`
`T-Lymphocyte Entry Into the Central
`Nervous System
`
`W.F. Hickey, B.L. Hsu, and H. Kimura
`Department of Pathology, Washington University School of Medicine, Saint Louis, Missouri
`(W.F.H., B.L.H.), and Department of Surgery, University of Pennsylvania School of Medicine,
`Philadelphia, Pennsylvania (H.K.)
`
`The entry of T-lymphocytes into the parenchyma of
`the central nervous system is a critical early feature in
`the pathogenesis of many experimental and sponta-
`neously occurring immune-mediated illnesses. The
`physiological mechanisms controlling this entry have
`not been elucidated. This study reports that T-cell
`entry into the rat CNS appears to be primarily de-
`pendent upon the activation state of the lymphocytes;
`T-lymphoblasts enter the CNS (and all other tissues
`examined) in an apparently random manner while T
`cells not in blast phase are excluded. Antigen speci-
`ficity, MHC compatibility, T-cell phenotype, and T-
`cell receptor gene usage do not appear related to the
`ability of cells to enter. This study demonstrates that
`when T-lymphoblasts are introduced into the circu-
`lation they rapidly appear in the CNS tissue. Their
`concentration in the CNS reaches a peak between 9
`and 12 hr, and lymphocytes which have entered, exit
`within 1 to 2 days. Cells capable of reacting with a
`CNS antigen remain in the tissue or cyclically reenter
`to initiate inflammation if they are able to recognize
`their antigen in the correct MHC context. This ob-
`servation also appears to pertain to the entry of acti-
`vated T cells into many other tissues, although their
`concentrations in these non-CNS sites was not quan-
`tita tied.
`Key words: T-lymphocyte, central nervous system,
`autoimmunity, encephalomyelitis, lymphocyte migra-
`tion, lymphocyte homing
`
`INTRODUCTION
`During the course of evolution the central nervous
`system (CNS) has become isolated from the circulation
`and other bodily organs behind a physiological wall
`termed the ‘ ‘blood-brain barrier” (Rapoport, 1976).
`Thus, the intricate functions of the nervous system are
`somewhat spared the episodic perturbations occurring in
`the remainder of the body. However, this arrangement
`presents a problem when a need arises for an immune
`response in the CNS to combat foreign or neoantigens.
`
`0 1991 Wiley-Liss, Inc.
`
`Current scientific belief is that there is no routine immu-
`nological “survey” of the brain and spinal cord. This is
`supported by a number of observations. In healthy mam-
`mals the CNS contains T-lymphocytes in such low con-
`centration that they are virtually undetectable by immu-
`nohistochemical methods (Barker and Billingham, 1977;
`Williams et al., 1980; Hauser et al., 1983; Cruzner et al.,
`1988; Hickey and Kimura, 1987; Hickey et al., 1985). in
`addition, molecules of the major histocompatibility com-
`plex (MHC) required for antigen presentation to T cells
`are expressed on constituent neural parenchymal cells
`at very low levels, if at all (Hauser et al., 1983; Hayes
`et al., 1987; Woodroofe et al., 1986; Lampson and
`Hickey, 1986; Traugott et al., 1985; Craggs and Web-
`ster, 1985). In distinction to this situation of “immuno-
`logical privilege,” it is well known that lymphocytes and
`MHC molecule expression are easily detectable during
`immune-mediated illnesses such as multiple sclerosis,
`postinfectious encephalomyelitis, the viral encephaliti-
`des, and experimental models of immune-mediated neu-
`rological disease (Paterson et al., 1981; Paterson and
`Day, 1982; Traugott, 1984; Hickey et al., 1985; Hayes et
`al., 1987; Lassmann et al., 1986; Hickey and Kimura,
`1987). In these processes T-cell entry into the CNS is
`one of the earliest events in their pathogenesis (Paterson
`and Day, 1982; Hickey et al., 1989; Lassmann et al.,
`1986; Traugott, 1984).
`Many recent reports have dealt with the homing of
`T-lymphocytes to selected tissues based upon the inter-
`action of molecules on the lymphocyte and endothelial
`cell membranes (Stoolman, 1989; Woodruff et al., 1987;
`Berg et a]., 1989). Yet to date the mechanisms by which
`T cells are normally excluded from the CNS, while being
`permitted to enter during illness or to induce disease,
`remain undefined. in view of the apparent pivotal role of
`
`Received June 28, 1990; revised September 6, 1990; accepted Sep-
`tember 7, 1990.
