365
Clinical Science (1983) 65, 365-371
The effect of the mammalian neuropeptide, gastrin-releasing
peptide (GRP), on gastrointestinal and pancreatic hormone
secretion in man
SUSAN M. WOOD, ROLAND T. JUNG, JOAN D. WEBSTER, MOHAMMED A. GHATEI,
THOMAS E. ADRIAN, NOBURU YANAIHARA*f, CHIZUKO YANAIHARA*
A N D STEPHEN R. BLOOM
Department of Medicine, Hammersmith Hospital, London, * Luboratory of Bimrganic Chemistry, Schizuoka College of
Pharmacy, Shizuoka, Japan, and t National Institute for Physiological Sciences, Okazaki, Aichi, Japan
(Received 29 November 1982118 March 1983; accepted 8 April 1983)
summary
1. Gastrin-releasing peptide, a newly isolated
mammalian peptide similar in its structure and
actions to the amphibian peptide, bombesin, has
recently been localized to nerves in the brain, gut
and pancreas. The present study investigates its
effects on gut and pancreatic peptides in man.
2. Intravenous infusion of 0.7 and 2.9 pmol
min-' kg-' produced significant .elevation of
plasma gastrin, cholecystokinin-like immunoreactivity and neurotensin. It was found also to
potentiate glucose-dependent insulin secretion.
3. Its specific location in nerve fibres in the
proximal gut and pancreas and its selective effect
on gastroenteropancreatic peptides may favour its
role as a physiological regulatory neuropeptide.
Key words: gastrin-releasing peptide, gastrointestinal hormones, pancreatic hormones.
Abbreviations: CCK, cholecystokinin; GIP, glucosedependent insulinotropic peptide; GRP, gastrinreleasing peptide.
Introduction
The autonomic nervous system plays a central
role in the regulation of the gastrointestinal tract
and pancreas [l]. Until recently it has been considered to consist exclusively of cholinergic and
Correspondence: Dr S. M. Wood, Department
of Medicine, Hammersmith Hospital, Du Cane
Road, London W12 OHS.
adrenergic nerve fibres; it is now evident, however,
that a large and complex peptidergic component
exists [2,31. The ganglia of the intrinsic plexuses
of the gut and pancreas are richly supplied by
these peptidergic fibres [4, 51. These appear to be
part of the local neural modulation of motility,
secretion and blood flow within the gut, being in
turn influenced by the extrinsic nerve supply [l].
A variety of putative peptide neurotransmitters
have been identified by radioimmunoassay and
immunocytochemistry in the gut and pancreas
[4,6, 71. One of these reacts with antibodies to
bombesin, a 14 amino acid peptide originally
isolated from the skin of the frog Bombina bombina [8]. This peptide has been shown to release
a number of other peptides in the mammal [9-121.
Bombesin-like immunoreactivity has been localized to intrinsic nerves throughout the mammalian
gut [13], and may function as a stimulatory
neuropeptide in a wide variety of sites. It has been
suggested for example that bombesinergic neurons
correspond to the postganghonic nerves of the
vagal innervation of the endocrine cells in the
stomach, where they may regulate the secretion of
such peptides as gastrin [ 141, and so modify gastric
acid and exocrine secretion.
More recently a 27 amino acid peptide, gastrinreleasing peptide (GRP), has been isolated from
porcine gut [ 151. This peptide shows remarkable
homology with bombesin at the biologically active
C-terminal region of the peptide (Table 1). Furthermore its effects on gastroenteropancreatic hormone
secretion are very similar to those of bombesin
[16]. Immunostaining with anti-GRP sera has
S. M. Wood et al.
366
TABLE 1. Sequences of porcine GRP and amphibian bombesin
Amino acid residues present in identical positions in both are enclosed in the continuous line.
PGRF'
Met
Tyr
Pro
Arg
Gly
Asn
His
Trp
Ah
Val
Gly
His
Leu
Met
NH,
Bornbesin
Pyr
Gln
Arg
Leu
Gly
Asn
Gln
Trp
Ala
Val
Gly
His
Leu
Met
NH,
revealed GRP nerve fibres in the myenteric nerve
plexus of the porcine duodenum and intrapancreatic ganglia [ 171. As bombesin's mammalian
counterpart, GRP is therefore a strong contender
for the stimulatory neurotransmitter or neuromodulator of the gut and pancreas.
