NeuroscienceVol. 57, No. 3, pp. 861-871, 1993 03064522/93 $6.00t 0.00 Printedin GreatBritain Pergamon Press Ltd 0 1993 IBRO

SYNERGISTIC ACTION OF ESTRADIOL AND MYELIN BASIC PROTEIN ON SECRETION AND BRAIN MYELIN CHANGES RESEMBLING EARLY STAGES OF DEMYELINATION

T. C. THP.OHARIDES,*V. DmmtIADOu,t R. LETOURNEAU,J.J. ROZNIECKI,~ H. VLIAGOFTI# and W. BOUCHER Department of Pharmacology and Experimental Therapeutics, Tufts University School of Medicine, 136 Harrison Avenue, Boston, MA 02111, U.S.A.

Abstract-Mast cells are known for their participation in immediate and, more recently, delayed hypersensitivity reactions. They have been found in the meninges and certain brain areas where they are strictly perivascular, in close apposition to neurons, and they are activated by direct nerve stimulation or by neuropeptides. Intracranial mast cells contain many vasoactive substances which can increase the permeability of the blood-brain barrier, proteolytic enzymes which can degrade myelin in vitro, as well as chemotatic molecules which can attract inflammatory molecules in uivo. Connective tissue mast cells, with which intracranial mast cells share many characteristics, contain cytokines which can cause inflammation directly. Multiple sclerosis is a human demyelinating disease of unknown etiology, with a high prevalence in women which results in penetration of blood-borne immune cells within the brain parenchyma and subsequent destruction of myelin. Here, we report that l7)-estradiol and myelin basic protein, a major suspected immunogen in multiple sclerosis, had a synergistic action on inducing mast cell secretion. This effect was more pronounced in Lewis rats, which arc susceptible to the development of experimental allergic encephalomyelitis, an animal model for multiple sclerosis, than in Sprague- Dawley rats, which are fairly resistant. Moreover, 18 h incubation of purified peritoneal mast cells with homogeneic slices of brain white matter in the presence of l’lfi-estradiol and myelin basic protein resulted in myelin changes resembling early stages of brain demyehnation, which were also more evident in Lewis rats than in Sprague-Dawley rats. These results support the notion that mast cells could participate in the pathophysiology of demyelinating diseases.

Mast cells are necessary for the development of to directly degrade myelin in virro.‘.‘*.ZSThere is also immediate, and possibly delayed, hypersensitivity strong evidence suggesting a functional interaction reactions.j3 They derive from the bone marrow and between mast cells and the nervous system, since enter the tissues as immature cells which differentiate peripheral mast cells secrete in response to some under microenviromnental factors such as interleukin neuropeptides,9J’,24,s6 are often in close apposition 3 released from local T-cells.18 Their activation is to neurons,4~‘3,36and are regulated by nerve growth known to be triggered by , which factor.4 Moreover, mast cells secrete in response to binds to specific surface receptors” and combines myelin basic protein (MBP) and induce peripheral*’ with antigen leading to the release of numerous and centra13Tj3demyelination. Mast cells, probably powerful mediators. 45,4’Mast cell vasoactive amines originating from the neural crest,” have also been could modulate the integrity of the blood-brain identified intracranially,23 where they show a strict barrier, while proteolytic enzymes” have been shown perivascular location3’ and where they secrete vaso- active amines in response to substance P,” compound 48/80 (C48/80) or carbachol.‘6a32 *To whom correspondence should be addressed. TPresent address: Laboratory of Physiology, School of Certain autoimmune disorders such as multiple Pharmacy, Universite Rene Descartes, Paris V, 4, avenue sclerosis (MS),66 irritable bowel disease and inter- de l’observatoire, 75270 Paris Cedex 06, France. stitial cystitis” show a high prevalence in women, SPermanent address: Department of Neurology, Medical suggesting that a hormonal component could be Academy of Lodz, 22 Kopcinskiego Street, Lodz 90153, Poland. _ implicated in the pathophysiology. It was, therefore, §Present address: Allergic Diseases Section, NlAID, of interest to examine the effect of 17j-estradiol and Bldg. 10, Rm 116205, Bethesda, MD 20892, U.S.A. MBP on mast cell secretion and brain myelin changes Abbrev~arions: C48/80, compound 48/80; EAE, experimen- in two strains of rats which exhibit different sensi- tal allergic enceohalomvelitis; HEPES, N-2-hvdroxv- ethylpip&azine-i?‘-2-eth&esuifonic acih; LDI-i, la&c tivities to the development of experimental allergic dehydrogenase; MBP, myelin basic protein; MS, encephalomyelitis (EAE), an established animal multiple sclerosis. model for MS.

