- Email: [email protected]
lpr Mice
Brain, Behavior, and Immunity 13, 348–360 (1999) Article ID brbi.1998.0535, available online at http://www.idealibrary.com on
Alterations in Hypothalamic–Pituitary–Adrenal Function Correlated with the Onset of Murine SLE in MRL ⫹/⫹ and lpr/lpr Mice N. Shanks, P. M. Moore,* P. Perks, and S. L. Lightman Department of Medicine, University of Bristol, BRI Laboratories, Marlborough Street, Bristol BS2 8HW, United Kingdom; and *Department of Neurology, Wayne State University School of Medicine, Detroit, Michigan Systemic lupus erythematosus (SLE) is a spontaneously occurring, chronic autoimmune disease that can manifest neuropsychiatric abnormalities. The pathways mediating these central changes are not known; however, neuroendocrine alterations associated with inflammation may play a role. Predisposition to and progression of autoimmune disease has been associated with altered hypothalamic–pituitary–adrenal (HPA) function and inflammation has been reported to alter hypothalamic regulation of HPA responses. We investigated whether disease progression in a murine model of systemic lupus erythematosus (MRL ⫹/⫹. MRL lpr/lpr) resulted in altered expression of HPA regulatory peptides at the level of the hypothalamus and how these alterations related to circulating levels of corticosterone, corticosterone binding globulin, and autoantibody titers. We report that as MRL ⫹/⫹ and MRL lpr/lpr mice age and circulating levels of autoantibodies increase, there is a decrease in hypothalamic CRH mRNA expression and finally an increase in AVP mRNA expression. We also report that associated with increased autoantibody levels, disease progression, and altered hypothalamic peptide expression there is an increase in circulating levels of corticosterone and a trend for levels of corticosterone binding globulin to decrease. Our data complement previous observations of altered peptidergic regulation of the HPA axis and increased HPA activity during chronic inflammation in exogenously induced rodent models of chronic inflammation and indicate that similar processes may occur in spontaneous murine models of SLE. 1999 Academic Press Key Words: hypothalamic–pituitary–adrenal axis; SLE; murine; CRH; AVP; corticosterone; CBG.
INTRODUCTION
Systemic lupus erythematosus (SLE) is a spontaneously occurring chronic autoimmune disease of humans and some animals, where both genetic and exogenous factors are thought to contribute to the manifestations of disease. The disease is characterized by the appearance of particular patterns of autoantibodies and although the clinical tempo varies widely, renal, cutaneous, hematological, and joint disease are typical, while neurobehavioral abnormalities are also a frequent but enigmatic feature. Of the murine models of SLE, the MRL/lpr and NZB/WF1 hybrids are particularly well studied. In these, as in the human disease, autoantibodies develop to a variety of nuclear, cytoplasmic, and cell surface components. Certain autoantibodies (antiDNA, anti-Sm) are pathogenic in both the renal and the cutaneous abnormalities, largely by immune complex deposition and recruitment of an inflammatory cascade (Bloom, Davingnon, Cohen, Eisenberg, & Clarke, 1993; Isenberg, 1997; Pisetsky, 1993; Vlahakos, Foster, Adams, Katz, Ucci, Barrett, Datta, & Madaio, 1992). Other autoantibodies exert their effects by binding to cognate antigens on the surfaces of platelets and lymphocytes resulting in thrombocytopenia and leukopenia. Autoantibodies may play a role in determining neuropsychiatric abnormalities, however, despite associations of several autoantibodies with learning disorders and emotional 348 0889-1591/99 $30.00
Copyright 1999 by Academic Press All rights of reproduction in any form reserved.
ENDOCRINE CORRELATES OF MURINE SLE
349
anxiety behaviours (Hutchinson, Nehall, & Simeon, 1996; Vogelweid, Wright, Johnson, Hewett, & Walker, 1992) the pathways and pathogenic mechanisms in the development of the neurobehavioral SLE are not yet defined. It is now established, however, that the brain is a prominent regulator of immune responses through both neuroendocrine and central activation. Indeed, predisposition to and severity of autoimmune disease is thought to be partially dependent on hypothalamic–pituitary–adrenal (HPA) activity and circulating levels of glucocorticoids. These studies report that elevated levels of circulating glucocorticoids and adrenocorticotropic hormone (ACTH) during chronic inflammation limit inflammatory processes (Aguilera, 1994; Harbuz & Lightman, 1992; Sternberg, Young, & Bernadini, 1989; Sternberg, Wilder, Gold, & Chrousos, 1990) and that rat strains with a hyporesponsive HPA axis are more susceptible to inflammation. Additionally, there is evidence that in animal models of exogenously induced inflammation (adjuvant arthritis (AA), experimental allergic encephalomyelitis (EAE)) that adrenalectomy exacerbates, while corticosterone administration alleviates, symptoms (Harbuz, Rees, & Lightman, 1993; MacPhee, Antoni, & Mason, 1989). Interactions between inflammatory and endocrine responses are not, however, limited to the periphery as there are alterations in HPA regulation at the level of the hypothalamus during chronic inflammatory responses (Harbuz, Leonard, Lightman, & Cuzner, 1993; Harbuz, Rees, Eckland, Jessop, Brewerton, & Lightman, 1992). In particular, it has been reported that as inflammation progresses in AA and during the active phase of EAE, circulating levels of corticosterone are elevated and CRH mRNA levels decline. A similar pattern of changes in HPA function has also been observed in B 10 mice infected with the parasite Leischmania, wherein decrements in paraventricular CRH mRNA expression were associated with the inflammatory response; on the other hand no such changes in CRH mRNA expression were observed in a congenic strain that was not susceptible to infection with the parasite (Harbuz, Jessop, Chowdrey, Blackwell, Larsen, & Lightman, 1995; Ma, Levy, & Lightman, 1997). Reductions in CRH mRNA during inflammation at first appear paradoxical in light of the concurrent elevation of plasma corticosterone and suggest that some other ACTH secretagogue(s) must be involved in the activation of the HPA axis. Indeed, it has been found that during chronic inflammation levels of AVP mRNA are elevated and it has been suggested that peptidergic control of the HPA axis is shifted to a predominantly AVP-dependent mechanism and that this shift may maintain HPA responsivity during elevated glucocorticoid negative feedback (Chowdrey, Larsen, Harbuz, Jessop, Aguilera, Eckland, & Lightman, 1995a). The potential role of the HPA axis in spontaneous autoimmune models of SLE is not well studied, although there is some evidence that HPA function may be altered. In the NZB/NZW model of murine lupus it has been reported that disease onset is associated with a dampened HPA stress responsivity and decreased adrenal sensitivity to ACTH (Chesnokova, Ivanova, Karyagina, Chukhlib, & Ivanonva, 1995). Others report, however, increased levels of circulating corticosterone in both NZB/NZW and MRL lpr/lpr mice as the disease progresses (Wilder, 1995) and that these mice are able to mount an HPA response to IL-1β challenge similar to that seen in nonautoimmune strains of mice (Lechner, Hu, Jafarian-Tehrani, Dietrich, Schwarz, Herold, Haour, & Wick, 1996). As would be expected, however, the HPA activating effects of IL-1β were less pronounced the more advanced the inflammation in the auto-
350
SHANKS ET AL.