Its effects on gastrointestinal and pancreatic
hormone secretion have not previously been investigated in man. We have therefore studied these
effects at basal plasma glucose concentrations, and
during glucose infusions adjusted to mimic the
postprandial rise in plasma glucose.
Methods
Studies were carried out o n six healthy volunteers
(four females and two males, mean age 24k0.6
years, and mean weight 60.6 f 3.5 kg). All subjects
gave their informed consent to take part in the
investigations, which had been approved by the
Royal Postgraduate Medical School Ethical Committee.
Each subject was studied on five different
mornings after an overnight fast. Abbott butterfly
cannulae (Abbott Laboratories, North Chicago,
U.S.A.) were inserted into arm veins, one for
infusion of peptide and glucose, the other for
blood sampling.
Glucose was infused at a rate of 0.3 g/min for
30 min to mimic the normal postprandial plasma
glucose profie. On a separate occasion synthetic
gastrin-releasing pe tide was infused at a rate of
0.7 pmol min-' kg- and on a further occasion at
a rate of 2.9 pmol min-' kg-' each over 30 min.
Simultaneous infusions of glucose (0.3 glmin)
and GRP, at each infusion rate, were also performed. The infusions were carried out in random
order.
The GRP used in the study was a synthetic
preparation of the porcine peptide, which coeluted as single peak with the natural 27 amino
acid peptide on column chromatography [18].
For infusion it was dissolved in sterile 0.9%
sodium chloride solution (1 54 mmol/l) containing
P
0.5% (70 pmol/l) human serum albumin (Lister
Institute, Elstree, Herts, U.K.), the latter to reduce
adsorption of the peptide to the plastic syringe
and tubing of the infusion equipment.
Basal blood samples were taken before and then
at 5 min intervals during each 30 min infusion
period, and finally at 15 and 30 min after the end
of infusion.
For estimation of the gastroenteropancreatic
hormones lOml of blood was taken into chilled
tubes containing 4000 kallikrein inactivator units
(KIU) of aprotonin and 100 units of heparin and
immediately centrifuged. The plasma was removed
and stored at -20°C until the time of assay.
Additional samples (2 ml) were collected into
sodium fluoride/sodium oxalate tubes for glucose
estimation. GRP infusate samples were collected
at the beginning and end of each infusion for
measurement of GRP to assess losses of peptide
and allow estimation of the actual concentration
of GRP infused.
The gastroenteropancreatic hormones were
measured by previously described radioimmunoassays for insulin [ 191, pancreatic polypeptide,
glucagon, enteroglucagon, glucose-dependent insulinotropic peptide (GIP), neurotensin and gastrin
[20]. These assays showed no cross-reactivity between peptides, with the exception of enteroglucagon with glucagon. In the last-named enteroglucagon was calculated by subtraction of pancreatic
glucagon (measured using a specific C-terminal
directed antiserum, RCSS, which gave zero readings after total pancreatectomy) from total
N-terniinal glucagon immunoreactivity (measured
by antiserum R59, which fully reacted with pure
porcine enteroglucagon, glicentin [20]).
A subtraction assay was also used to measure cholecystokinin (CCK)-like immunoreactivity.
Rabbit antiserum raised to the sulphated C-terminal octapeptide, CCK-8, cross-reacted with CCK-8,
CCK-33 and the gastrins. When used at a final antiserum dilution of 1:160 000 with '2SI-labelled
unsulphated CCK-8, and sulphated CCKS as
standard, the assay could measure changes of
Gastrin-releasingpeptide and pancreatic and gut hormone secretion
1 pmol of cholecystokinin-like immunoreactivity/l
between adjacent tubes with 95%confidence. The
gastrin antibody (Gas-1 79) cross-reacted with all
forms of gastrin but showed less than 2% crossreactivity with the CCK. Therefore plasma concentrations of CCK-like immunoreactivity were obtained by subtracting gastrin, as measured by
Gas-179, from the total CCK and gastrin measurements determined by the CCK-like immunoreactivity assay. These subtraction assays were
used, as specific single assays for enteroglucagon
and CCK are not available.
