GASTROENTEROLOGY 2000;119:13731381

Nerve Growth Factor Expression Is Up-regulated in the Rat

Model of L-ArginineInduced Acute Pancreatitis
HIROKI TOMA, JOHN WINSTON, MARIAADELAIDE MICCI, MOHAN SHENOY, and
PANKAJ J. PASRICHA
Enteric Neuromuscular Disorders and Pain Laboratory, Division of Gastroenterology and Hepatology, Department of Internal Medicine,
University of Texas Medical Branch, Galveston, Texas

Background & Aims: In somatic pain models, increases

in nerve growth factor (NGF) are linked to the development of pain and hyperalgesia. The aim of this study
was to examine a rat model of acute necrotizing pancreatitis for changes in NGF expression. Methods: NGF
protein and messenger RNA (mRNA) levels in the pancreas were correlated with histopathologic changes during the course of acute necrotizing pancreatitis in rats
induced by the intraperitoneal injection of L-arginine.
Immunohistochemistry for NGF localization was performed on the pancreatic tissue. Results: Two phases of
NGF production were observed in the inflamed pancreas: an early release from pancreatic islets at 2 and 6
hours and a later increase in mRNA (18-fold at maximum) at 3 days and in protein levels (7-fold at maximum) at 5 days coinciding with maximum parenchymal
necrosis. The intense NGF-like immunoreactivity was
observed predominantly in the ductal cells in pancreas
from rats with pancreatitis at 5 days. Conclusions: The
development of acute necrotizing pancreatitis in this
model leads to a significant increase in NGF production
and appears to shift the major cellular sites of NGF
production from the islets to the ductal cells. It is conceivable that NGF production in the inflamed pancreas is
responsible for plastic changes in the sensory neurons
that mediate peripheral sensitization and contribute to
the generation of pain.

feature of both acute and chronic pancreatitis is the

presence of an ongoing inflammatory response. Inflammatory mediators generated during this response can
affect the pathogenesis of the disease and may contribute
to the generation of pain.1 Although pain is a cardinal
feature of pancreatitis, its mechanism is poorly understood and its treatment remains difficult.2 4 Inflammation is known to cause significant changes in the somatic
sensory nervous system affecting peripheral nerves, spinal
cord, and supraspinal structures.5 Several biologic agents
induced by inflammation such as neurotrophins (e.g.,
nerve growth factor [NGF]), prostanoids, bradykinin,
and various cytokines can sensitize peripheral nocicep-

tors.1,5 This leads to an increase in afferent signaling,

which in turn may cause sensitization at the central level,
resulting in amplification and persistence of pain. However, the contribution of inflammation in general or that
of specific inflammatory agents to the etiology of pain in
pancreatitis is not known.
NGF, a mediator of the inflammatory response in
various tissues, regulates the sensitivity of peptidergic
nociceptors and is up-regulated in somatic inflammatory
states, where it is believed to play an important role in
the pathogenesis of persistent pain.6 11 When NGF is
injected at the periphery, it induces thermal and mechanical hyperalgesia in adult rats.8,9 Up-regulation of
NGF has also been shown in inflammatory conditions,
e.g., experimental arthritis10 and cystitis.11 A recent
study of human chronic pancreatitis showed increased
expression of NGF messenger RNA (mRNA) in 42% of
human patients examined.12 Together, these findings
suggest that the intense inflammation seen in pancreatitis will also be associated with an increase in NGF levels
in pancreatic tissues. However, with the exception of this
human study, little else is known about the expression of
NGF in the pancreas during the course of acute or
chronic pancreatitis in humans or in any animal models
in which the potential impact of NGF on pancreatic
sensory afferents can be evaluated more easily. To determine whether NGF expression is altered during the
course of pancreatitis, we measured NGF levels in a rat
model of acute necrotizing pancreatitis. We were able to
correlate changes in the levels and cellular distribution of
NGF expression with the histopathologic changes observed during the development of pancreatitis in the
model. Our findings suggest 2 phases of NGF producAbbreviations used in this paper: CGRP, calcitonin gene-related
peptide; ELISA, enzyme-linked immunosorbent assay; IL, interleukin;
NGF, nerve growth factor; NGF-IR, NGF-like immunoreactivity; SP, substance P; TNF, tumor necrosis factor.
2000 by the American Gastroenterological Association
0016-5085/00/$10.00
doi:10.1053/gast.2000.19264

1374

TOMA ET AL.

tion: an early shift from pancreatic islets, followed several

days later by an increase in NGF mRNA and protein.

Materials and Methods

Animals
Male SpragueDawley rats weighing 200 350 g were
fed water and laboratory chow ad libitum. The experimental
protocol was approved by the Institutional Animal Care and
Use Committee at the University of Texas Medical Branch,
Galveston.

