| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Cedars-Sinai Research Institute-UCLA School of Medicine, Los Angeles, California 90048
Address all correspondence and requests for reprints to: Shlomo Melmed, M.D., Academic Affairs, Room 2015 Cedars-Sinai Medical Center, 8700 Beverly Boulevard, Los Angeles, California 90048. Email: melmed@csmc.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We recently demonstrated pituitary LIF gene expression in human fetal (12) and adult pituitary tissue (13) and in murine corticotrophs (12). LIF potently enhances CRH induction of POMC transcription, induces ACTH secretion in vitro (14) and in vivo (15), and is essential for psychological stress-activation of the hypothalamic pituitary adrenal (HPA) axis (15). LIF and LIF receptor are constitutively expressed in the normal murine hypothalamus and pituitary and are induced in response to lipopolysaccharide endotoxin (LPS) injection (16). Taken together, these data indicate that LIF is an inducible hypothalamo-pituitary proinflammatory cytokine that functions as either an autocrine or paracrine ACTH regulator.
HPA axis activation during inflammation is an important protective
mechanism, as the neuroendocrine stress response and resultant
induction of endogenous corticosteroids suppress immune reaction
(17, 18, 19, 20, 21, 22) including experimental allergic encephalomyelitis
(23), adjuvant-induced arthritis (24, 25),
and local inflammatory process (26). We recently showed
that LIF provides a sustained activation of the HPA axis response to
acute psychological stress (15). Inflammation is
considered a stressor in terms of HPA axis stimulation. However,
different mechanisms are involved in the activation of the HPA axis in
inflammatory compared with psychological stress. Several inflammatory
cytokines such as IL 1ß, IL 6, and TNF
are major mediators of the
inflammatory response and activate the HPA axis both at the periphery,
as well as centrally (27, 28, 29). We have previously shown
that LIF modulates IL 1ß-induced activation of this system
(30). In this study, we sought to determine whether LIF
potentiates the HPA axis response to inflammation by acting as an
inducer of POMC expression and ACTH secretion in the course of an
inflammatory process.
In rats, complete Freund’s adjuvant (CFA) injected sc induces hyperactivation of cellular immune function and a systemic cell-mediated inflammatory reaction during the first 7 days after injection. Within 2 weeks, this process results in development of autoimmune adjuvant-induced arthritis (31). Although mice are less susceptible to the effect of CFA and do not develop autoimmune disease, CFA stimulates murine cellular immune mechanisms (32) and the early phase of inflammation leading to HPA axis activation (15). We studied mice with a disrupted LIF gene (LIFKO) to examine the involvement of hypothalamic and pituitary LIF in activating the HPA axis in response to systemic immune stimulation induced by injection of complete CFA, or by acute local inflammation induced by turpentine administration. We assessed levels of LIF gene expression in the hypothalamus and pituitary of WT mice in both models. Recently, a family of suppressors of cytokine-signaling (SOCS) has been described (33). We have shown that a member of this family, SOCS 3, is a LIF-inducible intracellular regulator of POMC gene expression and ACTH secretion (34). Here we studied hypothalamic and pituitary SOCS 3 expression as an additional indicator of LIF gene stimulation. In wild-type (WT) mice hypothalamic LIF, pituitary SOCS 3 and POMC expression and blood ACTH and corticosterone levels were elevated during the course of inflammation in both models. In contrast, LIFKO mice did not mount an ACTH response and demonstrated blunted POMC and corticosterone responses to both CFA and turpentine injection. Thus, LIF appears critical for mediating HPA axis response to inflammation.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Injection of CFA
Mice were given a single 0.2 ml sc tail base injection of a 10
mg/ml suspension of ground heat-killed Micobacterium butiricum (Difco,
Detroit, MI) in paraffin oil (Fluka Chemica-Biochemica,
Buchs, Switzerland). Animals were killed 0, 1, 3, and 5 days after
injection. In a preliminary experiment control, WT animals were
injected sc with 0.2 ml normal saline.
Induction of local inflammation
Sterile local inflammation was induced by left hind limb muscle
injection of 50 µl/100 g BW turpentine. This produced pronounced
swelling visible 7–24 h after injection. Animals were killed at time 0
(baseline), and 1, 7, 18, and 24 h after treatment. In preliminary
experiments, control WT animals was injected im with 50 µl normal
saline.
LIF replacement
Murine LIF (5 µg) in 0.2 ml normal saline, generously provided
by Dr. R. Klupacs (AMRAD, Victoria, Australia) was injected ip to LIFKO
mice simultaneously with im injection of 50 µl/100 g BW turpentine.