Antisera to bombesin were raised in rabbits by
using a synthetic Lys3-bombesin analog conjugated
to bovine serum albumin as antigen. A radiolabelled antigen ('2SI-labelled bombesin) was
prepared from Tyr5-bombesin C-terminal nonapeptide analog by the chloramine-T oxidation
technique. The product was purified on a Sephadex G-25 column eluted with formic acid (0.1
mol/l) containing 2.0%human serum albumin and
40KIU of aprotonin/ml. Synthetic porcine GRP
was used as standard. This assay measured bombesin and pure porcine GRP with equal potency,
detecting 5pmol/l with 95% confidence [16].
Plasma glucose was measured by a standard
glucose oxidase method.
Results are expressed as means and SEM. Statistical significance was assessed by using Student's
paired t-test.
loo
367
r
75
c
i
150
2c3
25
L
-
0
60
30
Time (min)
FIG. 1. Plasma GRP concentrations during 30 min
infusions of: 10 g of glucose (m); GRP 0.7 pmol
min-' kg-' alone (0)and in the presence of the
10 g of glucose infusion ( 0 ) ;GRP 2.9 pmol min-'
kg-' alone (A) and with 10 g of glucose (A).
Infusion period is shown by the bar.
200
TT
Results
150
Gastrin-releasingpeptide
GRP produced no subjective or objective side
effects at either dose. The concentration of GRP
measured in the infusates at the end as compared
with the start of each infusion demonstrated a
variable loss of peptide of between 1 5 and 20%
of the total during any 30min infusion. The
infusion rates were therefore calculated from the
mean of the infusate concentrations at the beginning and end of each infusion for each subject
and were 0.7 f 0.02 and 2.9 f 0.1 5 pmol min-'
kg-' for the low and high dose infusions respectively. These resulted in elevation of plasma GRP
from 7.5 f 0.7 pmol/l to 24.2 1.7 pmol/l and
79.8f 11.9 pmol/l for the low and high dose
respectively. A plateau concentration of plasma
GRP was reached at 10 min in either case (Fig. 1).
Intravenous glucose had no effect on the GRP rise.
*
Pancreatic peptides
There was no change in plasma insulin during
infusion of GRP at fasting plasma glucose concen-
2
a ioa
5
9
CI
5c
I
0
I
I
30
I
I
60
Time (min)
FIG. 2. Plasma insulin concentrations during
30 min infusions of: l o g of glucose (m); GRP
0.7 pmol min-' kg-' alone (0)and in the presence
of the 10 g of glucose infusion ( 0 ) ;GRP 2.9 pmol
min-' kg-' alone (A) and with 10 g of glucose (A).
Infusion period is shown by the bar.
S. M. Wood et al.
368
Gastrointestinal peptides
GRP at both infusion rates produced large rises
in plasma gastrin (P< 0.01) and cholecystokininlike immunoreactivity (P< 0.01) (Figs. 4 and 5),
and a smaller rise in plasma neurotensin (P<0.05)
(Fig. 6). There was no significant change in enteroglucagon or GIP.
Discussion
In the present study gastrin-releasing peptide
(GRP)was found potently to release a number of
I
1
0
I
I
30
Time (min)
I
I
60
gastrointestinal hormones, including neurotensin,
gastrin and cholecystokinin-like immunoreactivity,
when circulating concentrations of the peptide
were as low as 24 pmol/l. The higher infusion rate
of GRP resulted in a circulating concentration of
80 pmol/l, but this did not significantly enhance
the peptide responses, suggesting that the maximal
response had already been reached with the lower
dose. The effect of GRP on these gut peptides was
more potent than that previously reported for
synthetic amphibian bombesin, although circu-
FIG. 3. Plasma glucose concentrations during
30 min infusions of: 10 g of glucose (m); GRP
0.7 pmol min-' kg-' alone (0) and in the presence
of the 10 g of glucose infusion ( 0 ) ;GRP 2.9 pmol
min-' kg-' alone (A) and with 10 g of glucose (A).
Infusion period is shown by the bar.
40
trations; however, when plasma glucose was raised
by 2.6f0.2 mmol/l during a l o g of glucose
infusion, GRP (0.7 pmol min-' kg-') resulted in
a peak insulin release of 156.8 f 14 pmol/l, which
was significantly greater than the insulin response
t o glucose alone (121 f 16 pmol/l, P< 0.05)
(Fig. 2). This was paralleled by a lower plasma
glucose profile compared with that seen during
infusion of glucose alone (P<O.O2) (Fig. 3),
despite an identical glucose load and infusion
rate in each case. A similar degree of enhancement
of glucose-induced insulin secretion (1 70 f 16
pmol/l) was observed during the higher GRP
infusion rate (2.9 pmol min-' kg-') (Fig. 2). GRP
produced neither a significant release of glucagon
nor a reduction in its suppression by intravenous
glucose.