`Address reprint requests to Dr. Wm. F. Hickey, Neuropathology, BOX
`81 18, Washington University School of Medicine, 660 South Euclid
`Avenue, Saint Louis, MO 631 10.
`
`MYLAN PHARMS. INC. EXHIBIT 1091 PAGE 1
`
`
`
`CNS lymphocyte entry in the induction of many dis-
`eases, this study was undertaken to test the hypothesis
`that T-lymphocytes circulating in the activated state are
`able to enter the CNS regardless of antigen specificity,
`T-cell phenotype, or recognition of any one major his-
`tocompatibility complex molecule.
`
`MATERIALS AND METHODS
`An i m a 1 s
`Lewis rats (MHC type RT-1’) were obtained from
`Charles River Breeders of Wilmington, Massachusetts.
`DA rats (MHC type RT- 1 “) were supplied by Bantin and
`Kingman of Fremont, California. All rats were used as
`adults, and were housed three to a cage with food and
`water provided ad lib.
`
`Cell Culture
`T-lymphocyte culture was carried out using a me-
`dium composed of RPMI-1640 solution to which was
`added 5% NCTC- 109 (MA Bioproducts, Walkersville,
`MD), 5 X
`lop5 M 2-mercaptoethanol, 2.0 pM glu-
`tamine, and 50 units/ml penicillin-streptomycin. For cell
`types 1-111 10% fetal bovine serum (Hyclone Labordto-
`ries, Logan, Utah) was added as a protein source, as was
`5.0 pg/ml of concanavalin A to stimulate T-cell blast
`formation. For cell type IV, 1% Lewis rat serum was
`added, and no stimulatory lectin was included. The cells
`were incubated at 37°C in a humidified atmosphere con-
`taining 6.5% CO,. When ready for injection, the viable
`cells were selected by density separation using a discon-
`tinuous gradient of 1.077 specific gravity. T-lympho-
`blasts were further enriched by separation on a fluid with
`density of 1.066 specific gravity. Following separation
`the cells were washed three times in Dulbecco’s modi-
`fied phosphate-buffered saline, and the cell numbers ad-
`justed to the appropriate concentration.
`
`T-cell Types
`T-lymphocytes from Lewis rats (RT-1’) were pre-
`pared in four different fashions which could potentially
`affect their ability to enter the nervous system. (I) T-
`lymphocyte lines of CD4+, CD8-, RT-IB’ restricted,
`myelin basic protein specific T cells were derived (Ben
`Nun et al., 1981; Vendenbark et al., 1985; Hickey et al.,
`1987; Hickey and Kimura, 1988). These lymphocytes
`induced experimental allergic encephalomyelitis (EAE),
`an autoimmune CNS inflammation, 4 to 6 days after
`injection into the circulation of syngeneic rats. They
`were not alloreactive with DA rat antigens. Since the
`antigen for this cell line is a component of CNS myelin,
`these cells should be expected to show selective “hom-
`
`T-cell Entry Into the CNS
`
`255
`
`ing” to the brain and spinal cord, if such a mechanism is
`a necessary feature of CNS entry (Heber-Katz and Acha-
`Orbea, 1989). (11) A line of T cells was derived with the
`same phenotype and restriction element as the first, but
`having calf thymus histone (Sigma Chemicals, Saint
`Louis, MO) as their antigen. These cells cause no known
`illness or inflammation in syngeneic rats. This “irrele-
`vant” antigen was selected because it has a similar mo-
`lecular weight and pl as myelin basic protein, the en-
`cephalitogenic antigen used above. For consistency these
`first two types of cells were activated to the lymphoblast
`stage by 72 hrs of in vitro culture with the mitogenic
`lectin concanavalin A prior to use. This type of activa-
`tion does not diminish the encephalitogenic ability of T
`cells specific for myelin basic protein (Panich, 1980,
`Peters and Hinrichs, 1982; Takenaka et al., 1986), and
`thus would not interfere with the entry of T cells into the
`CNS. (111) The third type of T cells was obtained from
`healthy, nonimmunized Lewis rats by bulk activation of
`unselected splenocytes and lymph node cells with con-
`canavalin A. Following 72 hr of in vitro activation, cells
`in this pool were predominantly T-lymphoblasts and con-
`tained both CD4- and CD8-positive populations. Theo-
`retically these cells represented a wide spectrum of ad-
`mixed clones with
`random antigen
`specificities,
`requiring a variety of MHC restriction elements and em-
`ploying a spectrum of T-cell receptor gene combinations.
