Oligodendrocytes Expressing the CC Chemokine CCL21 in ...

Central Nervous System Inflammation and Neurological Disease in Transgenic Mice Expressing the CC Chemokine CCL21 in Oligodendrocytes Shu-Cheng Chen,* Michael W. Leach,1‡ Yuetian Chen,† Xiao-Yan Cai,† Lee Sullivan,* Maria Wiekowski,* B. J. Dovey-Hartman,‡ Albert Zlotnik,2§ and Sergio A. Lira3* To study the biological role of the chemokine ligands CCL19 and CCL21, we generated transgenic mice expressing either gene in oligodendrocytes of the CNS. While all transgenic mice expressing CCL19 in the CNS developed normally, most (18 of 26) of the CCL21 founder mice developed a neurological disease that was characterized by loss of landing reflex, tremor, and ataxia. These neurological signs were observed as early as postnatal day 9 and were associated with weight loss and death during the first 4 wk of life. Microscopic examination of the brain and spinal cord of CCL21 transgenic mice revealed scattered leukocytic infiltrates that consisted primarily of neutrophils and eosinophils. Additional findings included hypomyelination, spongiform myelinopathy with evidence of myelin breakdown, and reactive gliosis. Thus, ectopic expression of the CC chemokine CCL21, but not CCL19, induced a significant inflammatory response in the CNS. However, neither chemokine was sufficient to recruit lymphocytes into the CNS. These observations are in striking contrast to the reported activities of these molecules in vitro and may indicate specific requirements for their biological activity in vivo. The Journal of Immunology, 2002, 168: 1009 –1017.

L

ymphocytes normally enter and survey the CNS constantly in small numbers (1, 2). Their presence in the CNS increases in various diseases, such as multiple sclerosis (3, 4) and viral infections (5), but the mechanisms involved in their trafficking under both physiological and pathological conditions are poorly understood. Candidate molecules mediating lymphocyte trafficking have been recently identified (6, 7). Among these molecules are small, secreted cytokines with chemotactic function known as chemokines (for review see Refs. 8 –10). Evidence supporting a role for chemokines in mediating leukocyte recruitment into the CNS is derived from studies in transgenic mice which demonstrate that influx of neutrophils (11) and macrophages (12) occurs following CNS-specific expression of the chemokines KC and JE. Other studies have documented expression of chemokines in autoimmune and posttraumatic brain models, which are associated with leukocyte infiltration (13–16). Leukocyte recruitment into the CNS has also been demonstrated by intracerebroventricular injection of the chemokine C10 (16). These in vivo models clearly demonstrate that the CNS is permeable to leukocytes, particularly neutrophils and macrophages, when the appropriate chemokines are expressed. Recently, the chemokines CCL19 and CCL21 have been implicated in the migration of lymphocytes in

Departments of *Immunology and †Bioanalysis, Schering-Plough Research Institute, Kenilworth, NJ 07033; ‡Department of Drug Safety and Metabolism, ScheringPlough Research Institute, Lafayette, NJ 07848; and §Department of Immunology, DNAX Research Institute, Palo Alto, CA 94304 Received for publication August 13, 2001. Accepted for publication November 28, 2001. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1

Current address: Abbott Bioresearch Center, Worcester, MA 01605.

2

Current address: Eos Biotechnology, South San Francisco, CA 94080.

3

Address correspondence and reprint requests to Dr. Sergio A. Lira, Department of Immunology, Schering-Plough Research Institute, 2015 Galloping Hill Road, Kenilworth, NJ 07033. E-mail address: [email protected] Copyright © 2002 by The American Association of Immunologists

vitro and in vivo. In this report we study the ability of these chemokines to mediate the influx of lymphocytes into the CNS. CCL21, also known as 6Ckine (17), secondary lymphoid chemokine (18), Exodus-2 (19), and thymus-derived chemotactic agent-4 (20), is a CC chemokine whose amino acid sequence exhibits an unusual pattern of six conserved cysteines, two of which are located in an unique, highly charged, carboxyl-terminal extension (17–20). CCL21 is constitutively expressed at high levels in secondary lymphoid organs, particularly lymph nodes and spleen in both human and mouse (17–20). Expression of CCL21 has also been localized to the high endothelium venules of lymph nodes and Peyer’s patches (21, 22). Recombinant murine CCL21 is chemotactic in vitro for thymocytes, naive T cells, mature dendritic cells, and, at high concentrations, naive B cells, but not for macrophages or neutrophils (17, 18, 20, 21, 23, 24). CCL21 binds and induces calcium flux in cells transfected with CCR7 (25). Binding to CXCR3-transfected cells has also been reported for murine CCL21 (26) but not for human CCL21 (27). Interaction of CCL21 with receptors on lymphocytes is thought to account for the rapid integrin-dependent arrest of lymphocytes rolling under physiological shear (28, 29). CCL19, also known as macrophage-inflammatory protein (MIP)4-3␤ (30) or EBV-induced gene-1 ligand chemokine (31), is another ligand for CCR7. It is expressed in thymus and secondary lymphoid organs (30, 31). CCL19, similar to CCL21, attracts thymocytes, T and B lymphocytes, and mature dendritic cells (24, 32, 33). Both CCL19 and CCL21 are reportedly expressed by stromal cells in the T cell zone of lymph nodes (34). In the mouse genome, CCL19 and CCL21 are encoded by multiple genes. CCL21 is represented by at least two genes encoding for two different forms of CCL21 protein (CCL21a and CCL21b) which differ by one amino acid at position 65 (35, 36). Mutations in CCL19 and CCL21 genes are associated with specific defects in 4 Abbreviations used in this paper: MIP, macrophage-inflammatory protein; MBP, myelin basic protein; GFAP, glial fibrillary acidic protein; P, postnatal day; BBB, blood-brain barrier.

0022-1767/02/$02.00

1010

TRANSGENIC EXPRESSION OF CCL19 AND CCL21 IN THE CNS

lymphocyte trafficking. The mouse strain, paucity of lymph node T cells ( plt), which lacks CCL21a and the only functional copy of CCL19 (34 –36), has deficits in the migration of both T lymphocytes and dendritic cells into lymph nodes (37). In addition, mice lacking CCR7, a receptor for both CCL19 and CCL21, have impaired migration of lymphocytes into secondary lymphoid organs and develop a phenotype that is similar to that presented by the plt mice (38). To determine whether CCL19 or CCL21 could alter lymphocyte influx/trafficking into the CNS, we generated transgenic mice expressing either of these molecules in oligodendrocytes.