861 862 T. C. THEOHARIUES cl (I/

EXPERIMENTAL PROCEDURES recorded and the IMEXEC program was exited. images were then read into the DISPEX program for the measure- Masf cell purification and activation ment of areas with selected threshold values, i.e. myelin Mast cells were collected from the peritoneal cavities of areas. These values were also stored on disk files. Editing 32553508 male Sprague-Dawley or Lewis rats (Taconic, became necessary to remove densities from profiles which Inc., Germantown, NY) in HEPES-buffered Locke’s sol- were not myelin related. ution (pH 7.2) containing 2 mM Cal +s4 and were purified The axons studied within each group were numbered over metrizamide.” Purity was better than 95% as judged for positive identification. The perimeter of each axon was by Toluidine Blue staining (0.25% at pH 2.5). They were measured in pixels and converted to a micrometer length loaded with [3H]serotonin (1S-30 Ci/mmol, New England value by dividing a correction factor dependent upon mag- Nuclear, Boston, MA) for 1h at 37”C, washed twice nification (28.004 pixels/pm in this study). The myelin area and then incubated at 37-C (IO’ cells/tube) with different for each axon was determined by thresholding and editing concentrations of MBP (Sigma, St. Louis, MO) for 10 min of non-myelin densities. This measure yielded area in pixels. either alone or following 90 min or 18 h preincubation with The total pixel areas for each axon were then divided by a different concentrations of 178-estradiol (Sigma); 17a-estra- correction factor (784.22 pixels/pm2) which converted pixel diol (Sigma) was used only at 5 x 10m5 M. Stock concen- areas to micrometers. Finally, the myelin area in pm* was trations of estradiol (lo-*M) were prepared in 95% divided by the axon perimeter to yield an answer of square ethanol; the final ethanol dilution had no effect on mast micrometers of myelin area per micrometer length of axon cell secretion. [3H]Serotonin was assayed in a scintillation perimeter. This conversion is important in that it provides counter, while the endogenous was measured a meaningful number by which to compare the axon sheaths fluorometrically.2R from axons of differing caliber, i.e. differing perimeter. A two-tailed t-test was applied to this data for determi- Brain dissection and slice incubation nation of significant differences in area axon sheath per unit Brain slices were obtained from male 32553.50 g Sprague- perimeter between treatment groups. Dawley or Lewis rats (Taconic, Germantown, NY) which Cell viability during isolated mast cell, as well as were killed by CO, vapor and decapitation. The skull was brain slice, experiments was evaluated by measuring lactic opened, the brain was rapidly excised at 4°C and brain dehydrogenase (LDH) release spectrophotometrically,’ transverse slices, approximately 1 mm thick, were then cut using an assay kit obtained from Sigma (St. Louis, MO). using a brain slicer and were introduced into a solution ,411 data were evaluated with a two-tailed l-test and any containing Eagle’s minimal essential medium supplemented level of significance where the probability was less than with 10% calf serum (Grand Island Biological, LI), 300 0.001 were simply listed as P i 0.001. units penicillin and 250 mg streptomycin per 100 ml. Con- trol slices were incubated with estradiol alone, MBP alone or estradiol and MBP for 18 h in a CO, incubator. Other slices were also treated with either 5 x IO-’ M 17P-estra- diol, 2 x 10-s M MBP or both, but in the presence of mast cells (2 x IO’ per 3 ml). They were then fixed overnight at 4°C with medium composed of 2% paraformaldehyde, 2.5% glutaraldehyde, with 0.5% tannic acid, in Sorenson’s phosphate buffer adjusted to pH 7.4. Samples from later experiments were fixed in 3% glutaraldehyde with 0.5% tannic acid in 0.1 M cacodylate buffer. Then, the white matter was carefully dissected and samples taken for elec- tron microscopy were postfixed in 1% osmium tetroxide (0~0,) at room temperature for 2 h, dehydrated in graded alcohols and embedded in Epon. Sections were then cut in an LKB or Servall ultramicrotomes, counterstained with many1 acetate and lead nitrate and studied using a Jeol-1005 or a Phillips 300 transmission electron microscope. Brain slices from three incubation experiments were processed for electron microscopy. Forty blocks were used and brain sections were photographed and examined by image analy- sis to determine the myelin thickness in relation to the axon diameter. Any axons showing myelin lamellae changes, such as breaks, swelling and/or myelin thinning, were scored as “affected”.

Image analysis The effects of the conditions described above were studied by electron microscopic computerized image analysis. Imaging studies were carried out at the Image Analysis Laboratory at Tufts-New England Medical Center. Electron Fig. 1. Effect of 17/I-estradiol on MBP-induced mast cell microscope negatives, along with a distance calibration serotonin release from Sprague+Dawley rats. Preincubation standard, were digitized from a Gordon variable intensity for 90 min at 37°C with (filled diamonds) or without (open table using the imaging program IMEXEC running on a squares) l7b-estradiol (5 x 1O-5 M). Cells were stimulated Digital System 780 computer. Digitized images stored on with MBP for 10min at 37°C. Mean and standard devi- disk were recalled into the IMEXEC program for a 256 grey ations were derived from four experiments (n = 4), which scale - 1 slope transformation which converted the negative were performed in duplicate except at the concentration of image into a positive image. The transformed image was 10e5 M MBP, where values from a total of nine (n = 9) used in the analyses. The perimeters of axon membranes experiments were used. Probability values are as shown: were measured. Thereafter, the pixel threshold value, which (*) P (0.001; (*‘) P

RESULTS bation with 17fl-estradiol increased from 19.9 + 2.0% to 37.6 + 1.6% (P < 0.01). Preincubation with Incubation for 10min of purified peritoneal 5 x 10e5 M 17p-estradiol followed by a lower dose mast cells from Sprague-Dawley rats with MBP (10 p M) of MBP (n = 9) increased serotonin release at 37°C caused a dose-dependent net release of from 10.6 + 5.9% to 23.12 k 3.6% (P

Fig 2. High power light micrographs of mast cells from Sprague-Dawley or Lewis rats treated v Iith 17p-estradiol and MBP. Mast cells were purified and used (10’ cells/tube) as described in Experimental Procedures. Mast cells were pretreated with 5 x lo-’ M I7/3-estradiol for 90 min before treatment with 2 x 10 -‘M MBP for an additional 10 min (n = 4). Mast cells were stained with Giemsa. Over 20 different fields were examined from each condition and one respective representative field is shown below. Upper panel: mast cells from Sprague-Dawley rats. Note that only a few granules from a small portion of mast cells have released their content. Magnification: x 1000. Lower panel: mast cells from Lewis rats. Note that most of the granules have released their content in practically every mast cell shown. Magnification: X 1000. 864 T. C. THEOHARIDESet al.