Induction of Experimental Pancreatitis

Experimental pancreatitis was induced by the procedure described by Takacs et al.13 Each rat received an intraperitoneal injection of 250 mg/100 g body wt of L-arginine
(Sigma Chemical Co., St. Louis, MO) as a 20% solution in 0.15
mol/L NaCl twice, with an interval of 1 hour. Control rats
received an equal volume of 0.15 mol/L NaCl alone. In the
entire course of the procedure, rats were allowed to take water
ad libitum. After the second injection, laboratory chow was
made available to the rats again. Groups of rats (L-arginine
injected, n 7; saline injected, n 4) were killed under
anesthesia with a mixture of xylazine and ketamine at 2 and 6
hours and 1, 2, 3, and 5 days after injection. Pancreatic tissues
were immediately processed for total RNA, fixed for histology,
or quickly frozen in liquid nitrogen for protein quantification
by enzyme-linked immunosorbent assay (ELISA).

Extraction of Total RNA

Total RNA was extracted by the modified guanidinium thiocyanatephenol chloroform extraction method of
Chomczynski and Sacchi.14 Pancreatic tissues were quickly
homogenized in RNAzol (a solution of 4 mol/L guanidinium
thiocyanate [Fluka, Buchs, Switzerland], 0.5% sarkosyl
[Fisher Scientific, Fair Lawn, NJ], 0.025 mol/L sodium citrate
pH 7.0, 0.2 mol/L 2-mercaptoethanol [Sigma], 1/10 volume of
3 mol/L sodium acetate [pH 5.2] and equal volume of phenol
[Sigma], saturated with water and heated to 65C). Chloroform (Sigma), 100 mL, was added to 1 mL of the homogenate
and centrifuged at 12,000g for 10 minutes at 4C. The aqueous phase was combined with an equal volume of isopropanol
and incubated at 20C for 1 hour. Sedimentation was performed at 12,000g for 15 minutes at 4C. The resulting pellet
was washed with 75% ethanol, air-dried, and dissolved in
RNase-free water (Ambion Inc., Austin, TX). The optical
density was measured at 260 nm to determine the concentration of total RNA. Total RNA was stored at 80C until use.

RNase Protection Assay

The steady-state levels of mRNA of NGF in pancreatic
tissues were analyzed using the Riboquant Multi-Probe RNase
Protection Assay System (PharMingen, San Diego, CA). The
antisense RNA probe was synthesized from the template set
of rat neurotrophin and labeled with -[32P]deoxyuridine

GASTROENTEROLOGY Vol. 119, No. 5

triphosphate according to the manufacturers protocol. Twenty

micrograms of total RNA was precipitated and dissolved in 10
L of hybridization buffer containing the probe at a concentration of 2.3 105 cpm/L. Samples were heated to 90C for
5 minutes and then hybridized at 56C for 16 hours. After
hybridization and RNase treatment, protected fragments were
resolved on a 4.8% sequencing gel. Radioactive bands on the
dried gels were detected and quantified using the Cyclone
Storage Phosphor System (Packard, Meriden, CT). NGF values
were normalized to those measured for the housekeeping gene,
L32.

ELISA
NGF protein content was measured using an NGF
Emax immunoassay system ELISA kit (Promega, Madison, WI).
Frozen pancreatic tissues were homogenized in 5 volume of
0.2N perchloric acid. Insoluble material was removed by centrifugation at 10,000g for 5 minutes at 4C, and samples were
neutralized by addition of an equal volume of 1 mol/L sodium
borate. Various dilutions of the samples, along with NGF
standards, were run in the ELISA, and the amount of NGF
detected was normalized to wet tissue weight.

Histology
Fresh specimens of rat pancreas were fixed in 10%
formaldehyde (Sigma) in phosphate-buffered saline (PBS; pH
7.4) containing 1 mmol/L MgCl2 at 4C overnight. Sections
from paraffin-embedded specimens were stained with H&E
and observed under a light microscope (BX60; Olympus,
Tokyo, Japan). Evaluation of the histopathologic changes was
based on the scale described by Tito et al.15 Details are given
in Table 1. All sections were evaluated by 2 investigators.

Immunohistochemistry for NGF Localization

Deparaffinized sections from rat pancreas were incubated in blocking solution (5% normal goat serum in PBS) for
1 hour. A rabbit polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA) against rat NGF (sc-548) was used at
1/100 dilution in PBS. Sections were incubated with the
primary antibody overnight at 4C. They were then washed in
PBS, incubated with goat anti-rabbit biotinylated immunoglobulin (Ig) G (Vectastain ABC kit; Vector Laboratories,
Burlingame, CA) for 30 minutes at room temperature, washed
again in PBS, and incubated with peroxidase-conjugated
streptavidin (Vectastain ABC kit) for 30 minutes at room
temperature. The reaction products were labeled brown after
incubation with diaminobenzidine. When needed, slides were
counterstained with hematoxylin. Control slides were incubated with the primary antibody that was preabsorbed with
the antigen peptide.

