| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
U.S. Department of Agriculture, Agricultural Research Service (T.H.E., S.K., R.F.), Beltsville, Maryland 20705; the Department of Physiology and Pharmacology, Auburn University (J.L.S.), Auburn, Alabama 36849; and the Department of Cell and Cancer Biology, National Cancer Institute, National Institutes of Health (A.M., L.M., R.P., M.J.M., F.C.), Bethesda, Maryland 20892
Address all correspondence and requests for reprints to: Dr. Ted H. Elsasser, U.S. Department of Agriculture, Agricultural Research Service, Beltsville, Maryland 20705. E-mail: elsasser{at}lpsi.barc.usda.gov
| Abstract |
|---|
|
|
|---|
, insulin, and AM
and up-regulated iNOS activity. These acute complications rapidly
progress into a more chronic state characterized by diminished insulin
response to feeding stimulus and colocalized increases in pancreatic
islet AM and iNOS. The pancreatic responses in AM and iNOS may play a
major role in mediating prolonged disturbances in nutrient use by
tissues through their influences on temporal patterns of pancreatic
hormone secretion during chronic illness. | Introduction |
|---|
|
|
|---|
During the acute phase response, cytokine-driven induction of nitric oxide synthase (iNOS) increases NO production in many cell types (1). The use of arginine analogs to disrupt NOS generation of NO also attenuates the actions of AM (13, 14), further suggesting a link between AM and NO in the regulation of cellular processes. Although several cardiovascular organ perfusion functions of AM have been linked to NO-dependent mechanisms (13, 17) during lethal endotoxemia, few, if any, studies have examined the systemic and localized changes in AM and iNOS in low level disease states, and none has explored the involvement of AM in the metabolic perturbations seen in disease stress.
Recently, we suggested that AM plays a significant role in regulating metabolism through AMs ability to modulate insulin responses in vivo and in vitro (8). The aims of the present study were to 1) determine whether AM is affected by nonlethal, chronic parasitic disease that has underlying elements of insulin secretion impairment and is further altered in response to challenge with low level endotoxin when additive to this chronic disease, and 2) identify the tissue site(s) where AM and iNOS responses might be localized.
| Materials and Methods |
|---|
|
|
|---|
On the day immediately preceding the LPS and saline challenges, each animals feed was removed at 1600 h. New feed was offered each animal at 1400 h on the day of challenge, 6 h after administration of the challenge. Refeeding after fasting was used as a physiological stimulus for food-induced insulin release. Feed intake after refeeding was measured in individual animals.
Blood sampling and tissue collection
On days 2930 postinfection, a sterile Teflon cannula
(Abbocath, Abbott, North Chicago, IL) was inserted into a jugular vein
of each calf for blood sampling and administration of endotoxin or
saline. After an initial baseline blood sample (8 ml, EDTA
anticoagulant), calves were injected with either 3 ml pyrogen-free
saline or LPS. Additional blood samples were obtained via cannula up
through 24 h relative to the administration of LPS. After the last
blood sample, calves were moved to the abattoir, at which time they
were killed by captive bolt and exsanguination. Samples of liver, lung,
and pancreas were obtained and immediately placed into Bouins
fixative or frozen in liquid nitrogen and maintained at -80 C.
Analytical
Hormones, cytokines, and metabolites. Insulin (3), tumor
necrosis factor-
(TNF
) (3, 22), and AM (23) were measured by
validated RIAs. Plasma concentrations of nitrite were measured using
standard methods in the Greiss reaction after enzymatic conversion of
plasma nitrate to nitrite with bacterial nitrate reductase (22).
Glucose was measured using a solid phase membrane glucose oxidase-based
analyzer (3) (YSI, Inc., Yellow Springs, OH).