immune strains. These data confirm that there may be alterations in neuroendocrine function associated with autoimmune disease in susceptible mouse strains. Increasing attention is now turning to the endocrine correlates of SLE and their interaction with immune factors within the CNS. Alterations in central CRH expression may have implications for observed neuropsychiatric symptoms sometimes associated with clinical SLE. Data from murine models indicate that behavioral alterations are associated with early stages of disease onset (Sakic, Szechtman, Talangbayan, Denburg, Cabotte, & Denburg, 1994; Schrott & Crnic, 1996) and that behavioral abnormalities can be prevented with reduction of antibody levels by immunosuppression (Sakic, Szechtman, Denburg, & Denburg, 1995); however, no direct or indirect pathogenic role for autoantibodies is established. The pathways and pathogenic mechanisms in the development of neurobehavioral disorders in murine SLE are not defined. We were interested in determining whether hypothalamic peptide expression was altered in a spontaneously developing autoimmune disease model and whether these changes were associated with disease onset or progression. We chose to investigate neuroendocrine correlates of disease progression in the MRL model as it has less CNS disease involvement (Sherman, Galaburda, Behan, & Rosen, 1987). METHODS
Animals MRL ⫹/⫹ and MRL lpr/lpr male mice were obtained from Harlan, UK, at 4 weeks of age and group housed upon arrival. The mice were maintained under standard housing conditions with food and water available ad lib and a light cycle 0500– 0700. Procedure A total of 50 animals of each of the strains were sacrificed at 6 (n ⫽ 8), 8 (n ⫽ 8), and 12 weeks (n ⫽ 8 ⫹/⫹, n ⫽ 10 lpr/lpr) of age. Urine samples were collected immediately prior to sacrifice and tested for the presence of protein using diagnostic test sticks (BM Test 5L, Boehringer Mannheim). Trunk blood was collected in heparinized tubes and brains were quickly dissected and frozen immediately on dry ice. Assays Antibody ELISAs. Circulating autoantibody titers were determined by ELISA against double-stranded DNA, ribosomal proteins, and lupus brain antigen 1 (19–21 Mer proteins, Chiron Peptides). ELISAs were performed against 50 ng of peptide/ well with serum at a 1 : 40 dilution. The secondary antibody was alkaline-phosphatase conjugated goat anti-mouse IgG and optical densities were determined at 405 nm. Values are expressed as optical densities following subtraction of mean blank values (i.e., Balb serum). Corticosterone. Plasma was separated by centrifugation, 3500 rpm for 15 min, then aliquoted and frozen at ⫺20°C until analysis for corticosterone by radioimmunoassay. Total plasma corticosterone concentrations were measured directly in plasma using a citrate buffer at pH 3.0 to denature the binding globulin (1 µl plasma fraction diluted in 100 µl buffer), antiserum kindly provided by Professor G. Makara (Institute of Experimental Medicine, Budapest, Hungary) and 125I-corticosterone (ICN Biomed-
ENDOCRINE CORRELATES OF MURINE SLE
351
icals, Irvine, CA) with a specific activity of 2 to 3 mCi/µg. All samples were analyzed in one assay, and intra-assay variation was 5%. Corticosterone binding globulin (CBG). CBG levels were determined by singlepoint assays. Plasma samples (20 µl) were stripped of endogenous corticosterone using LH-20 columns and eluted with 500 µl of 30 mM Tris, 1 mM EDTA, 10 mM sodium molybdate, 10% v/v glycerol, and 1 mM dithiothreitol (TEDGM). Sample aliquots (150 µl) were incubated with a saturating concentration (100 nM) of [3 H]corticosterone (Amersham) (total binding) or [3H]corticosterone and a 200-fold excess cold corticosterone (nonspecific binding). Following incubation, bound and unbound steroid were separated with LH-20 chromatography. For each sample triplicate determinants of total and nonspecific binding were made by washing 100 µl of incubates into columns with 100 µl TEDGM. The columns were eluted 30 min later with 500 µl of TEGM into 1.5-ml Eppendorf tubes. Scintillation cocktail (0.5 ml, Optiphase Supermix, Wallac) was added to each sample and subsequently counted in a 1450 Wallac Microbeta plus counter at a 40% efficiency. Protein content for the stripped aliquots was determined using the Bradford method (Bio-Rad) adapted to a 96-well plate and final values are expressed pmol [3H]corticosterone bound/mg protein. In situ hybridization. In situ hybridization was performed on 20-µm cryostat sections of the hypothalamic paraventricular nucleus (PVN) as described previously (8) using 48-mer oligonucleotides complimentary to exonic mRNA sequences coding for CRH (CAG TTT CCT GTT GCT GTG AGC TTG CTG AGC TAA CTG CTC TGC CCT GGC) and another probe complimentary to exonic mRNA sequences coding for the last 16 amino acids for AVP (GTA GAC CCG GGG CTT GGC AGA ATC CAC GGA CTC TTG TGT CCC AGC CAG) (Hara, Battey, & Gainer, 1990; Schmale, Heinsohn, & Richter, 1983). The probes were labeled using terminal deoxytransferase to add a 35S-labeled deoxy-ATP (1000 Ci/mmol) tailed to the 3′-end of the probe. The autoradiographic images together with 35S-labeled standards (to compensate for the nonlinear response of the film) were measured using an image analysis system. For better visualization of alterations in AVP mRNA expression, after exposure to film the slides were dipped in emulsion (Ilford Scientific Products K-5gel, Mobberly, Cheshire, UK). After 10–12 days the slides were developed (Kodak) and sections counterstained with cresyl fast violet. Statistics. All data were analyzed by a two-way analysis of variance (Strain ⫻ Age) and Scheffe post hoc comparisons were used to analyze significant interactions and age main effects. Some samples were lost during in situ hybridization assays resulting in unequal ns between groups, but never below an n ⫽ 5 per group. RESULTS
We observed that circulating levels of antibody against double-stranded DNA, F(2,43) ⫽ 15.94, p ⫽ .0001, and anti-ribp antibody titers, F(2,43) ⫽ 12.11, p ⫽ .0001, increased significantly with age and were higher in the MRL lpr/lpr mice (antiDNA F(1,43) ⫽ 22.41, p ⫽ .0001; anti-ribp F(2,43) ⫽ 19.35, p ⫽ .0001) (see Fig. 1). The Strain ⫻ Age interaction approached statistical significance for both these measures ( p ⫽ .06). A significant Strain ⫻ Age interaction was observed for the anti-lba1 titers, F(2,43) ⫽ 3.56, p ⫽ .037, with greater increment in titers over age in the MRL lpr/lpr mice. Detectable urine proteins were significantly increased by 8 and 12 weeks in the MRL lpr/lpr mice and less so in MRL ⫹/⫹ mice of this age (see Fig. 2), Strain ⫻ Age, F(2,39) ⫽ 12.56, p ⫽ .0001.
352
SHANKS ET AL.
FIG. 1. Mean (⫾ SEM) serum autoantibody titers against double-stranded DNA (anti-DNA), ribosomal p protein (anti-ribp), and lupus brain antigen 1 (anti-lba1) at a 1 : 40 dilution, measured by ELISA at 405 nm.
These data confirm that by 8 weeks of age there is evidence of autoimmune pathology in MRL lpr/lpr mice and, to a less extent and of later onset, in the MRL ⫹/⫹ mice. Resting levels of plasma corticosterone increased with age, F(2,44) ⫽ 4.31, p ⫽ .02, but there were no significant differences observed between the mouse strains, F(1,44) ⫽ 1.68, p ⫽ .20. Scheffe’s comparisons revealed that corticosterone levels were significantly elevated at 8 and 12 weeks relative to 6 weeks of age (see Fig. 2). No statistically significant differences were observed for CBG levels across treatment conditions (see Fig. 2); however, an overall decrease with age approached statistical significance, F(2,41) ⫽ 2.61, p ⫽ .09.
ENDOCRINE CORRELATES OF MURINE SLE
353
FIG. 2. Mean (⫾ SEM) plasma corticosterone (ng/ml), corticosterone binding globulin as measured by bound [3H]corticosterone (pmol) per mg plasma protein and urine protein levels (g/L).
Analysis of hypothalamic CRH mRNA in the PVN demonstrated that CRH mRNA was significantly decreased in both strains associated with increasing age and with the onset of autoimmune disease, F(2,32) ⫽ 4.53, p ⫽ .02 (see Fig. 3). Post hoc analysis revealed CRH mRNA levels were significantly reduced by 8 weels in the MRL lpr/lpr mice and by 12 weeks in the MRL ⫹/⫹ mice when compared with 6 weeks of age. Analysis of hypothalamic AVP mRNA demonstrated an increase in mRNA expression between ages 8 and 12 weeks in both strains, F(2,19) ⫽ 4.69, p ⫽ .02, (see Fig. 4). Analysis of circulating autoantibody levels demonstrated autoantibody titers and urine protein levels increased with age, confirming an active disease process in these mice. Group means demonstrate alterations in endocrine function with disease onset/
354
SHANKS ET AL.
FIG. 3. Representative autoradiograph images of PVN CRHmRNA expression in MRL ⫹/⫹ and MRL lpr/lpr mice at 6, 8, and 12 weeks of age as measured by in situ hybridization and mean (⫾ SEM) arbitrary optical density units of CRH mRNA expression.
progression; however, to analyze the associations between observed changes in basal HPA function and disease onset Pearson correlation coefficients were calculated for all the data. Autoantibody titers (anti-DNA, r (37) ⫽ ⫺.46, anti-ribp, r (37) ⫽ ⫺.50, anti-lba1, r (37) ⫽ ⫺.48, all significant at p ⬍ .01) appeared to be negatively associated with altered CRH mRNA expression. These correlations suggest that increments in autoantibody titers were associated with decreased CRH mRNA expression. Urine protein levels would be expected to increase after autoantibodies are detected; indeed this was the case, wherein autoantibodies are detected in the MRL lpr/lpr mice as early as 6 weeks and become as evident at lower levels in the MRL ⫹/⫹ mice at 8 weeks. Urine proteins are markedly increased in the MRL lpr/lpr mice by both 8 and 12 weeks and to a lesser extent in the MRL ⫹/⫹ mice. The correlation between
ENDOCRINE CORRELATES OF MURINE SLE
355
FIG. 4. Representative photomicrograph of PVN AVPmRNA expression in MRL ⫹/⫹ and MRL lpr/lpr mice at 8 and 12 weeks of age, as measured by in situ hybridization and mean (⫾ SEM) arbitrary optical density units of AVP mRNA expression.