12 + 5.9% to 17.6*4.8% (P <0.04), while 17a- Table 1. Effect of 18 h incubation with 17P-estradiol on estradiol (n = 9) did not increase histamine secretion myelin basic protein-induced mast cell secretion at all (12.0*5.9% vs 11.1 +3.5%). Conditions Serotonin release (% total) Preincubation for 18 h was evaluated for sero- Control 16.4 + 4.9 tonin secretion only (n = 18). Basal secretion was MBP (IO-jM) 23.3 k ?.6* 16.4 + 4.9%, while secretion with MBP alone was 17a-Estradiol f MBP 37.9 I8.8t l7p-Estradiol + MBP 52.1 + 19.7tj 23.3 &-7.6% (P < 0.002). Augmentation to 52.1 f 19.7% by 17/?-estradiol was highly significant Mast cells from Sprague-Dawley rats were preincubated (P < 0.001). Augmentation by 17a-estradiol was with 5 x 10e5M of 17a- or 17jestradiol for 18h at 37°C and were then stimulated with 10eSM MBP smaller (37.9 k 8.8%), but was still significant for an additional 10min. The results represent the (P < 0.001). Nevertheless, the difference between mean k S.D. (n = 18). *P -c0.001 compared to control; the amount of serotonin secreted in the presence of tP < 0.001 compared to MBP; $P < 0.008 compared to 17/?-estradiol was significantly higher over that seen 17~ -estradiol and MBP. with 17a-estradiol (P < 0.008) (Table 1). Interestingly, the amount of serotonin released appeared normal with numerous filopodia and many with 2 x lo-’ M MBP alone was 35.4 + 1.8% for homogeneous, electron-dense granules (Fig. 3B,F). Lewis vs 19.9 + 2.0% for Sprague-Dawley .rats, In contrast, cells stimulated with 2 x lo- ’ M MBP while after 90 min preincubation with 5 x 10e5 M (n = 50) showed some loss of filopodia and many of l7b-estradiol followed by 2 x IO- 5M MBP was swollen granules which had lost their ejectron density 45.3 f 0.8% for Lewis vs 37.6 _t 1.8% for Sprague- (Fig. 3C,G). When mast cells were,pretreated with Dawley (P < 0.01) rats (Table 2). Mast cell secretion 5 x 10e5M 178-estradiol before stimulation with was higher in Lewis than in Sprague-Dawley rats, 2 x 1O-5 M MBP, degranulation was clearly en- which is consistent with the susceptibility to the hanced (Fig. 3D,H). Degranulation was definitively induction of EAE. There was no significant difference more marked in Lewis rats than that observed in in the minor amount of LDH release between control Sprague-Dawley rats, both when the cells were incu- and stimulated mast cells from either rat species bated with MBP alone (compare Fig. 3C and G) (results not shown). or with 17j-estradiol and MBP together (compare Light micrographs of purified mast cells from Fig. 3D and H). Sprague-Dawley and Lewis rats clearly showed (Fig. 2) that treatment with 5 x 10m5 M 17/3-estradiol Table 2. Effecectof 17/?-estradiol and myelin basic protein on followed by stimulation with 2 x 10-j MBP resulted mast cell secretion from different rat species in a greater degree of degranulation in the latter Serotonin release (% total) (lower panel) than the former (upper panel). Mast cells treated as in Fig. 2 were also studied with Conditions Sprague-Dawley Lewis transmission electron microscopy. Representative Control 2.4 + 0.8 2.8 & 0.4 cells from each condition are shown in Fig. 3 and the MBP (2 x 1O-5 M) 19.9 + 2.0 35.4 f 1.8t 17p-Estradiol + MBP 37.6 f 1.8: 45.3 * 0.8*t total number of cells examined is shown in paren- theses. These observations showed that control mast Mast ceils were preincubated with 5 x 10ms M 17B-estradiol for 90 min at 37°C and were then stimulated with cells (n = 20) from either Sprague-Dawley or Lewis 2 x 10m5 M MBP for an additional 10 min. The results rats (Fig. 3A,E), as well as mast cells incubated in represent the mean + S.D. (n = 4). *P < 0.01 compared the presence of 5 x 1O-5 M 17fi-estradiol (n = 20), to MBP; tP < 0.01 compared to Sprague-Dawley.

Fig. 3. Transmission electron micrographs of mast cells from Sprague-Dawley or Lewis rats treated with 17/?-estradiol and/or MBP. Mast cells were obtained and basically treated in four experiments (n = 4) as described in Fig. 2 and as further specified below. The individual cells shown are representative of the respective populations of mast cells with similar appearance, studied and noted in parentheses below. High magnification of single cells was chosen in order to show the extent of degranulation and because the mean and S. D. of secretion is reflected in Table 2. (A) Sprague-Dawley, peritoneal control mast cell with numerous electron-dense granules (n = 20 cells studied). (B) Sprague-Dawley mast cell treated with 5 x IO-‘M 17B-estradiol with unaltered electron-dense granules (n =20 cells studied). (C) Sprague-Dawley mast cell treated with 2 x 10m5M MBP. Note that the core of some granules is slightly altered, but most of the granules are still intact (n = 50 cells studied). (D) Sprague-Dawley mast cells pretreated with 5 x 10m5M l7fi-estradiol for 90min before treatment with 2 x lo-‘M MBP. Note that degranulation IS more pronounced (n = 75 cells studied). (E) Lewis peritoneal control mast cell containing many electron-dense granules (n = 20 cells studied). (F) Lewis mast cell treated with 5 x 10-j M l’lfl-estradiol. As with the Sprague-Dawley rata (Fig. 4B), no ultrastructural changes were observed (n = 20 cells studied). (G) Lewis mast cell treated with 2 x 10m5 M MBP. Note that degranu- lation is much more pronounced than that observed with Sprague-Dawley rats under the same conditions used in Fig. 4C (n = 50 cells studied). (H) Lewis mast cell pretreated with 5 x 10-j M 17,?-estradiol for 90 min before treatment with 2 x 10 _ 5 M MBP. Note again. that degranulation is much more pronounced idcomparison to the Sprague-Dawley mast cells under the-same conditions used in Fig. 4D (n = 75 cells studied). Scale bars = 1 pm. Mast cell secretion and brain myelin changes 865

