low (respectively, 400 mosmol/kg and 1.3 pg/mL). Normally the plasma VP
concentration is higher than 8 pg/mL at this plasma osmolality. The boy
had more than a 50% increase in urinary osmolality after the
administration of exogenous VP, confirming the diagnosis of
neurohypophysial diabetes insipidus.
In response to the water deprivation test, subject A-III-18, the father of the boy, had a rise in plasma osmolality to 307 mosmol/kg with a maximum urinary osmolality of 305 mosmol/kg and plasma VP of 0.33 pg/mL (normally >5 pg/mL). His symptoms were relieved by the use of 1-desamino-8-D-arginine vasopressin.
Identification of the VP prohormone gene mutation in affected individuals
We previously identified a mutation in the VP prohormone gene by
sequencing all three exons of this gene for the two affected
individuals, A-III-18 and A-IV-21 (17). The DNA sequence
indicated heterozygosity for the normal sequence and a TG transition
mutation at nucleotide 2110 (codon 116) of the VP prohormone gene
(17). This mutation encodes a cysteineglycine
substitution at amino acid position 85 of the NP domain of the
preprohormone (NP85CG; Fig. 1). Samples from two unaffected
individuals from the family demonstrated homozygosity for the normal
DNA sequence for all three exons of the VP prohormone gene
(17).
The mutation detected by direct sequencing of the PCR-amplified VP
prohormone gene was confirmed by restriction endonuclease analysis. PCR
fragments of 257 bp containing exon 3 of the normal VP prohormone
allele have two Sau96I restriction sites and yield three
restriction fragments of, respectively, 160, 87, and 10 bp. The TG
mutation introduces an extra Sau96I restriction site
and produces four fragments of 160, 64, 23, and 10 bp
(17). Thirty-one members of branch A of the family were
tested by restriction analysis (17) (Fig. 2A).
In this study 3 additional branches of this family were screened,
encompassing 61 individuals (23 of branch B, 30 of branch C, and 8 of
branch D; Fig. 2, B–D). For 7 subjects, the resulting restriction
fragment patterns are displayed in Fig. 3. All affected subjects were
heterozygous for the mutation, whereas none of the healthy family
members displayed the mutation (Fig. 2). The results are consistent
with autosomal dominant inheritance of FNDI in this family.
Diabetes insipidus and growth in children
Subject A-IV-21 was a 2-yr-old boy who presented with
failure to thrive. Retrospectively, this subject showed decelerated
weight gain as an early sign from the age of 9 months and a retarded
length from the age of 1.5 yr (Fig. 4A).
Closer study of his diet revealed a shortage of calorie intake due to
excessive fluid intake. He drank 4 L/day. He was successfully treated
with DDAVP. Within 3 months after starting treatment his
weight rose dramatically to a normal level. His length curve normalized
slowly (Fig. 4A). Thus, the growth-related symptoms appear as early,
indirect signs of diabetes insipidus.
|
, lane 7). He had not presented in
the clinic because there were no complaints, although he had been
drinking progressively more for a few years. At the age of 7 yr he
drank 2.5–5 L/day. His growth curve already demonstrated growth
retardation from the age of 3 yr (Fig. 4B). Treatment of diabetes
insipidus with DDAVP normalized polydipsia and improved
his appetite and his weight gain (Fig. 4B).
The NP85CG mutation abolishes processing and secretion of the VP
prohormone
To prove that the NP85CG mutation results in a defective VP
prohormone and to investigate the dominant nature of FNDI caused by
this mutation, we examined intracellular transport and processing of
the mutant VP prohormone in cell lines. PC12/PC2 cells, stably
transfected with wild-type or NP85CG mutant VP prohormone gene, were
metabolically labeled, and NP-containing proteins were
immunoprecipitated and analyzed. PC12/PC2 cells possess a regulated
secretory pathway and are able to process the wild-type VP prohormone
(11) (Fig. 5).