CRH mRNA levels and urine protein levels was, therefore, not as marked as that seen with the antibodies, but reached statistical significance (r(37) ⫽ ⫺.40, p ⬍ .02). In contrast, correlations of autoantibody levels and urine proteins with plasma corticosterone levels did not reach statistical significance (p ⬎ .05). DISCUSSION
All three of the antibodies tested correlated with the decline in CRH expression; however, despite their widely differing specificity this is not surprising for several reasons: (1) The antibodies tended to occur in tandem so it was not possible to correlate the changes in hypothalamic CRH with individual antibodies. (2) Antibodymediated biologic changes need not be antigen specific as effector mechanisms can occur through the Fc portion of the molecule when the antibodies are in complexes or by nonspecific injury to the vasculature resulting in endothelial production of cytokines. Nonetheless, features of two of the antibodies are sufficiently relevant to
356
SHANKS ET AL.
neuropsychiatric SLE to warrant further study. The anti-ribosomal P antibodies have a clinical correlation with psychoses in patients with SLE, a feature that certainly suggests an endocrine component (Hutchinson et al., 1996). We recognize that a murine model may not be appropriate to explore these features; however, the antilba1 antibody is interesting because the target autoantigen is present only in the brain and there most notably in the cingulate gyrus and hippocampus—two regions with prominent communication with the hypothalamus (Moore, Vo, & Carlock, 1998). Our observations concerning altered peptide mRNA expression in the hypothalamus are similar to those previously reported in exogenously activated models of rodent autoimmune disease which develop reduced PVN CRH mRNA and increased AVP mRNA expression. These changes in hypothalamic gene expression occur at a very early stage in the development of murine SLE and do not appear to be a result of profound alterations in circulating steroid, kidney pathology, or cerebrovasculature immune complex deposition which occur later on. Evidence from studies in rats with AA suggest that the alterations in PVN peptidergic profiles are not a result of increased glucocorticoid feedback, but rather reflect a chronic activation of the HPA axis regulation during chronic inflammation. The increased contribution of AVP to HPA regulation may enable continued responsivity of the axis when faced with elevated levels of circulating glucocorticoids and may explain the diminution of HPA responsivity to restraint stress, but not to LPS or adrenalectomy seen in rats with AA (Harbuz et al., 1993). It would appear, therefore, that HPA responses ‘‘selectively adapt’’ to the chronic stress regimen, but maintain responsivity to salient stimuli or even become sensitized to immune–HPA pathways with preferential HPA responses to regulate disease progression. This explanation may also apply to animals with murine SLE which respond to acute challenge with IL-1β, but whose responsivity diminishes as the disease progresses (Lechner et al., 1996). Further investigation is required to determine whether these animals maintain HPA responses to other stressors. It could be argued that the alterations observed in these mouse strains reflect HPA development rather than an ongoing disease process. While less is known about HPA development in mice, it has been observed in rats that at 6 weeks of age corticosterone, CBG, and hypothalamic CRH and AVP peptide and mRNA levels are expressed at adult levels (see review, Meaney, O’Donnell, Viau, Bhatnagar, Sarrieau, Smythe, Shanks, & Walker, 1993). Furthermore, while increasing mRNA expression for central peptides with development would be expected, rarely would one expect to see a decrement in hypothalamic CRH mRNA expression associated with development. Consequently, we suggest that the observed alterations in mRNA expression for this peptide are in all likelihood a consequence of the ongoing inflammatory processes. However, the mechanisms that mediate the decreased hypothalamic CRH and increased AVP mRNA levels are unknown, and it has been reported that catecholamine levels within the PVN are altered in AA (Harbuz, Chover-Gonzalez, Biswas, & Lightman, 1994) and alterations in neurotransmitter input to the hypothalamus undoubtedly contribute to altered CRH mRNA expression. 6-Hydroxydopamine lesions of noradrenergic PVN neurons (75% reduction), however, did not significantly alter either the CRH mRNA or corticosterone responses to AA, although they did result in a more severe inflammatory response (Harbuz et al., 1994). On the other hand, serotonergic depletion at inflammation onset significantly reduced hind paw inflammation again with little effect on CRH mRNA (Harbuz, Perveen-Gill, Lalies, Jes-
ENDOCRINE CORRELATES OF MURINE SLE
357
sop, & Lightman, 1996). These data suggest roles for serotonin and catecholamines in the AA inflammatory response. Neuropeptides may also play a role in the regulation of CRH and AVP mRNA and substance P has been implicated in hypothalamic CRH changes associated with inflammation (Chowdrey, Larsen, Harbuz, Lightman, & Jessop, 1995b). Changes in parvocellular hypothalamic peptide mRNA expression associated with chronic inflammation may also relate to the differential activation of cells containing CRH alone or cells with colocalized CRH and AVP. Both CRH ⫹ and CRH/AVP ⫹ cell populations exhibit peptide release following IL-1 administration and it has been suggested that this may be why inflammatory stimuli continue to activate the HPA axis even when superimposed onto a chronically stimulated background and could explain why the response to inflammatory stimuli differs from other stressors (Whitnall, 1994). More recently it has been demonstrated that one peripheral injection of IL-1 promotes the colocalization of AVP within hypothalamic CRH neurons (Schmidt, Janszen, Wouterlood, & Tilders, 1995). It was observed that for up to 2 weeks post-IL-1β treatment animals remain ‘‘hyperresponsive’’ to subsequent stress exposure. It is thought that this may be due to a synergy of the peptide effects on HPA activity following increased colocalization. Furthermore, it was demonstrated that a single injection with IL-1β prior to induction of EAE reduced the vulnerability of Lewis rats to EAE and this was correlated with increased colocalization of AVP and CRH (Schmidt et al., 1995). Consequently, it appears that increasing colocalization of AVP and CRH may better enable the HPA axis to deal with inflammatory inputs to the CNS and subsequently improve immune regulation and dampening of inflammatory responses. The alterations in parvocellular AVP/CRH colocalization in autoimmune models may represent a hypothalamic adaptation to moderate disease progression or alternatively, decreased colocalization may dictate increased susceptibility to inflammation (Tilders, Schmidt, Aguilera, Kiss, Van Dam, Huitinga, & Dijkstra, 1996; Whitnall, Anderson, Lane, Mougey, Neta, & Perlstein, 1994). The present data do differ to some extent from endocrine profiles seen in rat models of inflammation, in that MRL mice did not exhibit such pronounced elevations in plasma corticosterone. It may be the case that during the gradual development of inflammation in spontaneously occurring pathologies such as this murine model of SLE, HPA responses are not as suddenly or severely stimulated as in exogenously induced animal models of inflammation. Indeed, in the present study circulating levels of corticosterone tended to gradually increase in association with increments in disease indicators and were also associated with decreased levels of circulating CBG. Individually these effects were relatively small, but together these responses may contribute to a physiologically relevant alteration in free steroid signal especially when considering the morning levels of circulating corticosterone. However, what these data do indicate is that alterations in resting levels of corticosterone (1) are not associated with the observed alterations in hypothalamic CRH mRNA expression and (2) are not elevated during the early stages of disease onset in murine SLE. We have also observed in the present investigation that the alterations seen in PVN CRH of MRL mice corresponded temporally with behavioral alterations previously observed in these mouse strains (Sakic et al., 1994; 1995). While it was not investigated directly, it is tempting to speculate that altered expression of ‘‘anxiety behaviors’’ or ‘‘emotionality’’ may be associated with changes in central CRH regulation in these animals. Administration of immunosuppressants can prevent behavioral alter-
358
SHANKS ET AL.
ations seen in the MRL mice prior to disease onset and has also been shown to disrupt the immune sequelae and HPA activation in adjuvant-induced autoimmune rat disease models (Stephanou, Sarlis, Knight, Lightman, & Chowdrey, 1992). Immune factors associated with disease are likely mediators of altered central neuroendocrine function in MRL mice, but whether or not alterations in hypothalamic CRH contribute to behavioral disturbances associated with disease onset in this murine model remains to be determined. Alterations in central endocrine and limbic functions resulting from signals of immune origin could contribute to neuropsychiatric manifestation of SLE. Furthermore, chronic inflammatory conditions that alter endocrine and CNS function may be etiological factors contributing to psychiatric and stress-related disease. ACKNOWLEDGMENTS The authors thank The Wellcome Trust (Grant Ref. 045362/Z/95/MP/JF) for their support of this project.