`And (IV) lymphocytes from the spleen and lymph nodes
`from nonimmunized Lewis rats were cultured for 24 hr in
`a nonstimulatory medium with 1% syngeneic rat serum
`as the protein source. Following the in vitro period this
`fourth type of cells was depleted of lymphoblasts to ob-
`tain resting, nonblast T cells by density separation (Ben
`Nun et al., 198 1) using a solution with specific gravity
`1.066; cells collecting at this interface are highly en-
`riched for lymphoblasts, while the population falling
`through the interface is extensively depleted of blast
`cells. This final group of cells would again have random
`antigen specificities and differed from the third group
`only by the activation state of the T-lymphocytes.
`
`Experimental Protocol
`T cells of each of these four types were injected via
`tail vein into syngeneic Lewis rats or allogeneic DA (RT-
`I”) rats. At specific times following injection, the rats
`were killed and their spinal cords rapidly removed and
`snap frozen for immunohistochemical staining (1 2). T
`cells in Lewis rat CNS tissue were identified by mono-
`clonal antibodies for rat CD4 (W3/25), CD8 (OX-8), and
`11-2 receptor (OX-39) molecules (Mason et al., 1983). T
`cells of Lewis rat origin were identified in the CNS of
`DA rats using a murine monoclonal antibody, 11-69
`(Kimura and Wilson, 1984) recognizing the Lewis RT-
`] A 1 molecule which the DA rat cannot produce.
`
`MYLAN PHARMS. INC. EXHIBIT 1091 PAGE 2
`
`
`
`Hickey et al.
`256
`Immunohistochemistry
`The method employed for immunohistochemical
`identification of the T cells in the CNS has been exten-
`sively detailed elsewhere (Hickey et al., 1983, 1985). In
`brief, 5 km thick cryostat sections were cut from longi-
`tudinally oriented portions of the rat lumbosacral spinal
`cord. A minimum of 7 sections was taken from the spinal
`cord of each test animal, and these sections were ob-
`tained at a minimum interval of 15 pm to prevent dupli-
`cate counting of any single cell in adjacent sections.
`These cryostat sections were fixed for 30-60 sec in ab-
`solute methanol at -2O"C, and then rinsed extensively in
`a 0.5 M trisma buffer at pH 7.6. Following these washes,
`the sections were incubated with the appropriate primary
`antibody overnight at 4°C. On the following day the
`staining procedure was completed using an biotinylated
`anti-mouse IgG antibody which had been previously ad-
`sorbed against rat serum proteins (Vector Laboratories,
`Burlingame, CA), and then coupled with an avidin-per-
`oxidase conjugate. The reaction was developed with
`0.03% hydrogen peroxide and 3,3'-diaminobenzidine as
`the chromagen.
`
`Quantitation
`The spinal cord sections were evaluated using the
`light microscope. Each positive cell in the tissue section
`was counted, and then the total aggregate area of the
`histological tissue sections was determined by morpho-
`the quantitation for each animal, * the standard devia-
`metric means (Hickey et al., 1985, 1987). The results of
`tion, are presented graphically in Figures 1 and 2 as the
`number of T cells in the CNS parenchyma per unit area
`as a function of time following lymphocyte injection.
`
`RESULTS
`At the time of injection, T-lymphoblasts of the first
`two types were >95% CD4+, C D K , IL-2 receptor-
`positive cells. The third group of cells were >85% T-
`lymphocytes with a slight predominance of CD4+ cells
`over CD8+ lymphocytes. A minor population of B cells
`was present (< 10%) while cells of the monocyte/macro-
`phage phenotype were always less than 5%. T-lympho-
`cytes predominated in the lymphoblast-depleted popula-
`tion, but the concentration of B cells and monocyte/
`macrophage was slightly increased.