Materials and Methods Transgene construction and microinjection Transgenes were generated by cloning the coding region of either a 441-bp murine CCL21b (17) or a 484-bp murine CCL19 (30) cDNA into the XbaISalI sites of the plasmid pMBP, generously provided by Dr. R. A. Lazzarini (Mount Sinai Medical Center, New York, NY). This plasmid contains 1.9 kb (⫺1907 to ⫹36 bp) of the promoter/enhancer of the myelin basic protein (MBP) gene. Additional genomic sequences include splice and polyadenylation signals supplied by exon 6, intron 6, and exon 7 of the human proteolipid gene (39). The resulting transgenes are referred to as MBPCCL21 and MBPCCL19, respectively. Both transgenes were isolated from the plasmids by restriction digest with NotI. Separation of the transgenes from vector sequences was accomplished by zonal sucrose gradient centrifugation as described (40). Fractions containing the transgene were pooled, microcentrifuged through Microcon-100 filters (Amicon, Beverly, MA), and washed five times with microinjection buffer (5 mM Tris-HCl (pH 7.4), 5 mM NaCl, 0.1 mM EDTA).

Generation of transgenic mice Each transgene was resuspended in microinjection buffer (5 mM Tris-HCl (pH 7.4), 5 mM NaCl, 0.1 mM EDTA) to a final concentration of 1–5 ng/␮l, microinjected into (C57BL/6J ⫻ DBA/2)F2 (The Jackson Laboratory, Bar Harbor, ME) eggs, and transferred into oviducts of ICR (Charles River Breeding Laboratories, Wilmington, MA) foster mothers, according to published procedures (41). At 10 days after birth, a piece of tail from the resulting animals was clipped for DNA analysis. Identification of transgenic founders was conducted by PCR analysis, as previously described (42). The following primer pairs were used to identify the transgenic mice by amplification of mouse tail DNA: for MBPCCL21, 5⬘-gcttcagaccatc caagaagacc-3⬘ and 5⬘-ttgcaccccttggagccctttcct-3⬘; for MBPCCL19, 5⬘-gcg gaagactgctgcctgtctgtg-3⬘ and 5⬘-gcctatttacttgccaagatcattc-3⬘. The endogenous ZP3 gene (5⬘-cagctctacatcacctgcca-3 and 5⬘- cactgggaagagacactcag3⬘) was used as an internal control for the amplification reactions. These primers amplify a 443-bp segment of the MBPCCL21 transgene, a 380-bp segment of the MBPCCL19 transgene, and a 511-bp segment of the ZP3 gene. PCR conditions were 95°C for 30 s, 60°C for 30 s, and 72°C for 60 s for 30 cycles. The resulting transgenic animals were kept under specific pathogen-free conditions. All animal experiments were performed following the guidelines of the Schering-Plough Animal Care and Use Committee.

RNA analysis RNA was extracted from tissues using Ultraspec RNA (Biotecx Laboratories, Houston, TX) following specifications from the manufacturer. Total RNA (20 ␮g) was denatured, separated by gel electrophoresis, and blotted onto GeneScreen membrane (NEN, Boston, MA). A 441-bp fragment of the murine CCL21 cDNA and a 484-bp fragment of the murine CCL19 cDNA were radiolabeled using a random-priming DNA-labeling kit (Stratagene, La Jolla, CA) and used as probes in the hybridization of Northern blots.

Histological analysis After euthanasia, tissues were either fresh-frozen with freezing medium for cryosection or fixed by immersion in 10% phosphate-buffered formalin and then processed for paraffin sections. Tissues for light microscopic examination were routinely processed, sectioned at 5 ␮m, and stained with H&E. For immunohistochemistry, fresh-frozen sections were first fixed with icecold acetone for 10 min and air-dried. Paraffin sections were deparaffined and rehydrated to 1⫻ PBS before use. The sections were then incubated sequentially with biotin solution (Vector Laboratories, Burlingame, CA), avidin solution (Vector Laboratories), and 10% normal goat or rabbit se-

rum (Vector Laboratories) for 30 min each. Primary Abs were incubated for either 1 h at room temperature or overnight at 4°C at dilutions as suggested by the manufacturer. Incubation with the appropriate secondary Abs was conducted for 30 min followed by immersion in avidin-biotinHRP complex (Vectastain Elite ABC kit; Vector Laboratories) for 30 min. The tissue sections were finally stained with diaminobenzidine (Vector Laboratories) and counterstained with hematoxylin. Primary Abs used are as follows: anti-B220, anti-CD3e, anti-CD4, anti-TCR␤ (BD PharMingen, San Diego, CA), anti-CCL19, anti-CCL21 (R&D Systems, Minneapolis, MN), anti-F4/80 (Serotec, Raleigh, NC), and anti-glial fibrillary acidic protein (GFAP; DAKO, Carpinteria, CA). To label apoptotic cells, an In Situ Cell Death Detection kit (Roche Molecular Biochemicals, Indianapolis, IN) was used for immunohistochemical staining of paraffin sections following the procedure described by the manufacturer. For Luxol fast blue staining of the myelin, paraffin sections were dewaxed and rehydrated through 95% ethanol. The sections were then stained with 0.1% Luxol fast blue solution overnight at 60°C followed by brief differentiation with 0.05% LiCl solution.

Electron microscopy Animals were perfused with McDowell-Trump solution in 0.1 M phosphate buffer. After fixation, samples of perfused brains were collected from the cerebrum, cerebellum, medulla, and spinal cord and were postfixed in sodium cacodylate-buffered 2.5% glutaraldehyde, then further fixed in 2% buffered osmium tetroxide. Samples were then rinsed, dehydrated through a graded series of ethanol, cleared in two changes of propylene oxide, and embedded in resin. Sections were cut at 60 –90 nm, stained with uranyl acetate and lead citrate, and examined on a Philips CM10 transmission electron microscope.

ELISA The concentration of cytokines and chemokines in serum and tissue homogenates was measured using ELISA kits purchased from commercial sources and run according to the manufacturers’ instructions. Tissue samples were placed in 0.5–1 ml of 20 mM Tris/1 mM EDTA buffer containing a protease inhibitor mixture (Roche Molecular Biochemicals). Samples were then homogenized and centrifuged, and the supernatants were removed for cytokine/chemokine assays. The protein concentration was determined using the Bio-Rad protein assay. ELISA kits for murine MIP-1␣ (sensitivity ⬍ 1.5 pg/ml), murine MIP-2 (sensitivity ⬍ 1.5 pg/ml), murine IFN-␥ (sensitivity ⬍ 2 pg/ml), murine TNF-␣ (sensitivity ⬍ 5.1 pg/ml), murine KC (sensitivity ⬍ 2 pg/ml), murine IL-2 (sensitivity ⬍ 3 pg/ml), and murine eotaxin (sensitivity ⬍ 3 pg/ml) were purchased from R&D Systems. Murine IL-1␣ ELISA kit (sensitivity ⬍ 6 pg/ml) was purchased from Endogen (Cambridge, MA).