REFERENCES Aguilera, G. (1994). Regulation of pituitary ACTH-secretion during chronic stress. Frontiers Neuroendocrinol. 4, 321–350. Bloom, D. D., Davingnon, J. L., Cohen, P. L., Eisenberg, R. A., & Clarke, S. H. (1993). Overlap of the anti-Sm and anti-DNA responses of MRL/Mp-lpr/lpr mice. J. Immunol. 150, 1579–1590. Chesnokova, V. M., Ivanova, A. N., Karyagina, A. Y., Chukhlib, V. L., & Ivanova, L. N. (1995). Patterns of interaction between the immune and endocrine systems in the manifestation of hereditary autoimmune disorders in NZB mice. Bull. Exp. Biol. Med. 119, 381–384. Chowdrey, H. S., Larsen, P. J., Harbuz, M. S., Jessop, D. S., Aguilera, G., Eckland, D. J. A., & Lightman, S. L. (1995a). Evidence for arginine-vasopressin as the primary activator of the HPA axis during adjuvant-induced arthritis. Br. J. Pharmacol. 116, 2417–2424. Chowdrey, H. S., Larsen, P. J., Harbuz, M. S., Lightman, S. L., & Jessop, D. S. (1995b). Endogenous substance-P inhibits the depression of corticotropin-releasing-hormone during a chronic inflammatory stress. Life Sci. 57, 2021–2029. Hara, Y., Battey, J. F., & Gainer, H. (1990). Structure of mouse vasopression and oxytocin genes. Mol. Brain Res. 8, 319–324. Harbuz, M. S., Chover-Gonzalez, A. J., Biswas, S., & Lightman, S. L. (1994). Role of central catecholamines in the modulation of corticotropin-releasing factor messenger-RNA during adjuvant-induced arthritis in the rat. Br. J. Rheumatol. 33, 205–209. Harbuz, M. S., Jessop, D. S., Chowdrey, H. S., Blackwell, J. M., Larsen, P. J., & Lightman, S. L. (1995). Evidence for altered hypothalamic CRF in immune-mediated diseases. Ann. N.Y. Acad. Sci. 771, 449–458. Harbuz, M. S., Leonard, J. P., Lightman, S. L., & Cuzner, M. L. (1993). Changes in hypothalamic corticotrophin releasing factor (CRF) and pituitary proopiomelanocortin (POMC) messenger RNA during the course of experimental allergic encephalomyelitis. J. Neuroimmunol. 45, 127–132. Harbuz, M. S., & Lightman, S. L. (1992). Stress and the hypothalamo–pituitary–adrenal axis: Acute, chronic and immunological activation. J. Endocrinol. 134, 327–339. Harbuz, M. S., Perveen-Gill, Z., Lalies, M. D., Jessop, D. S., & Lightman, S. L. (1996). The role of endogenous serotonin in adjuvant-induced arthritis in the rat. Br. J. Rheumatol. 35, 112–116. Harbuz, M. S., Rees, R. G., Eckland, D., Jessop, D. S., Brewerton, D., & Lightman, S. L. (1992). Paradoxical responses of hypothalamic CRF mRNA and CRF-41 peptide and adenohypohyseal POMC mRNA during chronic inflammatory stress. Endocrinology 130, 1394–1400. Harbuz, M. S., Rees, R. G., & Lightman, S. L. (1993). HPA axis responses to acute stress and adrenalectomy during adjuvant-induced arthritis in the rat. Am. J. Physiol. 264, R179–R185. Hutchinson, G. A., Nehall, J. E., & Simeon, D. T. 1996. Psychiatric-disorders in systemic Lupus-erythematosus. West Indian Med. J. 45, 48–50.
ENDOCRINE CORRELATES OF MURINE SLE
359
Isenberg, D. A. (1997). Autoantibodies: Markers of disease or pathogenic? Ann. N.Y. Acad. Sci. 823, 256–262. Lechner, O., Hu, Y., Jafarian-Tehrani, M., Dietrich, H., Schwarz, S., Herold, M., Haour, F., & Wick, G. (1996). Disturbed immunoendocrine communication via the hypothalamo–pituitary–adrenal axis in murine lupus. Brain Behav. Immun. 10, 337–350. Ma, X-M., Levy, A., & Lightman, S. L. (1997). Rapid changes in heteronuclear RNA for corticotrophinreleasing hormone and arginine vasopressin in response to acute stress. J. Endocrinol. 152, 81– 89. MacPhee, I. A., Antoni, F. A., & Mason, D. W. (1989). Spontaneous recovery of rats from experimental allergic encephalomyelitis is dependent on regulation of the immune system by endogenous adrenal corticosteroids. J. Exp. Med. 169, 431–435. Meaney, M. J., O’Donnell, D., Viau, V., Bhatnagar, S., Sarrieau, A., Smythe, J., Shanks, N., & Walker, C-D. (1993). Corticosteroids receptors in the rat brain and pituitary during development and hypothalamic-pituitary-adrenal function. In Zagon, I. S., & McLaughlin, P. J. (Eds.), Receptors in the Developing Nervous System, Vol 1, pp. 163–201. Chapman & Hall, London. Moore, P. M., Vo, T., & Carlock, L. R. (1998). Identification and cloning of a brain autoantigen in neurobehavioral SLE. J. Neuroimmunol. 82, 116–125. Pisetsky, D. S. (1993). Autoantibodies and their significance. Curr. Opin. Rheumatol. 5, 549–556. Sakic, B., Szechtman, H., Talangbayan, H., Denburg, S. D., Cabotte, R. M., & Denburg, J. A. (1994). Disturbed emotionality in autoimmune MRL-lpr mice. Physiol. Behav. 56, 609–617. Sakic, B., Szechtman, H., Denburg, S. D., & Denburg, J. A. (1995). Immunosuppressive treatment prevents behavioral deficit in autoimmune MRL-lpr mice. Physiol. Behav. 58, 797–802. Schmale, H., Heinsohn, S., & Richter, D. (1983). Structural organization of the rat gene for the arginine vasopression-neurophysin precursor. EMBO J. 2, 763–767. Schmidt, F. D., Janszen, A. W. J. W., Wouterlood, F. G., & Tilders, F. J. H. (1995). IL-1 induced long-lasting changes in hypothalamic corticotropin releasing hormone (CRH) neuron and hyperresponsiveness of the HPA axis. J. Neurosci. 15, 7417–7426. Schrott, L. M., & Crnic, L. S. (1996). Anxiety behavior, exploratory behavior, and activity in NZBxNZWF1 hybrid mice: Role of genotype and autoimmune disease progression. Brain Behav. Immun. 10, 260–274. Sherman, G. F., Galaburda, A. M., Behan, P. O., & Rosen, G. D. (1987). Neuroanatomical anomalies in autoimmune mice. Acta Neuropathol. 74, 239–242. Stephanou, A., Sarlis, N. J., Knight, R. A., Lightman, S. L., & Chowdrey, H. S. (1992). Effects of cyclosporine-A on the hypothalamic–pituitary–adrenal axis and anterior-pituitary IL-6 messengerRNA expression during chronic inflammatory stress in the rat. J. Neuroimmunol. 41, 215–222. Sternberg, E. M., Wilder, R. L., Gold, P. W., & Chrousos, G. P. (1990). A defect in the central component of the immune system–hypothalamic–pituitary–adrenal axis feedback loop is associated with susceptibility to experimental arthritis and other inflammatory diseases. Ann. N.Y. Acad. Sci. 594, 289– 291. Sternberg, E. M., Young, W. S., III, & Bernardini, R. (1989). A central nervous defect in the stress response is associated with susceptibility to streptococcal cell wall arthritis in Lewis rats. Proc. Natl. Acad. Sci. USA 86, 2374–2378. Theofilopoulos, A. N., & Dixon, F. J. (1985). Murine models of systemic lupus erythematosus. Adv. Immunol. 37, 269–390. Tilders, F. J. H., Schmidt, E. D., Aguilera, G., Kiss, A., Van Dam, A. M., Huitinga, I., & Dijkstra, C. D. (1996). Sustained effects of cytokines, neuroplasticity, stress and immunomodulation. 2nd Annual Meeting of the Brain-Immune Network Group, London. Vlahakos, D. V., Foster, M. H., Adams, S., Katz, M., Ucci, A. A., Barrett, K. J., Datta, S. K., Madaio, M. P. (1992). Anti-DNA antibodies form immune deposits at distinct glomerular and vascular sites. Kidney Int. 41, 1690–1700. Vogelweid, C. M., Wright, D. C., Johnson, J. C., Hewett, J. E., & Walker, S. E. (1992). Evaluation of memory, learning ability, and clinical neurologic function in pathogen-free mice with systemic lupus erythematosus. Arthritis Rheum. 37, 889–897.
360
SHANKS ET AL.