`The entry of activated T-lymphocytes into the rat
`CNS seems to follow a set time course following injec-
`tion. Cells were detectable within 3 hr. At this time the
`cells were scattered in the CNS parenchyma. A peak
`concentration of T cells was achieved between 9 to 12 hr,
`at which time cells were predominantly found near the
`meninges or blood vessels, a tendency that continued
`
`Fig. 1. The graph depicts the time-related alterations in the
`concentration of T cells per unit area in the lower spinal cord
`of syngeneic rats. Each bar depicts the mean and standard
`deviations of the T-cell concentration in the spinal cord of one
`animal. The quantity and type of cells injected are indicated.
`Controls were uninjected rats, or rats receiving an injection of
`physiological saline via tail vein. In determining the concen-
`tration of cells in the spinal cord of a rat approximately 1 cm2
`total area of 5 p m thick cryostat sections taken at least 15 k m
`apart was analyzed.
`
`t
`
`P 8 0.3 0'4
`
`11 I'
`
`HOURS POST INJECTION
`
`Fig. 2. The graph shows the time-related concentration of T
`cells in the spinal cord of DA rats injected with Lewis rat cells.
`
`thereafter. This time-concentration curve was followed
`whether the T cells were syngeneic or allogeneic to the
`host. These cells could be found at all levels of the CNS,
`although they were quantitated only in the spinal cord.
`The results presented in Figure 1 posed two theo-
`retical problems which prompted the repetition of the
`
`MYLAN PHARMS. INC. EXHIBIT 1091 PAGE 3
`
`
`
`analysis using Lewis cells injected into the allogeneic
`DA rats. By injecting activated Lewis rat T-lymphoblasts
`into Lewis recipients, it was impossible to be certain that
`the cells detected in the CNS were actually derived from
`the bolus of lymphocytes injected. Since activated T
`cells are known to elaborate a spectrum of stimulatory
`Iymphokines, it was possible that the injected cells were
`either inducing other host lymphocytes to enter the CNS,
`or it was possible that the injection of activated lympho-
`cytes totally altered the normal lymphocyte traffic
`through the CNS for all T cells. These two problems
`were eliminated by giving Lewis lymphocytes to DA rats
`and seeking them in the CNS with a monoclonal anti-
`body specific only for Lewis-derived cells. Thus, any
`cell found in the DA CNS must have derived from the
`bolus of activated or “resting”
`(i.e., unstimulated)
`Lewis rat T cells given at the specific initiating time. The
`results of T-cell entry into the CNS of DA rats shown in
`Figure 2 parallel that noted in the Lewis rat. In both the
`Lewis and DA rats it was confirmed that T-lymphocytes
`depleted of lymphoblasts exhibited little or no ability to
`enter the nervous system tissues. As constructed, these
`studies comparing cells of type I11 (activated) with those
`of type IV (unstimulated) would result in lymphocytes
`which varied only in their activation state, while being
`otherwise comparable in their spectrum of antigen spec-
`ificities, MHC restriction requirements, and T-cell re-
`ceptor gene usage.
`In histological sections of spinal cord from both
`Lewis and DA rats the concentration of lymphocytes was
`not uniform on each slide examined. Some rats exhibited
`a wide spectrum of lymphocyte numbers from one level
`to the next. No anatomic explanation for this diversity
`could be found. Also, it was noted that the concentration
`of T cells found in the CNS of allogeneic rats was
`slightly below that of the syngeneic rats. This may be
`due to allogeneic elimination and/or sequestration of a
`portion of the injected cells by the host.
`
`DISCUSSION
`The data shown in Figures 1 and 2 indicate that
`activated T cells (lymphoblasts) appear to enter the CNS
`in a random manner. The antigen specificity of the in-
`jected cells, the ability of the T cell to “recognize” its
`antigen among the CNS constituents, and the MHC com-
`patibility of the lymphocytes with the host nervous sys-
`tem, all seem to have no necessary bearing on the ability
`of the cells to enter. Both CD4- and CDS-positive T-cell
`phenotypes were encountered in the tissue. If the cells
`injected were not specific for a CNS antigen, or they
`were unable to recognize their antigen because of MHC
`incompatibility (Hickey and Kimura, 1988), the cell con-
`
`T-cell Entry Into the CNS
`
`257
`
`centration reached the same level as myelin basic pro-
`tein-specific T cells able to induce CNS inflammation in
`an MHC compatible host. However, the concentration of
`T cells unable to “find” their antigen in the CNS re-
`turned to baseline levels between 24 and 48 hr from the
`time of injection. T-lymphocytes able to recognize their
`specific antigen in the CNS of the host persisted in or
`cyclically reentered the ‘‘target organ” at numbers sig-
`nificantly above baseline levels beyond 72 hr (Fig. 1).
