Journal of Autoimmunity 21 (2003) 353–363
`
`www.elsevier.com/locate/issn/08968411
`
`VEGF and vascular changes in chronic neuroinflammation
`
`S.L. Kirk a*, S.J. Karlik a,b,c
`
`aDepartment of Pathology, The University of Western Ontario, 1151 Richmond Street, London, ON, Canada, N6A 5C1
`bDepartment of Diagnostic Radiology and Nuclear Medicine, The University of Western Ontario, 1151 Richmond Street, London, ON, Canada,
`N6A 5C1
`cDepartment of Physiology, The University of Western Ontario, 1151 Richmond Street, London, ON, Canada, N6A 5C1
`
`Received 14 April 2003; revised 3 July 2003; accepted 4 August 2003
`
`Abstract
`
`A vascular component has long been associated with the pathological changes in multiple sclerosis (MS) and its animal model,
`experimental allergic encephalomyelitis (EAE). Despite the codependence of angiogenesis and many chronic inflammatory
`disorders, only circumstantial evidence is available to support the existence of angiogenesis in MS or EAE. To determine if
`angiogenesis occurs in conjunction with clinical and pathological signs of CNS inflammatory disease we evaluated temporal and
`spatial blood vessel counts, VEGF immunoreactivity, and histopathological changes in the spinal cord of guinea pigs with
`chronic-progressive (CP)-EAE (day 0–90 post-immunization, n=64) and controls (n=17). The number of vessels per section
`increased in infiltrated and demyelinated lesions by day 15 post-immunization and remained significantly higher than controls
`throughout the course of the disease. The number of vessels correlated with both clinical and pathological scores for inflammation,
`infiltration and demyelination. Vascular endothelial growth factor (VEGF) expression increased during acute disease peaking at day
`26, which was the transition from the acute-inflammatory to chronic-demyelinating phase, before gradually returning to baseline
`levels. These observations implicate angiogenesis as a component of chronic neuroinflammation and demyelination and may suggest
`alternative therapeutic strategies for multiple sclerosis.
` 2003 Elsevier Ltd. All rights reserved.
`
`Keywords: Angiogenesis; EAE; Multiple sclerosis; VEGF
`
`1. Introduction
`
`Angiogenesis and vascular remodeling are important
`elements in the pathophysiology of cancer and several
`chronic inflammatory diseases including rheumatoid
`arthritis, psoriasis
`and osteoarthritis
`[1,2]. New
`blood vessels serve to transport inflammatory cells,
`nutrients and oxygen to the site of inflammation. The
`increased endothelial surface area also enhances the
`production of cytokines, adhesion molecules and other
`inflammatory stimuli
`[3]. Recognizing the important
`contribution of vascular changes to disease progression
`has
`led researchers
`to focus on anti-angiogenic
`treatment in both cancer and chronic inflammatory
`disease.
`
`* Corresponding author. Tel.: +1-519-663-3648;
`fax: +1-519-663-3544.
`E-mail address: skarlik@imaging.robarts.ca (S.J. Karlik).
`
`0896-8411/03/$ - see front matter  2003 Elsevier Ltd. All rights reserved.
`doi:10.1016/S0896-8411(03)00139-2
`
`Multiple sclerosis (MS) is a chronic inflammatory,
`demyelinating disease of the central nervous system
`(CNS), once considered to have a vascular disease
`etiology [4,5]. Today, contemporary hypotheses suggest
`that an interplay of both environmental and multiple
`genetic factors result in the lesions characteristic of MS
`[6,7]. Although inflammatory changes, demyelination,
`axonal and oligodendrocyte loss represent characteristic
`pathological features of MS lesions [8], an association
`between abnormal blood vessels and MS lesions is well
`recognized, and was described as early as 1872 [9].
`Lesions are typically centered on one or more veins in
`the white matter, which exhibit increased permeability
`[10,11]. Dawson’s fingers, thought to represent the in-
`vasion of
`the demyelinating process
`into normal-
`appearing white matter (NAWM), often extend from
`lesions and directly follow the course of veins or venules
`[12,13]. Similarly, MR-Venography found both the form
`
`Apotex v. Novartis
`IPR2017-00854
`NOVARTIS 2043
`
`