Western blot confirmation of AM and iNOS. Extracts of the pancreatic tissue were prepared by thawing the tissue on ice and homogenizing accurately weighed samples in ice-cold homogenization buffer [5:1, vol/wt; 50 mM HEPES, 1 mM EDTA, and 1 mM dithiothreitol, pH 7.2, containing soy trypsin inhibitor (10 µg/ml), antipain, pepstatin, aprotinin, leupeptin (1 µg/ml), and phenylmethylsulfonylfluoride (100 µg/ml)] using a Polytron (Brinkmann Instruments, Inc., Westbury, NY) in two bursts of 20 sec each at maximum speed. The homogenate was centrifuged at 100,000 x g for 30 min, and the supernatant was collected. Homogenate supernatant protein content was measured using the bicinchoninic acid procedure (Pierce Chemical Co., Rockford, IL) after trichloroacetic acid precipitation and NaOH resolubilizing of the extract protein.
For iNOS Western blots, protein from pancreas extracts (100 µg/well) was electrophoretically separated on 38% gradient polyacrylamide Tris-acetate gels (Novex, San Diego, CA) 120 V for 1.5 h under reducing conditions. Proteins in the gels were transferred to nitrocellulose by semimoist transfer blotting. Nonspecific binding was blocked by incubating the nitrocellulose overnight with 5% nonfat milk containing 0.1% Tween-20 in PBS at 4 C. The membrane was probed for iNOS immunoreactivity with rabbit antimouse macrophage iNOS specific for the carboxyl-terminal NADPH binding region (Transduction Laboratories, Inc., Lexington, KY). Protein bands were visualized by chemiluminescence excitation of autoradiographic film using an avidin-biotin-peroxidase complex (Vector Laboratories, Inc., Burlingame, CA) and ECL Plus Western blotting reagents (Amersham Pharmacia Biotech, Piscataway, NJ). Band intensities were quantified by densitometry using the ChemiImager 4000 (Alpha Inntech Co., San Leandro, CA).
Western blot analysis of AM was performed on the same pancreatic extracts using 1020% acrylamide gradient Tricine gels as previously performed by Miller et al. (7).
Immunocytochemistry
Light microscopy. For immunohistochemistry, Bouins fixed
tissues from each calf were embedded in paraffin, sectioned at 6 µm,
and mounted on poly-L-lysine-coated glass slides (8). After
deparaffinizing in xylene, tissue sections were rehydrated, blocked
with normal goat serum, and incubated overnight at 4 C with either
antimouse macrophage iNOS (Transduction Laboratories, Inc.; 5
µg/ml anti-iNOS) or rabbit anti-AM [serum 2343 (8); 1:1000
dilution]. Initial determinations of cell-specific localizations of AM
and iNOS were obtained using serial cut sections of the
paraffin-embedded tissues from calves challenged with infection and
LPS. The serial sections were immunostained for AM, iNOS, somatostatin,
glucagon, and pancreatic polypeptide. Immunoreactivity was visualized
using the avidin-biotin horseradish peroxidase complex method
(Vector Laboratories, Inc.), followed by a light nuclear
counterstain with Gills hematoxylin. Percentages of AM- and
iNOS-immunopositive cells within islets were obtained by direct
counting of cells in multiple islets in 20 fields under a x20
objective.
Confocal microscopy. Colocalizations were further studied using confocal microscopy. Paraffin sections were dewaxed and rehydrated through a graded ethanol series. Sections were blocked with normal donkey serum (1:30 in PBS) for 30 min and then incubated overnight at 4 C in a mixture of three antibodies: guinea pig antibovine insulin (CAPPE/Labs, Inc., King of Prussia, PA) at 1:2000 dilution, mouse monoclonal anti-iNOS (Transduction Laboratories, Inc.) at 1:200 dilution, and rabbit anti-AM at 1:1000 dilution. The second layer consisted on a mixture of Cy5-antiguinea pig (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA), Bodipy antirabbit (Molecular Probes, Inc., Eugene, OR), and biotinylated horse antimouse serum (Vector Laboratories, Inc.), each at a final dilution of 1:200. A third layer containing lissamine-rhodamine-streptavidin (Jackson Immuno-Research Laboratories, Inc.) at 1:200 dilution was used to detect iNOS. Sections were observed with a Carl Zeiss confocal microscope (New York, NY) equipped with four lasers.