Watanabe-Fukunaga, R., Brannan, C. I., Copelan, N. G., Jenkins, N. A., & Nagata, S. (1992). Lymphoproliferation disorder in mice explained by defects in Fas antigen that mediates apoptosis. Nature 356, 314–317. Wilder, R. L. (1995). Neuroendocrine-immune system interactions and autoimmunity. Annu. Rev. Immunol. 13, 307–338. Whitnall, M. H. (1994). Regulation of the hypothalamic corticotropin-releasing hormone neurosecretorysystem. Prog. Neurobiol. 40, 573–629. Whitnall, M. H., Anderson, K. A., Lane, C. A., Mougey, E. H., Neta, R., & Perlstein, R. S. (1994). Decreased vasopression content in parvocellular CRH neurosecretory-system of Lewis rats. Neuroreport 5, 1635–1637. Received June 16, 1998
Alterations in Hypothalamic–Pituitary–Adrenal Function Correlated with the Onset of Murine SLE in MRL ⫹/⫹ and lpr/lpr Mice N. Shanks, P. M. Moore,* P. Perks, and S. L. Lightman Department of Medicine, University of Bristol, BRI Laboratories, Marlborough Street, Bristol BS2 8HW, United Kingdom; and *Department of Neurology, Wayne State University School of Medicine, Detroit, Michigan Systemic lupus erythematosus (SLE) is a spontaneously occurring, chronic autoimmune disease that can manifest neuropsychiatric abnormalities. The pathways mediating these central changes are not known; however, neuroendocrine alterations associated with inflammation may play a role. Predisposition to and progression of autoimmune disease has been associated with altered hypothalamic–pituitary–adrenal (HPA) function and inflammation has been reported to alter hypothalamic regulation of HPA responses. We investigated whether disease progression in a murine model of systemic lupus erythematosus (MRL ⫹/⫹. MRL lpr/lpr) resulted in altered expression of HPA regulatory peptides at the level of the hypothalamus and how these alterations related to circulating levels of corticosterone, corticosterone binding globulin, and autoantibody titers. We report that as MRL ⫹/⫹ and MRL lpr/lpr mice age and circulating levels of autoantibodies increase, there is a decrease in hypothalamic CRH mRNA expression and finally an increase in AVP mRNA expression. We also report that associated with increased autoantibody levels, disease progression, and altered hypothalamic peptide expression there is an increase in circulating levels of corticosterone and a trend for levels of corticosterone binding globulin to decrease. Our data complement previous observations of altered peptidergic regulation of the HPA axis and increased HPA activity during chronic inflammation in exogenously induced rodent models of chronic inflammation and indicate that similar processes may occur in spontaneous murine models of SLE. 1999 Academic Press Key Words: hypothalamic–pituitary–adrenal axis; SLE; murine; CRH; AVP; corticosterone; CBG.
INTRODUCTION
Systemic lupus erythematosus (SLE) is a spontaneously occurring chronic autoimmune disease of humans and some animals, where both genetic and exogenous factors are thought to contribute to the manifestations of disease. The disease is characterized by the appearance of particular patterns of autoantibodies and although the clinical tempo varies widely, renal, cutaneous, hematological, and joint disease are typical, while neurobehavioral abnormalities are also a frequent but enigmatic feature. Of the murine models of SLE, the MRL/lpr and NZB/WF1 hybrids are particularly well studied. In these, as in the human disease, autoantibodies develop to a variety of nuclear, cytoplasmic, and cell surface components. Certain autoantibodies (antiDNA, anti-Sm) are pathogenic in both the renal and the cutaneous abnormalities, largely by immune complex deposition and recruitment of an inflammatory cascade (Bloom, Davingnon, Cohen, Eisenberg, & Clarke, 1993; Isenberg, 1997; Pisetsky, 1993; Vlahakos, Foster, Adams, Katz, Ucci, Barrett, Datta, & Madaio, 1992). Other autoantibodies exert their effects by binding to cognate antigens on the surfaces of platelets and lymphocytes resulting in thrombocytopenia and leukopenia. Autoantibodies may play a role in determining neuropsychiatric abnormalities, however, despite associations of several autoantibodies with learning disorders and emotional 348 0889-1591/99 $30.00
Copyright 1999 by Academic Press All rights of reproduction in any form reserved.
ENDOCRINE CORRELATES OF MURINE SLE
349
anxiety behaviours (Hutchinson, Nehall, & Simeon, 1996; Vogelweid, Wright, Johnson, Hewett, & Walker, 1992) the pathways and pathogenic mechanisms in the development of the neurobehavioral SLE are not yet defined. It is now established, however, that the brain is a prominent regulator of immune responses through both neuroendocrine and central activation. Indeed, predisposition to and severity of autoimmune disease is thought to be partially dependent on hypothalamic–pituitary–adrenal (HPA) activity and circulating levels of glucocorticoids. These studies report that elevated levels of circulating glucocorticoids and adrenocorticotropic hormone (ACTH) during chronic inflammation limit inflammatory processes (Aguilera, 1994; Harbuz & Lightman, 1992; Sternberg, Young, & Bernadini, 1989; Sternberg, Wilder, Gold, & Chrousos, 1990) and that rat strains with a hyporesponsive HPA axis are more susceptible to inflammation. Additionally, there is evidence that in animal models of exogenously induced inflammation (adjuvant arthritis (AA), experimental allergic encephalomyelitis (EAE)) that adrenalectomy exacerbates, while corticosterone administration alleviates, symptoms (Harbuz, Rees, & Lightman, 1993; MacPhee, Antoni, & Mason, 1989). Interactions between inflammatory and endocrine responses are not, however, limited to the periphery as there are alterations in HPA regulation at the level of the hypothalamus during chronic inflammatory responses (Harbuz, Leonard, Lightman, & Cuzner, 1993; Harbuz, Rees, Eckland, Jessop, Brewerton, & Lightman, 1992). In particular, it has been reported that as inflammation progresses in AA and during the active phase of EAE, circulating levels of corticosterone are elevated and CRH mRNA levels decline. A similar pattern of changes in HPA function has also been observed in B 10 mice infected with the parasite Leischmania, wherein decrements in paraventricular CRH mRNA expression were associated with the inflammatory response; on the other hand no such changes in CRH mRNA expression were observed in a congenic strain that was not susceptible to infection with the parasite (Harbuz, Jessop, Chowdrey, Blackwell, Larsen, & Lightman, 1995; Ma, Levy, & Lightman, 1997). Reductions in CRH mRNA during inflammation at first appear paradoxical in light of the concurrent elevation of plasma corticosterone and suggest that some other ACTH secretagogue(s) must be involved in the activation of the HPA axis. Indeed, it has been found that during chronic inflammation levels of AVP mRNA are elevated and it has been suggested that peptidergic control of the HPA axis is shifted to a predominantly AVP-dependent mechanism and that this shift may maintain HPA responsivity during elevated glucocorticoid negative feedback (Chowdrey, Larsen, Harbuz, Jessop, Aguilera, Eckland, & Lightman, 1995a). The potential role of the HPA axis in spontaneous autoimmune models of SLE is not well studied, although there is some evidence that HPA function may be altered. In the NZB/NZW model of murine lupus it has been reported that disease onset is associated with a dampened HPA stress responsivity and decreased adrenal sensitivity to ACTH (Chesnokova, Ivanova, Karyagina, Chukhlib, & Ivanonva, 1995). Others report, however, increased levels of circulating corticosterone in both NZB/NZW and MRL lpr/lpr mice as the disease progresses (Wilder, 1995) and that these mice are able to mount an HPA response to IL-1β challenge similar to that seen in nonautoimmune strains of mice (Lechner, Hu, Jafarian-Tehrani, Dietrich, Schwarz, Herold, Haour, & Wick, 1996). As would be expected, however, the HPA activating effects of IL-1β were less pronounced the more advanced the inflammation in the auto-
350
SHANKS ET AL.