`

`354
`
`S.L. Kirk, S.J. Karlik / Journal of Autoimmunity 21 (2003) 353–363
`
`and orientation of oval MS lesions to follow the course
`of veins [13].
`Many factors already implicated in the pathogenesis
`of MS have been shown to act either directly or in-
`directly to promote angiogenesis. Matrix metallopro-
`teases (MMP) -1, -2, -3 and -9, intercellular cell adhesion
`molecule (ICAM)-1, vascular cell adhesion molecule
`(VCAM)-1 and E-selectin facilitate the entry of mono-
`nuclear cells through the blood–brain barrier (BBB) in
`MS [6,14]. The same enzymes and cell adhesion mol-
`ecules participate in the breakdown of the basement
`membrane in angiogenesis and induce further release of
`growth factors and angiogenic signaling molecules from
`the basement membrane [15]. Inflammatory mediators
`found in MS such as interferon (IFN)-♤ and tumor
`necrosis factors (TNF)-♡/-♢ [6], are also capable of
`enhancing angiogenesis [16]. In contrast, IFN-♡ and -♢
`that have shown to be beneficial in MS [6,17] have been
`found to inhibit angiogenesis [16]. Whereas, nitric oxide
`levels are elevated in MS patients and correlate well with
`clinical and MR markers of disease progression [18],
`elevated nitric oxide contributes both directly and
`indirectly to angiogenesis in inflammatory, vascular dis-
`eases and tumor expansion [19]. Another signaling pep-
`tide, endothelin-1 (ET-1), that induces angiogenesis in
`cultured endothelial cells and stimulates neovasculariz-
`ation in concert with vascular endothelial growth factor
`(VEGF), has been reported to be significantly elevated
`in MS patients [20]. The presence of ET-1 correlates with
`tumor vascularity and malignancy and induces MMP-2
`[21], and an ET-1 receptor antagonist ameliorates acute
`EAE [22].
`Vascular endothelial growth factor (VEGF) signals
`the proliferation and migration of endothelial cells in
`angiogenesis [3]. VEGF expression was recently found
`to be associated with inflammatory cells in the lesions of
`both MS patients and animals with acute EAE [23].
`Furthermore, an intracerebral infusion of VEGF in an
`acute model of EAE induced an inflammatory response
`in the brain suggesting that neuroinflammatory disease
`may be exacerbated by the over-expression of VEGF
`[23].
`Further evidence for the possible presence of neovas-
`cularization in MS is observed with contrast-enhanced
`MRI in the appearance of “ring enhancement” at the
`periphery, but not at the center of chronic lesions [24].
`Early enhancing lesions were nodular but progressed to
`ring enhancement that grew in size over time, supporting
`the belief that enhancement and possibly demyelination
`occurred from the center outwards [25]. This ring
`enhancement is very similar to that seen for certain
`growing CNS tumors [26] suggesting that the demyeli-
`nating lesion may grow larger by an angiogenic process
`analogous to that in tumor growth. Additional evidence
`comes from another MR imaging study, which found an
`increase in cerebral perfusion in MS patients compared
`
`to controls [27]. This finding is consistent with an
`increase in the number of blood vessels in inflammatory
`lesions.
`Taken together, the pathological, biochemical and
`MRI results suggest that, as in other chronic inflam-
`matory disorders, angiogenesis may play a role in the
`persistence of disease. Herein we investigated the role of
`angiogenesis in a chronic model of demyelination
`(chronic-progressive experimental allergic encephalomy-
`elitis, CP-EAE). We sought to: (1) quantify the number
`of blood vessels over the course of CP-EAE; (2) examine
`the relationship between disease severity and number of
`blood vessels; (3) determine if there is VEGF expression
`in the spinal cord of CP-EAE animals; (4) assess the
`temporal and spatial relationship between inflammation,
`demyelination, number of blood vessels and VEGF
`expression.
`
`2. Materials and methods
`
`2.1. EAE induction and monitoring
`
`Eighty-three female Hartley guinea pigs (200–250 g)
`(Charles River Canada, St Constant, PQ, Canada) were
`maintained in a light and temperature controlled
`environment and allowed food and water ad libitum. We
`induced EAE in 64 animals by intradermal nuchal
`injection of 0.6 ml of a 1:1 mixture of homogenized
`isologous CNS tissue and complete Freud’s adjuvant
`(CFA) (Difco, Detroit, MI, USA), with the addition of
`10 mg/ml
`inactivated Mycobacterium tuberculosis
`(Difco). Two animals, immunized with CFA alone, were
`used as controls. The remaining 17 guinea pigs were not
`immunized and were used as age-matched non-EAE
`controls. Each animal was weighed and assessed daily
`according to the following 4-point clinical scale: 0: no
`abnormality; 0.5: more then one day of weight loss; 1:
`hind limb weakness, poor righting reflex; 2: paresis,
`urinary incontinence, fecal impaction; 3.0: paralysis; 4.0:
`terminal paralysis. The experimental protocol was
`approved by the Animal Use Subcommittee and
`conforms to the guidelines of the Canadian Council on
`Animal Care.
`
`2.2. Tissue collection
`
`CNS samples were collected over a 90-day period
`post-immunization. At the time of sacrifice, the spinal
`cord was fixed in 10% formalin. The spinal cord was
`divided into two pieces: lumbar–thoracic and thoracic–
`cervical, and further divided into several 5 mm pieces
`that were embedded in two separate paraffin blocks.
`Five μm sections were stained with hematoxylin-eosin
`(H&E) and were scored by a blinded observer on a
`4-point scale for inflammation in the meninges (M),
`
`

