| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Pancreas Transplant Unit, Prince of Wales Hospital and University of New South Wales, Sydney, New South Wales 2031, Australia
Address all correspondence and requests for reprints to: Bernard Tuch M.D., Ph.D., Pancreas Transplant Unit, Department of Endocrinology, Diabetes, and Metabolism, Prince of Wales Hospital, High Street, Randwick, New South Wales 2031, Australia. E-mail: b.tuch{at}unsw.edu.au
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
As an alternative to monitoring insulin levels we have examined the possibility of measuring the hormones produced by the pancreatic polypeptide (PP) cell in the pancreas. The PP cell is the first and only mature pancreatic endocrine cell to develop in the pig before birth (8). All other pancreatic hormones are colocalized in one or more cells during this time. It is only after birth that these other endocrine hormones are found in separate cells. Peptide hormones are generated through posttranslational modification of larger precursor molecules. This processing results in the formation of several stable peptide products. Two hormones are produced in the PP cell, PP and pancreatic icosapeptide (PI), from a 95-amino acid precursor (9, 10).
We have previously shown that levels of porcine PP can probably be used to monitor the viability of fetal pig pancreatic tissue transplanted into mice (11). Levels of PP were measured by RIA using an antibody that bound more to porcine than to mouse PP; these two peptides are six amino acids different from each other. Such an RIA cannot be used to measure levels of porcine PP in human serum, because pig and human peptides differ by only two amino acids. Because of the cross-over in binding, we examined the possibility of measuring the more heterogeneous PI, which should be secreted in an equimolar basis with PP.
The structures of human, ovine, and canine, but not porcine, PI have been described previously (1, 9). Human PI differs from sheep and dog PI by six and nine amino acids, respectively. Theoretically, the composition of porcine PI should differ markedly from that of human PI. It is generally believed that the structure of a biologically relevant peptide is better preserved through evolution than a peptide with a limited action. Thus, the structure of insulin is reasonably well preserved between species, but that of C peptide is not.
We set out to determine the sequence of porcine PI, develop an assay to measure it, and show how it is secreted physiologically. We also determined its usefulness in monitoring the viability of xenotransplanted fetal pig pancreatic cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Raising antibodies to porcine PI
The amino acid composition of porcine PI was deduced from the
cDNA sequence determined, and the peptide was then synthesized (Auspep,
Parkville, Australia). Next, polyclonal antibodies to synthetic porcine
PI were raised in rabbit and sheep (Institute of Medical and Veterinary
Science, Adelaide, Australia). The peptide was conjugated to keyhole
limpet hemocyanin before immunization. Preimmune and immune sera were
tested for binding with the synthetic peptide by a direct ELISA. For
this, 96-well microtiter plates (Maxisorp immunoplates, Nunc
International, Naperville, IL) were coated at 4 C overnight with 10
µg/ml synthetic peptide in coating buffer (0.1 M
NaHCO3 buffer, pH 8.3). The wells were blocked
for 1 h at room temperature using PBS containing 1% BSA (blocking
buffer). Sera diluted to 102,
103, 104, and
105 by PBS containing 0.25% BSA and 0.1% Tween
(assay buffer) were added to the wells for 2 h at room
temperature, followed by a similar incubation with horseradish
peroxidase (HRP) conjugated to rabbit or sheep IgG as the secondary
antibody. The wells were washed before and after incubation with PBS
containing 0.1% Tween (wash buffer). Color development was achieved
with 3,3',5,5'-tetramethylbenzidine peroxidase substrate (KPL
Laboratories, Inc., Gaithersburg, MD), with 0.18 M
H2SO4 added to stop the
development (stop solution). Color intensity was measured using a
Microplate Reader (Multiscan, Labsystems, Helsinki, Finland).