Results Generation of transgenic mice expressing CCL19 or CCL21 in the CNS The murine MBP promoter was used to direct transgene expression to oligodendrocytes (39) (Fig. 1A). A total of seven founders carrying the MBPCCL19 transgene and 26 founders carrying the MBPCCL21 transgene were generated. These transgenic mice are referred to as MBPCCL19 and MBPCCL21 mice, respectively. Transgene expression was initially assessed by Northern hybridization. As shown in Fig. 1, several lines of either MBPCCL19 (Fig. 1B) or MBPCCL21 (Fig. 1C) showed transgene expression in their brain tissues. Consistent with previous reports (12, 43), transgene expression driven by the 2-kb murine MBP promoter was developmentally regulated, peaking at 2–3 wk postnatally (data not shown). Expression of CCL19 or CCL21 protein in the CNS of the transgenic mice was analyzed by immunohistochemical staining. Intense cellular staining was predominantly detected in the white matter of the CNS of transgenic but not control mice (Fig. 2). Neurological phenotype in MBPCCL21 but not in MBPCCL19 transgenic mice Five of seven independent lines of MBPCCL19 mice showed transgene expression. All mice in these five lines developed to maturity without showing any abnormal clinical signs (n ⫽ 174).

The Journal of Immunology

1011

FIGURE 1. MBPCCL19 and MBPCCL21 transgenes and analysis of their expression in the CNS. A, Diagram of the MBPCCL21 and MBPCCL19 transgenes. The MBP promoter drives the expression of either murine CCL19 (upper panel) or CCL21b (lower panel) cDNA fused to the genomic DNA of the human proteolipid gene (hPLP). This genomic human proteolipid gene DNA provides introns, exons, and polyadenylation sequences. Translation of the hybrid mRNA terminates at the stop codon of the cDNAs. B and C, Northern hybridization. Brain RNAs isolated from wild-type, MBPCCL19 (B), or MBPCCL21 (C) mice were used to examine the transgene expression. cDNAs of CCL19 (B) or CCL21 (C) were used as probes in the Northern hybridization. The lane numbers indicate the transgenic line numbers.

In contrast, 18 of 26 MBPCCL21 founders (or their progeny) developed a distinct neurological phenotype, characterized by loss of the righting reflex, tonic tremor during standing or walking, and ataxia. Most of the MBPCCL21 founders or offspring that developed neurological signs had reduced body weight and were either found dead within the first 4 wk of life or were infertile. Macroscopic and microscopic examination of non-neural tissues revealed no obvious cause of death except for wasting. To investigate whether the neurological signs were associated with expression of the MBPCCL21 transgene, we examined the transgene expression in the brain of the MBPCCL21 animals by Northern blot analysis. Transgene expression was detected in the affected founders and correlated with the intensity of the signs. Animals showing high expression of the transgene showed more severe clinical abnormalities, whereas those showing low levels of expression demonstrated a nearly normal righting reflex and were generally more active than those expressing higher levels of the

FIGURE 2. Immunohistochemical analysis of CCL19 and CCL21 expression in the CNS. Shown are fresh-frozen (A and B) or paraffin (C and D) sections from wild-type (A and C), MBPCCL19 (B), and MBPCCL21 (D) brains stained with anti-CCL19 (A and B) or anti-CCL21 (C and D) Abs. No CCL19 or CCL21 staining was detected in wild-type control brains. cc, Corpus callosum; CA, pyramidal cell layer of the hippocampus. Bar ⫽ 80 ␮m.

transgene. Founders that had no transgene expression did not show neurological signs (data not shown). Due to poor health and premature death it was difficult to generate lines from most of the MBPCCL21 founders. However, transgenic offspring were derived from four founders that showed little or no disease (nos. 171, 296, 299, and 322). Interestingly, animals derived from founders 299 and 322 were mildly affected clinically and expressed very low levels of the transgene (Fig. 1C). Founders 171 and 296 had mild disease but were most likely mosaic, because their progeny were highly affected and expressed high levels of the transgene. The studies reported in this work were mostly conducted with animals derived from founders 171 and 296. In progeny from both 171 and 296 founders, the first signs of neurological deficit were detected at postnatal day (P) 9. Similar to what was observed with most founders, these progeny also presented with severe neurological disease and died within the first month of life, preventing further expansion of the lines.

1012


FIGURE 3. Absence of leukocytic infiltrates in the CNS of MBPCCL19 mice. H&E-stained paraffin sections of the medullar and overlying cerebellum (A) and spinal cord (B) from MBPCCL19 transgenic mice showed no leukocyte infiltration. Bar ⫽ 400 ␮m in A and 80 ␮m in B.

Leukocyte infiltration in the CNS of MBPCCL21 mice Routine histologic analysis of the H&E-stained paraffin sections was used to examine the CNS of transgenic mice at the light microscopic level. Brain and spinal cord sections from MBPCCL19 (five independent lines) and wild-type mice did not show any abnormalities. There were no inflammatory cell infiltrates, and both the gray and white matter appeared normal (Fig. 3). In contrast, in MBPCCL21 mice (15 founders or their offspring), histologic changes were evident in the medulla, white matter of the cerebellum, and spinal cord, and included inflammatory cell infiltrates (Fig. 4). Inflammatory infiltrates were first detected at P11 and consisted of minimal to mild infiltrates of neutrophils and/or eosinophils, accompanied by rare mononuclear cells. Leukocytes were found in small clusters of up to 15–20 cells in the parenchyma (Fig. 4A), as individual scattered cells in the parenchyma (Fig. 4B), or in small clusters adjacent to blood vessels (Fig. 4C). Gemistocytic astrocytes were also seen in a few mice.