Statistical analysis
Data were statistically analyzed using regression analysis in
the general linear models (24) procedure of SAS (SAS Institute, Cary,
NC) with nonorthogonal contrasts to differentiate among the effects of
LPS, infection, and additivity. Area under the concentration-time curve
responses for hormones and metabolites were estimated by trapezoidal
summation of the total area with the specific response presented as the
response above baseline after subtraction of the associated baseline
area after time zero.
| Results |
|---|
|
|
|---|
The immune challenge model used in this study clearly demonstrated that some aspects of the infection-response process in the host can be compartmentalized into two distinct phases. When a host is challenged by immune stimuli that elicit the customary acute phase response, an early onset phase initiates, marked by and easily quantified through rapid, large changes in plasma concentrations of hormones, and further significantly affected by the additivity of concurrent stresses. Secondarily, a more chronic, subtle phase can be delineated when a quantification of the response is shifted to tissue-specific markers of metabolic perturbation and perturbed regulation of hormone secretion normally regulated by physiological secretagogues such as feeding.
Concentrations of glucose, nitrate/nitrite, hormones, and TNF
in
plasma were measured in samples collected through the 24-h point after
LPS challenge. Mean plasma concentrations of glucose were not different
between experimental animal groups before LPS administration (Table 1
). After LPS, plasma glucose
concentrations changed in a biphasic pattern; the immediate period of
hyperglycemia peaked at 1 h after LPS and was numerically higher
in LPS + PI than in LPS alone, but the difference was not statistically
significant. By 3 h post-LPS the hypoglycemic period ensued. In
calves challenged with saline, plasma glucose concentrations remained
stable throughout the postchallenge sampling period and were not
affected by refeeding at the 6 h point. In calves challenged with
LPS, plasma glucose responses were at a nadir at the 3 h point
after LPS challenge and slowly returned to prechallenge concentrations
by 12 h.
|
|
levels increased in all calves challenged with LPS, with
numerically greater responses occurring in PI + LPS calves than in
calves receiving LPS only (Table 2
|
|
|
|
Western blot analysis of pancreatic homogenate proteins for iNOS
demonstrated protein bands migrating at a relative molecular mass of
117 kDa under the reducing conditions employed. A moderate degree of
band density was present in all control calves (Fig. 5
). However, the band intensity increased
up to 3.5-fold in samples from animals with underlying infection
challenged with LPS. The relative intensity of the iNOS bands for each
animal tested was significantly correlated (r2 = 0.64;
P < 0.02) with the increasing percentage of islet
cells staining positive for iNOS between treatments. Western blot
analysis of the pancreatic extracts indicated that AM was present
largely as the higher molecular mass precursor forms, pro-AM and
prepro-AM, with relative molecular weights of 14 and 18 kDa,
respectively (data not shown).
|
|
|
| Discussion |
|---|
|
|
|---|
Although the coordination and relationship between the early acute response and the late phase chronic response need to be further explored, the linear correlation of TNF and AM responses to the LPS challenge are consistent with literature citations suggesting a cytokine-driven mechanism for the increase in circulating plasma AM after LPS challenge within the early phase of the response (13, 14, 15, 16). With regard to a potential source for the AM measured in plasma after LPS challenge, Isumi et al. (25) suggested that the AM response to cytokines released post-LPS in vascular endothelial and smooth muscle cells is sufficient to account for the measured increases in plasma AM, which peak 34 h after LPS.
Although the acute AM response is easily measured in plasma AM changes from baseline, it appears that the more chronic phase of the host response to these immune stimuli may be more of a localized paracrine nature not readily apparent in terms of plasma concentrations of AM. Western blot analysis of extracts of calf pancreas indicated that immunoreactive AM was present in the higher molecular mass precursor forms rather than the 6-kDa form to which biological activity is ascribed. The absence of 6-kDa AM is not uncommon, in that authentic AM is rapidly secreted from sites of processing but accumulates in cell culture fluid (7, 25), in temporal agreement with the induction of AM mRNA after stimulation.