Fig. 3. 866 T. C. THEUHAKILNS er al.

Brain slices (three slices of 1 mm thickness in each (Fig. 4, upper panel). In contrast. the myelin sheaths dish) from Sprague-Dawley or Lewis rats were also surrounding axons of the tissues incubated with incubated for 18 h at 37°C with or without purified 17/?-estradiol and MBP in the presence of mast cells homogenic peritoneal mast cells (2 x 10’ per 3 ml). were of irregular thickness and density; myelin Control samples incubated with 5 x 10 _ ’ M estradiol lamellae were separated with clear spaces between and 2 x lo-‘M MBP without mast cells contained them and there were frequent sites where the myelin myelinated axons where the myelin thickness was sheath bulged out into “smudged” irregular shapes uniform around any given axon; myelin lamellae were with sharply reduced electron density (Fig. 4, lower easily resolved and myelin density was homogeneous panel). Some axons, as well as some mitochondria,

Fig. 4. Transmission electron micrographs of cortical brain slices from Lewis rats incubated for 18 h with purified homogeneic rat peritoneal mast cells. Upper panel: brain parenchyrna of a brain cortical cross-section incubated with unstimulated mast cells in the presence of 17B-estradiol as described in Experimental Procedures, showing myelinated axons. Note that the myelin sheaths remain well stacked with homogeneous density (solid arrowhead); the axons show some swelling; one mitochondrion present is also swollen, but Intact. Magnification: x 13,750. Lower panel: brain parenchyma of a bram cortical cross-section incubated with mast cells in the presence of 178-estradiol and MBP as described in Experimental Procedures, showing myelinated nerve fibers. Note areas where myelin lamellae have lost their periodicity and are separated with clear spaces between them (white bounded solid arrowhead) or have swollen up and formed “smudged” knot-like expansions with myelin dissolution-like changes (asterisk): one mitochondrion present is swollen, but intact. Magnification: x 16,500. Mast cell secretion and brain myelin changes 867 were swollen in both the control and stimulated were still abnormal when examined at very high samples, but most of them were intact (Fig. 4). magnification, These axons contained areas with Brain slices incubated in the presence of mast cells normal periodic dense lines containing a single intra- trigger1:d with 17P-estradiol and MBP contained periodic line which actually consists of two “intact -looking” axons, but with myelin sheaths that branes fused together (bilaminar) (Fig. 5, 4-w

Fig. 5. Transmission electron micrograph of myelin sheaths from brain cortical slices from Lewis rats treated as in Fig. 4. Upper panel: a myelin sheath from control tissue as shown in Fig. 4, upper panel. Arrows indicate the periodic major dense fines containing bilaminar intraperiod dense lines. This figure is typical of normal myelin. Magnification: x 211,000. Lower panel: a myelin sheath from tissue treated with mast cells incubated with MBP and l’lfi-estradiol as shown in Fig. 4, lower panel. Arrows indicate major dense lines between which lies the bilamellar intraperiod dense line (center) plus two adjacent dense lines. Quadrilaminar intraperiod structure is typical for myelin swelling, which is indicative of early stages of demyelination. Magnification: x 211,000. 868 T. C. THE~HARIUESet ul panel). However, there were also abnormal areas with the endogenous mast cells expected to be present in three intraperiodic lines (Fig. 5. lower panel). Since brain slices used. In addition, the reactivity of the the ‘middle line was still bilaminar, the overall endogenous mast cells to MBP and estradiol may bc appearance was quadrilaminal. different to that of the exogenous peritoneal mast Thcsc myclin changes were evaluated objectively cells added. The axonal swelling and degeneration by comparing the density of myelin sheath on high which was also observed in control slices could. magnification transmission electron micrographs of therefore, be due to the large number of secreting samples from both Sprague-Dawley and Lewis rats mast cells and the length of incubation. However. incubated with 17/?-estradiol and MBP as described the brain tissue was still grossly intact, as evidenced above. In each instance, the micrometer square area by the fact that the mitochondria present were only of myelin per micrometer of axon perimeter was somewhat swollen. Estradiol and MBP without greater in tissues treated with 17/l-estradiol and MBP added mast cells did not have a significant effect on with mast cells as compared to 17P-estradiol MBP myelin integrity, probably because MBP is too basic without mast cells (Table 3). It is also important to to easily penetrate the brain slices. For instance, in note that these changes were statistically significant studies using C48/80 to stimulate brain mast cell only in the case of Lewis and not Sprague-Dawley secretion, the concentration of C4SjSO required was rats (Table 3). 100 times higher than that required for peritoneal mast cells.32 Endogenous mast cells may not have been exposed to estradiol and MBP long enough DISCUSSION for complete demyelination to occur, since EAE The present data not only confirm previous usually requires 14-21 days to develop. However, studies showing that MBP can trigger mast cell mast cells with extensive morphological changes m secretion,3~25~3’~53but also provide additional direct their granules indicating secretion (“activated mast in vitro evidence of mast cell participation in myelin cells”) were found in brain of rats3*j3 and monkeys changes resembling brain demyelination.3,53 More- with EAE.j5 over, our present work shows that 17/l-estradiol Quadrilaminar myelin sheaths were a common augments mast cell secretion induced by MBP. This finding in our samples treated with 17/l-estradiol effect is clearly not due to cell toxicity because and MBP, but not in control samples. Such quadri- 17/l-estradiol, or the final ethanol concentration laminar myelin structure is not non-specific but in which it was dissolved, had no effect on their has been regarded exclusively as an early stage of own. Moreover, there was no significant LDH demyelination both in EAE and MS. For instance, release and the electron micrographs of treated the appearance of a quadrilaminar myelin sheath mast cells showed no evidence of damage. Mast pattern has been reported in EAE4’ where the nor- cells may, therefore, not only permit leukocyte entry mally intraperiod line was increased to four electron- into parenchyma by increasing blood-brain barrier dense leaflets; this pattern was also accompanied by permeability, but may also participate directly in abrupt breaks and angulated regions of the myelin demyelination. sheath.’ The appearance of a quadrilaminar myelin The number of mast cells in whole rat brain varies sheath in our study was often followed by swelling from about 1500 to 19,000, most of which are found (hydration) of the myelin and finally fragmentation in the thalamus.** If we were to assume that an of the outer lamellae with further expansion, loss average of 10,000 mast cells are evenly distributed of periodicity and electron density leading to throughout the rat brain, then there would be - 1000 “smudged” knot-like protrusions. Smudged myelin mast cells per each of the 10 slices into which each sheaths were observed in EAE plaques8 before the brain was cut, or 3000 mast cells per three slices outer lamellae separate and their thickness decreases placed in each dish. In this study, the number of mast considerably to that of only a few lamellae.4’ Our cells added to brain slices was 2 x 105/3 ml containing experiments would be expected to contain many such three brain slices; it was, therefore, much greater than areas because of the impact of numerous stimulated