The rise in pancreatic polypeptide seen in many
subjects during GRP infusion failed t o reach
significance due to inter-subject and inter-infusion
variation in concentrations of pancreatic polypeptide.
= 30
-
z
3
G
3 20 c1
10
-
-
Gastrin-releasingpeptide and pancrpeatic and gut hormone secretion
2
25
r
20
.
15
-
T
4
8u 10 '
5 '
1 -
I
I
0
60
30
Time (mid
FIG. 5. Plasma cholecystokinin (CCK) concentrations in response to infusions of GRP: 0.7 pmol
min-' kg-' (0)and 2.9 pmol min-' kg-' (A).
r
1
I
I
0
I
I
30
I
I
60
Time (min)
FIG. 6. Plasma neurotensin concentrations in
response to infusions of GRP: 0.7 pmol min-'
kg-' (0)and 2.9 pmol min-' kg-' (A).
369
lating plasma concentrations of bombesin-like
immunoreactivity were virtually identical in the
two studies [21]. This is perhaps predictable as
GRP is the mammalian form of the peptide. The
possible physiological relevance of these actions
needs to be considered in relation to the location
of GRP in the gut. By using a recently developed
specific radioimmunoassay for GRP [17], high
concentrations have been detected in the mucosa
and muscle of the body of the porcine stomach,
in both layers of the antmm where the majority
of gastrin cells are localized and in jejunal muscle
[17]. The two molecular forms of GRP identified
at these sites included one coeluting with synthetic
porcine GRP and the other with porcine GRP(14-27) [ 171. Immunohistochemical studies have
shown numerous GRP immunoreactive nerve
fibres in the myenteric nerve plexus of the porcine
duodenum [17]. In the light of this distribution
and the action of GRP on gastrin release reported
here in man, it is tempting to speculate that GRP
plays a part in the neural control of gastrin release
and gastric acid secretion. Many studies have
shown that amphibian bombesin and more
recently porcine GRP potently stimulate gastric
acid secretion in animals and man [22,23,27], an
effect which appears to be dependent on gastrin
release [21].
As has been suggested for bombesin, GRP may
play a part in the regulation of pancreatic exocrine
secretion [22, 251 and gall bladder function via
the release of cholecystokinin [16, 25-27]. The
recent localization of GRP to nerves in the porcine
duodenum and jejunum strengthens such a concept. Bornbesin's reported action on gastrointestinal
motility [23,28] may also be shared by GRP, particularly if this effect is found to be mediated by
peptides such as CCK or neurotensin.
In the present study no stimulation of enteroglucagon release by GRP was found. This differs
from the studies on GRP in the dog [16] (N.
Yanaihara, personal communication), which may
be explained by species variation in the response.
GRP was found in this study to have no significant effect on basal insulin secretion but to
enhance glucose-induced insulin secretion. This
agrees with the reported effects of bombesin on
isolated perfused rat pancreas [29] and its infusion
in the dog [3]. SurprisinglyGRPhas been reported
to stimulate basal insulin secretion in the dog [16].
The possible physiological significance of the
action of GRP on glucose-inducedinsulin secretion
g a s more relevance in the light of the recent
localization of GRP to porcine intrapancreatic
neural ganglia [ 171. Nerve fibres from these ganglia
have been shown to innervate the islets of Langerham. This is further supported by our recent
S. M. Wood et al.
370
finding of GRP like-immunoreactivity in human
pancreas (8.2 2.8 pmol/g, n =4).
There was no significant effect of GRP on the
other islet peptides, glucagon and pancreatic polypeptide. This contrasts with the reported action of
bombesin on these peptides in the dog and man
[16, 21,301, and of GRPin the dog [16]. In many
of these studies there seems to be marked differences in responses between dog and man. For
instance, the effect of bombesin on secretion of
pancreatic polypeptide in man is small compared
with its effect in the dog, and it has been suggested
that bombesin might have a dual inhibitorystimulatory effect on secretion of pancreatic polypeptide in man [3 11. This may be a reason for the
indefinite pancreatic polypeptide responses in this
study.