Amylase Assay
Blood samples were collected by cardiac puncture at
the time of death and allowed to clot. They were centrifuged
at 10,000g for 10 minutes at room temperature, and serum
samples were stored at 20C until use. The -amylase

November 2000

UPREGULATION OF NGF IN RAT PANCREATITIS

1375

Table 1. Histopathologic Grading Scale for Pancreatitis

Edema
0
No edema
0.5 Focal expansion of interlobar septa
1
Diffuse expansion of interlobar septa
1.5 Same as 1 focal expansion of interlobular septa
2
Same as 1 diffuse expansion of interlobular septa
2.5 Same as 2 focal expansion of interacinar septa
3
Same as 2 diffuse expansion of interacinar septa
3.5 Same as 3 focal expansion of intercellular spaces
4
Same as 3 diffuse expansion of intercellular spaces
Parenchymal necrosis
0
No necrosis
0.5 Focal occurrence of 14 necrotic cells/HPF
1
Diffuse occurrence of 14 necrotic cells/HPF
1.5 Same as 1 focal occurrence of 510 necrotic cells/HPF
2
Diffuse occurrence of 510 necrotic cells/HPF
2.5 Same as 2 focal occurrence of 1116 necrotic cells/HPF
3
Diffuse occurrence of 1116 necrotic cells/HPF
3.5 Same as 3 focal occurrence of 16 necrotic cells/HPF
4
16 Necrotic cells/HPF
Inflammation and perivascular infiltrate
0
No inflammation
0.5 25 Leukocytes/HPF
1
610 Leukocytes/HPF
1.5 1115 Leukocytes/HPF
2
1620 Leukocytes/HPF
2.5 2125 Leukocytes/HPF
3
2630 Leukocytes/HPF
3.5 30 Leukocytes/HPF or focal microabscesses
4
35 Leukocytes/HPF or confluent microabscesses
Vacuolization
0
No vacuolization
0.5 Less than 18 of cells had vacuolization
1
Between 18 and 28 of cells had vacuolization
1.5 Between 28 and 38 of cells had vacuolization
2
Between 38 and 48 of cells had vacuolization
2.5 Between 48 and 58 of cells had vacuolization
3
Between 58 and 68 of cells had vacuolization
3.5 Between 68 and 78 of cells had vacuolization
4
All cells had vacuolization
HPF, high-power field.
Reprinted from Tito et al.,15 Am J Surg 1993;165:690 696, with
permission from Excerpta Medica Inc.

activity in serum was measured by using 4,6-ethylidene (G7)p-nitrophenyl (G1)-, D-maltoheptaside as the substrate in
Sigma Diagnostic Amylase Reagent (Sigma). The concentration was expressed as units per liter.

Statistics
Quantitative data were expressed as an average SD
from different animals in each group, and comparison was
made using the 2-tailed Student t test. P values of 0.05 were
considered statistically significant.

Results
Experimental Pancreatitis
The severity of pancreatitis induced by L-arginine
injections was assessed by scoring of H&E-stained pan-

Figure 1. Serum amylase levels in control (F) and L-arginineinjected

(E) rats. Serum levels of amylase were significantly higher up to 1 day,
then were lower beginning 2 days after the second injection in
L-arginineinjected rats than in control rats. *P 0.05; **P 0.01.

creatic sections on the histopathologic scale (Table 1).

Serum levels of -amylase were also measured (Figure 1)
because acute elevation of this enzyme appears to be a
good marker for successful induction of pancreatitis in
rat models.13,16 Serum levels of -amylase were significantly elevated over controls 2 hours after the second
injection of L-arginine (P 0.01) and reached a maximum at 24 hours (P 0.05). The levels of -amylase in
L-arginineinjected rats were significantly lower than
those of control rats at 2 days (P 0.01). Although the
difference was not statistically significant, serum levels of
-amylase in L-arginineinjected rats were lower than
those in control rats at 3 and 5 days.
The gross appearance of the pancreas from L-arginine
injected rats varied among groups at different time
points. No obvious change was observed at 2 and 6
hours, but gross interlobular edema was prominent from
1 to 3 days. The edema was almost gone at 5 days, but
the pancreas appeared atrophic compared with controls.
Microscopic findings on sections from L-arginineinjected rats were quantified as shown in Table 2. The most
Table 2. Quantification of Histopathologic Changes
in Experimental Pancreatitis
Edema
2
6
1
2
3
5

hours
hours
day
days
days
days

Necrosis

Inflammation Vacuolization

0.07 0.19 0.71 0.27 0.57 0.19

0.1 0.22
0.7 0.27
0.6 0.22
2.33 0.29 2.33 1.61
1.5 0
1.38 1.75 2.13 1.49 2.25 0.87
3.67 0.29
4.0 0
4.0 0
1.92 0.8
3.5 0.77
3.0 0.45

3.93 0.19
4.0 0
00
1.38 1.44
1.0 0
0.17 0.41

NOTE. Intense vacuolization is prominent in the pancreas 2 and 6

hours after the second injection of L-arginine. Edema, necrosis, and
inflammation reached a maximum at 3 days.