IgG was purified from sera using protein A (for rabbit sera) or protein
G (for sheep sera). Sera were mixed 1:1 with 0.05 M Tris
buffer, pH 7.2, and centrifuged for 10 min at 10,000 rpm. Supernatant
was loaded on to a column of protein A or protein G conjugated to
agarose (Pierce Chemical Co., Rockford, IL). The column
was then washed with 10 column vol using the same buffer to remove
unbound and nonspecifically bound proteins and other components. IgG
bound to the column was eluted by using 0.05 M glycine, pH
2.5. Fractions of 1 ml were collected, and protein levels in the
fractions were determined by Bradford assay. Typically, 80% of the
eluted IgG was present in a single fraction, and all IgG was present in
three fractions. These three fractions (3 ml) were pooled and
concentrated by centrifuging in a Centriprep-30 filter (Millipore Corp., Bedford, MA) at 1,500 x g. The
concentrated volume (
0.3 ml) was dialyzed against PBS using a
dialysis cassette (Pierce Chemical Co., Rockford, IL) for
1 d. Protein levels were determined in the purified IgG, followed
by another direct ELISA to confirm the activity.
Antigen inhibition assay for PI
An antigen inhibition assay was developed to measure PI levels
in samples using the biotin-neutravidin system for amplification of the
signal. For this, IgG, purified as described above, was conjugated to
biotin according the manufacturer’s instructions (Pierce Chemical Co.).
The antigen inhibition assay was set up as follows. Ninety-six-well microtiter plates were coated overnight using 10 µg/ml of the peptide in coating buffer. Plates were blocked for 1 h at room temperature using the blocking buffer. A 2-fold dilution series of PI standards (staring from 1000 pg/ml) was prepared using the assay buffer. One hundred microliters of the standards or samples and 400 µl biotin-IgG (1.2 µg/ml) were mixed in microfuge tubes for 2 h at room temperature on a rotator. The concentration of biotin-IgG required for the antigen inhibition assay was determined by a dilution assay. Triplicates of 100 µl/well were added to the plate (after blocking) and incubated for 1 h at room temperature. This was followed by a 1-h room temperature incubation with a Neutravidin (1.0 µg/ml) and biotin-HRP (250 ng/ml) mixture (100 µl/well). Washing of wells before and after incubations was carried out with the washing buffer. The substrate, 3,3',5,5'-tetramethylbenzidine peroxidase, was added (100 µl/well), and the color development reaction was allowed to proceed for 1 h. Then the stop solution was added, and OD was measured at 450 nm.
Pig studies
The following experiments were carried out with the approval of
the animal care and ethics committee of our institution. Juvenile Large
White Landrace pigs were anesthetized, and a catheter inserted into the
jugular vein (12, 13). Animals were fasted overnight, and
the next morning an insulin tolerance test was performed with an iv
injection of 0.1 U soluble insulin/kg body weight. Blood was collected
at 0, 5, 10, 15, 20, 30, 35, 40, 45, 50, and 60 min for measurement of
glucose, PP, and PI levels. PP levels were determined using an RIA kit
(Phoenix Pharmaceuticals, Inc., Belmont, CA), which uses a
human standard (77% cross-reactivity with porcine PP).
Fetal pig pancreas
Fetal pigs were removed from freshly killed Large White Landrace
sows of gestational age 75–100 d (normal period of gestation, 118
d). The fetal pigs were air-freighted on ice to reach the investigators
in Sydney within 4 h of death. Pancreases were removed from the
fetuses and digested to produce islet-like cell clusters (ICCs) by
exposure to collagenase P, as described previously (2, 14). These clusters of cells round up after 2–3 d in culture
medium RPMI containing 11.1 mM glucose supplemented with
antibiotics and 10% FCS.
Measurement of PI in fetal pancreatic tissue
Fetal pancreatic tissue was homogenized in acid-ethanol,
followed by overnight incubation at 4 C. After centrifugation at 10,000
rpm for 10 min, the supernatant was neutralized by adding 1.0
mM Na2CO3, pH
8.5, and assayed for levels of PI.