Animals were killed 18 h after treatment. Control animals received
no injection, or turpentine injection only.
Tail biopsy, blood collection, and hormone assay
Five millimeters of mouse tail were cut under light isoflurane
narcosis (Isofluran, Abbott Laboratories, North Chicago,
IL) and genomic DNA extracted (Purigene, Genetech, Inc.,
Research Triangle Park, NC). Whole blood was obtained immediately after
decapitation and plasma collected between 1000 h and 1200 h,
in ice-chilled tubes containing 0.1% EDTA, separated and stored at
-70 C until assayed. Plasma ACTH (Nichols Institute Diagnostics, San Juan Capistrano, CA) and plasma corticosterone
(ICN Biomedicals, Inc., Costa Mesa, CA) were measured by
commercially available RIAs. Sensitivity of ACTH and corticosterone
assay were 10 pg/ml and 25 ng/ml, respectively. Inter and intraassay
variability for ACTH was 7.3 and 3.1%, respectively; inter and
intraassay of variability for corticosterone was 4.4 and 6.5%,
respectively.
Tissue dissection and RNA isolation
Mice were decapitated, pituitary and hypothalamus dissected, and
tissue immediately frozen on dry ice and kept at -70 C until RNA
extraction. Total tissue RNA was extracted with Trizol reagent
(Life Technologies, Inc., Gaithersburg, MD) according to
manufacturer’s instructions, in which 5–100 mg frozen tissue were
immersed in 1 ml Trizol solution and homogenized with a Polytron
homogenizer, dissolved in diethylpyrocarbonate water, and RNA
concentration spectrometrically quantitated and quality checked by gel
electrophoresis.
Northern blot hybridization analysis
Northern analysis was performed using 2–25 µg total RNA/lane.
Samples were electrophoresed through a 1% agarose gel containing 2.2
M formaldehyde, transferred to a nylon membrane (Nytran,
228, Schleicher & Schuell, Keene, NH), membranes
UV-cross-linked and blots prehybridized in 1 M NaPO4, 20%
SDS and 0.1% BSA for 1 h at 65 C. 106
cpm/ml 32P-labeled specific probes were added and
membranes hybridized overnight at 65 C. Membranes were then washed in 2
xSSC, 0.1% SDS for 30 min; 1 x SSC. 0.05%, SDS for 30 min;
0.5 x SSC, 0.025% SDS for 1 h, 0.1 x SSC, 0.005% SDS
for 1 h at 65 C, and exposed to Kodak Biomax film for
1–24 h for POMC, SOCS 3, ß actin or 72–96 h for CRH and LIF at -70
C.
Northern blot data analysis
The experiment with LIF replacement was performed two times, and
the results of both Northern blots are presented (see Fig. 7
). Other
experiments were performed three times. Autoradiographs of each
Northern blot show results of one representative experiment (from 3) in
pools of 7–8 pituitaries or 7–8 hypothalami per sample.
|
Plasmids and templates
Mouse ß actin (mouse DECA probe template, Ambion, Inc. Austin, TX) is a 1.076-kb fragment of the mouse cytoplasmic
ß actin gene. The EcoRI-XbaI fragment of the
mLIF complementary DNA (cDNA) spanning the entire coding sequence of
mLIF (2–631 bp; GenBank accession number A01690; provided by Dr. Tracy
Willson, Walter and Eliza Hall Institute of Medical Research,
Melbourne, Australia) was cloned into a pcDNA3 vector, isolated and
electrophoresed in 1.2% agarose gel, and extracted with Quiaex II.
Murine SOCS 3 cDNA (19–610 bp; GenBank accession number U88328; 20-bp
primers) was isolated in our laboratory by RT-PCR of murine pituitary
mRNA (34). Before use as a template for random priming,
the specificity of the RT-PCR product was verified by multiple
restriction enzyme analysis. 0.6-kb fragment of murine POMC cDNA,
encoding the 3' half of exon 3 was kindly provided by Dr. Malcolm J.
Low (Portland, OR). Mouse CRH (pGEM vector containing a 578 bp of Exon
II) was a generous gift of Dr. Audrey Seasholtz (Ann Arbor, MI). Probes
were labeled by random priming with RadPrime DNA Labeling System
(Life Technologies, Inc., Gaithersburg, MD).