1376

TOMA ET AL.

GASTROENTEROLOGY Vol. 119, No. 5

Figure 2. Paraffin sections of rat pancreas stained with H&E. The pancreas was removed and processed for paraffin embedding at (A) 2 hours,
(B) 6 hours, (C) 3 days, and (D) 5 days after either L-arginine or saline treatment. (A and B) Intense vacuolization is observed in acinar cells at
2 and 6 hours in L-arginineinjected rats. (C) The tissue was severely necrotic, with loss of exocrine pancreas at 3 days. (D) Widespread areas
of pancreas were replaced by connective tissue with inflammatory cells at 5 days. Bar 100 mm.

prominent finding was intense vacuolization in acinar

cells in the pancreas from L-arginineinjected rats at 2
and 6 hours (Figure 2A and B; Table 2).
The entire parenchyma on all sections examined was
affected. In accordance with the gross appearance, interstitial edema was not obvious, and neither parenchymal
necrosis nor inflammatory cells were prominent at 2 and
6 hours. The vacuoles in acinar cells were less prominent,
but significant edema and a large inflammatory infiltrate
consisting primarily of polymorphonuclear leukocytes
were observed from 1 day. The tissue was severely necrotic, with loss of exocrine pancreas at 3 days (Figure
2C). These inflammatory changes reached a maximum at
3 days (Table 2 ). Widespread areas of pancreas were
replaced by connective tissue with inflammatory cells at
5 days (Figure 2D). No histologic abnormalities were
observed in the islets of Langerhans at any time point
examined (data not shown). In sections from control, no
histopathologic findings were observed. Thus, L-arginine
treatment induced acute necrotizing pancreatitis consistent with previous reports.13,16 Elevated -amylase levels

at 2, 6, and 24 hours were correlated with histopathologic evidence of pancreatitis.

NGF Expression
To determine whether the histopathologic changes
in the rat pancreas correlated with changes in NGF expression, NGF mRNA and protein were measured at various
time points after L-arginine injection. In RNase protection
assay, bands corresponding to the expected size of the
-NGF (322 base pairs [bp])protected fragment were
observed in both L-arginineinjected and control rats at all
time points examined (Figure 3A).
At 2 and 6 hours, no significant difference between
L-arginineinjected and control rats (Figure 3B) was
observed. At 3 days, 10 18-fold increases in NGF
mRNA levels were observed in L-arginineinjected rats
compared with controls (Figure 3B). The levels of NGF
mRNA remained significantly higher in L-arginineinjected rats at 5 days (P 0.01), ranging from 3.2- to
6.0-fold increases over control values (Figure 3B). These
findings, together with the histopathologic results, indi-

November 2000

UPREGULATION OF NGF IN RAT PANCREATITIS

1377

Figure 3. Increase in NGF mRNA expression in pancreas from L-arginineinjected rats. (A ) RNase protection assay on samples from L-arginineinjected (E)
or control rats 2 hours, 6 hours, 3 days, and 5 days after the second injection.
The bands corresponding to -NGF (322 bp) were weak in pancreas from control
rats at all time points and in pancreas from L-arginineinjected rats at 2 and 6
hours. The strong signals were obtained in pancreas from L-arginineinjected
rats at 3 and 5 days. (B) Normalized NGF mRNA expression as fold increase over
control values. The intensities of NGF mRNA bands obtained by RNase protection assay were densitometrically analyzed and normalized by signals of the
housekeeping gene, L32. Note the dramatic increase of NGF mRNA expression
in pancreas at 3 and 5 days. **P 0.01.

cate an increase in steady-state NGF mRNA during the

necrotizing phase of rat pancreatitis in this model.
To confirm that changes in mRNA levels were reflected by increases in NGF protein, NGF protein content in the pancreatic tissue was measured. No significant differences between L-arginineinjected and control
rats were observed at any time point before 3 days
(Figure 4). NGF protein levels in L-arginineinjected
rats were 2-fold higher than control levels (although this

difference did not reach the statistical significance) at 3

days and 4-fold higher than controls at 5 days (P
0.05). NGF levels in controls were lower at 3 days and
were 2-fold lower at 5 days than control values at 2 and
6 hours. NGF levels in L-arginineinjected rats remained
relatively constant. The 37-fold difference in NGF protein levels at 5 days was consistent with the 3 6-fold
difference in mRNA levels observed at 5 days. The
increase in NGF mRNA in L-arginineinjected rats compared with controls at 3 days was not accompanied by an
increase of similar magnitude in NGF protein at 3 days.
Whether the magnitude of the difference in NGF
mRNA levels between control and L-arginineinjected
rats was matched by a similar difference in NGF protein
levels appeared to depend on the time point examined
during the development of pancreatitis in this model.
Immunohistochemistry

Figure 4. NGF protein levels in control () and L-arginineinjected ()

rats. No difference was observed between control and L-arginine
injected rats at 2 and 6 hours. NGF protein levels were higher in
pancreas from L-arginineinjected rats than in those from control rats
at 3 and 5 days. The difference was statistically significant at 5 days
*P 0.05.