immune strains. These data confirm that there may be alterations in neuroendocrine function associated with autoimmune disease in susceptible mouse strains. Increasing attention is now turning to the endocrine correlates of SLE and their interaction with immune factors within the CNS. Alterations in central CRH expression may have implications for observed neuropsychiatric symptoms sometimes associated with clinical SLE. Data from murine models indicate that behavioral alterations are associated with early stages of disease onset (Sakic, Szechtman, Talangbayan, Denburg, Cabotte, & Denburg, 1994; Schrott & Crnic, 1996) and that behavioral abnormalities can be prevented with reduction of antibody levels by immunosuppression (Sakic, Szechtman, Denburg, & Denburg, 1995); however, no direct or indirect pathogenic role for autoantibodies is established. The pathways and pathogenic mechanisms in the development of neurobehavioral disorders in murine SLE are not defined. We were interested in determining whether hypothalamic peptide expression was altered in a spontaneously developing autoimmune disease model and whether these changes were associated with disease onset or progression. We chose to investigate neuroendocrine correlates of disease progression in the MRL model as it has less CNS disease involvement (Sherman, Galaburda, Behan, & Rosen, 1987). METHODS
Animals MRL ⫹/⫹ and MRL lpr/lpr male mice were obtained from Harlan, UK, at 4 weeks of age and group housed upon arrival. The mice were maintained under standard housing conditions with food and water available ad lib and a light cycle 0500– 0700. Procedure A total of 50 animals of each of the strains were sacrificed at 6 (n ⫽ 8), 8 (n ⫽ 8), and 12 weeks (n ⫽ 8 ⫹/⫹, n ⫽ 10 lpr/lpr) of age. Urine samples were collected immediately prior to sacrifice and tested for the presence of protein using diagnostic test sticks (BM Test 5L, Boehringer Mannheim). Trunk blood was collected in heparinized tubes and brains were quickly dissected and frozen immediately on dry ice. Assays Antibody ELISAs. Circulating autoantibody titers were determined by ELISA against double-stranded DNA, ribosomal proteins, and lupus brain antigen 1 (19–21 Mer proteins, Chiron Peptides). ELISAs were performed against 50 ng of peptide/ well with serum at a 1 : 40 dilution. The secondary antibody was alkaline-phosphatase conjugated goat anti-mouse IgG and optical densities were determined at 405 nm. Values are expressed as optical densities following subtraction of mean blank values (i.e., Balb serum). Corticosterone. Plasma was separated by centrifugation, 3500 rpm for 15 min, then aliquoted and frozen at ⫺20°C until analysis for corticosterone by radioimmunoassay. Total plasma corticosterone concentrations were measured directly in plasma using a citrate buffer at pH 3.0 to denature the binding globulin (1 µl plasma fraction diluted in 100 µl buffer), antiserum kindly provided by Professor G. Makara (Institute of Experimental Medicine, Budapest, Hungary) and 125I-corticosterone (ICN Biomed-
ENDOCRINE CORRELATES OF MURINE SLE
351
icals, Irvine, CA) with a specific activity of 2 to 3 mCi/µg. All samples were analyzed in one assay, and intra-assay variation was 5%. Corticosterone binding globulin (CBG). CBG levels were determined by singlepoint assays. Plasma samples (20 µl) were stripped of endogenous corticosterone using LH-20 columns and eluted with 500 µl of 30 mM Tris, 1 mM EDTA, 10 mM sodium molybdate, 10% v/v glycerol, and 1 mM dithiothreitol (TEDGM). Sample aliquots (150 µl) were incubated with a saturating concentration (100 nM) of [3 H]corticosterone (Amersham) (total binding) or [3H]corticosterone and a 200-fold excess cold corticosterone (nonspecific binding). Following incubation, bound and unbound steroid were separated with LH-20 chromatography. For each sample triplicate determinants of total and nonspecific binding were made by washing 100 µl of incubates into columns with 100 µl TEDGM. The columns were eluted 30 min later with 500 µl of TEGM into 1.5-ml Eppendorf tubes. Scintillation cocktail (0.5 ml, Optiphase Supermix, Wallac) was added to each sample and subsequently counted in a 1450 Wallac Microbeta plus counter at a 40% efficiency. Protein content for the stripped aliquots was determined using the Bradford method (Bio-Rad) adapted to a 96-well plate and final values are expressed pmol [3H]corticosterone bound/mg protein. In situ hybridization. In situ hybridization was performed on 20-µm cryostat sections of the hypothalamic paraventricular nucleus (PVN) as described previously (8) using 48-mer oligonucleotides complimentary to exonic mRNA sequences coding for CRH (CAG TTT CCT GTT GCT GTG AGC TTG CTG AGC TAA CTG CTC TGC CCT GGC) and another probe complimentary to exonic mRNA sequences coding for the last 16 amino acids for AVP (GTA GAC CCG GGG CTT GGC AGA ATC CAC GGA CTC TTG TGT CCC AGC CAG) (Hara, Battey, & Gainer, 1990; Schmale, Heinsohn, & Richter, 1983). The probes were labeled using terminal deoxytransferase to add a 35S-labeled deoxy-ATP (1000 Ci/mmol) tailed to the 3′-end of the probe. The autoradiographic images together with 35S-labeled standards (to compensate for the nonlinear response of the film) were measured using an image analysis system. For better visualization of alterations in AVP mRNA expression, after exposure to film the slides were dipped in emulsion (Ilford Scientific Products K-5gel, Mobberly, Cheshire, UK). After 10–12 days the slides were developed (Kodak) and sections counterstained with cresyl fast violet. Statistics. All data were analyzed by a two-way analysis of variance (Strain ⫻ Age) and Scheffe post hoc comparisons were used to analyze significant interactions and age main effects. Some samples were lost during in situ hybridization assays resulting in unequal ns between groups, but never below an n ⫽ 5 per group. RESULTS
We observed that circulating levels of antibody against double-stranded DNA, F(2,43) ⫽ 15.94, p ⫽ .0001, and anti-ribp antibody titers, F(2,43) ⫽ 12.11, p ⫽ .0001, increased significantly with age and were higher in the MRL lpr/lpr mice (antiDNA F(1,43) ⫽ 22.41, p ⫽ .0001; anti-ribp F(2,43) ⫽ 19.35, p ⫽ .0001) (see Fig. 1). The Strain ⫻ Age interaction approached statistical significance for both these measures ( p ⫽ .06). A significant Strain ⫻ Age interaction was observed for the anti-lba1 titers, F(2,43) ⫽ 3.56, p ⫽ .037, with greater increment in titers over age in the MRL lpr/lpr mice. Detectable urine proteins were significantly increased by 8 and 12 weeks in the MRL lpr/lpr mice and less so in MRL ⫹/⫹ mice of this age (see Fig. 2), Strain ⫻ Age, F(2,39) ⫽ 12.56, p ⫽ .0001.
352
SHANKS ET AL.
FIG. 1. Mean (⫾ SEM) serum autoantibody titers against double-stranded DNA (anti-DNA), ribosomal p protein (anti-ribp), and lupus brain antigen 1 (anti-lba1) at a 1 : 40 dilution, measured by ELISA at 405 nm.
These data confirm that by 8 weeks of age there is evidence of autoimmune pathology in MRL lpr/lpr mice and, to a less extent and of later onset, in the MRL ⫹/⫹ mice. Resting levels of plasma corticosterone increased with age, F(2,44) ⫽ 4.31, p ⫽ .02, but there were no significant differences observed between the mouse strains, F(1,44) ⫽ 1.68, p ⫽ .20. Scheffe’s comparisons revealed that corticosterone levels were significantly elevated at 8 and 12 weeks relative to 6 weeks of age (see Fig. 2). No statistically significant differences were observed for CBG levels across treatment conditions (see Fig. 2); however, an overall decrease with age approached statistical significance, F(2,41) ⫽ 2.61, p ⫽ .09.
ENDOCRINE CORRELATES OF MURINE SLE
353
FIG. 2. Mean (⫾ SEM) plasma corticosterone (ng/ml), corticosterone binding globulin as measured by bound [3H]corticosterone (pmol) per mg plasma protein and urine protein levels (g/L).
Analysis of hypothalamic CRH mRNA in the PVN demonstrated that CRH mRNA was significantly decreased in both strains associated with increasing age and with the onset of autoimmune disease, F(2,32) ⫽ 4.53, p ⫽ .02 (see Fig. 3). Post hoc analysis revealed CRH mRNA levels were significantly reduced by 8 weels in the MRL lpr/lpr mice and by 12 weeks in the MRL ⫹/⫹ mice when compared with 6 weeks of age. Analysis of hypothalamic AVP mRNA demonstrated an increase in mRNA expression between ages 8 and 12 weeks in both strains, F(2,19) ⫽ 4.69, p ⫽ .02, (see Fig. 4). Analysis of circulating autoantibody levels demonstrated autoantibody titers and urine protein levels increased with age, confirming an active disease process in these mice. Group means demonstrate alterations in endocrine function with disease onset/
354
SHANKS ET AL.
FIG. 3. Representative autoradiograph images of PVN CRHmRNA expression in MRL ⫹/⫹ and MRL lpr/lpr mice at 6, 8, and 12 weeks of age as measured by in situ hybridization and mean (⫾ SEM) arbitrary optical density units of CRH mRNA expression.
progression; however, to analyze the associations between observed changes in basal HPA function and disease onset Pearson correlation coefficients were calculated for all the data. Autoantibody titers (anti-DNA, r (37) ⫽ ⫺.46, anti-ribp, r (37) ⫽ ⫺.50, anti-lba1, r (37) ⫽ ⫺.48, all significant at p ⬍ .01) appeared to be negatively associated with altered CRH mRNA expression. These correlations suggest that increments in autoantibody titers were associated with decreased CRH mRNA expression. Urine protein levels would be expected to increase after autoantibodies are detected; indeed this was the case, wherein autoantibodies are detected in the MRL lpr/lpr mice as early as 6 weeks and become as evident at lower levels in the MRL ⫹/⫹ mice at 8 weeks. Urine proteins are markedly increased in the MRL lpr/lpr mice by both 8 and 12 weeks and to a lesser extent in the MRL ⫹/⫹ mice. The correlation between
ENDOCRINE CORRELATES OF MURINE SLE
355
FIG. 4. Representative photomicrograph of PVN AVPmRNA expression in MRL ⫹/⫹ and MRL lpr/lpr mice at 8 and 12 weeks of age, as measured by in situ hybridization and mean (⫾ SEM) arbitrary optical density units of AVP mRNA expression.