Statistical analysis
Results were analyzed using two-way ANOVA to assess differences
between genotypes, and one-way ANOVA within genotypes, and are
presented as mean ± SE. A probability of
P < 0.05 was considered significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Hypothalamic and pituitary LIF and SOCS 3 expression are both very low
in intact WT animals. Treatment with CFA led to increased LIF gene
expression in the hypothalamus on day 1 (2.2 ± 0.0.25-fold,
P < 0.05), day 3 (2.0 ± 0.18-fold,
P < 0.05) and day 5 (1.8 ± 0.08-fold,
P < 0.05) after injection. In the pituitary, LIF mRNA
levels were weakly expressed 1 day after CFA injection (Fig. 1, a and b). Administration of CFA
induced pituitary SOCS 3 gene expression. Strong induction of SOCS 3
was apparent on day 3 (4.3 ± 0.5-fold, P < 0.05)
and day 5 (3.8 ± 0.28-fold, P < 0.05) after
treatment (Fig. 1, a and c).
|
).
|
). In contrast,
pituitary POMC levels increased at day 1 (3.8 ± 0.3-fold), at day
3 (5.5 ± 0.46-fold) and at day 5 (3.5 ± 0.08-fold) after
CFA inoculation (P < 0.05 compared with untreated
control animals). In LIFKO mice, although POMC expression was
moderately induced, this induction did not reach levels of statistical
significance. Overall, pituitary POMC induction in LIFKO mice was lower
than in WT littermates in response to CFA (Fig. 3, b and c).
|
Northern blot hybridization analysis of hypothalamic RNA derived from
untreated WT mice (time 0), and for up to 24 h after turpentine
injection revealed that LIF expression increased at 1 h (1.67
± 0.06-fold), peaked at 7 h (3.2 ± 0.25-fold,
P < 0.05) and remained persistently elevated for
24 h after injection, compared with untreated mice. Pituitary LIF
gene expression was negligible after turpentine injection (Fig. 4, a and b).
|
, a and c).
Plasma ACTH and corticosterone responses to turpentine injection in
WT and LIFKO mice
Basal plasma ACTH and corticosterone levels were similar in WT and
LIFKO mice (28 ± 5 vs. 29 ± 8 pg/ml ACTH; and
166 ± 23.2 vs. 127.1 ± 55 ng/ml CS). In WT mice,
im turpentine injection resulted in brisk ACTH and corticosterone
responses 1 h after injection, and these elevated levels were
sustained for up to 24 h. Both ACTH and corticosterone responses
were biphasic, with initial peaks observed at 1 h (180 ± 45
vs. 28 ± 5 pg/ml ACTH; P < 0.001; and
600 ± 65 vs. 166 ± 23 ng/ml CS,
P < 0.05) followed by decreased levels of both
hormones observed 7 h after injection. The second peak occurred at
18 h (150 ± 35 pg/ml ACTH and 432 ± 59 ng/ml CS) and
was followed by subsequent declines in plasma ACTH and corticosterone
concentrations 24 h after turpentine treatment. In contrast,
turpentine injection did not elicit plasma ACTH responses in LIFKO
mice. Although plasma corticosterone was elevated above pretreatment
control levels at 1 (337 ± 43 vs. 127 ± 55
pg/ml, P < 0.05) and 7 h (249 ± 30
vs. 127 ± 55 pg/ml, NS) after injection, this
elevation was less profound than that observed in WT animals at the
same time periods. By 18 h after turpentine, corticosterone levels
had returned to basal pretreatment values in LIFKO animals (Fig. 5).
|
). Pituitary POMC mRNA
transcripts were induced (3.5 ± 0.43-fold, P <
0.05) in WT mice at 7 h and remained high (2.7 ± 0.5,
P < 0.05) 24 h after treatment, whereas
turpentine-injected LIFKO mice showed more modest increases in POMC
mRNA at 7 and 24 h (1.7 ± 0.45-fold, and 1.8 ±
0.3-fold, respectively, both NS) during the same time period (Fig. 6, b
and c).
|
. In both experiments modest
induction in POMC mRNA levels were noted in LIFKO mice 18 h after
treatment with turpentine only as compared with untreated control.