To identify the cellular sites of increased NGF production in the inflamed pancreas, paraffin sections were
immunostained with an anti-NGF antibody. In pancreatic
tissue sections from control rats, NGF-like immunoreactivity (NGF-IR) was observed only in the islets of Langerhans
at all time points examined (Figure 5A). The intense
NGF-IR was observed in most of the islet cells (Figure 5B).
In sections of pancreas from L-arginineinjected rats at 2
and 6 hours, NGF-IR showed widespread distribution in

1378

TOMA ET AL.

GASTROENTEROLOGY Vol. 119, No. 5

Figure 5. Immunohistochemistry for NGF localization in pancreas paraffin sections from control and L-arginineinjected rats. (A) NGF-IR in a cross
section of pancreas from a control rat (saline-injected). The intense immunoreactivity is present in the islet of Langerhans (arrow). Bar 100
mm. (B) Higher-magnification view of the islet of Langerhans from control rat pancreas. NGF-IR is found in the majority of islet cells. Bar 25
mm. (C) NGF-IR in a cross section of pancreas from an L-arginineinjected rat (2 hours after treatment). NGF-IR is lost in the islet of Langerhans
(arrow). Strong immunoreactivity is present in the surrounding parenchyma. Bar 100 mm. (D) Higher-magnification view of an islet of
Langerhans from L-arginineinjected rat pancreas showing loss of NGF-IR in the majority of islet cells. Bar 25 mm. (E and F ) NGF-IR in pancreas
paraffin sections from (E ) saline-injected and (F ) L-arginineinjected rats 5 days after injection. NGF-IR is predominantly associated with
pancreatic ducts 5 days after induction of pancreatitis (F, arrow). Bar 25 mm.

November 2000

the parenchyma and was observed in the cytoplasm of

exocrine pancreatic tissues, including acinar and ductal cells
(Figure 5C and D). These findings demonstrated a shift in
NGF localization from the endocrine to the exocrine pancreas. In sections of pancreas from L-arginineinjected rats
at 5 days, NGF-IR was scattered in the parenchyma, observed predominantly in ductal cells (Figure 5F, arrow).
Less intense staining was also present in inflammatory cells
(data not shown) and the remaining acinar cells (Figure 5F ).

Discussion
This study documents changes in NGF expression
in a rat model of acute necrotizing pancreatitis. Our
histopathologic findings are in agreement with previous
reports of the induction of pancreatitis by L-arginine in
the rat. Tani et al.16 observed vacuoles in the acinar cells
at 6 hours that are identified as swollen mitochondria and
autophagic vacuoles containing damaged cytoplasmic organelles. The maximal necrotic damage with inflammatory cell infiltration occurred at 3 days,16 as reported in
the present study. Damage appears to be caused by
disruption of acinar cell metabolism, leading to apoptosis
and eventually destruction of the exocrine pancreas.17 A
recent report has shown that L-arginine produces an
increase in apoptosis in rat pancreatic acinar AR4-2J
cells, although the mechanism by which this increase
occurs is still unclear.18 This cell death is accompanied
by inflammatory cell infiltration. Significant increases in
serum levels of -amylase observed during the first 24
hours confirm previous findings that serum amylase level
can be a reliable indicator of successful induction of
pancreatitis in the rat model,13,16 although it does not
appear to be a reliable marker of the severity of acute
pancreatitis in humans.19 Pancreatitis in humans is accompanied by inflammatory cell infiltration. Thus, this
model produces both inflammation and necrosis, factors
that inevitably contribute to severe pain in human pancreatitis.
In the rat model of L-arginineinduced pancreatitis,
changes in NGF distribution and increases in NGF
expression occurred as a function of the pathologic state
of the pancreas. Although no net increase in the levels of
NGF mRNA or proteins was detected at 2 and 6 hours,
the distribution of NGF-IR appeared to shift from the
islets to the parenchyma. It is possible that this change in
staining pattern represents a release of stored NGF from
the islets or a shift in synthesis from the islets to parenchyma with no net increase in total amount produced.
Pancreatic islet cells contain and secrete NGF,20 and
NGF immunostaining increases in exocrine and endocrine tissue after injury to the islets cells.21 Thus, changes