Analysis of NP-containing proteins in cells stably expressing the
wild-type prohormone demonstrates that most of the prohormone has been
processed to the processing intermediate VP-NP or to the final product,
NP (Fig. 5, lane 3). Of these proteins, NP is selectively retained
intracellularly, indicating sorting into and storage in the regulated
secretory pathway (Fig. 5, lane 4). In contrast, in cells expressing
the NP85CG mutant VP prohormone, the only NP-containing protein
present is the intact prohormone (Fig. 5, lane 5). No processing could
be detected. In addition, whereas part of the wild-type VP prohormone
was secreted via the constitutive secretory pathway, this secretion was
abolished by the NP85CG mutation (Fig. 5, lanes 4 and 6). Although
the partial missorting of regulated secretory proteins to the
constitutive pathway is an imperfection of cell lines (11, 18, 40, 41, 42), the absence of secretion in the case of the NP85CG
mutant VP prohormone suggests an intracellular transport defect for
this prohormone.
|
G VP prohormone
Recently, defective processing and intracellular transport of an
FNDI VP prohormone with a mutation in the signal sequence were
demonstrated to coincide with the formation of disulfide-linked
multimers by this mutant prohormone (24). Because the
NP85CG mutant VP prohormone possesses an uneven number of cysteines
and thus at least one unpaired cysteine, we analyzed whether this
mutant forms disulfide-linked multimers. Analysis of NP-containing
proteins by nonreducing gel electrophoresis gave similar results as
that by reducing gel electrophoresis, indicating the absence of
disulfide-linked dimers or multimers (Fig. 5). The only higher
molecular weight bands observed after nonreducing gel electrophoresis
were also obtained from the medium of PC12/PC2 cells that had been
transfected with the empty vector, indicating that these are background
bands.
Comparison of the processing and secretion defect of the NP85CG
VP prohormone and other FNDI VP prohormones
Recently, we demonstrated different FNDI mutant VP prohormones to
vary in the degree of processing deficiency in PC12/PC2 cells
(11). Whereas no processing was observed for NP14GR and
NP65GV prohormones, which both have their mutations within the
ß-pleated sheet domains of the NP moiety, some processing was
observed for the NP47
E and NP47EG mutant prohormones, which have
their mutations within the -helix. The processing defect was the
least strong for the NP57GS mutation, which is located within the
loop connecting the -helix and the second ß-pleated domain. Thus,
the processing defect of the NP85CG mutant is among the most severe
that we have observed to date. To examine whether this secretion and
processing defect is as severe in a cell line that expresses higher
levels of proprotein convertases (43) and thus processes
the VP prohormone more efficiently, we expressed wild-type and mutant
prohormones in the cell line AtT-20. Whereas in this cell line the
majority of the wild-type VP prohormone had been processed to VP-NP and
NP during the metabolic labeling, no processing was observed for the
NP85CG mutant (Fig. 6, lanes
3–6). In addition, the NP85CG mutation abolished secretion of the
VP prohormone. Similar results were obtained for the NP14GR mutant
VP prohormone (Fig. 6, lanes 7 and 8). The NP47EG mutant prohormone
demonstrated some processing and secretion, albeit at much reduced
amounts compared with the wild type. These data indicate that the
processing and secretion defect observed for the NP85CG mutant
prohormones is among the most severe observed for FNDI mutant
prohormones.
|
G VP prohormone is retained in the ER
Because other FNDI VP prohormones have been demonstrated to be
retained in the ER (11, 18, 24, 42, 44), we examined
whether this was also the case for the NP85CG mutant. NP-containing
proteins were immunoprecipitated from the cell lysates of PC12/PC2
transfectants and either mock-treated or treated with endoglycosidase
H. Whereas most of the wild-type prohormone was resistant to
endoglycosidase H digestion, all of the NP85CG mutant
prohormone was deglycosylated (Fig. 7).