CRH mRNA levels and urine protein levels was, therefore, not as marked as that seen with the antibodies, but reached statistical significance (r(37) ⫽ ⫺.40, p ⬍ .02). In contrast, correlations of autoantibody levels and urine proteins with plasma corticosterone levels did not reach statistical significance (p ⬎ .05). DISCUSSION
All three of the antibodies tested correlated with the decline in CRH expression; however, despite their widely differing specificity this is not surprising for several reasons: (1) The antibodies tended to occur in tandem so it was not possible to correlate the changes in hypothalamic CRH with individual antibodies. (2) Antibodymediated biologic changes need not be antigen specific as effector mechanisms can occur through the Fc portion of the molecule when the antibodies are in complexes or by nonspecific injury to the vasculature resulting in endothelial production of cytokines. Nonetheless, features of two of the antibodies are sufficiently relevant to
356
SHANKS ET AL.
neuropsychiatric SLE to warrant further study. The anti-ribosomal P antibodies have a clinical correlation with psychoses in patients with SLE, a feature that certainly suggests an endocrine component (Hutchinson et al., 1996). We recognize that a murine model may not be appropriate to explore these features; however, the antilba1 antibody is interesting because the target autoantigen is present only in the brain and there most notably in the cingulate gyrus and hippocampus—two regions with prominent communication with the hypothalamus (Moore, Vo, & Carlock, 1998). Our observations concerning altered peptide mRNA expression in the hypothalamus are similar to those previously reported in exogenously activated models of rodent autoimmune disease which develop reduced PVN CRH mRNA and increased AVP mRNA expression. These changes in hypothalamic gene expression occur at a very early stage in the development of murine SLE and do not appear to be a result of profound alterations in circulating steroid, kidney pathology, or cerebrovasculature immune complex deposition which occur later on. Evidence from studies in rats with AA suggest that the alterations in PVN peptidergic profiles are not a result of increased glucocorticoid feedback, but rather reflect a chronic activation of the HPA axis regulation during chronic inflammation. The increased contribution of AVP to HPA regulation may enable continued responsivity of the axis when faced with elevated levels of circulating glucocorticoids and may explain the diminution of HPA responsivity to restraint stress, but not to LPS or adrenalectomy seen in rats with AA (Harbuz et al., 1993). It would appear, therefore, that HPA responses ‘‘selectively adapt’’ to the chronic stress regimen, but maintain responsivity to salient stimuli or even become sensitized to immune–HPA pathways with preferential HPA responses to regulate disease progression. This explanation may also apply to animals with murine SLE which respond to acute challenge with IL-1β, but whose responsivity diminishes as the disease progresses (Lechner et al., 1996). Further investigation is required to determine whether these animals maintain HPA responses to other stressors. It could be argued that the alterations observed in these mouse strains reflect HPA development rather than an ongoing disease process. While less is known about HPA development in mice, it has been observed in rats that at 6 weeks of age corticosterone, CBG, and hypothalamic CRH and AVP peptide and mRNA levels are expressed at adult levels (see review, Meaney, O’Donnell, Viau, Bhatnagar, Sarrieau, Smythe, Shanks, & Walker, 1993). Furthermore, while increasing mRNA expression for central peptides with development would be expected, rarely would one expect to see a decrement in hypothalamic CRH mRNA expression associated with development. Consequently, we suggest that the observed alterations in mRNA expression for this peptide are in all likelihood a consequence of the ongoing inflammatory processes. However, the mechanisms that mediate the decreased hypothalamic CRH and increased AVP mRNA levels are unknown, and it has been reported that catecholamine levels within the PVN are altered in AA (Harbuz, Chover-Gonzalez, Biswas, & Lightman, 1994) and alterations in neurotransmitter input to the hypothalamus undoubtedly contribute to altered CRH mRNA expression. 6-Hydroxydopamine lesions of noradrenergic PVN neurons (75% reduction), however, did not significantly alter either the CRH mRNA or corticosterone responses to AA, although they did result in a more severe inflammatory response (Harbuz et al., 1994). On the other hand, serotonergic depletion at inflammation onset significantly reduced hind paw inflammation again with little effect on CRH mRNA (Harbuz, Perveen-Gill, Lalies, Jes-
ENDOCRINE CORRELATES OF MURINE SLE
357
sop, & Lightman, 1996). These data suggest roles for serotonin and catecholamines in the AA inflammatory response. Neuropeptides may also play a role in the regulation of CRH and AVP mRNA and substance P has been implicated in hypothalamic CRH changes associated with inflammation (Chowdrey, Larsen, Harbuz, Lightman, & Jessop, 1995b). Changes in parvocellular hypothalamic peptide mRNA expression associated with chronic inflammation may also relate to the differential activation of cells containing CRH alone or cells with colocalized CRH and AVP. Both CRH ⫹ and CRH/AVP ⫹ cell populations exhibit peptide release following IL-1 administration and it has been suggested that this may be why inflammatory stimuli continue to activate the HPA axis even when superimposed onto a chronically stimulated background and could explain why the response to inflammatory stimuli differs from other stressors (Whitnall, 1994). More recently it has been demonstrated that one peripheral injection of IL-1 promotes the colocalization of AVP within hypothalamic CRH neurons (Schmidt, Janszen, Wouterlood, & Tilders, 1995). It was observed that for up to 2 weeks post-IL-1β treatment animals remain ‘‘hyperresponsive’’ to subsequent stress exposure. It is thought that this may be due to a synergy of the peptide effects on HPA activity following increased colocalization. Furthermore, it was demonstrated that a single injection with IL-1β prior to induction of EAE reduced the vulnerability of Lewis rats to EAE and this was correlated with increased colocalization of AVP and CRH (Schmidt et al., 1995). Consequently, it appears that increasing colocalization of AVP and CRH may better enable the HPA axis to deal with inflammatory inputs to the CNS and subsequently improve immune regulation and dampening of inflammatory responses. The alterations in parvocellular AVP/CRH colocalization in autoimmune models may represent a hypothalamic adaptation to moderate disease progression or alternatively, decreased colocalization may dictate increased susceptibility to inflammation (Tilders, Schmidt, Aguilera, Kiss, Van Dam, Huitinga, & Dijkstra, 1996; Whitnall, Anderson, Lane, Mougey, Neta, & Perlstein, 1994). The present data do differ to some extent from endocrine profiles seen in rat models of inflammation, in that MRL mice did not exhibit such pronounced elevations in plasma corticosterone. It may be the case that during the gradual development of inflammation in spontaneously occurring pathologies such as this murine model of SLE, HPA responses are not as suddenly or severely stimulated as in exogenously induced animal models of inflammation. Indeed, in the present study circulating levels of corticosterone tended to gradually increase in association with increments in disease indicators and were also associated with decreased levels of circulating CBG. Individually these effects were relatively small, but together these responses may contribute to a physiologically relevant alteration in free steroid signal especially when considering the morning levels of circulating corticosterone. However, what these data do indicate is that alterations in resting levels of corticosterone (1) are not associated with the observed alterations in hypothalamic CRH mRNA expression and (2) are not elevated during the early stages of disease onset in murine SLE. We have also observed in the present investigation that the alterations seen in PVN CRH of MRL mice corresponded temporally with behavioral alterations previously observed in these mouse strains (Sakic et al., 1994; 1995). While it was not investigated directly, it is tempting to speculate that altered expression of ‘‘anxiety behaviors’’ or ‘‘emotionality’’ may be associated with changes in central CRH regulation in these animals. Administration of immunosuppressants can prevent behavioral alter-
358
SHANKS ET AL.
ations seen in the MRL mice prior to disease onset and has also been shown to disrupt the immune sequelae and HPA activation in adjuvant-induced autoimmune rat disease models (Stephanou, Sarlis, Knight, Lightman, & Chowdrey, 1992). Immune factors associated with disease are likely mediators of altered central neuroendocrine function in MRL mice, but whether or not alterations in hypothalamic CRH contribute to behavioral disturbances associated with disease onset in this murine model remains to be determined. Alterations in central endocrine and limbic functions resulting from signals of immune origin could contribute to neuropsychiatric manifestation of SLE. Furthermore, chronic inflammatory conditions that alter endocrine and CNS function may be etiological factors contributing to psychiatric and stress-related disease. ACKNOWLEDGMENTS The authors thank The Wellcome Trust (Grant Ref. 045362/Z/95/MP/JF) for their support of this project.