UPREGULATION OF NGF IN RAT PANCREATITIS

1379

in NGF localization may constitute an early response of

the pancreas to damage. Secretion of NGF by cultured
pancreatic cells is increased by glucose and KCl stimulation, indicating that the secretion rate may be regulated by external factors.20 The factor(s) responsible for a
possible increase in secretion in this model is not clear,
although a direct effect of L-arginine on the islet cells
cannot be ruled out. It would be useful to determine
whether production of other islet proteins, e.g., insulin,
is also affected.
The increase in NGF mRNA and protein levels at 3
and 5 days that occurred at the time of almost total
parenchymal necrosis suggests that NGF is regulated by
mechanisms distinct from those at 2 and 6 hours. NGF
up-regulation is a well-documented early event in other
inflammatory conditions. Increases of NGF content in
tissues and afferent nerves occur within a few hours of
initiation of inflammation,10,22,23 and mRNA increases
are detected in inflamed bladder tissues within 2 hours.24
NGF is produced in a number of different cell types in
response to a number of cytokines, including tumor
necrosis factor (TNF)- and interleukin (IL)-1,7 whose
levels are increased in both animal models and human
pancreatitis.25 In the rat model of L-arginineinduced
pancreatitis, inflammation is well advanced and serum
levels of cytokines capable of inducing NGF expression
such as IL-6 and TNF- are increased within 20 hours of
induction and remain elevated for at least 48 hours.13 In
our study, significant increases in tissue levels of NGF
were not evident when inflammatory infiltration was first
observed at 24 hours but appeared at 72 hours, at the
time of maximum tissue necrosis. Our immunohistochemical findings of NGF staining predominantly observed in ductal cells in the inflamed pancreas at 5 days
are similar to those reported for human specimens of
chronic pancreatitis. The highest levels of NGF mRNA
staining in human pancreatitis have been shown in ductal cells and in some degenerating acinar cells, with less
intense staining observed in fibroblasts and inflammatory
cells.12 These data indicate that ductal cells are a possible
major site of increased NGF production during the acute
necrotizing phase of L-arginineinduced pancreatitis.
Mast cells could also be the other cellular source of NGF
in the pathologic condition because they are known to
have the ability to migrate to the site of inflammation
and to release NGF.26
The pancreas in this rat model undergoes a gradual
regeneration over a period of 12 weeks16 (Toma and
Pasricha, unpublished observation). NGF up-regulation
and expression in ductal cells may be part of the regenerative response of the pancreas.21 Cells of the pancreatic

1380

TOMA ET AL.

duct transiently express high levels of trkA and NGF

during development, and NGF is also expressed in precursors of exocrine and endocrine cells located at the
ducts.21 NGF can function as a mitogen/development
factor for pancreatic cells.27 Although high levels of NGF
observed in the ducts in both the rat model and human
chronic pancreatitis12 may be induced as part of a regenerative program, high levels of NGF in the vicinity of
afferent nerves would be expected to affect their function.
Weaver et al.28 observed no change in intrapancreatic
neural elements in this model, but their study did not
directly assess primary afferent nerve function. The extent to which the afferent nerve function is altered by
elevated NGF levels may be clarified by further investigation of the relationship of NGF-producing cells to
afferent nerves both in the rat model and in sections from
human chronic pancreatitis. Whether these pancreatic
afferents show alterations in sensitivity consistent with a
response to NGF also needs to be investigated.
It has been suggested that changes in nerves in human
chronic pancreatitis studies are consistent with a response
to increases in tissue NGF. A previous study found
marked intensification of immunostaining for the NGFresponsive neuropeptides calcitonin gene-related peptide
(CGRP) and substance P (SP)10,29 31 (which were coexpressed) in nerve fibers in pancreatic specimens from
chronic pancreatitis.32 Increased expression of growthassociated protein 43, an NGF-responsive marker of neuronal plasticity,31 has also recently been shown in
pancreatic specimens from chronic pancreatitis.33,34 Although these studies do not specify the type of neurons
in which the changes were observed, it is likely that
spinal sensory neurons that express both CGRP and SP
are affected. There is a well established connection between enhanced tissue levels of NGF, alterations in sensory nerves, including up-regulation of neuropeptides
CGRP and SP, and the development of hyperalgesia in
other models of inflammation.7 This connection, combined with the findings of Friess et al.12 and the increase
in NGF-responsive neuropeptide levels, suggests that
NGF production in the inflamed human pancreas induces plastic changes in the sensory neurons that could
lead to peripheral sensitization. Two potential causes of
pain in human chronic pancreatitis are increased ductal/
parenchymal pressure3537 and continued exposure of
damaged nerves to sensitizing factors generated by recurrent episodes of inflammation/necrosis.2,33,38 The detection of increased levels of NGF in rat and human
pancreatitis is consistent with the inflammation/necrosis
model and may be one of the factors responsible for the
generation of pain. However, Friess et al.12 found an