`Even at peak concentration the number of cells per unit
`area of spinal cord was, nevertheless, still two to three
`orders of magnitude below the concentration of T cells
`during acute EAE (Hickey et al., 1983, 1987). This ob-
`servation would suggest that a significant degree of “re-
`cruitment” of T cells not specific for an immunogenic
`antigen occurs once inflammation is initiated. Our find-
`ings in this quantitative analysis fail to confirm previous
`observations that EAE-inducing cells progressively and
`selectively continue to accumulate in the CNS up to the
`onset of clinical EAE (Naparstek et al., 1983). This dis-
`crepancy may be due to a difference in the methodology
`used to detect the presence of T cells.
`When T cells depleted of lymphoblasts (i.e., “rest-
`ing” T cells) were injected, and their presence sought at
`a time that should have corresponded to the peak T-cell
`concentration in the CNS, they were virtually undetect-
`abIe (Fig. 2). Nevertheless, rare cells were identified on
`some sections. One possible explanation is that density
`separation of blasts from resting cells is highly efficient,
`but not absolute. Therefore, some lymphoblasts would
`have been present in the population of resting cells in-
`jected and did enter the CNS.
`In the past it has been observed that T-lymphocytes
`capable of inducing EAE, such as the anti-MBP lympho-
`cyte lines used here, fail to produce disease unless the
`cells are stimulated prior to injection (Panitch, 1980;
`Peters and Hinrichs, 1982; Holda et al., 1980; Ben Nun
`et al., 1981). This fjnding might now be explained by the
`relative inability of unstimulated cells to reach the target
`organ. It has also been reported by Wekerle et al. (1986)
`and in the work of Meyermann and colleagues (1986)
`that rat T cells specific for an irrelevant (non-CNS) an-
`tigen could be detected in the CNS following peripheral
`injection into syngeneic hosts. Consistent with these
`findings, our results provide evidence that T-lymphocyte
`entry is independent of antigen specificity, MHC restric-
`tion element, and T-cell phenotype. As can be seen from
`Figure 1, the control rats “normally” had a low but
`detectable number of T cells in the CNS. This would be
`the expected situation if animals were continually acti-
`vating T-lymphocytes in their immune organs in re-
`sponse to normal environmental challenges. A sampling
`of all these activated T cells would be encountered in the
`CNS and other organs.
`
`MYLAN PHARMS. INC. EXHIBIT 1091 PAGE 4
`
`
`
`258
`
`Hickey et al.
`Based on the examination of other organs in DA
`rats given Lewis T-lymphoblasts, it appears that T-cell
`activation empowers them to enter many organs and tis-
`sues which typically exclude them (including peripheral
`nerve, eye, muscle, and the medulla of the thymus).
`Since entry occurs rapidly after injection of activated
`cells, it is proposed that alterations in the normal pattern
`of tissue entry may be the result of a general, nonclono-
`typic, activation-related lymphocyte membrane change
`rather than rapidly occurring variations in endothelial
`receptors. It is also postulated that the endothelium in
`most organs express a molecule which is recognized by
`corresponding molecules elaborated on the membrane of
`activated T cells. The information currently available
`does not permit us to conclude which, if any, of the
`changes known to occur in T cells during activation are
`responsible for their acquisition of the ability to enter the
`CNS. At this time, lymphocyte-endothelial receptor sys-
`tems have not be defined which are specific for T-cell
`homing to the nervous system.
`Some investigators have reported that activated
`lymphocytes produce elevated levels of heparan sulfate
`endoglycosidase and that substances inhibiting this en-
`zyme prevent the development of EAE which is depen-
`dent on T-cell entry (Naparstek et al., 1984; Lider et al.,
`1989). This may prove to be an essential feature of the
`migration properties of activated T cells, but alone may
`not fully explain it. In both rats and mice cells specific
`for the antigen which causes EAE show a strong prede-
`lection to use a very narrow group of T-cell receptor
`V-beta chain genes (Urban et al., 1988; Burns et a].,
`1989; Zamvil et al., 1988). in a recent review of this
`highly selected use of T-cell receptor genes by lympho-
`cytes specific for myelin basic protein, it was postulated
`that the ability of T cells to reach the CNS parenchyma
`may be a consequence of this specific T-cell receptor
`gene usage (Heber-Katz and Achea-Orbea, 1989). The
`findings reported here argue against this hypothesis.