`

`S.L. Kirk, S.J. Karlik / Journal of Autoimmunity 21 (2003) 353–363
`
`355
`
`Table 1
`Pathological scoring scale
`
`M: Inflammatory reaction in the meninges
`0
`no changes
`1
`perivascular and/or meningeal infiltration by mononuclear cells,
`1–3 vessels in any one section involved
`4–6 vessels involved
`6+ vessels involved
`dense infiltration of meninges and nearly all or all blood vessels
`involved
`
`2
`3
`4
`
`P: Parenchymal perivascular infiltration
`0
`no changes
`1
`1–3 parenchymal vessels infiltrated in Virchow–Robin spaces in
`any one section
`4–6 vessels involved
`6+ vessels involved
`virtually all or nearly all vessels involved
`
`2
`3
`4
`
`E: Encephalitis or myelitis
`0
`no invasion of inflammatory cells or migration of microglia in
`the neural parenchyma
`a few scattered cells
`invasion by cells from several perivascular cuffs
`large areas of neural parenchyma involved
`virtually the entire section is infiltrated
`
`1
`2
`3
`4
`
`D: Demyelination, remyelination and myelin debris
`0
`no demyelination
`1
`single focus of subpial demyelination or myelin debris
`2
`several small foci of demyelination
`3
`one large confluent area of demyelination on a single section
`4
`several large confluent areas of demyelination in any one
`section
`
`parenchymal perivascular infiltration (P) and invasion of
`the neural parenchyma (myelitis) (E). Demyelination
`(D) scores were evaluated in solochrome-R-cyanin (ScR)
`stained sections (Table 1) [28].
`
`2.3. Immunohistochemistry and analysis
`
`Immunohistochemistry was also performed on the
`5 μm formalin-fixed paraffin-embedded sections. Endog-
`enous peroxidase was blocked with 3% methanolicper-
`oxide and sections to be stained for factor VIII were
`pretreated with pepsin at 37 (C for 30 min. After
`blocking with 10% horse serum for rabbit anti-human
`Von Willebrand factor (factor VIII) antibody (A-082,
`DAKO, Mississauga, ON, USA) or 5% bovine serum
`for goat anti-human VEGF antibody (#AF-293-NA,
`R&D Systems, Minneapolis, MN, USA), sections were
`incubated with the primary antibody diluted with serum
`for 60 min at room temperature for factor VIII (1:300)
`or at 4 (C overnight for VEGF (1:10). Sections were
`then incubated with the corresponding non-immune
`immunoglobulin G diluted in serum. Antibody binding
`was visualized by staining with the Vectastain ELITE
`
`ABC kit (Vector Laboratories, Burlington, ON, USA)
`for 35 min. Diaminobenzidine was used for chromagen
`and Harris’s hematoxylin as counterstain.
`Immunostaining for factor VIII has been shown to be
`useful as an endothelial cell marker in inflammatory and
`neoplastic disease [29]. Toi et al.
`[30] found a close
`correlation between CD-31 staining and factor VIII
`staining,
`suggesting both are reliable methods of
`quantifying angiogenesis. Counting of microvessels is
`reproducible [31] and factor VIII staining provides
`good contrast between microvessels and other tissue
`components [32].
`The number of blood vessels was quantified both
`temporally and spatially. To quantify the number of
`blood vessels over time, we counted all white matter
`microvessels staining for factor VIII (without knowledge
`of condition or day, at 250 magnification) in a mean
`of eight spinal cord sections per animal to calculate the
`average number of blood vessels per spinal cord cross-
`section. To assess the spatial relationship between blood
`vessels and lesions we counted the number of factor VIII
`stained vessels in two 13.2 μm2 areas (at 250 magnifi-
`cation) in the same spinal cord section in 18 animals
`with a clinical score of 2 or greater. From each CP-EAE
`animal one area was counted from a demyelinating
`lesion, while the other area was chosen from NAWM.
`Demyelinated lesions and NAWM areas were chosen
`on ScR stained sections and counted on the serial factor
`VIII stained section. For comparison corresponding
`areas were also counted from 17 non-EAE controls.
`Vessel
`lumen, although present
`in some, were not
`necessary for a structure to be counted as a microvessel
`[33].
`To measure VEGF expression, four images (captured
`at the dorsal, ventral and two lateral funiculi) from each
`of eight tissue sections per animals were captured (at
`100 magnification) with a digital camera (total area/
`image was 48 μm2). The proportion of stained area was
`quantified using Sigma Scan (SPSS Inc., Chicago, IL,
`USA) yielding an average area (μm2) of VEGF expres-
`sion per image for each animal. Additionally, a system-
`atic qualitative analysis of VEGF expression was
`undertaken by mapping out the location of M, P, E, D
`and VEGF in a mean of eight tissue sections in each of
`20 CP-EAE animals.
`
`2.4. Statistical analysis
`
`Statistical analysis was performed using Sigma Stat
`v2 software (SPSS Inc. Chicago, IL, USA). Clinical and
`pathological scores, number of vessels and VEGF were
`assessed with Kruskal–Wallis ANOVA on ranks,
`Dunn’s method. A paired t-test was used to compare the
`number of blood vessels in lesion area versus NAWM
`within an animal. The number of vessels between
`
`