This indicates that the NP85CG mutant prohormone still carries the
high mannose-type sugar that is present on proteins within the
ER. In contrast, most of the wild-type prohormone has left the ER, and
its glycan has been converted to a mature type glycan in the Golgi
apparatus. These data indicate that the NP85CG mutant VP prohormone
is retained in the ER.
|
G VP prohormone induces large
accumulates in the ER and a changed morphology of this organelle
Because the magnocellular neurons that synthesize VP in the
supraoptic nucleus and paraventricular nucleus of the
hypothalamus express very high amounts of the VP prohormone, we
examined the effect of very high level expression of the NP85CG
mutant VP prohormone. We did so by immunocytochemistry of Neuro-2A
neuroblastoma cells transiently transfected with plasmids encoding
either wild-type or mutant prohormone. The NP-containing proteins in
cells transiently expressing wild-type prohormone were homogeneously
stained throughout the entire cytoplasm of the cells. In contrast, in
approximately 25% of cells transiently expressing the NP85CG mutant
prohormone, staining concentrated at distinct sites within the
cytoplasm (Fig. 8A). Costaining with an
antibody against the ER marker protein-disulfide-isomerase demonstrated
that the large sites of accumulation of the NP85CG prohormone were
located within the ER (Fig. 8B). In addition, in cells displaying these
large accretions of mutant prohormone, a changed morphology of the ER
was observed. Whereas the ER marker demonstrated the normal reticulate
staining in nontransfected cells, in transfected cells the marker
concentrated in the same large accretions as the mutant prohormone
(Fig. 8B, middle panel). We hypothesize that these large
accretions might also form in magnocellular neurons expressing the
NP85CG mutant VP prohormone and that this will result in severe
dysfunction of the cells or even cell death due to the strong
disturbance of the ER in these cells.
|
In this study we investigated the mutant NP85CG VP prohormone
from individuals of a Dutch kindred with FNDI. We recently identified
one branch of this family (17). We now extend the data to
three other branches and prove the cell biological deficit of this
mutant precursor. The kindred we identified is the largest kindred with
FNDI reported to date. Heterozygosity for a mutation in the VP
prohormone gene was identified, and was confirmed in all affected
individuals by restriction endonuclease digestion.
The mutation cosegregated with clinical symptoms of the disease. Although some family members believed they had no complaints, personal history revealed polydipsia, and in the children signs of growth retardation were found. The mutation affects males and females and was transmitted to approximately 50% of those at risk. These findings are consistent with an autosomal dominant trait.
This family demonstrates a typical clinical presentation of FNDI, with onset of symptoms of polyuria and polydipsia in early infancy. It is well known that the excessive drinking associated with diabetes insipidus can decrease food and thus caloric intake and so influence growth and development (45, 46, 47). However, this effect is most severe when the diabetes insipidus starts at birth and is usually unnoticed in patients with FNDI. Of the four generations of FNDI patients we describe, only one patient presented in the clinic with a failure to thrive. Nonetheless, we demonstrate that a deceleration in growth can be an early sign in childhood of FNDI and that this deceleration can also be measured in a patient in whom growth retardation was not problematic. Symptoms and growth recover during treatment with DDAVP. In the clinical management of children with an inherited FNDI mutation, this deceleration in growth can be employed as an early symptom. It may mark the moment that substitution therapy with DDAVP is required.
To proof that the NP85CG mutation results in a defective VP
prohormone and thus underlies the symptoms of FNDI in this kindred, the
cell biological properties of the mutant prohormone were evaluated.
Analysis of the intracellular transport of the mutant VP prohormone in
cell lines demonstrated that the mutation abolished ER exit and thus
processing of the VP prohormone. The ER retention caused by the
NP85CG mutation was almost absolute and stronger than that caused by
FNDI mutations in the -helix (NP47EG and NP47E) or in the loop
connecting the -helix with the second ß-pleated sheet domain
(NP57GS) that were examined previously (11). It was
comparable in severity with the retention caused by two mutations
within a ß-pleated sheet domain (NP14GR and NP65GV)
(11). This suggests that the folding defect caused by the
NP85CG mutation is more severe than that caused by the NP47EG,
NP47E, and NP57GS mutations.