The localization of GRP to the enteric plexi of
the gut and pancreas [17] and its effect on a
limited number of peptides, gastrin, CCK, neurotensin and insulin at low circulating concentrations
may imply a physiological role for it as a neurotransmitter or neuromodulator. It appears to have
a more selective effect on peptide release than its
amphibian counterpart, possibly suggesting a
greater specificity for receptor sites, perhaps by
virtue of its greater molecular size. It is of interest,
however, that significant quantities of a smaller
molecular form of GRP, GRP(14-27), similar to
bombesin has been identified in porcine gut and
pancreas [ 171 together with GRP(1-27).
*
Acknowledgment
S.M.W. is a British Diabetic Association R. D.
Lawrence Research Fellow.
References
1. Gabella, G. (1979) Innervation of the gastrointestinal
tract. International Review of Cytology, 59, 129-193.
2. Baumgarten, H.G., Holstein, A.F. & Owman, C.H.
(1970) Auerbach's plexus of mammals and man:
electron microscopic identification of three different
types of neuronal processes in myenteric ganglia of
the large intestine from rhesus monkeys, guinea pigs
and man. Zeitschrift fur Zellforschung und Mikroskopische Anatomie, 106, 376-397.
3. Bishop, A.E., Ferri, G.L., Probert, L., Bloom, S.R. &
Polak, J.M. (1982) Peptidergic nerves. Scandinavian
Journal of Gastroenterology, 17 (Suppl. 71), 44-59.
4. Furness, J.B. & Costa, M. (1980) Types of nerves in
the enteric nervous system. Neuroscience, 5,1-20.
5. Larsson, L.l. & Rehfeld, J.F. (1979) Peptidergic and
adrenergic innervation of pancreatic ganglia. Scandinavian Journal of Gastroenterology, 14,433-437.
6. Fumess, J.B., Costa, M., Franco, R. & LlewellynSmith, J.J. (1980) In: Advances in Biochemical
Psychopkarmaco~o~y, Neuronal Communications,
vol. 22, p. 608. Ed. Costa, E. & Trabucchi, M. Raven
Press, New York.
7. Larsson, L.4. (1979) Innervation of the pancreas by
substance P, enkephalin, vasoactive intestinal polypeptide and gastrin/CCK immunoreactive nerves.
Journal of Histochemistry and Cytochemistry, 27,
1283-1284.
8. Anastasi, A., Erspamer, V. & Bucci, M. (1971) Isolation of bombesin and alytesin, two analogous active
peptides from the skin of the European amphibians,
Bombina and Alytes. Experientia, 27,166-167.
9. Becker, H.D., Borger, H.W., Schafmayer, A. &Werner,
M. (1979) Bombesin releases GIP in dogs. Scandinavian Journalof Gastroenterology, 13 (Suppl. 49), 14.
lO.Taylor, I.L., Walsh, J.H., Carter, D., Wood, J.R. &
Grossman, M.I. (1979) Effects of atropine and bethanechol o n bombesin stimulated release of pancreatic
polypeptide and gastrin in the dog. Gastroenterology,
77,714-718.
11. Lezoche, E., Ghatei, M.A., Carlei, F., Blackbum,
A.M., Basso, N., Adrian, T.E., Speranza, V. & Bloom,
S.R. (1979) Gut hormone responses t o bombesin in
man. Gastroenterology, 76, 1185.
12. Basso, N., Lezoche, E. & Speranza, V. (1979) Studies
with bombesin in man. World Journal o f Surgery, 3,
579-5 85.
13. Dockray, G.J., Vaillant, C. & Walsh, J.H. (1979) The
neuronal origin of bombesin like immunoreactivity in
the rat gastrointestinal tract. Neuroscience, 4, 15611568.
14. Uvnas-Wallensten, K. (1981) In: Cellular and Chemical
Messengers in the Digestive System, p. 299. Ed.
Grossman, M.I., Brazier, M.A.B. & Lechago, J.
Academic Press, London, New York.
15. MacDonald, T. J., Jornvall, H., Nilsson, G., Vagne, M.,
Ghatei, M., Bloom, S.R. & Mutt, V. (1979) Characterisation of a gastrin releasing peptide from porcine
non-antral gastric tissue. Biochemical and Biophysical
Research Communications, 90,227-233.