`CNS entry was equivalent for Lewis rat T-cell blasts
`specific for myelin basic protein and random, unselected
`splenic T-lymphoblasts having a wide diversity of T-cell
`receptor gene usage.
`These results potentially have great significance for
`the pathogenesis of CNS inflammatory conditions. It ap-
`pears that any T-lymphocyte clone or group of clones
`which is activated in the tissues of the immune system
`will gain access to the CNS in a random fashion shortly
`after entering the circulation. If true, this would mean
`that there is no strict “immunologic privilege” for the
`nervous system. Any antigen which stimulates a T-lym-
`phoblast response will give rise to cells that will seek that
`antigen anywhere, including the CNS. This contention is
`supported experimentally by studies in which a graft of
`allogeneic tissue, which is well tolerated by the host
`
`when placed in the CNS, will be rapidly rejected when
`the host is exposed to the same alloantigens in the pe-
`riphery (Mason et al., 1986).
`Unfortunately, the experimental design does not
`permit us to address two critical issues intimately related
`to initiation of inflammation in the CNS. What is the
`defect in T-cell lines or clones which is specific for a
`CNS antigen, but is not encephalitogenic? Based on the
`above data, it would be presumed that such clones would
`enter the CNS, but in some manner fail to induce inflam-
`mation. However, the possibility that their lack of en-
`cephalitogenicity reflects a failure in the process of entry
`into the target organ cannot be excluded. Second, what is
`the fate of T cells which enter the CNS during inflam-
`mation, but then enter a resting phase? Are they reacti-
`vated in situ, or do they continue to migrate under dif-
`ferent restraining criteria? These issues will undoubtedly
`be the focus of future investigations.
`A correllary of the hypothesis we set forth here is
`that T cells recovered from the cerebrospinal fluid of
`experimental animals or human patients represent a ran-
`dom sample of the T-cell clones which have been acti-
`vated by the immune system. They do not necessarily
`represent only T-lymphocytes specific for some antigen
`located within the CNS.
`In addition to T-cell entry, the present data docu-
`ment the arrest or repeated return of activated T-lympho-
`cytes to an anatomic site where their antigen is encoun-
`tered; otherwise, the cells will “survey” the organ and
`exit the tissue leaving it morphologically undisturbed.
`This proposed mechanism can account for the ability of
`a peripheral antigenic challenge to induce CNS autoim-
`mune attack, as in EAE, and provides a plausible expla-
`nation for the well-known phenomenon of multiple scle-
`rosis relapses associated with stress or illnesses where
`potentially a large group of T cells is activated in the
`tissues of the immune system. Finally, via the mecha-
`nism of “molecular mimicry” (Fujinami and Oldstone,
`1985; Fujinami et al., 1983), our hypothesis regarding
`T-cell migration into privileged sites could account for
`the induction and/or recurrence of CNS (or other organ
`specific) autoimmune reactions in organs far removed
`from the site of foreign antigenic challenge and T-cell
`activation.
`
`ACKNOWLEDGMENTS
`The authors express their appreciation to Drs. Emil
`Unanue, Kenneth Tung, Judith Kapp-Pierce, and Paul
`Allen for helpful comments and advice. These studies
`were supported by awards from the National Institutes of
`Mental Health (NS-27321) and the National Multiple
`Sclerosis Society (RG-2244).
`
`MYLAN PHARMS. INC. EXHIBIT 1091 PAGE 5
`
`
`
`REFERENCES
`Barker CF, Billingham RE (1977): Immunologically privileged sites.
`Adv Immunol 25: 1-6.
`Ben-Nun A, Wekerle H, Cohen, 1R (1981): The rapid isolation of a
`clonable antigen-specific T-lymphocyte lines capable of medi-
`ating EAE. Eur J Immunol 11:195-199.
`Berg EL, Goldstein LA, Jutila MA, Nakach M, Picker LT, Streeter P
`R, Wu N W, Zhou D, Butcher E C (1989): Immunol Rev
`108:5-18.