Secretion of PI from pig tissue in vitro
ICCs were removed on the third day of culture and placed in
groups of 200 in Eppendorf tubes. They were washed twice with
HEPES-buffered Earle’s medium, which consists of 124 mM
NaCl, 5.4 mM KCl, 1.8 mM
CaCl2, 0.8 mM
MgSO4, 1 mM
NaH2PO4, 14.3
mM NaHCO3, and 2 mM
glucose supplemented with 0.2% BSA (Sigma) and 10
mM HEPES. The ICCs were then washed with medium containing
the stimulus 10 mM carbamylcholine chloride (carbachol;
Sigma) before being exposed to it for 2 h. The tubes
were then centrifuged, and the supernatant was collected for
measurement of PI.
Transplantation of pancreatic explants into nonobese
diabetic-severe combined immune deficiency (NOD-SCID) mice
The following experiments were carried out with the approval of
the animal care and ethics committee of our institution. Uncultured
1-mm3 fragments of fetal pig pancreas were
transplanted beneath the renal capsule of NOD-SCID mice as described
previously (6). The weight of tissue grafted beneath the
renal capsule was 24.05 ± 8.86 mg (mean ± SD;
n = 22). Animals were euthanized with CO2 at
different times after transplantation ranging from 4 d to 3 wk,
with blood collected from the beating heart for measurement of PI and
porcine C peptide levels. To optimize the secretion of PI in the
grafted mice, an ip injection of 4.1 mol/kg carbachol was given 2 min
before the animals were killed (11). After death, the
kidneys were exposed, and the grafts, which were pearly white, slightly
raised masses on the surface of the kidneys, were shelled out. PI and
insulin were extracted by homogenization of the tissue in acid-ethanol
and overnight incubation at 4 C. Porcine C peptide was measured by RIA
(Linco Research, Inc., St. Charles, MO), and insulin was
determined by an in-house RIA using a human standard (Novo, Bagsvaerd,
Denmark).
Statistics
Data were analyzed using one-way ANOVA, with Duncan’s test
(P < 0.05) applied to separate the groups, or
t test. The computer statistical package NCSS was used for
all analysis (15).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
; GenBank accession no. AF203915). This
sequence contained the C-terminal part of PP followed by the rest of
the gene. The deduced amino acid sequence of the peptide is shown in
Fig. 1B. The three amino acids, glycine-lysine-arginine, which separate
PP and PI (and thus provide cleavage sites) were similar in both pig
and human (Fig. 1B). Therefore, the porcine PI sequence was assumed to
start with the next amino acid, Asp, as occurs in the peptides from
sheep and dog (Fig. 1B). Human, ovine, and canine PI are 20 amino acids
long with a C-terminal cleavage site between arginine and glutamine.
Porcine PI was also assumed to end at this site, thus making it 19
amino acids long with the last 6 amino acids being mostly homologous
with PI of the other three species. The sequence of PI in pigs was
47%, 58%, and 53% homologous with that in humans, sheep, and dogs,
respectively (10).
|
). The interassay coefficient of
variability was 6.8 ± 3.4% (n = 5), and the intraassay
coefficient of variability was 7.8 ± 3.4% (n = 5). At
concentrations below 15 pg/ml (7.2 pmol/liter), the OD readings were
the same as the background reading of 0.35 ± 0.01 (n = 12).
Therefore, 15 pg/ml (OD, 0.31 ± 0.01; n = 12) was considered
the lower limit of detection of the assay, and OD readings above 0.31
were thus considered undetectable.
|
Effect of hypoglycemia on PI levels in pigs
Hypoglycemia induced in pigs (Fig. 3A) caused a counterregulatory increase
in the level of PP, as expected (Fig. 3B). As PI should be secreted on
an equimolar basis with PP, levels of PI should also rise in response
to hypoglycemia. This was observed (Fig. 3C), with the increase in PI
being parallel to that in PP. The subsequent fall in PI levels also was
parallel to that in PP. The ratio of PP/PI throughout the test was
0.96 ± 0.07 (mean ± SEM; n = 32). These
data show that PI is secreted in response to the same stimulus as PP
and support the concept that these two hormones are secreted on an
equimolar basis.