16. MacDonald, T.J., Ghatei, M.A., Bloom, S.R., Track,
N.S., Radziuk, J., Dupre, J. & Mutt, V. (1981)
A qualitative comparison of canine plasma gastroenteropancreatic hormone responses to bombesin and
the porcine gastrin releasing peptide (GRP). Regulatory Peptides, 2,293-304.
17. Yanaihara, N., Yanaihara, C., Mochizuki, T., Iwahara,
K., Fujita, T. & Iwanaga, T. (1981) lmmunoreactive
GRP. Peptides, 2 (Suppl. 2), 185-191.
18. Mochizuki, T., Yanaihara, C., Inoue, A., Kubota, M.,
Yanaihara, N., McDonald, T.J. & Mutt, V. (1981)
In: Peptide Chemistry, p. 171. Ed. Okawa, K. Protein
Research Foundation, Osaka.
19. Albano, J.D.M., Ekins, R.P., Maritz, G. & Turner,
R.C. (1972) A sensitive precise radioimmunoassay of
serum insulin relying on charcoal separation of
bound and free moieties. Acta Eiidocrinologica
(Copenhagen), 70,487-509.
20. Bloom, S.R. & Long, R.G. (1982) Radioimmunoassay
of' Gut Regulatory Peptides, 1st edn. W.B. Saunders,
London.
21.Ghatei, M.A., Jung, R.T., Stevenson, J.C., Hillyard,
C.J., Adrian, T.E., Lee, Y.C., Christofides, N.D.,
Sarson, D.L,, Mashiter, K., Mclntyre, I. & Bloom,
S.R. (1982) Bombesin: Action o n gut hormones and
calcium in man. Journal of Clinical Endocrinology
and Metabolism, 54,980-985.
22. Basso, N., Giri, S., Improta, G., Lezoche, E., Melchiorri, P., Peroca, M. & Speranza, V. (1975) External
pancreatic secretion after bombesin infusion in man.
Gut, 16,994-998.
23. Vamer, A.A., Modlin, I.M. & Walsh, J.H. (1981) High
potency of bombesin for stimulation of human
Gastrin-releasingpeptide and pancreatic and gut hormone secretion
gastrin release and gastric acid secretion. Regulatory
Peptides, 1,289-296.
24. Walsh, J.H., Maxwell, V., Ferrori, J. & Vaner, A.A.
(1981) Bombesin stimulates human gastric function
by gastrindependent and independent mechanism.
Peptides, 2,193-198.
25. Erspamer, V., Improta, V., Melchiorri, P. & Sopranzi,
N. (1974) Evidence of cholecystokinin release by
bombesin in the dog. British Journal of Pharmacology, 52,227-232.
26. Corazziani, Jorsoli, J.A., Melchiorri, P. & Delle Fave,
G.Y. (1974) Effect of bombesin on human gall
bladder emptying. Rendic Gastroenterology, 6,52.
27. McDonald, T.J., Nilsson, G., Vagne, M., Bloom, S.R.,
Ghatei, M. & Mutt, V. (1978) Purification of a porcine intestinal peptide with bombesin like properties.
Scandinavian Journal of Gastroenterology, 13 (Suppl.
49), 119.
371
28. Bertaccini, G., Impicciatore, I., Molina, E. & Zappin,
L. (1976) Action of bombesin on human gastrointestinal motility. Rendic Gastroenterology, 6 , 4 5 4 1.
29. Martindale, R., Levin, S. & Mi-Slater, R. (1982)
Effects of caerulein and bombesin on insulin and
glucagon secretion from the isolated, perfused rat
pancreas. Regulatory Peptides, 3,313-324.
30. Schusdziarra, V., Rouiller, D., Harris, V., Pfeiffer,
E.F. & Unger, R.H. (1980) Effect of bombesin upon
plasma somatostatin-like immunoreactivity, insulin
glucagon in normal and cheniically sympathectomized
dogs. Regulatory Peptides, 1,89-96.
31. de Magistris, L., Delle Fave, G., Kohn,A. & Schwa-,
T.W. (1981) Differential stimulation of pancreatic
polypeptide and gastrin secretion by bombesin in
man. Life Sciences, 28,2617-2621.