The coordinated increase in pancreatic islet AM and iNOS can be implicated as a key factor participating in the pancreatic response to disease that complicates insulin and glucose homeostasis. These data suggest for the first time that the pancreas develops a shock organ-type response in calves challenged with LPS or infection that resolves to a coordinated localized increase in both iNOS and AM. Interpreting the immunohistochemical data raises some interesting questions regarding whether the pancreatic response to either low level infection or LPS is more sensitive to disease stress factors that affect pancreatic hormone secretion regulation or whether the pancreas has a rather unique capacity to maintain these AM and iNOS responses for a period of time considerably longer than those in the lung or liver. This is particularly striking considering that the dose of LPS per unit BW used in the present study was approximately 1/10,000th the typical dose used to stimulate iNOS or AM responses in several rodent models and was clinically relevant because human sensitivity to LPS is also far greater than that of the rat (22).
In terms of tissue response, the interesting feature about the pancreas appears to be that the AM and iNOS responses are different from those that develop in liver or lung, the more traditional shock organs in disease. In the present study, pancreas specimens harvested 24 h after LPS challenge still maintained AM- and iNOS-immunopositive staining, whereas the liver parenchyma was negative, and the lung showed no more immunostaining than that normally observed in control animals. These localized paracrine effects are typical of NO-mediated responses due to the chemical properties of tissue diffusion potential and relative half-life of NO (26) and the autocrine and paracrine character of AM actions in development, growth, metabolism, and neoplasia (27). Data from our laboratory (unpublished) demonstrated that doses of LPS greater than 1.0 µg/kg administered iv to calves similar to the normal calves used here could cause an up-regulation of immunohistochemically identifiable iNOS in lung epithelia as well as in liver and an increase in AM staining on infiltrating monocytes at 56 h postchallenge. This immunostaining subsided within a few hours. The transient expression of these effector molecules in lung and liver is consistent with observations by others on the timing of the induction and loss of mRNA for AM and iNOS in lung, liver, and monocytes (13, 28, 29). The data here suggest that the pancreatic response to inducers of AM and iNOS (LPS in this case) is either more sensitive or of longer duration than that characteristically seen in lung and liver. The contributions that localized changes in cell AM responses make to deleterious (hypotension, metabolic shunting) as well as beneficial (antiapoptotic, antimicrobial peptide) (27) biological processes remain open for debate.
Contained in our hypothesis was the concept that changes in local tissue and pancreatic islet cell function can evolve during low level chronic disease stress and predispose animals to greater pathophysiological responses when further subjected to acute challenges from bacterial toxins. The present report integrates and defines new elements of a pancreatic-immune axis that appear to play a significant role in the metabolic perturbations of disease stress. The coordinated up-regulation of AM and iNOS as components of the pancreatic-immune axis may mediate changes in pancreatic function observed during low level chronic disease stress. These observations complement the mechanism originally suggested by Mandrup-Poulsen et al. (30) and extended by McDaniel (31) and others (32) that nitric oxide generation by pancreatic cells limits the release of insulin and contributes to perturbations of the islets. That this response may be particular to specific hormones in the pancreas such as insulin and not simply a generalized occurrence is supported by the facts that 1) plasma glucagon responses after LPS in these calves were neither temporally nor concentration different between LPS and PI + LPS (data not presented); 2) we previously demonstrated that AM inhibited insulin release from isolated cultured islets (8); and 3) recently, NO has been shown to confer aspects of differential hormone regulation in the pancreas by inhibiting insulin and stimulating glucagon secretion in experiments where N-nitro-L-arginine methyl ester was used to inhibit NOS (33). We had little indication in the present study that the iNOS response was greatly manifest in the ß-cells, although light immunostaining was evident in ß-cells in calves experiencing both immune stresses. Our data are consistent with the capability for islet cells themselves to generate NO in the iNOS cascade, separate and apart from that which might be generated from inflammatory cells that migrate to or reside in the pancreas (31, 32). Light microscopy and monocyte staining procedures failed to demonstrate the presence of infiltrating immune cells in the pancreas of calves at the time of tissue fixation.