Ultrastructurally, significant changes were present in the cerebrum, cerebellum, medulla, and spinal cord. Perivascular infiltrates of neutrophils and eosinophils were present, as were individual neutrophils and eosinophils scattered throughout the parenchyma (Fig. 4D). A few macrophages were also seen adjacent to blood vessels and scattered throughout the parenchyma. Rare necrotic cells were also seen in the brain. Hypertrophic astrocytes containing bundles of intermediate filaments were sometimes observed. Astrocyte processes, especially adjacent to blood vessels, were also swollen. To further investigate the cellular infiltrates in the CNS of both MBPCCL19 and MBPCCL21, immunohistochemical staining with several Abs was conducted. We were not able to detect B220positive cells within the CNS parenchyma of either MBPCCL19 or MBPCCL21 transgenic mice (data not shown). Rare CD4- and CD3e-positive cells were detected in the parenchyma and perimeningeal areas in the brains of MBPCCL21 mice (Fig. 4E). Therefore, no significant infiltration of lymphocytes, either T or B

FIGURE 4. Neutrophil and eosinophil infiltration in the CNS of MBPCCL21 transgenic mice. A, A cluster of neutrophils and eosinophils in the brain. Bar ⫽ 15 ␮m. B, Neutrophils scattered throughout the neuropil of the medulla. Bar ⫽ 25 ␮m. C, Perivascular infiltrate of eosinophils in the medulla. Bar ⫽ 10 ␮m. D, Transmission electron micrograph of the brain of a MBPCCL21 mouse. There is a prominent perivascular eosinophilic infiltrate. Note the lobulated nuclei and numerous cytoplasmic granules. Bar ⫽ 1.7 ␮m. D, inset, Higher magnification of granules, showing elongated granules with a central core characteristic of eosinophils. E, Anti-CD4 immunohistochemical staining of a MBPCCL21 brain. No CD4-positive cells were detected in an area with significant inflammatory infiltrates. Bar ⫽ 93 ␮m. Samples were derived from 2to 3-wk-old MBPCCL21 animals.

The Journal of Immunology cells, was found in the CNS of MBPCCL19 and MBPCCL21 transgenic animals. Gliosis in the CNS of MBPCCL21 mice Inflammatory conditions of the CNS are often associated with changes in the activation of microglia. To determine whether there were changes in this cell population, we used an Ab against the activation marker F4/80 and studied the CNS of wild-type and MBPCCL21 mice between P8 and P18. In wild-type mice, a few lightly stained cells expressing F4/80 were detected throughout the brain and spinal cord parenchyma (Fig. 5A). In contrast, numerous F4/80-positive cells with morphological characteristics of ramified microglia/macrophages were observed in the spinal cord and brain of the MBPCCL21 mice. In the spinal cord, the F4/80-positive cells were first observed at P10 and the majority of these cells were located in the white matter. In animals at P11 and older, F4/80positive cells were detected in both gray and white matter of the brain and spinal cord (Fig. 5B). Both the intensity of staining and the number of F4/80 positive cells increased with age. The expression of F4/80 closely mirrored the levels of transgene expression as determined by anti-CCL21 immunostaining and Northern analysis. To determine whether astrocytes were activated in the MBPCCL21 mice, an anti-GFAP Ab was used for immunohistochemical staining. A marked increase in GFAP staining was found as early as P8 in the gray matter of the spinal cord. Increased numbers of GFAPpositive cells were observed in the medulla at the age of P10 and older. Like F4/80 staining, an age-related increase in the intensity

FIGURE 5. Gliosis in the MBPCCL21 CNS. Paraffin-embedded spinal cord sections of wild-type (A and C) or MBPCCL21 mice (B and D) were stained with F4/80 (A and B) or GFAP (C and D) Abs. Bar ⫽ 15 ␮m in A and B and 20 ␮m in C and D.

1013 of staining and numbers of GFAP-positive cells was also observed. At P17 GFAP staining could be seen in several areas of the CNS, particularly in the medulla and in deep cerebellar nuclei. As shown in Fig. 5, C and D, very intense GFAP staining was observed in both the gray and white matter of the spinal cord as compared with the wild-type control. The temporal and spatial pattern of GFAP expression was similar to that of F4/80. Taken together these results indicate that expression of CCL21 in the CNS resulted in significant gliosis. Expression of chemokines and cytokines in the CNS of the MBPCCL21 mice Neutrophils and eosinophils were the primary infiltrating cells in the CNS of the MBPCCL21 mice. However, CCL21 has not been reported to promote chemotaxis of these cell populations (Refs. 17–20 and our unpublished results), suggesting that indirect mechanisms accounted for their presence in the CNS. To identify these mechanisms, we evaluated the levels of chemokines and cytokines in the brains of the transgenic mice. A representative series of four wild-type and 14 transgenic brain homogenates were tested for the expression of the cytokines IL-1␣, IL-4, IL-5, TNF-␣, and IFN-␥. We did not detect expression of IL-4, IL-5, or IFN-␥ in wild-type or transgenic brains (data not shown). However, elevated concentrations of IL-1␣ and TNF-␣ were observed in one third of the transgenic brain homogenates, but not in serum, suggesting a localized rather than systemic inflammatory process. Interestingly, a consistent and significant increase was observed in the levels of the

1014


FIGURE 6. Elevated levels of chemokines in the MBPCCL21 brain homogenates. Bars represent average ⫾ SE. ⴱ, p ⬍ 0.05.

chemokines KC, eotaxin, and MIP-1␣ in the transgenic brain homogenates (Fig. 6).

with myelin alterations appeared normal. A few macrophages containing lipid and myelin debris were adjacent to blood vessels and were scattered throughout the parenchyma, consistent with breakdown and phagocytosis of myelin. Luxol fast blue staining for myelin demonstrated a decrease in the intensity of staining in brain and spinal cord of MBPCCL21 mice, compared with wild-type, at about P14 and older ages (Fig. 8, A and B), suggesting hypomyelination. The white matter of the spinal cord, cerebellum, and medulla were most severely affected. TUNEL staining was used next to investigate whether the apoptosis of oligodendrocytes could account for the hypomyelination observed in MBPCCL21 mice. Brain sections taken from mice P14 and older showed that the number of apoptotic cells in the transgenic mice was slightly higher than in their littermates (data not shown). However, the majority of these apoptotic cells were not associated with the white matter, where hypomyelination was most significant. Therefore, oligodendrocyte cell death may not be the primary factor causing hypomyelination.

Spongiosis and hypomyelination in the CNS of MBPCCL21 mice

Discussion

In addition to leukocyte infiltration and gliosis, mild spongiosis was also observed in the cerebellum, medulla, and spinal cord of P11 and older MBPCCL21 mice (Fig. 7A). Inflammatory cells were often, but not always, found in areas with spongiosis. In the brain, spongiosis was frequently detected in areas associated with control of motor function and balance, such as deep cerebellar nuclei and vestibular nuclei. Ultrastructurally, the spongiosis observed microscopically correlated with a spongiform myelinopathy (Fig. 7, B and C). Myelin sheaths contained vacuoles and were split in the intraperiod line. Defects in the myelin sheaths often contained irregular whorls of myelin debris. On occasion, axons were dilated and sometimes contained clusters of organelles. However, most axons