Markedly higher induction was observed in the pituitary POMC gene
expression in LIFKO mice simultaneously injected with LIF and
turpentine (5.6-fold, experiment I; 6-fold, experiment II). Due to the
limited ability of LIKO mice, LIF-only control injection were not
performed in this experiment. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our results describe the involvement of hypothalamic and pituitary LIF in the functioning of the HPA axis during an inflammatory perturbation. Thus, in WT mice injected with CFA, there is a substantial induction of hypothalamic LIF gene expression for up to 5 days after treatment. In the turpentine model, levels of hypothalamic LIF mRNA increased within 7–24 h of turpentine treatment. In both experimental models, LIF induction was more pronounced in the hypothalamus than in the pituitary. These observations are in concordance with our previous work showing, by competitive RT-PCR, induction of the diffusible form of murine LIF in the hypothalamus and pituitary after LPS injection (16). We previously demonstrated that LIF protein is expressed in human corticotroph cells (13). We also showed the stimulatory effect of LIF on POMC gene expression and ACTH secretion in vitro (12, 13, 34, 38) and in vivo (15). Despite the fact that LIF expression is higher in the hypothalamus, the pituitary is the major site of LIF action. We suggest that hypothalamic LIF may induce pituitary POMC gene expression and ACTH secretion in synergy with hypothalamic CRH as we have demonstrated in vitro (12, 14).
We have recently shown that SOCS 3, a cytokine-inducible signaling inhibitor, is expressed in the hypothalamus and pituitary and is potently stimulated in vivo by LIF (34). SOCS 3 functions as an intracellular regulator of POMC gene expression and ACTH secretion, acting as a negative feedback mediator of the cytokine-induced HPA axis activity (38, 39, 40). After CFA injection, we observed a moderate induction of hypothalamic SOCS 3, and a striking increase of pituitary SOCS 3. Similar observations were made in the turpentine model where WT mice demonstrated a slight elevation of hypothalamic SOCS 3 mRNA levels, and significant pituitary induction of this gene was evident for up to 24 h. These results suggest that LIF acts at the pituitary where it stimulates corticotroph function and ACTH secretion and induces pituitary SOCS 3 production as a feedback regulator of POMC gene expression.
Compelling evidence favoring the involvement of LIF in the HPA axis response to immune stress was obtained from the experiments using the LIF knockout mice. Overall plasma ACTH and corticosterone levels were much higher in response to both CFA and turpentine treatments in WT compared with LIFKO animals. In WT animals, plasma levels of ACTH were elevated 3 days after CFA injections and remained persistently high for up to 5 days. Plasma corticosterone levels more than doubled within a day after injection and remained high throughout the observation. In LIFKO mice, however ,CFA treatment did not produce significant changes either in plasma ACTH or in corticosterone concentrations. Local inflammation in WT mice caused biphasic ACTH and corticosterone responses. A similar biphasic pattern of hormone levels after turpentine injection was previously shown by others (26) in rats, where it was postulated that the initial increase of hormone levels occurred as a result of nonspecific activation of injection site afferents, whereas the second rise coincides with onset of the inflammatory process (41). In LIFKO mice, levels of both hormones were lower than in WT littermates at each tested point, corticosterone peaked 1 h after treatment, and no biphasic HPA axis activation was noted in these animals. These results are in agreement with our recent results showing that in the absence of LIF animals mount a sufficient corticosterone response to short acute psychological stress, although decreased hypothalamo-pituitary reserve diminishes the ability of LIFKO mice to endure chronic stressful situations (15).
In our experiments, we detected changes in CRH mRNA levels only in the turpentine model. Thus, in spite of higher ACTH axis activity, we observed a decreased hypothalamic CRH gene expression in WT animals starting 7 h after turpentine injection. In contrast, the inflammatory process had no impact on hypothalamic CRH mRNA levels in LIFKO mice. The CRH decline observed in WT animals could occur as a result of a negative feedback regulation of hypothalamic CRH levels by high concentrations of stress-induced circulating glucocorticoids (42, 43). Of the two inflammation models we used, injection of turpentine evoked a much higher elevation of plasma corticosterone than did CFA treatment (608 ± 65 vs. 268 ± 60 ng/ml peak levels) in WT animals. A marked elevation of the circulating inflammatory cytokine, IL 6, in mice after im turpentine injection was demonstrated (26). IL 6 induces ACTH axis responses at the level of both the pituitary and adrenals (29) and therefore can be an additional driving force for the ACTH axis stimulation in turpentine-treated animals.
Levels of POMC gene expression in WT pituitaries were significantly elevated within 1–5 days after CFA and from 7–24 h after turpentine injection, in concordance with the observed elevated plasma ACTH and corticosterone levels in these animals during the same time periods. In the absence of LIF, baseline pituitary POMC gene expression is low, and responses to both inflammatory challenges are less pronounced. These results are in agreement with our previous observation that a deficiency of pituitary LIF during development might lead to impaired POMC gene expression in LFKO mice (15, 44).