Table 3. Effect of 17P-estradiol and myelin basic protein on myelin changes resembling demyelination Extent of myelin changest Conditions* Sprague-Dawley Lewis 17/I-Estradiol + MBP (control) 0.14+0.71 (n =21) 0.28*0.11 (n =66) 17/?-Estradiol + MBP + mast cells 0.16 + 0.77 (n = 54)$ 0.67 + 0.37 (n = 42)s *Mast cells (2 x 10-j ml) were co-incubated with homogeneic brain slices either in the absence (control) or presence of mast cells with 5 x tom5 M 17P-&radio1 to 5 x lo-‘M MBP for 18 h at 37°C. TMyelin sheath area per unit perimeter of axon limiting membrane (pm’/pm); n = samples analysed. fNot statistically significant compared to the respective control. §P i 0.05 compared to the respective control. Mast cell secretion and brain myelin changes 869 mast cells. Our quantitation method records myelin- trations used were over lOOO-fold higher than the related density over a certain threshold and does not peak 17j-estradiol levels observed in humans prior include areas devoid of myelin. It does, therefore, to ovulation or during treatment for osteoporosis. exclude clear areas seen in non-specific swelling. The augmentation of mast cell secretion and hyelin The increased secretion from mast cells@’ and changes, therefore, occur at pharmacological 17fl- higher degree of myelin changes from Lewis, as estradiol concentrations. However, such high hor- compared to Sprague-Dawley, rats is intriguing in mone concentrations were necessary in our studies view of the fact that the former are more susceptible possibly because: (a) human rather than rat mast to the induction of EAE.“8 Such experiments will cells may be more sensitive to the hormone action; have to be repeated using Brown-Norway rats, which (b) brain mast cells may be more sensitive to the are resistant to EAE.‘* Even though the number of action of hormones than peritoneal mast cells; brain mast cells varies among these species:6 such (c) estrogen receptor may be damaged during the variations could not be=a factor in the present studies peritoneal mast cell purification procedure; or (d) since the number of homogeneic mast cells added was concurrent presence of other sex hormones may kept constant. Increased susceptibility of brain mast reduce the amount of each required to stimulate mast cells and/or neurons to the actions of 17P-estradiol cells; for instance, progesterone was recently shown and MBP is a more plausible explanation. For to cause selective mast cell secretion.65 Finally, the instance, we recently showed that trigeminal nerve effect of estradiol may not be through estrogen stimulation can directly cause mast cell secretionI receptors, but through estrogen-binding proteins, and subsequent increased vascular permeability,” possibly linked to protein kinase C, which have thus suggesting that susceptibility to nerve stimu- higher binding constants for estradioLM Future lation may be an important factor. Finally, there may studies using female Lewis rats at different stages of be differences in brain tissue susceptibility to proteo- their estrous cycle will help answer some of these lytic enzymes released from mast cells. It is, there- questions, as female mast cells may already be fore, important that the proteolytic enzyme tryptase, “primed” to have a higher response to MBP because which is a specific marker for human mast cells, was of prolonged exposure to 178 -estradiol. found to be elevated only in cerebrospinal fluid of MS The tricyclic drugs amitriptyline and hydroxyzine patients.43 have been shown to inhibit mast cell secretion54,58and Most of the studies investigating the cellular mast cell-induced demyelination,” while hydroxy- infiltrate in areas of demyelination have focused zinej9 and the mast cell inhibitor proxicromil inhibit on involvement of T-lymphocytes and mono- EAE.” Therefore, it may not be unreasonable to cytes,M,2m” and their role in EAE@ or experimental suggest that inhibitors of brain mast cell secretion allergic neuritis. In fact, mast cells are present within could be of therapeutic value in the treatment of the brain parenchyma,23,30 and have been localized in demyelinating diseases, such as MS.j2 In fact, a MS plaques by light microscopy.2Y~37~6’Also, prelimi- double blind study using hydroxyzine is in its third nary electron microscopic studies using brains from year at the Massachusetts General Hospital. monkeys with EAE show secreting mast cells in areas of demyelination.s5 CONCLUSIONS Mast cells also secrete during experimental allergic neuritis,6 the development of which is delayed when Our results indicate that estradiol and MBP have mast cells have been depleted of their biogenic a synergistic effect on mast cell secretion which is amines.’ Finally, mounting evidence indicates that more pronounced in cells obtained from Lewis, as a delayed T-cell response may depend on early release compared to those from Sprague-Dawley, rats. of mast cell mediator&@ and that T-cell products can Moreover, incubation of peritoneal mast cells acti- cause mast cell secretion. 27 Exposure of MBP from vated by estradiol and MBP with homogeneic brain demyelinated axons may in turn trigger mast cells slices resulted in myelin changes indicative of early leading to a cyclic reaction. stages of demyelination. These changes were again A number of studies suggest that MS may be more pronounced in Lewis rats, which are more caused by the immune response to some latent viral susceptible to EAE. Brain mast cell activation may, infection.42”s6’ It is therefore, of particular interest therefore, be involved in the pathophysiology of that mast cells arc involved in virus-induced inflam- demyelinating diseases such as MS. mation of the CNS,3r and that mast cell secretion is triggered by certain viruses,38,50while virus-induced Acknow[edgemenrs-The present work was supported in encephalitis does not develop in mast cell-deficient part by a fellowship from Muro Pharmaceutical Tewksbury, mice.” MA, the National Multiple Sclerosis Society (RG 1961-A-1) The augmentation of mast cell secretion induced by and KOS Pharmaceuticals, Miami, FL. Thanks are due to MBP was seen with both 17a- and 17/?-estradiol. Mr Harry Maskell for his excellent photographic services, MS Amy R. Lapidou for her thorough literature com- However, the extent of this effect seen with 17fi- puter searches, and to MS Barbara Graczyk, MS Kathleen estradiol was significantly greater (P < 0.008) than Jerome and MS Linda L. Tamulaites for their expert word- that with 17cl-estradiol. The 17/?-estradiol concen- processing skills. X70 T. C THEOHAKIIXS ri cd