GASTROENTEROLOGY Vol. 119, No. 5

average 13-fold increase in NGF mRNA in 42% of the

patients studied but no relationship between NGF
mRNA levels and the intensity or duration of the pain.
This finding does not rule out a role for NGF as a factor
in the generation of pain in human chronic pancreatitis
but suggests that other factors may contribute to the
pain. It is also possible that absolute levels of NGF may
not be as important as the site of production (e.g., in
ductal cells or in inflammatory cells in juxtaposition to
sensory nerves). The similarities in the pattern of NGF
staining between human chronic pancreatitis and this rat
model of acute necrotizing pancreatitis suggest that similar mechanisms of NGF regulation may operate in both
conditions. The rat model may prove useful in investigating the role of NGF in pancreatic inflammation and
pain.
Based on the evidence from other organs, it is conceivable that NGF production in the inflamed pancreas is
responsible for plastic changes in the sensory neurons
that mediate peripheral sensitization. This has important
implications for the treatment of pain in this condition.

References
1. Dray A. Inflammatory mediators of pain. Br J Anaesth 1995;751:
125131.
2. DiMagno EP. Toward understanding (and management) of painful
chronic pancreatitis. Gastroenterology 1999;116:12521257.
3. Warshaw AL, Banks PA, Fernandez-Del Castillo C. AGA technical
review: treatment of pain in chronic pancreatitis. Gastroenterology 1998;115:765776.
4. Adler G, Schmid RM. Chronic pancreatitis: still puzzling? Gastroenterology 1997;112:17621765.
5. Konttinen YT, Kemppinen P, Serberberg M, Hukkanen M, Rees R,
Santavirta S, Sorsa T, Polak JM. Peripheral and spinal neural
mechanisms in arthritis, with particular reference to treatment of
inflammation and pain. Arthritis Rheum 1994;37:965982.
6. Woolf CJ, Safieh-Garabedian B, Ma Q-P, Crilly P, Winter J. Nerve
growth factor contributes to the generation of inflammatory sensory hypersensitivity. Neuroscience 1994;62:327331.
7. McMahon SB. NGF as a mediator of inflammatory pain. Phil Trans
R Soc Lond B 1996;351:431 440.
8. Lewin GR, Rueff A, Mendell LM. Peripheral and central mechanisms
of NGF-induced hyperalgesia. Eur J Neurosci 1994;6:19031912.
9. Amann R, Schuligoi R, Herzeg G, Donnerer J. Intraplantar injection
of nerve growth factor into the rat hind paw: local edema and
effects on thermal nociceptive threshold. Pain 1995;64:323
329.
10. Donnerer J, Shuligoi R, Stein C. Increased content and transport
of substance P and calcitonin gene-related peptide in sensory
nerves innervating inflamed tissue: evidence for a regulatory
function of nerve growth factor in vivo. Neuroscience 1992;49:
693 698.
11. Andreev NY, Bennett D, Priestley J, Rattray M, McMahon SB. NGF
mRNA is increased by experimental inflammation of adult rat
urinary bladder. Soc Neurosci Abstr 1994;19:108.
12. Friess H, Zhu ZW, di Mola FF, Kulli C, Graber HU, Andren-Sandberg A, Zimmermann A, Korc M, Reinshagen M, Buchler MW.
Nerve growth factor and its high-affinity receptor in chronic pancreatitis. Ann Surg 1999;230:615 624.
13. Takacs T, Czako L, Jarmay K, Farkas G Jr, Mandi Y, Lonovics J.