`

`356
`
`S.L. Kirk, S.J. Karlik / Journal of Autoimmunity 21 (2003) 353–363
`
`Fig. 1. Time course for clinical disease. By day 15 and continuing throughout the study, clinical scores of EAE animals (j) increased to levels
`significantly greater than controls (,) (P<0.05, Kruskal–Wallis ANOVA on ranks, Dunn’s). Results are expressed as the meanSEM at each day
`of sacrifice. The number of animals per group for both CP-EAE and control, respectively, are indicated above each time point.
`
`controls and NAWM areas were assessed with a t-test.
`Relationships were analyzed using a Spearman corre-
`lation. On all comparisons P<0.05 was considered
`statistically significant.
`
`3. Results
`
`3.1. Clinical time course
`
`CP-EAE animals showed a typical clinical pattern
`that began with acute signs of disease day 9 post-
`immunization (Fig. 1 and references [34,35]). Clinical
`onset resulted in weight loss, hind limb weakness and an
`abnormal righting reflex. More severe clinical signs were
`seen in a few acute cases exhibiting hind limb paresis or
`complete hind limb paralysis before day 20. Thereafter,
`all animals continued on to a chronic disease course
`where no clinical recovery was seen [36]. As we immu-
`nized many groups of guinea pigs that were randomly
`assigned for sacrifice at different times post immuniz-
`ation, there was some variability in mean disease severity
`at the different time points.
`
`3.2. Pathological changes
`
`A temporal sequence was observed for the pathologi-
`cal changes evaluated in axial spinal cord sections
`stained with H&E and ScR (Fig. 2). Meningeal inflam-
`mation was the first sign of disease, apparent in some
`
`CP-EAE animals as early as day 7 post-immunization
`and was significantly greater than controls from day 15
`onwards (Fig. 2A). Parenchymal perivascular infiltra-
`tion (cuffed vessels) was the next feature to appear,
`beginning on day 9 (Fig. 2B). By day 15, inflammatory
`cells had invaded the neural parenchyma (Fig. 2C) and
`there was evidence of limited demyelination in some
`animals (Fig. 2D). Infiltration, myelitis and demyelin-
`ation were all significantly greater than controls from
`day 26 onwards. Pathological scores were abnormal for
`all
`four categories and extensive inflammation and
`demyelination were evident
`throughout
`the chronic
`phase.
`Fig. 3 illustrates the sequence of pathological changes
`seen in CP-EAE including ScR- and immuno-staining
`for factor VIII and VEGF.
`
`3.3. Blood vessel counts
`
`The number of blood vessels in CP-EAE was signifi-
`cantly higher than in controls by day 15 and remained
`high throughout the course of disease (Fig. 4A). Ani-
`mals immunized with CFA alone had the same number
`of blood vessels as non-immunized controls. In controls,
`factor VIII-immunoreactivity identified primarily radial
`vessels (Fig. 3B), whereas in the CP-EAE animals, the
`staining was profuse and concentrated in areas of
`inflammation and demyelination (Fig. 3H and K). In an
`area of 13.2 μm2 the mean number of blood vessels in
`
`