Our study was designed to investigate the ability of two lymphocyte chemoattractant chemokines (CCL19 and CCL21) to promote influx of lymphocytes into the CNS. To this end, we expressed these transgenes in the oligodendrocytes. This approach has been used to investigate the function of other chemokines (44). In this study we show that constitutive expression of the chemokine CCL21 in the CNS leads to the development of a severe neuropathological condition and death in the majority of the animals. The earliest abnormality in the CNS of these transgenic mice was astrocyte activation, as indicated by increased expression of the glial cell marker GFAP in the spinal cord at P8. The up-regulation of GFAP preceded the onset of neurological signs, which were first detected at P9 –P10, and the onset of neurological signs preceded

FIGURE 7. Spongiosis and myelination defects in MBPCCL21 brains. A, Spongiosis in the neuropil below the cerebellum (molecular layer of cerebellum at upper left). Myelin can be seen within some of the vacuoles, sometimes appearing to encircle axons. Bar ⫽ 25 ␮m. Transmission electron micrographs of the brain of a MBPCCL21 mouse show myelination defects (B and C). B, Vacuoles within myelin sheaths. Note the axons (arrows) appear normal. Several other myelinated axons that lack vacuolation are also present. Bar ⫽ 1.3 ␮m. C, Higher magnification from a different area, showing myelin splitting (arrow) at the intraperiod line. A longitudinal section of the axon (ⴱ) is on the left, and a portion of an oligodendrocyte is in the upper right. There is myelin debris (arrowhead) in the interlamellar space. Bar ⫽ 0.19 nm.

The Journal of Immunology

1015

FIGURE 8. Luxol fast blue staining of MBPCCL21 spinal cord. Paraffin-embedded spinal cord sections of wild-type or MBPCCL21 mice were stained with Luxol fast blue. A, Dense blue staining of normal myelin in the white matter (ⴱ) of wild-type mice. B, Note the loss of blue staining associated with the white matter (ⴱ) of MBPCCL21 mice. Arrows indicate the boundary between gray and white matter. Bar ⫽ 20 ␮m.

the activation of microglia, spongiosis, and inflammatory infiltrates, first detected at P10 –P12. In contrast, animals expressing the closely related chemokine CCL19 did not show any phenotypic abnormality. The phenotypic differences observed between MBPCCL21 and MBPCCL19 transgenic mice are very intriguing because these two chemokines are not only functionally related but also can bind to the same two cell surface receptors, CCR7 and CCR11 (25, 31, 45, 46). Because expression of CCL19 did not lead to any neurological phenotype, it is reasonable to suspect that the neurological phenotype observed in the MBPCCL21 mice could be due to the interaction of CCL21 with a receptor other than CCR7 and CCR11 in the CNS. One such receptor could be CXCR3. Murine CCL21, but not murine CCL19 or human CCL21, binds to CXCR3 (26, 27). Furthermore, expression of CXCR3, but not CCR7, has been detected in ischemic mouse brain, cultured microglia, cultured primary astrocytes, and normal and diseased human CNS (47– 49). CCL21 is not expressed normally in the CNS, but its expression has been reported in ischemic brain (48). It has also been shown that CCL21 induces chemotaxis and intracellular calcium mobilization in cultured microglia, and that these effects can be crossdesensitized by a ligand for CXCR3, CXCL10 (48). These observations have led to the suggestion that CCL21 and CXCR3 may mediate neuron-glia interactions during disease conditions. Thus, CCL21 expressed by oligodendrocytes may have interacted with CXCR3 expressed in astrocytes and/or microglia, inducing gliosis and production of inflammatory mediators including cytokines and chemokines, which in turn may have facilitated the influx of inflammatory cells into the CNS. Indeed, we have observed production of inflammatory cytokines and of neutrophil and eosinophil chemoattractant chemokines in the MBPCCL21 brains (Fig. 6). The expression of these chemokines may have been important to promote the influx of eosinophils and neutrophils into the CNS. Influx of these leukocytes into the CNS may have also contributed to the spongiosis and hypomyelination observed in the MBPCCL21 mice. Eosinophils, for instance, have been shown to contain a neurotoxin that causes spongiosis in the white matter of the cerebellum, medulla, and spinal cord following intracerebral or intrathecal injection (50, 51). Moreover, eosinophils have been found in the spinal cord of mice with experimental autoimmune encephalomyelitis and are suspected to play a role in nervous system damage observed in experimental autoimmune encephalomyelitis (52, 53). Alternatively, expression of CCL21 may have induced gliosis and disease by a receptor-independent mechanism, such as direct toxicity of oligodendrocytes (54 – 61). The paradoxical lack of lymphocyte infiltration in tissues expressing potent lymphocyte chemoattractants is arguably the most provocative finding of this study. We first considered the possibility that the activation state of the endothelium could have influ-

enced entry of lymphocytes responding to the CCL19 or CCL21 produced in the CNS. Lymphocyte entry into the CNS has been shown to depend on the activation of the endothelium by several cytokines (62). IFN-␥ and TNF-␣ have been shown to activate brain endothelium in vitro (63), and transgenic mice expressing IFN-␥ and TNF-␣ present lymphocytic infiltrates (60, 64). Interestingly, despite elevated levels of TNF-␣ in some of the MBPCCL21 brains, we failed to detect lymphocytes in the CNS parenchyma or in the perivascular space. Other parameters of endothelial function, such as the permeability of the blood-brain barrier (BBB) were also considered, because the integrity of the BBB has been shown to be yet another parameter regulating entry of the lymphocytes into the CNS (65). Immunohistochemical staining with an anti-mouse IgG Ab of the MBPCCL21 brains showed that the BBB was disrupted in areas heavily infiltrated with PMNs (data not shown), but no lymphocytes were detected in these areas. Finally, we considered the possibility that high levels of CCL19 or CCL21 may have prevented lymphocyte influx. The current paradigm postulates that leukocytes migrate toward a chemokine gradient and that high chemokine concentrations prevent further migration (for review, see Ref. 9). Thus, high levels of CCL19 or CCL21 in the transgenic brains could actually have prevented, rather than promoted, lymphocyte migration. However, no lymphocytes were observed in the CNS of transgenic mice expressing low levels of CCL19 or CCL21. Thus, it is unlikely that high levels of CCL19 or CCL21 explain the lack of lymphocyte infiltration in these models. In summary, expression of the T cell chemoattractant chemokine CCL21 in the CNS led to the development of a striking neurological phenotype and premature death. The pathological findings in the CNS included reactive gliosis, spongiosis, hypomyelination, and parenchymal infiltration by polymorphonuclear cells but, surprisingly, no lymphocytic infiltration. Lack of lymphocyte infiltration was also observed in mice expressing the closely related molecule CCL19. In light of these findings, we conclude that 1) CCL21 is not sufficient to promote lymphocyte influx into the CNS under basal conditions (as can be appreciated by the analysis of the transgenic mice expressing low levels of CCL21), 2) CCL21 is not sufficient to promote lymphocyte influx into the CNS under inflammatory conditions (as can be appreciated by the analysis of the transgenic mice expressing high levels of CCL21), and 3) the closely related chemokine CCL19 does not promote lymphocyte recruitment into the CNS. Fan et al. (66) have recently shown that ectopic expression of CCL21 in the pancreas leads to the formation of lymphoid structures. We have also found that expression of CCL21 in the pancreas induces recruitment of lymphocytes and formation of lymph node-like structures, but that this effect cannot be reproduced when CCL21 is overexpressed in the skin (67). Taken together these results suggest that CCL21 may