REFERENCES

Amador E, Dorfman L. E. and Wacker W. E. C. (1963) Serum lactic dehydrogenase: an analytical assessment ofcurrent assays. Clin. Chem. 9, 391-399. Askenase P. W. and Van Loveren H. (1983) Delayed-type hypersensitivity: activation of mast cells by antigen-specific T-cell factors initiates the cascade of cellular interactions. Immun. Today 4, 259-264. Baloyannis S. J. and Theoharides T. C. (1982) Mast cells may induce demyelination of rat brain shces in vitro. Clin. Res. 30, 407 (Abstract). Bienenstock J., Tomioka M.. Matsuda H., Stead R. H., Quinonez Cr., Simon G. T., Coughlin M. D. and Denburg J. A. (1987) The role of mast cells m inflammatory processes: evidence for nerve mast cell interactions. Id. Archs appl. Immun. 82, 238-243. 5. Bornstein M. B. and Raine C. S. (1976) The initial structural lesion in serum-induced demyelination in z$,ro. Lab. Inoest. 35, 391-401. 6. Brosnan C. F., Lyman W. D., Tansey F. A. and Carter T. H. (1985) Q uantitation of mast cells in experimental allergic neuritis. .I. Neuropath. exp. Neural. 44, 196-203. I. Brosnan C. F. and Tansey F. A. (1984) Delayed onset of experimental allergic neuritis in rats treated with reserpine. J. Neuropath. exp. Neural. 43, X4-93. 8. Brown A., McFarlin D. E. and Raine C. S (1982) Chronologic neuropathy of relapsing experimental allergic encephalomyelitis in the mouse. Lub. Invr~/. 46, 171&185. 9. Carraway R., Cochrane D. E.. Lansman J. B.. Leeman S. E., Paterson B. M. and Welch H. J. (1982) Neurotensin stimulates exocytotic histamine secretion from rat mast cells and elevates plasma histamine levels. J. F’hysiol. 323, 403-414. 10. Coutts S. M., Nehring R. E. and Jariwala N. U. (1980) Purification of rat peritoneal mast cells: occupation of IgE receptors by IgE prevents loss of the receptors. J. Immun. 124, 2309-2315. Il. Dietsch G. N. and Hinrichs D. J. (1989) The role of mast cells in the elicitation of experimental allergic encephalomyelitis. J. Immun. 142, 1476-1481, 12. Dietsch G. N. and Hinrichs D. J. (1991) Mast cell proteases hberate encephalitogenic fragments from intact myelin. Cell. Immun. 135, M-548. 13. Dimitriadou V., Aubineau P., Taxi J. and Seylaz J. (1987) Ultrastructural evidence for a functional unit between nerve fibers and type II cerebral mast cells in the cerebral vascular wall. Neuroscience 22, 621-630. 14. Dimitriadou V., Buzzi M. G., Moskowitz M. A. and Theoharides T. C. (1991) Trigeminal sensory fiber stimulation induces morphologic changes reflecting secretion in rat dura mast cells. Neuroscience 44, 97-I 12. IS. Dimitriadou V., Buzzi M. G., Theoharides T. C. and Moskowitz M. A. (1992) Ultrastructural evidence for neurogenically mediated changes in blood vessels of the rat dura mater and tongue following antidromic trigeminal stimulation. Neuroscience 48, 187-203. 16. Dimitriadou V., Lambracht-Hall M., Reichler _I. and Theoharides T. C. (1990) Histochemical and ultrastructural characteristics of rat brain perivascular mast cells stimulated with compound 48180 and carbachol. Neuroscience 39, 209-224. 17. Fewtrell C. M. S., Foreman J. C., Jordan C. C., Oehme P., Renner H. and Stewart J. M. (1982) The effects of substance P on histamine and 5-hydroxytryptamine release in the rat. J. Physiol. 330, 393-411. 18. Galli S. J. (1990) New insights into “the riddle of the mast cells”: microcnvironmcntal regulation of mast cell development and phenotypic heterogeneity. Lab. Inwst. 62, S--33. 19. Griffin D. E. and Mendoza Q. P. (1986) Identification of the inflammatory cells present in the central nervous system of normal and mast cell-deficient mice during Sindbis virus encephalitis. Cell. Immun. 97, 454-459. 20. Hammann K. P. and Hopf H. C. (1986) Monocytes constitute the only peripheral blood cell population showing an increased burst activity in multiple sclerosis patients. Inr. Archs Allergy uppl. Zmmun. 81, 230-234. 21. Hauser S. L., Bhan A. K., Gilles F., Kemp M , Kerr C. and Weiner H. L. (1986) Immunohistochemical analysis of the cellular infiltrate in multiple sclerosis lesions. Ann. Neural. 19, 57X%587. 22. Hough L. B., Goldschmidt R. C., Glick S. D. and Padawer J. (1984) Mast cells m rat brain: characterization, localization and histamine content. In Frontiers in Hisramine Research (eds Ganellin C. R. and Schwartz J. C.). pp. 131-140. Pergamon Press, New York. 23. Ibrahim M. Z. (1974) The mast cells of the mammalian central nervous system. Part I. Morphology, distribution and histochemistry. J. neural. Sci. 21, 431-478. 24. Johnson A. R. and Erdiis E. G. (1973) Release of histamine from mast cells hy vasoactive peptides. Proc. Ser. PIP_ Biol. Med. 142, 1252-1256. 25. Johnson D., Seeldrayers P. A. and Weiner H. L. (1988) The role of mast cells in demyelination. I. Myehn proteins are degraded by mast cell proteases and myelin basic protein and P, can stimulate mast cell degranulation. Bruin Res. 444, 195-198. 26. Johnson D., Yasui D. and Seeldrayers P. (1991) An analysis of mast cell frequency in the rodent nervous system: numbers vary between different strams and can be reconstituted in mast cell-deficient mice. /. Neuropath. exp. Neural. 50, 227.-234. 27. Kaplan A. P.. Reddinari S.. Baeza M. and Kuna P. (1991) Histamine releasing factors and cytokine-dependent activation of basophils and mast cells. Adu. lmmun. 50, i37-i60. 28. Kremzner L. T. and Wilson J. B. (1961) A procedure for the determination of histamine. Biochim. bioahvs.. Acfa SO, 364-367. 29. Kriiger P. G., Bo L., Myhr K. M., Karlsen A. E., Taule A., Nyland H. I. and Mork S. (1990) Mast cells and multiple sclerosis: a light and electron microscopic study of mast cells in multiple sclerosis emphasizing staining procedures. Acra neurol. stand. 81, 3 l-36. 30. Lambracht-Hall M., Dimitriadou V. and Theoharides T. C. (1990) Migration of mast cells in the developing rat bram Devl. Brain Res. 56, 151-159. 31. Lambracht-Hall M., Holdridge M.. Dimitriadou V. and Theoharides T. C. (1990) Myelin basic protein (MBP), substance P (SP), and carbachol (C) induce histamine release from rat dura mast cells. Fedn Proc. Fedn Am. Soc.~ exp. Biol. 4, 904 (Abstract) Mast cell secretion and brain myelin changes 871