`Burns FR, Li X, Shen N, Offner H, Chou YK, Vandenbark AA,
`Heber-Katz E (1989): Both rat and mouse T-cell receptors spe-
`cific for the encephalitic determinant of myelin basic protein
`use similar V-alpha and V-beta chain genes even though the
`major histocompatibility complex and encephalitic determinant
`are different. J Exp Med 169:27-39.
`Craggs RI, Webster H deF (1985): Ia antigens in the normal rat ner-
`vous system and in lesions of EAE. Acta Neuropathol 68:
`263-212.
`Cruzner ML, Hayes GM, Newcombe J, Woodroofe MN ( I 988): The
`nature of inflammatory components during demyelination in
`multiple sclerosis. J Neuroimmunol 20:203 -209.
`Fujinami RS, Oldstone MBA (1985): Amino acid homology between
`the encephalitogenic site of myelin basic protein and virus: A
`mechanism for autoimmunity. Science 230: 1043-1045.
`Fujinami RS, Oldstone MBA, Wroblewska Z, Frankel ME, Kop-
`rowski H (1983): Moleculer mimicry in viral injection: Cross
`reaction of measles virus phosphoprotein or of herpes virus
`protein with human intermediate filaments. Proc Natl Acad Sci
`USA 80~2346-2350.
`Hauser SL, Bhan AK, Gilles FH, Hoban CJ, Reinherz EL, Weiner HL
`(1983): Immunohistochemical staining of human brain with
`monoclonal antibodies that identify lymphocytes, monocytes
`and the Ia antigen. J Neuroimmunol 5:197-205.
`Hayes GM, Woodroofe MN, Cruzner ML (1987): Microglial are the
`major cell type expressing MHC class I1 antigens in human
`white matter. J Neurol Sci 80:25-37.
`Heber-Katz E, Acha-Orbbea H (1989): The V-region disease hypoth-
`esis: Evidence from autoimmune encephalomyelitis. Immunol
`Today 10:164-169.
`Hickey WF, Cohen JA, Burns JB (1987): A quantitative immunohis-
`tochemical comparison of actively versus adoptively induced
`experimental allergic encephalomyelitis in the Lewis rat. Cell
`Immunol 109:272-228.
`Hickey WF, Gonatas NK, Kimura H, Wilson DB (1983): Identifica-
`tion and quantitation of T-lymphocyte subsets found in the
`spinal cord of Lewis rats during acute EAE. J Immunol 131:
`2805-2809.
`Hickey WF, Hsu BL, Kimura H (1989): T-cell entry into the rat
`central nervous system. FASEB J 3 (Suppl. I):A482 (abstract).
`Hickey WF, Kimura H (1987): Graft-versus-host disease elicits ex-
`pression of class I and class I1 histocompatibility antigens and
`the presence of scattered T-cells in rat central nervous system.
`Proc Natl Acad Sci USA 84:2082-2087.
`Hickey WF, Kimura H (1988): Penvascular microglial cells of the
`CNS are bone marrow derived and present antigen in vivo.
`Science 239:290-292.
`Hickey WF, Osborn JP, Kirby WM (1985): Expression of Ia mole-
`cules by astrocytes during acute EAE in the Lewis rat. Cell
`Immunol9 1528-533.
`Holda JH, Welch AM, Swanborg RH (1980): Autoimmune effector
`cells I. Transfer of EAE with lymphoid cells cultured with
`antigen. Eur J Immunol 10:657-659.
`Kimura H, Wilson DB (1984): Anti-ideotypic T-cells in rats with graft
`versus host disease. Nature 308:463-466.
`
`T-cell Entry Into the CNS
`
`259
`
`Lanipson LA, Hickey WF (1986): Monoclonal antibody analysis of
`MHC expression in human brain biopsies. J Immunol 136:
`4054-4062.
`Lassmann H, Vass K, Brunner C, Seitelberger F (1986): Character-
`ization of inflammatory infiltrates in experimental allergic en-
`cephalomyelitis. Prog Neuropathol 6:33-62.
`Lider 0, Baharav E, Mekon Y, Miller T, Naparstek Y, Vladovsky 1,
`Cohen IR (1989): Suppression of experimental allergic enceph-
`alomyelitis and prolongation of allograft survival by treatment
`of animals with low dose heparin. J Clin Invest 83:752-756.