These findings indicate that the hypothalamic-pituitary-adrenal axis in LIFKO mice maintains a hypothalamic component but has insufficient subsequent pituitary reserve leading to attenuation of the HPA axis inflammatory stress response. The experiments with LIF substitution in LIFKO mice support this conclusion. Thus, administration of exogenous murine LIF together with turpentine resulted in a significant induction of pituitary POMC gene expression compared with turpentine-only injected animals. The POMC gene appears sensitive to effects of exogenous LIF and we observed long-lasting POMC gene induction in this experiment. It is more likely that this induction is not acute, as our preliminary experiments (not shown) demonstrated that in both WT and LIFKO mice POMC was induced only for up to 5 h after LIF treatment (5 µg). We suggest that the effect of LIF is presumably permissive as LIF injected before the onset of inflammatory process can sensitize pituitary corticotrophs (45) making them more vulnerable to the effects of other cytokines released in the course of local inflammation. Although LIF-only control injection is absent in this experiment and final conclusion cannot be made, these results suggest that LIF replacement to LIFKO animals potentiates a brisk POMC-response to turpentine that is similar to what observed in WT mice.
In this paper, we demonstrate the involvement of hypothalamic LIF in activating the ACTH axis. Although we could not detect significant changes in the pituitary LIF expression, the role of pituitary LIF should not be underestimated. Thus, earlier we showed potent induction of both hypothalamic and pituitary LIF in response to IL 1ß, one of the major circulating cytokines in sepsis (30). Pituitary LIF could be involved in the inflammatory stress reaction under threatening inflammatory conditions such as septic shock (11). Taken together, these results show that hypothalamic and pituitary LIF contribute to pituitary POMC gene expression and ACTH secretion under inflammatory stress and may play an important role in the neuroimmune modulation of HPA axis function in response to inflammation.
|
Footnotes |
|---|
Received February 8, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  L. Robb, K. Boyle, S. Rakar, L. Hartley, J. Lochland, A. W. Roberts, W. S. Alexander, and D. Metcalf Genetic reduction of embryonic leukemia-inhibitory factor production rescues placentation in SOCS3-null embryos but does not prevent inflammatory disease PNAS, November 8, 2005; 102(45): 16333 - 16338. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Kariagina, S. Zonis, M. Afkhami, D. Romanenko, and V. Chesnokova Leukemia inhibitory factor regulates glucocorticoid receptor expression in the hypothalamic-pituitary-adrenal axis Am J Physiol Endocrinol Metab, November 1, 2005; 289(5): E857 - E863. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. A. Abbud, R. Kelleher, and S. Melmed Cell-Specific Pituitary Gene Expression Profiles after Treatment with Leukemia Inhibitory Factor Reveal Novel Modulators for Proopiomelanocortin Expression Endocrinology, February 1, 2004; 145(2): 867 - 880. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Kariagina, D. Romanenko, S.-G. Ren, and V. Chesnokova Hypothalamic-Pituitary Cytokine Network Endocrinology, January 1, 2004; 145(1): 104 - 112. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Beishuizen and L. G. Thijs Review: Endotoxin and the hypothalamo-pituitary-adrenal (HPA) axis Innate Immunity, February 1, 2003; 9(1): 3 - 24. [Abstract] [PDF]
|
|
|
|
 |
|
 A. Ben-Shlomo, I. Miklovsky, S.-G. Ren, W. H. Yong, A. P. Heaney, M. D. Culler, and S. Melmed Leukemia Inhibitory Factor Regulates Prolactin Secretion in Prolactinoma and Lactotroph Cells J. Clin. Endocrinol. Metab., February 1, 2003; 88(2): 858 - 863. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Metcalf The Unsolved Enigmas of Leukemia Inhibitory Factor Stem Cells, January 1, 2003; 21(1): 5 - 14. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Bousquet, V. Chesnokova, A. Kariagina, A. Ferrand, and S. Melmed cAMP Neuropeptide Agonists Induce Pituitary Suppressor of Cytokine Signaling-3: Novel Negative Feedback Mechanism for Corticotroph Cytokine Action Mol. Endocrinol., November 1, 2001; 15(11): 1880 - 1890. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Chesnokova, A. Kariagina, and S. Melmed Opposing effects of pituitary leukemia inhibitory factor and SOCS-3 on the ACTH axis response to inflammation Am J Physiol Endocrinol Metab, May 1, 2002; 282(5): E1110 - E1118. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