November 2000

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

Cytokine level changes in L-arginineinduced acute pancreatitis in

rat. Acta Physiol Hung 1996;84:147156.
Chomczynski P, Sacchi N. Single-step method of RNA isolation by
acid guanidinium thiocyanate-phenol-chloroform extraction. Anal
Biochem 1987;162:156 159.
Tito J, Rudnicki M, Jones DH, Alpern HD, Gold MS. Peptide YY
ameliorates cerulein-induced pancreatic injury in the rat. Am J
Surg 1993;165:690 696.
Tani S, Itoh H, Okabayashi Y, Nakamura T, Fujii M, Fujisawa T,
Koide M, Otsuki M. New model of acute necrotizing pancreatitis
induced by excessive doses of arginine in rats. Dig Dis Sci
1990;35:367374.
Otani T, Idekita M, Tanuma S, et al. Subcutaneous arginine
administration causes apoptosis of rat pancreas (letter). Pancreas 1997;14:316 318.
Motoo Y, Kazuyuki T, Su S-B, Xie M-J, Sawabu N. Arginine induces
apoptosis and gene expression of pancreatitis-associated protein (PAP) in rat pancreatic acinar AR4-2J cells. Pancreas 2000;
20:61 66.
Lankisch PG, Burchard-Reckert S, Lehnick D. Underestimation of
acute pancreatitis: patients with only a small increase in amylase/lipase levels can also have or develop severe acute pancreatitis. Gut 1999;44:542544.
Rosenbaum T, Vidaltamayo R, Sanchez-Soto MC, Zentella A,
Hiriart M. Pancreatic beta cells synthesize and secrete nerve
growth factor. Proc Natl Acad Sci U S A 1998;95:7784 7788.
Teitelman G, Guz Y, Ivkovic S, Ehrlich M. Islet injury induces
neurotrophin expression in pancreatic cells and reactive gliosis
of peri-islet Schwann cells. J Neurobiol 1998;34:304 318.
Aloe L, Tuveri MA, Levi Montalcini R. Studies on carrageenaninduced arthritis in adult rats: presence of nerve growth factor
and role of sympathetic innervation. Rheumatol Int J 1992;12:
213216.
Safieh-Garabedian B, Poole S, Allchorne A, Winter J, Woolf CJ.
Contribution of interleukin-1-beta to the inflammation induced
increase in nerve growth factor levels and inflammatory hyperalgesia. Br J Pharm 1995;115:12651275.
Oddiah D, Anand P, McMahon SB, Rattray M. Rapid increase of
NGF, BDNF and NT-3 mRNAs in inflamed bladder. Neuroreport
1998;9:14551458.
Denham W, Yang J, Fink G, Denham D, Carter G, Ward K, Norman
J. Gene targeting demonstrates additive detrimental effects of
interleukin 1 and tumor necrosis factor during pancreatitis. Gastroenterology 1997;113:17411746.
Leon A Buriani A, Dal Toso R, Fabris M, Romanello S, Aloe L.
Mast cells synthesize, store, and release nerve growth factor.
Proc Natl Acad Sci U S A 1994;91:3739 743.
Bold RJ, Ishizuka J, Rajaraman S, Perez-Polo JR, Townsend CM Jr,

UPREGULATION OF NGF IN RAT PANCREATITIS

28.

29.

30.

31.

32.

33.

34.

35.
36.
37.

38.

1381

Thompson JC. Nerve growth factor as a mitogen for a pancreatic

carcinoid cell line. J Neurochem 1995;64:2622 628.
Weaver C, Bishop AE, Polak JM. Pancreatic changes elicited by
chronic administration of excess L-arginine. Exp Mol Pathol 1994;
60:71 87.
Kuraishi Y, Nanayama T, Ohno H. CGRP increases in the dorsal
root ganglia of the adjuvant arthritic rat. Peptides 1989;10:447
452.
Lindsay RM. The role of neurotrophic factors in functional maintenance of mature sensory neurons. In: Scott SA, ed. Sensory
neurons: diversity, development, and plasticity. New York: Oxford
University Press, 1989:404 420.
Leslie TA, Emson PC, Dowd PM, Woolf CJ. Nerve growth-factor
contributes to the up-regulation of growth-associated protein-43
and preprotachykinin-A messenger-RNAs in primary sensory neurons following peripheral inflammation. Neuroscience 1995;67:
753761.
Buchler M, Weihe E, Friess H, Malfertheiner P, Bockman E, Muller
S, Nohr D, Berger HG. Changes in peptidergic innervation in
chronic pancreatitis. Pancreas 1992;7:183192.
Di Sebastiano P, Fink T, Weihe E, Friess H, Innocenti P, Beger
HG, Buchler M. Immune cell infiltration and growth-associated
protein 43 expression correlate with pain in chronic pancreatitis.
Gastroenterology 1997;112:1648 1655.
Fink T, Di Sebastiano P, Buchler MW, Beger HG, Weihe E. Growthassociated protein-43 and protein gene product 9.5 innervation
in human pancreas: changes in chronic pancreatitis. Neuroscience 1994;63:249 266.
Bladley EL. Pancreatic duct pressure in chronic pancreatitis. Am J
Surg 1982;114:313316.
Jalleh RP, Aslam M, Williamson RC. Pancreatic tissue and ductal
pressures in chronic pancreatitis. Br J Surg 1991;8:12351237.
Karanjia ND, Widdison AL, Leung F, Alvarez C, Lutrin FJ, Reber
HA. Compartment syndrome in experimental chronic obstructive
pancreatitis: effect of decompressing the main pancreatic duct.
Br J Surg 1994;81:259 264.
Hunger RE, Mueller C, Zgraggen K, Friess H, Buchler MW. Cytotoxic cells are activated in cellular infiltrates of alcoholic chronic
pancreatitis. Gastroenterology 1997;112:1656 1663.

Received December 30, 1999. Accepted June 21, 2000 .

Address requests for reprints to: Pankaj J. Pasricha, M.D., Department
of Internal Medicine, University of Texas Medical Branch, 301 University
Boulevard, Galveston, Texas 77555. e-mail: jpasrich@utmb.edu; fax:
(409) 772-4789.
Funded through a commitment account.