Although the NP85CG mutant VP prohormone contains 15 cysteine
residues and at least 1 unpaired cysteine, it did not form
disulfide-linked homo- or heterodimers. This excludes dimer formation
as the cause of ER retention of the NP85CG VP prohormone. This is
remarkable, because formation of disulfide-linked oligomers was proven
for and implicated as the cause of ER retention of the
SP(M1P2D3T4) FNDI VP prohormone (24). In the gene
encoding this mutant, the G of the ATG start codon has been deleted.
Protein translation now starts at Met5 of the SP moiety, resulting in a
VP preprohormone with a truncated signal peptide that is not cleaved.
Due to the presence of a Cys residue at position 8 of the SP moiety,
the SP(M1P2D3T4) VP preprohormone contains an uneven number of
cysteine residues, like the NP85CG VP prohormone we investigated.
However, whereas the majority of the SP(M1P2D3T4) VP preprohormone
was present as disulfide-linked oligomers or aggregates, no
disulfide-linked NP85CG VP prohormone was detected. We conclude the
existence of at least two different causes of ER retention of FNDI VP
(pre)prohormones with at least one unpaired cysteine residue. Beuret
et al. (24) elegantly demonstrated that
disulfide-linked aggregate formation by the SP(M1P2D3T4) VP
preprohormone was probably due to a failure in formation of the
disulfide bridge within the VP moiety. NP folding seemed to be intact
in this mutant. In our mutant (and in other prohormones with mutations
within the NP moiety) it is probable that the folding problem is
located within the NP moiety.
The mutant VP prohormone demonstrates an enhanced tendency to form
large accumulates within the ER when it is expressed at very high
levels. Similar accretions have been observed in vivo in the
magnocellular neurons of rat expressing a mutant prohormone resulting
from the nonhomologous crossing over of the VP and oxytocin genes
(48, 49). They have been demonstrated to consist of
accumulations of globular aggregates in dilated saccules of the rough
ER (48, 49). A similar change in ER morphology has
recently been observed in the magnocellular neurons of transgenic rats
expressing a human diabetes insipidus prohormone (50).
Because accumulation of mutant VP prohormone causes an aberrant ER
morphology (11, 48, 49, 50), it is likely to disturb the
function and possibly even the viability of the cell. A disturbed
function or viability of magnocellular neurons caused by FNDI mutant VP
prohormones would explain the autosomal dominant inheritance of FNDI.
In this light, FNDI may even be considered a neurodegenerative disease.
Moreover, the delayed onset of the disease at only a few months or
years of age could be explained by the time required for the mutant
prohormone to accumulate in high enough amounts to change ER morphology
and block cell function. Several observations support this possible
mechanism for FNDI. First, a strong reduction in the amount of
magnocellular neurons in the paraventricular and supraoptic
nuclei of the hypothalamus was observed in autopsies of FNDI
patients (19, 20, 21, 22). Second, human FNDI VP prohormones were
demonstrated to cause cytotoxicity in differentiated Neuro-2A cells and
to decrease the secretion of the wild-type VP prohormone (18, 51). Third, cytotoxicity caused by ER accumulation of a mutant
protein has been proven for another disease,
1-antitrypsin deficiency (52, 53). The proposed mechanism for the dominance of FNDI would
predict that the only recessive FNDI mutation identified (VP7PL),
would not result in ER retention of the mutant VP prohormone. Instead,
the prohormone would be normally transported, and the symptoms in
homozygotes would be due to the reduced biological activity of the
mutant VP peptide (14). In agreement with this, the
mutation changes a residue within VP that is not required for NP
binding (25, 26, 27) and, in addition, is oriented toward the
solvent in the x-ray crystals of NP II and oxytocin (27),
suggesting a minor influence of this residue on folding of the oxytocin
and VP prohormones. It would be interesting to determine whether the
VP7PL mutant VP prohormone is indeed transported efficiently out of
the ER.