Constitutive forms, in contrast to the inducible form, of NOS
have been identified in several cell types in the pancreas, most
notably as neuronal NOS colocalized specifically with somatostatin
-cells (34) and some ß-cells (35). However, the present report is
the first to suggest that the inducible isoform can be present in the
AM/pancreatic polypeptide cells of islets and is coordinately
up-regulated with AM by low level parasitic infection as well as LPS in
both chronic and acute disease stress scenarios. A connection between
AM and NO was initially speculated to exist in the cardiovascular
system, i.e. the demonstration that the hypotensive effects
of AM were blocked in the presence of
N-nitro-L-arginine methyl ester, a competitive
inhibitor of NOS (36). Similarly, data from So et al. (13)
suggested that a significant part of AM function may be mediated by
AMs ability to induce NO production via the iNOS pathway. In
addition, AM has been implicated as having immunomodulatory capability
through its effects on interleukin and TNF
expression in activated
macrophages (37) and conversely as demonstrated by the effects of LPS
and the LPS-mimetic taxol to stimulate AM mRNA content and peptide
release from macrophages (38). Thus, the paracrine/autocrine function
of AM in the pancreas may be related to the modulation of NO production
within the cells in which AM and iNOS colocalize.
The relative importance of potential endocrine actions of AM (as
reflected in the acute rise and fall in plasma after LPS) compared with
the impact of paracrine effects (consistent with the increased
pancreatic localization of AM in the infected calves) is not clear.
Previous data from our laboratory demonstrated that this sarcocystis
infection causes a blunting of insulin responses to provocative
arginine challenge (39). The lag in insulin response to refeeding most
evident in infected calves further challenged with LPS is consistent
with the capacity for AM to blunt insulin release as previously
demonstrated in isolated cultured islets and in vivo after
glucose challenge (8). Furthermore, patterns of insulin measured in
plasma after LPS treatment were consistent with the concentration and
temporal changes in plasma TNF
and glucose during the hyperglycemic
and hypoglycemic phases of response. A significant NO response,
mirrored in the changes in nitrate measurable in plasma, is similarly
consistent with the acute induction of the high output NO pathway
mediated through iNOS in the cytokine response cascade after LPS.
Issues associated with more chronic changes in plasma glucose after LPS
and in association with infection are more difficult to interpret in
these experiments. The interpretation of the consequences associated
with and mechanisms underlying changes in tissue fluxes as well as
plasma concentrations of glucose are made difficult due in no small
part to the simultaneous alterations in pancreatic hormone release,
hepatic glucose-glycogen turnover, and paradoxical differences in
TNF-driven insulin resistance and glucose uptake in muscle (40, 41).
Therefore, the plasma concentrations of glucose present during the
postfeeding events may or may not be reflective of the tissue-specific
impact of the altered endocrine-immune milieu on glucose and energy
metabolism, even though significant differences in the insulin response
to feeding were present between some treatments.
As reviewed previously (5, 42), the degree to which TNF
and iNOS are
up-regulated by infection and LPS is highly dependent on the intensity
of the immune challenge, the nature of the cytokine milieu evoked, and
the type of tissue in which the response is studied. The ability of
cytokines such as interleukin-1 and TNF
to induce AM and iNOS
expression in nonimmune cells, such as vascular endothelial cells (43),
is consistent with the observed local islet cell response in the
endocrine pancreas. The coordinated amplification and colocalization of
AM and iNOS in islet cells as demonstrated in the present study
suggests some common functional linking of pancreatic tissue responses
to AM and the immunoeffector NO. The data further suggest that the
previously observed capacity for AM/NO to inhibit insulin secretion in
normal animals may be exacerbated in disease processes that result in
localized NO (and AM) up-regulation in the pancreas. Although previous
reports illustrated that frank cytotoxicity associated with
overproduction of NO in islets as driven by specific cytokine induction
can result in ß-cell destruction (31), the present report underscores
the sensitivity of the pancreas to low level infection and the apparent
paracrine mediation of hormone secretion in the pancreas via AM and
NO.