|
Secretion of porcine PI in grafted NOD-SCID mice
Porcine PI was measurable in the blood of NOD-SCID mice
transplanted with fetal pig ICCs for at least 3 wk after the tissue was
transplanted (Fig. 4A). These levels were
measured after stimulation with carbachol. Levels were highest at
4 d (P < 0.001; 19.69 ± 1.00 pmol/liter;
n = 6), with the values at 1, 2, and 3 wk relatively similar. The
blood of untransplanted mice gave readings in the PI assay no different
from background. Porcine C peptide was undetectable in all blood
samples of the transplanted mice.
|
). Levels of insulin were not
significantly different from one another (Fig. 4C). These data confirm
that the grafts remained viable for the duration of the experiment.
However, the secretion of these hormones from the grafts did not
correlate with their respective hormonal content. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The sequence heterology of PI may be comparable to that observed in the C peptide, which is cleaved off from proinsulin during insulin synthesis. Human and porcine insulin sequences differ by a single amino acid, whereas the C peptides show only 70% homology. Human and porcine PP sequences differ by two amino acids, whereas the cleaved icosapeptides show only 47% homology. Further, PP does have a physiological function in inhibiting exocrine enzyme secretion and bile acid secretion (16), whereas no functions are known for PI or C peptide.
This sequence heterology of PI is of potential importance in xenotransplantation of pig tissue. Xenotransplantation of fetal pancreatic pig cells is being explored by a number of investigators as an alternative method for the treatment of type 1 diabetes (1, 2, 3, 4, 5, 6). The period taken to achieve normalization of blood glucose levels is 1–5 months, during which time maturation of the ß-cell occurs. The only mature endocrine cell at the time of transplantation is the PP cell (8). Maturation of the graft is accompanied by a gradual increase in the number of ß-cells and a corresponding decrease in PP cells. From our previous immunohistochemical studies, PP cells, which comprise 9% of all epithelial cells in the graft at the time of transplantation, increased immediately after transplantation (17% at 4 d and at 3 wk, 21%) (11). Thereafter, the percentage of PP cells diminished (9% at 12 wk). In contrast, ß-cells, which comprised 12% of the epithelial cells in the graft at the time of transplantation, did not increase immediately thereafter (13% at 4 d). It was only at 3 wk and beyond that the percentage of these cells rose, reaching 34% at 12 wk. For this reason and because the fetal ß-cell is immature in its ability to secrete insulin in response to glucose, it is understandable that porcine C peptide is not detectable in host serum during the first few weeks after transplantation (porcine insulin cannot be fully distinguished from mouse insulin because of sequence homology). Thus, porcine insulin or C peptide cannot be used to monitor the viability of the graft.
We hypothesized that hormones secreted from the mature PP cell could be used for this purpose. Previously we have shown that PP secretion could probably be used as a marker of viability at this time (11). However, because of sequence homology between species, it was not possible to fully separate pig PP from that of other species, especially mouse and human. Porcine PI, another hormone secreted from the PP cell, can be fully distinguished from mouse and human PI. Its measurement in the blood of mice transplanted with fetal pig pancreatic tissue confirms our hypothesis that monitoring secretion of hormones from the PP cell is useful to monitor the viability of the grafted tissue.
Physiologically PP (and PI) is secreted from PP cells in the pancreas in response to a vagal stimulus, such as hypoglycemia or eating food (16). As vagal innervation of pancreas transplants does not occur, the secretion of PP (and PI) from such grafts cannot be enhanced by vagal stimulation (17). Exposure of the grafts to a cholinergic agent, carbachol, directly stimulated PP release from the graft (11). Therefore, the same stimulus, carbachol, was used in the current study to measure PI release from the graft.