REFERENCES Aguilera, G. (1994). Regulation of pituitary ACTH-secretion during chronic stress. Frontiers Neuroendocrinol. 4, 321–350. Bloom, D. D., Davingnon, J. L., Cohen, P. L., Eisenberg, R. A., & Clarke, S. H. (1993). Overlap of the anti-Sm and anti-DNA responses of MRL/Mp-lpr/lpr mice. J. Immunol. 150, 1579–1590. Chesnokova, V. M., Ivanova, A. N., Karyagina, A. Y., Chukhlib, V. L., & Ivanova, L. N. (1995). Patterns of interaction between the immune and endocrine systems in the manifestation of hereditary autoimmune disorders in NZB mice. Bull. Exp. Biol. Med. 119, 381–384. Chowdrey, H. S., Larsen, P. J., Harbuz, M. S., Jessop, D. S., Aguilera, G., Eckland, D. J. A., & Lightman, S. L. (1995a). Evidence for arginine-vasopressin as the primary activator of the HPA axis during adjuvant-induced arthritis. Br. J. Pharmacol. 116, 2417–2424. Chowdrey, H. S., Larsen, P. J., Harbuz, M. S., Lightman, S. L., & Jessop, D. S. (1995b). Endogenous substance-P inhibits the depression of corticotropin-releasing-hormone during a chronic inflammatory stress. Life Sci. 57, 2021–2029. Hara, Y., Battey, J. F., & Gainer, H. (1990). Structure of mouse vasopression and oxytocin genes. Mol. Brain Res. 8, 319–324. Harbuz, M. S., Chover-Gonzalez, A. J., Biswas, S., & Lightman, S. L. (1994). Role of central catecholamines in the modulation of corticotropin-releasing factor messenger-RNA during adjuvant-induced arthritis in the rat. Br. J. Rheumatol. 33, 205–209. Harbuz, M. S., Jessop, D. S., Chowdrey, H. S., Blackwell, J. M., Larsen, P. J., & Lightman, S. L. (1995). Evidence for altered hypothalamic CRF in immune-mediated diseases. Ann. N.Y. Acad. Sci. 771, 449–458. Harbuz, M. S., Leonard, J. P., Lightman, S. L., & Cuzner, M. L. (1993). Changes in hypothalamic corticotrophin releasing factor (CRF) and pituitary proopiomelanocortin (POMC) messenger RNA during the course of experimental allergic encephalomyelitis. J. Neuroimmunol. 45, 127–132. Harbuz, M. S., & Lightman, S. L. (1992). Stress and the hypothalamo–pituitary–adrenal axis: Acute, chronic and immunological activation. J. Endocrinol. 134, 327–339. Harbuz, M. S., Perveen-Gill, Z., Lalies, M. D., Jessop, D. S., & Lightman, S. L. (1996). The role of endogenous serotonin in adjuvant-induced arthritis in the rat. Br. J. Rheumatol. 35, 112–116. Harbuz, M. S., Rees, R. G., Eckland, D., Jessop, D. S., Brewerton, D., & Lightman, S. L. (1992). Paradoxical responses of hypothalamic CRF mRNA and CRF-41 peptide and adenohypohyseal POMC mRNA during chronic inflammatory stress. Endocrinology 130, 1394–1400. Harbuz, M. S., Rees, R. G., & Lightman, S. L. (1993). HPA axis responses to acute stress and adrenalectomy during adjuvant-induced arthritis in the rat. Am. J. Physiol. 264, R179–R185. Hutchinson, G. A., Nehall, J. E., & Simeon, D. T. 1996. Psychiatric-disorders in systemic Lupus-erythematosus. West Indian Med. J. 45, 48–50.
ENDOCRINE CORRELATES OF MURINE SLE
359
Isenberg, D. A. (1997). Autoantibodies: Markers of disease or pathogenic? Ann. N.Y. Acad. Sci. 823, 256–262. Lechner, O., Hu, Y., Jafarian-Tehrani, M., Dietrich, H., Schwarz, S., Herold, M., Haour, F., & Wick, G. (1996). Disturbed immunoendocrine communication via the hypothalamo–pituitary–adrenal axis in murine lupus. Brain Behav. Immun. 10, 337–350. Ma, X-M., Levy, A., & Lightman, S. L. (1997). Rapid changes in heteronuclear RNA for corticotrophinreleasing hormone and arginine vasopressin in response to acute stress. J. Endocrinol. 152, 81– 89. MacPhee, I. A., Antoni, F. A., & Mason, D. W. (1989). Spontaneous recovery of rats from experimental allergic encephalomyelitis is dependent on regulation of the immune system by endogenous adrenal corticosteroids. J. Exp. Med. 169, 431–435. Meaney, M. J., O’Donnell, D., Viau, V., Bhatnagar, S., Sarrieau, A., Smythe, J., Shanks, N., & Walker, C-D. (1993). Corticosteroids receptors in the rat brain and pituitary during development and hypothalamic-pituitary-adrenal function. In Zagon, I. S., & McLaughlin, P. J. (Eds.), Receptors in the Developing Nervous System, Vol 1, pp. 163–201. Chapman & Hall, London. Moore, P. M., Vo, T., & Carlock, L. R. (1998). Identification and cloning of a brain autoantigen in neurobehavioral SLE. J. Neuroimmunol. 82, 116–125. Pisetsky, D. S. (1993). Autoantibodies and their significance. Curr. Opin. Rheumatol. 5, 549–556. Sakic, B., Szechtman, H., Talangbayan, H., Denburg, S. D., Cabotte, R. M., & Denburg, J. A. (1994). Disturbed emotionality in autoimmune MRL-lpr mice. Physiol. Behav. 56, 609–617. Sakic, B., Szechtman, H., Denburg, S. D., & Denburg, J. A. (1995). Immunosuppressive treatment prevents behavioral deficit in autoimmune MRL-lpr mice. Physiol. Behav. 58, 797–802. Schmale, H., Heinsohn, S., & Richter, D. (1983). Structural organization of the rat gene for the arginine vasopression-neurophysin precursor. EMBO J. 2, 763–767. Schmidt, F. D., Janszen, A. W. J. W., Wouterlood, F. G., & Tilders, F. J. H. (1995). IL-1 induced long-lasting changes in hypothalamic corticotropin releasing hormone (CRH) neuron and hyperresponsiveness of the HPA axis. J. Neurosci. 15, 7417–7426. Schrott, L. M., & Crnic, L. S. (1996). Anxiety behavior, exploratory behavior, and activity in NZBxNZWF1 hybrid mice: Role of genotype and autoimmune disease progression. Brain Behav. Immun. 10, 260–274. Sherman, G. F., Galaburda, A. M., Behan, P. O., & Rosen, G. D. (1987). Neuroanatomical anomalies in autoimmune mice. Acta Neuropathol. 74, 239–242. Stephanou, A., Sarlis, N. J., Knight, R. A., Lightman, S. L., & Chowdrey, H. S. (1992). Effects of cyclosporine-A on the hypothalamic–pituitary–adrenal axis and anterior-pituitary IL-6 messengerRNA expression during chronic inflammatory stress in the rat. J. Neuroimmunol. 41, 215–222. Sternberg, E. M., Wilder, R. L., Gold, P. W., & Chrousos, G. P. (1990). A defect in the central component of the immune system–hypothalamic–pituitary–adrenal axis feedback loop is associated with susceptibility to experimental arthritis and other inflammatory diseases. Ann. N.Y. Acad. Sci. 594, 289– 291. Sternberg, E. M., Young, W. S., III, & Bernardini, R. (1989). A central nervous defect in the stress response is associated with susceptibility to streptococcal cell wall arthritis in Lewis rats. Proc. Natl. Acad. Sci. USA 86, 2374–2378. Theofilopoulos, A. N., & Dixon, F. J. (1985). Murine models of systemic lupus erythematosus. Adv. Immunol. 37, 269–390. Tilders, F. J. H., Schmidt, E. D., Aguilera, G., Kiss, A., Van Dam, A. M., Huitinga, I., & Dijkstra, C. D. (1996). Sustained effects of cytokines, neuroplasticity, stress and immunomodulation. 2nd Annual Meeting of the Brain-Immune Network Group, London. Vlahakos, D. V., Foster, M. H., Adams, S., Katz, M., Ucci, A. A., Barrett, K. J., Datta, S. K., Madaio, M. P. (1992). Anti-DNA antibodies form immune deposits at distinct glomerular and vascular sites. Kidney Int. 41, 1690–1700. Vogelweid, C. M., Wright, D. C., Johnson, J. C., Hewett, J. E., & Walker, S. E. (1992). Evaluation of memory, learning ability, and clinical neurologic function in pathogen-free mice with systemic lupus erythematosus. Arthritis Rheum. 37, 889–897.
360
SHANKS ET AL.
Watanabe-Fukunaga, R., Brannan, C. I., Copelan, N. G., Jenkins, N. A., & Nagata, S. (1992). Lymphoproliferation disorder in mice explained by defects in Fas antigen that mediates apoptosis. Nature 356, 314–317. Wilder, R. L. (1995). Neuroendocrine-immune system interactions and autoimmunity. Annu. Rev. Immunol. 13, 307–338. Whitnall, M. H. (1994). Regulation of the hypothalamic corticotropin-releasing hormone neurosecretorysystem. Prog. Neurobiol. 40, 573–629. Whitnall, M. H., Anderson, K. A., Lane, C. A., Mougey, E. H., Neta, R., & Perlstein, R. S. (1994). Decreased vasopression content in parvocellular CRH neurosecretory-system of Lewis rats. Neuroreport 5, 1635–1637. Received June 16, 1998