Disease stress may participate in and increase the risk for long lasting metabolic disease to develop in the pancreas (44). In essence, the presence of low level and subclinical infection may be sufficient to exacerbate a pathological response to a secondary infection within the pancreas, with the end result being a chronic disease state coincident with a localized organ response. Host resistance and susceptibility to disease vary with the competence of the immune system and are compromised in instances such as human immunodeficiency virus infection, diabetes, and other endocrinopathies and in association with certain therapeutic interventions in cancer. Often, patients suffering from these conditions harbor chronic, occult infections, and when further challenged with a secondary low level infection, present clinically significant signs of multiple organ or metabolic dysfunction, which have been attributed to NO effects (41). With the link established between AM and NO, intervention strategies targeting modulation of both AM and NO relationships to disease may have clinical application and relevance similar to those proposed by others in multiple organ failure associated with sepsis.
| Acknowledgments |
|---|
| Footnotes |
|---|
2 Present address: Department of Histology and Anatomical Pathology,
University of Navarra, Pamplona 31080, Spain. ![]()
3 Recipient of a fellowship from Instituto de Salud Carloss III,
Ministerio de Sanidad y Consumo, Spain (Grant 98/9172). ![]()
Received December 2, 1998.
| References |
|---|
|
|
|---|
and plasma and urinary
concentrations of NO2 and NO3 in calves. Proc
Soc Exp Biol Med 215:370376
(TNF-
)-induced insulin resistance:
evidence for differential regulation of insulin signaling by TNF-
.
Endocrinology 139:49284935
on glucose metabolism in
cultured human muscle cells from nondiabetic and type-2 diabetic
subjects. Endocrinology 139:47934800This article has been cited by other articles:
![]() |
T. H. Elsasser, T. J. Caperna, C-J. Li, S. Kahl, and J. L. Sartin Critical control points in the impact of the proinflammatory immune response on growth and metabolism J Anim Sci, April 1, 2008; 86(14_suppl): E105 - E125. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. H. Elsasser, J. W. Blum, and S. Kahl Characterization of calves exhibiting a novel inheritable TNF-{alpha} hyperresponsiveness to endotoxin: associations with increased pathophysiological complications J Appl Physiol, June 1, 2005; 98(6): 2045 - 2055. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. Y. F. Wong, B. M. Y. Cheung, Y.-Y. Li, and F. Tang Adrenomedullin Is Both Proinflammatory and Antiinflammatory: Its Effects on Gene Expression and Secretion of Cytokines and Macrophage Migration Inhibitory Factor in NR8383 Macrophage Cell Line Endocrinology, March 1, 2005; 146(3): 1321 - 1327. [Abstract] [Full Text] [PDF] |
||||
![]() |
W. Kim, S.-O. Moon, S. Lee, M. J. Sung, S. H. Kim, and S. K. Park Adrenomedullin Reduces VEGF-Induced Endothelial Adhesion Molecules and Adhesiveness Through a Phosphatidylinositol 3'-Kinase Pathway Arterioscler. Thromb. Vasc. Biol., August 1, 2003; 23(8): 1377 - 1383. [Abstract] [Full Text] [PDF] |
||||
![]() |
O. Garmendia, M. P. Rodriguez, M. A. Burrell, and A. C. Villaro Immunocytochemical Finding of the Amidating Enzymes in Mouse Pancreatic A-, B-, and D-cells: A Comparison with Human and Rat J. Histochem. Cytochem., October 1, 2002; 50(10): 1401 - 1416. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Martinez, S. Kapas, M.-J. Miller, Y. Ward, and F. Cuttitta Coexpression of Receptors for Adrenomedullin, Calcitonin Gene-Related Peptide, and Amylin in Pancreatic {beta}-Cells Endocrinology, January 1, 2000; 141(1): 406 - 411. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Pio, A. Martinez, E. J. Unsworth, J. A. Kowalak, J. A. Bengoechea, P. F. Zipfel, T. H. Elsasser, and F. Cuttitta Complement Factor H Is a Serum-binding Protein for Adrenomedullin, and the Resulting Complex Modulates the Bioactivities of Both Partners J. Biol. Chem., April 6, 2001; 276(15): 12292 - 12300. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||