1016


require a tissue-specific environment to induce lymphocyte recruitment and accumulation. These requirements may include the expression of specific adhesion molecules, or cofactors which facilitate lymphocyte recruitment and lymphoid neogenesis. The nature of these requirements is unknown at present, but they are clearly satisfied in lymphoid tissues and in transgenic pancreas when CCL21 is ectopically expressed. It will be interesting to define whether ectopic expression of CCL19 in pancreas will also result in lymphocyte accumulation and lymphoid neogenesis. In conclusion, the results reported in this work and those by Chen et al. (67) reveal a novel finding in chemokine biology, namely that there are tissue-specific requirements for lymphocyte recruitment induced by a single chemokine. Defining the nature of these requirements will hopefully contribute to a better understanding of the role of chemokines in leukocyte trafficking and lymphoid neogenesis.

Acknowledgments

20.

21.

22.

23.

24.

25.

We thank Dr. Robert Lazzarini for the pMBP vector; Carl Gommoll, Linda Hamilton, David Kinsley, Margaret Monahan, Lisa Tardelli, Galya Vassileva, Channa Young, Petronio Zalamea, Wanda Sharif, Faith Solinger, Brian Lee, and the staff of Veterinary Sciences (Schering-Plough Research Institute) for their excellent technical support; and Drs. Shelby Umland and James Jakway for discussion. We also thank Drs. Joseph Hedrick, Jonathan Sedgwick, and Leslie McEvoy for their comments on this manuscript; and Drs. Bernie Jortner, Andrew S. Fix, and Satwant Narula for their support and encouragement for this project.

28.

References

29.

1. Wekerle, H., C. Linington, H. Lassmann, and R. Meyermann. 1986. Cellular immune reactivity within the CNS. TINS 9:271. 2. Hickey, W. F., B. L. Hsu, and H. Kimura. 1991. T-lymphocyte entry into the central nervous system. J. Neurosci. Res. 28:254. 3. Hafler, D. A., and H. L. Weiner. 1995. Immunologic mechanisms and therapy in multiple sclerosis. Immunol. Rev. 144:75. 4. Storch, M., and H. Lassmann. 1997. Pathology and pathogenesis of demyelinating diseases. Curr. Opin. Neurol. 10:186. 5. Griffin, D. E., B. Levine, W. R. Tyor, and D. N. Irani. 1992. The immune response in viral encephalitis. Semin. Immunol. 4:111. 6. Springer, T. A. 1994. Traffic signals for lymphocyte recirculation and leukocyte emigration: the multistep paradigm. Cell 76:301. 7. Butcher, E. C., and L. J. Picker. 1996. Lymphocyte homing and homeostasis. Science 272:60. 8. Hedrick, J. A., and A. Zlotnik. 1999. Chemokines and chemokine receptors in T-cell development. Chem. Immunol. 72:57. 9. Rollins, B. J. 1997. Chemokines. Blood 90:909. 10. Zlotnik, A., J. Morales, and J. A. Hedrick. 1999. Recent advances in chemokines and chemokine receptors. Crit. Rev. Immunol. 19:1. 11. Tani, M., M. E. Fuentes, J. W. Peterson, B. D. Trapp, S. K. Durham, J. K. Loy, R. Bravo, R. M. Ransohoff, and S. A. Lira. 1996. Neutrophil infiltration, glial reaction, and neurological disease in transgenic mice expressing the chemokine N51/KC in oligodendrocytes. J. Clin. Invest. 98:529. 12. Fuentes, M. E., S. K. Durham, M. R. Swerdel, A. C. Lewin, D. S. Barton, J. R. Megill, R. Bravo, and S. A. Lira. 1995. Controlled recruitment of monocytes and macrophages to specific organs through transgenic expression of monocyte chemoattractant protein-1. J. Immunol. 155:5769. 13. Godiska, R., D. Chantry, G. Dietsch, and P. Gray. 1995. Chemokine expression in murine experimental allergic encephalomyelitis. J. Neuroimmunol. 58:167. 14. Ransohoff, R. M., A. Glabinski, and M. Tani. 1996. Chemokines in immunemediated inflammation of the central nervous system. Cytokine Growth Factor Rev. 7:35. 15. Lane, T. E., V. C. Asensio, N. Yu, A. D. Paoletti, I. L. Campbell, and M. J. Buchmeier. 1998. Dynamic regulation of ␣- and ␤-chemokine expression in the central nervous system during mouse hepatitis virus-induced demyelinating disease. J. Immunol. 160:970. 16. Asensio, V. C., S. Lassmann, A. Pagenstecher, S. C. Steffensen, S. J. Henriksen, and I. L. Campbell. 1999. C10 is a novel chemokine expressed in experimental inflammatory demyelinating disorders that promotes recruitment of macrophages to the central nervous system. Am. J. Pathol. 154:1181. 17. Hedrick, J. A., and A. Zlotnik. 1997. Identification and characterization of a novel ␤ chemokine containing six conserved cysteines. J. Immunol. 159:1589. 18. Nagira, M., T. Imai, K. Hieshima, J. Kusuda, M. Ridanpaa, S. Takagi, M. Nishimura, M. Kakizaki, H. Nomiyama, and O. Yoshie. 1997. Molecular cloning of a novel human CC chemokine secondary lymphoid-tissue chemokine that is a potent chemoattractant for lymphocytes and mapped to chromosome 9p13. J. Biol. Chem. 272:19518. 19. Hromas, R., C. H. Kim, M. Klemsz, M. Krathwohl, K. Fife, S. Cooper, C. Schnizlein-Bick, and H. E. Broxmeyer. 1997. Isolation and characterization of

26.

27.

30.

31.

32.

33. 34.

35.

36.

37.

38.

39.

40. 41. 42.

43.

44.