As described above, the severity of ER retention of the mutant VP prohormone varies between different FNDI mutations. However, there is probably no linear correlation between the severity of ER retention and the severity of the disease (i.e. the time of onset of FNDI). The time needed for formation of large accretions of mutant VP prohormone in the ER, resulting in an aberrant ER morphology and proposed dysfunction of the cell, will not only be dependent on the extent of ER retention, but also on the expression level and the extent of intracellular degradation of the mutant prohormone. This also predicts that the onset of the disease will display individual variations. Drinking habits and thus VP prohormone production might vary between individuals. In addition, individual differences in expression levels of proteins involved in degradation of mutant proteins have been observed (52). Indeed, even for a certain FNDI mutation, large differences were found in the onset of the disease (54). This concerns both differences between two kindred expressing the same FNDI mutation and differences between the affected members of one kindred. In the clinical management of persons afflicted with the FNDI mutation it is essential to note early symptoms and to provide adequate substitution therapy. This study shows that observation of growth may provide early parameters.
Acknowledgments
We express our gratitude to Dr. Iain Robinson for providing us with the last batches of the D7 antiserum, to Dr. Sharon Tooze for her kind gift of the PC12/PC2 cells, to Dr. (Ph.D.) Bill North for the Henry-K antiserum, to Dr. Peter van der Sluijs for providing us with the 1D3 monoclonal, and Dr. Fuller for the permission to use it, to Dr. Jim Battey for his kind gift of the cloned human VP prohormone gene, and to Dr. Oscar Schoots for the pRSVhyg vector and AtT-20 cells.
Footnotes
1 This work was supported by Research Grant NWO-MW-903-46-150 from
the Council for Medical and Health Research of the Netherlands
Organization for Scientific Research (to M.N.) and Research Grant
GRN-94002 from the Glaxo Research Foundation Netherlands (to R.Z.).
Part of this work was supported by a grant from the Hersenstichting
Nederland to this consortium (Grant 8F00.09). 
2 M.N. and E.L.T.v.d.A. share first authorship.
3 Present address: Sophia Children’s Hospital, Department of
Pediatric Endocrinology, Rotterdam, The Netherlands.
Received November 22, 2000.
Revised February 27, 2001.
Accepted March 19, 2001.
References
Val in
neurophysin II. J Clin Endocrinol Metab. 82:3643–3646.1-antitrypsin Z, in the
endoplasmic reticulum requires proteasome activity. J Biol Chem. 271:22791–22795.Val in the signal peptide of the arginine
vasopressin/neurophysin II/copeptin precursor. J Clin Endocrinol
Metab. 82:51–56.This article has been cited by other articles:
 |
|
 |
|
  J. H. Christensen, C. Siggaard, T. J. Corydon, G. L. Robertson, N. Gregersen, L. Bolund, and S. Rittig Differential Cellular Handling of Defective Arginine Vasopressin (AVP) Prohormones in Cells Expressing Mutations of the AVP Gene Associated with Autosomal Dominant and Recessive Familial Neurohypophyseal Diabetes Insipidus J. Clin. Endocrinol. Metab., September 1, 2004; 89(9): 4521 - 4531. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. T. Wahlstrom, M. J. Fowler, W. E. Nicholson, and W. J. Kovacs A Novel Mutation in the Preprovasopressin Gene Identified in a Kindred with Autosomal Dominant Neurohypophyseal Diabetes Insipidus J. Clin. Endocrinol. Metab., April 1, 2004; 89(4): 1963 - 1968. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Rittig, C. Siggaard, M. Ozata, I. Yetkin, N. Gregersen, E. B. Pedersen, and G. L. Robertson Autosomal Dominant Neurohypophyseal Diabetes Insipidus due to Substitution of Histidine for Tyrosine2 in the Vasopressin Moiety of the Hormone Precursor J. Clin. Endocrinol. Metab., July 1, 2002; 87(7): 3351 - 3355. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
, 