32. Lambracht-Hall M., Konstantinidou A. D. and Theoharides T. C. (1990) Serotonin release from rat brain mast cells in vitro. Neuroscience 39, 199-207. 33. Metcalfe D. D., Kaliner M. and Donlon M. A. (1981) The mast cell. CRC Crit. R~u. Immun. 3, 23-74. 34. Metzger H., Alcazar G., Hohman R., Kinet J.-P., Pribluda V. and Quart0 R. (1986) The receptor with high affinity for immunoglobulin E. A. Rev. Immun. 4, 4199470. 35. Mokhtarian F. and Griffin D. E. (1984) The role of mast cells in virus-induced intlammation in the murine central nervous system. Ceil. fmmun. 86, 491-500. 36. Newson B., Dahlstrom A., Enerbbk L. and AhIman H. (1983) Suggestive evidence for a direct innervation of mucosal mast cells; an eiectron micro~opic study. ~~roscience IO, 565-570. 37. O&on Y. (1974) Mast cells in plaques of multiple sclerosis. Arta neurot. wand. 50, 611-618. 38. Padawer J. (1968) Uptake of virus by mast cells. J. rer~~~~~~or~e~. Sot. 5, 578 579. 39. Prineas J. W. and Wright R. G. (1978) Macrophages, lymphocytes, and plasma cells in the perivascular compartment in chronic multiple sclerosis. Lub. Invesr. 38, jo9--421. 40. Raine C. S. (1984) Analysis of autoimmune demyelination: its impact upon multiple sclerosis. Lab. hvesr. 50,60X-635. 41. Raine C. S.. Barnett L. B,, Brown A., Behar T. and McFarlin D. E. (1980) Neuropathology of experimental allergic encephalomyelitls in inbred strains of mice. Lab. Invest. 2, 150-157. 42. Reddy E., Sanberg-Wollheim M., Mettus R. V., Ray P. E., DeFreitas E. and Koprowski H. (1989) Amplification and molecular cloning of HTLV-1 sequences from DNA of multiple sclerosis patients. Science 243. 529-533. 43. Rozniecki J. J., Hauser S. L. and Thebharides T. C. (1992) Elevated mast cell-tryptase in cerebrospinal fluid (CSF) of patients with multiple sclerosis (MS). Fedn Proc. Fedn Am. Sots enp. Biof. 93 (Abstract). 44. Sam G. R. (1989) interstitial cystitis: pathophysiology, clinical evaluation and treatment. Ural. Ann. 3, 172-179. 45. Schwartz L. B. (1987) Mediators of human mast cells and human mast ccl1 subsets. Ann. Ariergy 58, 226-235. 46. Seil F. J. (19771 Tissue culture studies of demvelinatine disease: a critical review. Ann. Neural. 2. 3455355. 47. Serafln W. E. and Austen K. F. (I 987) Mediators of immediate hypersensitivity reactions. New Engf. >, Med. 317,30-34. 48. Singer D. E., Moore M. J. and Witlams M. R. (1981) EAE in rat bone marrow chimeras: analvsis of the ccllufar mechanism of BN resistance. J. Immupt.126, 1553-1557. 49. Sobel R. A., Btanchette B. W., Bhan A. K. and Colvin R. B. (1984) The immuno~tholo~y of experimental allergic en~phalomyelitis; quantitative analysis of inflammatory cells in situ. 1. lmmun. 132, 2393-2401. 50. Sugiyama K. (1977) Histamine release from rat mast cells induced by Sendai virus. Nurure 270, 6144615. 51. Theoharides T. C. (1987) Substance P-induced release of brain mast cell nociceptive mediators. Pain 40, 262-ahs. (Abstract). 