Exodus-2, a novel C-C chemokine with a unique 37-amino acid carboxyl-terminal extension. J. Immunol. 159:2554. Tanabe, S., Z. Lu, Y. Luo, E. J. Quackenbush, M. A. Berman, L. A. Collins-Racie, S. Mi, C. Reilly, D. Lo, K. A. Jacobs, and M. E. Dorf. 1997. Identification of a new mouse ␤-chemokine, thymus-derived chemotactic agent 4, with activity on T lymphocytes and mesangial cells. J. Immunol. 159:5671. Gunn, M. D., K. Tangemann, C. Tam, J. G. Cyster, S. D. Rosen, and L. T. Williams. 1998. A chemokine expressed in lymphoid high endothelial venules promotes the adhesion and chemotaxis of naive T lymphocytes. Proc. Natl. Acad. Sci. USA 95:258. Warnock, R. A., J. J. Campbell, M. E. Dorf, A. Matsuzawa, L. M. McEvoy, and E. C. Butcher. 2000. The role of chemokines in the microenvironmental control of T versus B cell arrest in Peyer’s patch high endothelial venules. J. Exp. Med. 191:77. Nagira, M., T. Imai, R. Yoshida, S. Takagi, M. Iwasaki, M. Baba, Y. Tabira, J. Akagi, H. Nomiyama, and O. Yoshie. 1998. A lymphocyte-specific CC chemokine, secondary lymphoid tissue chemokine (SLC), is a highly efficient chemoattractant for B cells and activated T cells. Eur. J. Immunol. 28:1516. Kellermann, S.-A., S. Hudak, E. R. Oldham, Y.-J. Liu, and L. M. McEvoy. 1999. The CC chemokine receptor-7 ligands 6Ckine and macrophage inflammatory protein-3␤ are potent chemoattractants for in vitro- and in vivo-derived dendritic cells. J. Immunol. 162:3859. Campbell, J. J., E. P. Bowman, K. Murphy, K. R. Youngman, M. A. Siani, D. A. Thompson, L. Wu, A. Zlotnik, and E. C. Butcher. 1998. 6-C-kine (SLC), a lymphocyte adhesion-triggering chemokine expressed by high endothelium, is an agonist for the MIP-3␤ receptor CCR7. J. Cell Biol. 141:1053. Soto, H., W. Wang, R. W. Strieter, N. G. Copeland, D. J. Gilbert, N. A. Jenkins, J. Hedrick, and A. Zlotnik. 1998. The CC chemokine 6Ckine binds the CXC chemokine receptor CXCR3. Proc. Natl. Acad. Sci. USA 95:8205. Jenh, C.-H., M. A. Cox, H. Kaminski, M. Zhang, H. Byrnes, J. Fine, D. Lundell, C.-C. Chou, S. K. Narula, and P. J. Zavodny. 1999. Species specificity of the CC chemokine 6Ckine signaling through the CXC chemokine receptor CXCR3: human 6Ckine is not a ligand for the human or mouse CXCR3 receptors. J. Immunol. 162:3765. Campbell, J. J., J. Hedrick, A. Zlotnik, M. A. Siani, D. A. Thompson, and E. C. Butcher. 1998. Chemokines and the arrest of lymphocytes rolling under flow conditions. Science 279:381. Pachynski, R. K., S. W. Wu, M. D. Gunn, and D. J. Erle. 1998. Secondary lymphoid-tissue chemokine (SLC) stimulates integrin ␣4␤7-mediated adhesion of lymphocytes to mucosal addressin cell adhesion molecule-1 (MAdCAM-1) under flow. J. Immunol. 161:952. Rossi, D. L., A. P. Vicari, K. Franz-Bacon, T. K. McClanahan, and A. Zlotnik. 1997. Identification through bioinformatics of two new macrophage proinflammatory human chemokines, MIP-3a and MIP-3␤. J. Immunol. 158:1033. Yoshida, R., T. Imai, K. Hieshima, J. Kusuda, M. Baba, M. Kitaura, M. Nishimura, M. Kakizaki, H. Nomiyama, and O. Yoshie. 1997. Molecular cloning of a novel human CC chemokine EBI1-ligand chemokine that is a specific functional ligand for EBI1, CCR7. J. Biol. Chem. 272:13803. Kim, C. H., L. M. Pelus, J. R. White, E. Applebaum, K. Johanson, and H. E. Broxmeyer. 1998. CK␤-11/macrophage inflammatory protein-3␤/EBI1-ligand chemokine is an efficacious chemoattractant for T and B cells. J. Immunol. 160:2418. Cyster, J. G. 1999. Chemokines and the homing of dendritic cells to the T cell areas of lymphoid organs. J. Exp. Med. 189:447. Luther, S. A., H. L. Tang, P. L. Hyman, A. G. Farr, and J. G. Cyster. 2000. Coexpression of the chemokines ELC and SLC by T zone stromal cells and deletion of the ELC gene in the plt/plt mouse. Proc. Natl. Acad. Sci. USA 97: 12694. Vassileva, G., H. Soto, A. Zlotnik, H. Nakano, T. Kakiuchi, J. A. Hedrick, and S. A. Lira. 1999. The reduced expression of 6Ckine in the plt mouse results from the deletion of one of two 6Ckine genes. J. Exp. Med. 190:1183. Nakano, H., and M. D. Gunn. 2001. Gene duplication at the chemokine locus on mouse chromosome 4: multiple strain-specific haplotypes and the deletion of secondary lymphoid-organ chemokine and EBI-I ligand chemokine genes in the plt mutation. J. Immunol. 166:361. Nakano, H., S. Mori, H. Yonekawa, H. Nariuchi, A. Matsuzawa, and T. Kakiuchi. 1998. A novel mutant gene involved in T-lymphocyte-specific homing into peripheral lymphoid organs on mouse chromosome 4. Blood 91:2886. Forster, R., A. Schubel, D. Breitfeld, E. Kremmer, I. Renner-Muller, E. Wolf, and M. Lipp. 1999. CCR7 coordinates the primary immune response by establishing functional microenvironments in secondary lymphoid organs. Cell 99:23. Gow, A., V. J. Friedrich, and R. A. Lazzarini. 1992. Myelin basic protein gene contains separate enhancers for oligodendrocyte and Schwann cell expression. J. Cell Biol. 119:605. Mann, J. R., and A. P. McMahon. 1993. Factor influencing production frequency of transgenic mice. Methods Enzymol. 225:771. Hogan, B., F. Constantini, and L. Lacy. 1986. Manipulating the Mouse Embryo. Cold Spring Harbor Lab. Press, Cold Spring Harbor, NY. Lira, S. A., R. Kinloch, S. Mortillo, and P. M. Wassarman. 1990. An upstream region in the sperm receptor gene directs the expression of firefly luciferase to growing oocytes of transgenic mice. Proc. Natl. Acad. Sci. USA 87:7215. Zeller, N. K., M. J. Hunkoler, A. T. Campagnoni, J. Sprague, and R. A. Lazzarini. 1984. Characterization of mouse myelin basic protein messenger RNAs with a myelin basic protein cDNA clone. Proc. Natl. Acad. Sci. USA 81:18. Owens, T., H. Wekerle, and J. Antel. 2001. Genetic models for CNS inflammation. Nat. Med. 7:161.