`Mason DW, Arthur RP, Dallman MJ, Green JR, Spickett GP, Thomas
`ML (1983): Functions of rat T-lymphocyte subsets isolated by
`means of monoclonal antibodies. Immuno Rev 74:57-82.
`Mason DW, Charleton HM, Jones AJ, Lavy CBD, Puklavek M, Sim-
`monds SJ (1986): The fate of allogeneic and xenogeneic neu-
`ronal tissue transplanted into the third ventricle of rodents.
`Neuroscience 19 :685 -694,
`Meyermann R, Korr J, Wekerle H (1986): Specific target retrieval by
`encephalitogenic T-line cells. Clin Neuropathol 5: 101 (ab-
`stract).
`Naparstek Y, Ben Nun A, Holoshitz J, Resehf T, Frenkel A, Rosen-
`berg M, Cohen IR (1983): T-lymphocyte lines producing or
`vaccinating against EAE. Eur J Immunol 13:418-423.
`Naparstek Y, Cohen IR, Fuks Z, Vlodavsky I (1984): Activated T-
`lymphocytes produce a matrix degrading heparan sulphate en-
`doglycosidase. Nature 3 10:241-244.
`Panitch HS (1980): Adoptive transfer of EAE with activated spleen
`cells: Comparison of in vitro activation by concanavalin A and
`myelir, basic protein. Cell Immunol 56:163-171.
`Paterson PY, Day ED (1982): Current perspectives of neuroimmuno-
`Multiple sclerosis and experimental allergic
`tis. Clin Immunol Rev I:%-697.
`Paterson PY, Day ED, Whitacre CC (1981): Neuroimniunologic dis-
`eases: Effector cell responses and immunoregulatory mecha-
`nisms. Immunol Rev 55:89-120.
`Peters BA, Hinrichs DJ (1982): Passive transfer of EAE in the Lewis
`rat with activated spleen cells: Differential activation with mi-
`togens. Cell Immunol 69: 175-185.
`Rapoport SI (1976): “Blood-Brain Barrier in Physiology and Medi-
`cine.” New York: Raven Press.
`Stoolman LM (1 989): Adhesion molecules controlling lymphocyte
`migration. Cell 56:907-Y 10.
`Takenaka A, Minegawa H, Kaneka K, Mori R, Itoyama Y (1986):
`Adoptive transfer of EAE with lectin activated spleen cells,
`part 2. Studies on T-cell subsets and interleukin 2 production.
`J Neurol Sci 72:337-345.
`Traugott U (1984): Characterization and distribution of lymphocyte
`subpopulations in multiple sclerosis plaques versus autoim-
`mune demyelinating lesions. Spring Sem Immunopathol 8:7 I-
`95.
`Traugott U, Scheinberg LC, Raine CS (1985): On the presence of la
`positive endothelial cells and astrocytes in multiple sclerosis
`and its relevence to antigen presentation. J Neuroimmunol 8:
`1-9.
`Urban JL, Kumar V, Kono DH, Gomez C, Horvath SJ, Clayton J,
`Ando DG, Sercarz EE, Hood L (1988): Restricted use of the T
`cell receptor V genes in murine autoimmune encephalomyelitis
`raises possibilities for antibody therapy. Cell 54:577-592.
`Vandenbark AA, Gill T, Offner H (1985): A myelin basic protein
`specific T-cell line that mediates EAE. J Immunol 135:223-
`228.
`Wekerle H, Linnington C, Lassmann H, Meyermann R (1986): Cel-
`lular immune reactivity within the CNS. Trends Neurosci 9:
`27 1-277.
`
`MYLAN PHARMS. INC. EXHIBIT 1091 PAGE 6
`
`
`
`260
`
`Hickey et al.
`
`Williams K, Hart D, Fabre J, Morris P (1980): Distribution and quan-
`titation of HLA-A,B,C and DR(1a) antigens on human kidney
`and other tissues. Transplantation 29:274-280.
`Woodroofe MN, Bellamy AS, Feldmann M, Davison AN, Cruzner
`ML (1986): Immunocytochemical characterization of immune
`reactions in the central nervous system in multiple sclerosis:
`Possible role for microglia in lesion growth. J Neurol Sci 74:
`135-152.
`
`Woodruf