52. Theoharides T. C. (1990) Mast cells: the immune gate to the brain. Life Sci. 46, 607-617. 53. Theoharides T. C., Baloyannis S. J. and Manolidis L. S. (1991) Activated rat peritoneal mast cells can cause syngeneic brain demyelination in virro. Int. .I. Immunoputh. Pharmac. 4, 137-144 54. Theoharides T. C., Bondy P. K., Tsakalos N. D. and Askenase P. W. (1952) Differential release of serotonin and histamine from mast cells. Nature 297, 229.-231. 55. Theoharides T. C., Drmitriadou V., Massacesi L. and Hauser S. L. (1990) Ultrastructural evidence of mast cell involvement in monkey experimental allergic encephalomyelitis (EAE). In Proc. InternutianalCongress on In&nma?ion, Barcelona, p. 282. 56. Tbeoharides T. C. and Douglas W. W. (1978) Somatostatin induces histamine secretion from rat peritoneal mast cells. E~ldocri~ofogy102, 1637-1640. $7. Th~harides T. C. and Hauser S. L. (1988) Tricyclics inhibit brain mast cell secretion and neuronal demyelination. Proc. rnrernar~~~~~ Workshop on Diagnosiic and T~rape~l~c Triafs in ~ul~~fe Sclerosis (Ahstract). 58. Theoharides T. C., Kops S. K., Bondy P. K. and Askenase P. W. (1985) Differential release of serotomn wtthout comparable histamine under diverse conditions in the rat mast cell. Biochem. Pharmac. 34, 138991398. 59. Theoharides T. C., Lambracht-Hall M., Massacesi L., Dimitriadou V. and Hauser S. L. (1990) Reduced inadence of experimental allergic encephalomyelitis (EAE) with the tricyclic SEP-CH. Eur. J. Pharmac. 183, 626-627 (Abstract). 60. Theoharides T. C., Vliagoftis H. and Hauser S. L. (1989) Myelin basic protein induces mast cell secretion from EAE-susceptible Lewis, but not resistant Sprague-Dawley rats. Proc. 3rd fnrersci. World Canf. fnflamm.. Monte Carlo, p. 256 (Abstract). 61. Toms R., Weiner H. L. and Johnson D. (1990) Identification of IgE-positive cells and mast cells in frozen secttons of multiple sclerosis brains. J. Neuroimmun. 30, 169-177. 62. Vandiik B. and Norrby E. (1989) Paramyxovirus SV5 and multiple sclerosis. Narure 338, 769-771. 63. Van Loveren H., Kops S. K. and Askenase P. W. (1984) Different mechanisms of release of vasoactive amines by mast cells occur in T cell-dependent compared to &E-dependent cutaneous hypersensitivity responses. Eur. J. Imm;n. 14, 40-47. 64. Vliagoftis H., Dimitrjadou V., Boucher W., Romiecki J. J., Correia I., Raam S. and Theoharides T. C. (1992) Estradiol au~ents while tamoxifen inhibits rat mast cell secretion. Jnr. Archs Affergy Immun. 98, 398-409. 65. Vliagoftis H., Dimitriadou V. and Theoharides T. C. (1990) Progesterone triggers selective mast cell secretion of ~-hydroxytryptamine. Ins. Archs AifergJi cipltl. Immkn. 93, 113-l 19. 66. Waksman B. H. (1988) Autoimmunity in demyelinating diseases. Ann. N.Y. Acud. Sri. 540, 4-24. 67. Waksman B. H. (1989) Multiple sclerosis. Relationship to a retrovirus? Nafure 337, 599.

(Accepted 4 June 1993)