As expected, carbachol stimulation of the relatively large number of PP
cells in the graft during the experimental period enabled us to detect
porcine PI in mouse serum. Levels of PI were highest 4 d after
fetal pig pancreatic tissue was transplanted into the mice, but were
still measurable for at least the next 17 d (Fig. 4A). No attempt
was made to measure levels of PI earlier than 4 d, but based on
previous experiments it would be expected that PI would be measurable
earlier than this (11). Our current results are similar to
those reported previously when serum levels of PP were used as a marker
of graft viability for up to 3 wk after transplantation, but was absent
thereafter (11). In a separate study (unpublished)
by the authors in which diabetic mice were transplanted, we confirmed
that serum PI is not detectable after this time. In contrast, the
ß-cell does not secrete detectable levels of C peptide during this
time. Thus, the fetal pig PP cell, which is the least prevalent of all
pancreatic endocrine cells, secretes measurable levels of its hormones
for at least the first 3 wk after pancreatic tissue is transplanted.
The time at which the ß-cell becomes functional varies depending on
the glycemic status of the recipient. If the animal is normoglycemic,
serum C peptide is immeasurable before 12 wk (11), but if
the animal is hypoglycemic, C peptide is immeasurable from 3 wk
(unpublished data).
The secretion of PI from the graft did not correlate with the PI content of the grafts during the experimental period, probably indicating different signaling mechanisms required for synthesis/storage and regulated secretion. The same was true for insulin, where grafts showed its synthesis/storage (11).
The assay for porcine PI does not cross-react with human sera. This indicates that either human PI is not present in sufficient amounts in the sera to evoke a reaction or, if present in sufficient amounts, is noncross-reactive with the assay (due to significant heterology with porcine PI). Either way, monitoring levels of porcine PI may be useful in assessing the viability of fetal pig pancreatic tissue in the first few weeks after it is transplanted into humans. This is a noninvasive technique for monitoring graft viability, in contrast to biopsy, which is the method currently used (1).
To be of potential benefit, the measurement of PI should be possible with small amounts of sera, carried out rapidly, and be sensitive, specific, and reproducible. We have achieved these goals with an antigen inhibition ELISA using polyclonal antibodies against PI. This form of ELISA was used in preference to a sandwich ELISA because of its greater sensitivity. Antigen-antibody contact occurs in an Eppendorf tube and not in the well of a microtiter plate, thereby increasing contact between antigen and antibody. The amplification and detection system of biotin-neutravidin-HRP was selected because of its higher sensitivity compared with other systems, such as biotin-HRP and biotin-strepavidin-HRP. The ELISA developed allowed measurement of PI between 15–1000 pg/ml (7.2–480 pmol/liter), which is within the expected level of PI in pigs (12–68 pg/ml; 6–33 pmol/liter) (18, 19, 20, 21). The assay was rapid (performed within 6 h) and required only a small volume of sample (100 µl). The specificity of the assay for pig samples was confirmed when human and mouse sera were tested and gave readings similar to background. This also demonstrated that pig serum can be used directly for PI detection without the need to extract the peptide, a method that requires a large volume of sample, is time consuming, and introduces more variability.
Secretion of PP from PP cells is enhanced by a vagal stimulus, such as hypoglycemia or food intake, and by cholinergic agents such as carbachol (16, 17). As PP and PI are cleaved from the same precursor and secreted, PI secretion also should be enhanced by vagal stimuli and cholinergic agents. We found this to be true. Hypoglycemia induced in pigs caused a rise in levels of PI, whereas carbachol enhanced secretion of PI from ICCs in vitro. The secretion of PP and PI should be equimolar, as 1 mol PP precursor is cleaved to 1 mol PI and 1 mol PP. Our results in the pigs are consistent with this if it is assumed that the rate of clearance of the two peptides is similar, with the ratio of serum levels of PP to PI being 0.96.
In summary, we have sequenced porcine PI and developed a sensitive and specific ELISA to measure it. Levels of PI rise in response to hypoglycemia and a cholinergic stimulus and can be used to monitor the viability of xenotransplanted fetal pig pancreatic tissue.
|
Footnotes |
|---|
Abbreviations: HRP, Horseradish peroxidase; ICC, islet-like cell clusters; NOD-SCID, nonobese diabetic-severe combined immune deficiency; PI, pancreatic icosapeptide; PP, pancreatic polypeptide; 3'RACE, rapid amplification of cDNA 3'-ends.
Received March 7, 2001.
Accepted for publication June 12, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