The Journal of Immunology 45. Yoshida, R., M. Nagira, T. Imai, M. Baba, S. Takagi, Y. Tabira, J. Akagi, H. Nomiyama, and O. Yoshie. 1998. EBI1-ligand chemokine (ELC) attracts a broad spectrum of lymphocytes: activated T cells strongly up-regulate CCR7 and efficiently migrate toward ELC. Int. Immunol. 10:901. 46. Gosling, J., D. J. Dairaghi, Y. Wang, M. Hanley, D. Talbot, Z. Miao, and T. J. Schall. 2000. Cutting edge: identification of a novel chemokine receptor that binds dendritic cell- and T cell-active chemokines including ELC, SLC, and TECK. J. Immunol. 164:2851. 47. Wang, X., X. Li, D. B. Schmidt, J. J. Foley, F. C. Barone, R. S. Ames, and H. M. Sarau. 2000. Identification and molecular characterization of rat CXCR3: receptor expression and interferon-inducible protein-10 binding are increased in focal stroke. Mol. Pharmacol. 57:1190. 48. Biber, K., A. Sauter, N. Brouwer, S. C. Copray, and H. W. Boddeke. 2001. Ischemia-induced neuronal expression of the microglia attracting chemokine secondary lymphoid-tissue chemokine (SLC). Glia 34:121. 49. Goldberg, S. H., P. Van Der Meer, J. Hesselgesser, S. Jaffer, D. L. Kolson, A. V. Albright, F. Gonzalez-Scarano, and E. Lavi. 2001. CXCR3 expression in human central nervous system diseases. Neuropathol. Appl. Neurobiol. 27:127. 50. Durack, D. T., S. J. Ackerman, D. A. Loegering, and G. J. Gleich. 1981. Purification of human eosinophil-derived neurotoxin. Proc. Natl. Acad. Sci. USA 78:5165. 51. Durack, D. T., S. Mark Sumi, and S. J. Klebanoff. 1979. Neurotoxicity of human eosinophils. Proc. Natl. Acad. Sci. USA 76:1443. 52. Gladue, R. P., L. A. Carroll, A. J. Milici, D. N. Scampoli, H. A. Stukenbrok, E. R. Pettipher, E. D. Salter, L. Contillo, and H. J. Showell. 1996. Inhibition of leukotriene B4-receptor interaction suppresses eosinophil infiltration and disease pathology in a murine model of experimental allergic encephalomyelitis. J. Exp. Med. 183:1893. 53. Milici, A. J., L. A. Carroll, H. A. Stukenbrok, A. K. Shay, R. P. Gladue, and H. J. Showell. 1998. Early eosinophil infiltration into the optic nerve of mice with experimental allergic encephalomyelitis. Lab. Invest. 78:1239. 54. Turnley, A. M., G. Morahan, H. Okano, O. Bernard, K. Mikoshiba, J. Allison, P. F. Bartlett, and J. F. Miller. 1991. Dysmyelination in transgenic mice resulting from expression of class I histocompatibility molecules in oligodendrocytes. Nature 353:566. 55. Yoshioka, T., L. Feigenbaum, and G. Jay. 1991. Transgenic mouse model for central nervous system demyelination. Mol. Cell. Biol. 11:5479. 56. Turnley, A. M., and G. Morahan. 1995. Dysmyelination in class I MHC transgenic mice. Microsc. Res. Tech. 32:286.

1017 57. Corbin, J. G., D. Kelly, E. M. Rath, K. D. Baerwald, K. Suzuki, and B. Popko. 1996. Targeted CNS expression of interferon-␥ in transgenic mice leads to hypomyelination, reactive gliosis, and abnormal cerebellar development. Mol. Cell. Neurosci. 7:354. 58. Gravel, M., J. Peterson, V. W. Yong, V. Kottis, B. Trapp, and P. E. Braun. 1996. Overexpression of 2⬘,3⬘-cyclic nucleotide 3⬘-phosphodiesterase in transgenic mice alters oligodendrocyte development and produces aberrant myelination. Mol. Cell. Neurosci. 7:453. 59. Power, C., P. A. Kong, and B. D. Trapp. 1996. Major histocompatibility complex class I expression in oligodendrocytes induces hypomyelination in transgenic mice. J. Neurosci. Res. 44:165. 60. Horwitz, M. S., C. F. Evans, D. B. McGavern, M. Rodriguez, and M. B. Oldstone. 1997. Primary demyelination in transgenic mice expressing interferon-␥. Nat. Med. 3:1037. 61. Akassoglou, K., J. Bauer, G. Kassiotis, M. Pasparakis, H. Lassmann, G. Kollias, and L. Probert. 1998. Oligodendrocyte apoptosis and primary demyelination induced by local TNF/p55TNF receptor signaling in the central nervous system of transgenic mice. Am. J. Pathol. 153:801. 62. Male, D., G. Pryce, C. Hughes, and P. Lantos. 1990. Lymphocyte migration into brain modeled in vitro: control by lymphocyte activation, cytokines and antigen. Cell. Immunol. 127:1. 63. Hughes, C. C. W., D. K. Male, and P. L. Lantos. 1988. Adhesion of lymphocytes to cerebral microvascular cells: effects of interferon-␥, tumour necrosis factor and interleukin-1. Immunology 64:677. 64. Probert, L., K. Akassoglou, M. Pasparakis, G. Kontogeorgos, and G. Kollias. 1995. Spontaneous inflammatory demyelinating disease in transgenic mice showing central nervous system-specific expression of tumor necrosis factor ␣. Proc. Natl. Acad. Sci. USA 92:11294. 65. Raivich, G., L. L. Jones, C. U. A. Kloss, A. Werner, H. Neumann, and G. W. Kreutzberg. 1998. Immune surveillance in the injured nervous system: T-lymphocytes invade the axotomized mouse facial motor nucleus and aggregate around sites of neuronal degeneration. J. Neurosci. 18:5804. 66. Fan, L., C. R. Reilly, Y. Luo, M. E. Dorf, and D. Lo. 2000. Cutting edge: ectopic expression of the chemokine TCA4/SLC is sufficient to trigger lymphoid neogenesis. J. Immunol. 164:3955. 67. Chen, S.-C., G. Vassileva, D. Kinsley, S. Holzmann, D. Manfra, M. T. Wiekowski, N. Romani, and S. A. Lira. Ectopic expression of the murine chemokines CCL21a and CCL21b induces the formation of lymph node-like structures in pancreas, but not skin, of transgenic mice. J. Immunol. 168:1001.