`

`S.L. Kirk, S.J. Karlik / Journal of Autoimmunity 21 (2003) 353–363
`
`357
`
`Fig. 2. Time course for pathological changes in CP-EAE. All bars represent the meanSEM for CP-EAE animals at each day of sacrifice. There
`are no control bars shown, as pathological changes are exclusive to EAE animals. (A) Meningeal inflammation was significantly greater than
`controls by day 15. Infiltration (B), myelitis (C) and demyelination (D) were significantly higher than non-EAE controls by day 26 post-
`immunization. Pathological scores remained high throughout the course of the study. (P<0.05, Kruskal–Wallis ANOVA on ranks, Dunn’s).
`
`demyelinated lesions was 24.2; NAWM areas from the
`same section had 3.3 vessels (P<0.001). Non-EAE con-
`trols had a mean of 2 vessels/13.2 μm2 area (P<0.05,
`compared to NAWM areas) (Fig. 5).
`The number of blood vessels in the spinal cord of
`EAE animals strongly correlated with clinical score
`(r=0.810; P<0.01; n=64) and the pathological findings
`(r-values for M (0.841), P (0.850), E (0.818), D (0.772);
`P<0.01; n=64) (Fig. 6). The highest number of blood
`vessels was seen in guinea pigs that were chronic
`(>day 20 post-immunization), paralyzed (clinical score
`R2) and had spinal cord sections with the highest
`pathological scores.
`
`3.4. VEGF
`
`VEGF immunoreactivity was virtually absent in con-
`trol animals (Fig. 3C). In the acute phase of CP-EAE,
`VEGF appeared as early as day 7 post-immunization
`
`and were significantly higher than controls by day 15.
`VEGF expression increased during acute disease, peak-
`ing at day 26 post-immunization before gradually
`returning to baseline levels by day 70 (Fig. 4B and Fig.
`3F, I and L). A systematic qualitative analysis revealed
`VEGF staining was found predominantly in the vicinity
`of cuffed vessels (Fig. 3F) and at the border of demyeli-
`nating inflammatory lesions (Fig. 3I) but was absent
`from confluent demyelinated lesions (Fig. 3L). In con-
`trast to the cellular VEGF staining observed during
`acute EAE in rats [23], we observed “plaques” of
`parenchymal VEGF, surrounding intensely stained cells
`(Fig. 3F).
`There was no correlation between the area of VEGF
`expression and the number of vessels counted (r=0.144,
`Spearman, n.s.). This finding is consistent with the
`temporal sequence wherein peak VEGF expression (day
`26, Fig. 4B) preceded the appearance of the highest
`blood vessel counts (Fig. 4A). We had three cases
`
`

`

`358
`
`S.L. Kirk, S.J. Karlik / Journal of Autoimmunity 21 (2003) 353–363
`
`Fig. 3. Visualization of blood vessels and VEGF. All images are lumbar spinal cord sections at 400 magnification. Panels A, D, G and J are ScR
`stained sections, B, E, H and K are immuno-stained for factor VIII and C, F, I and L are immuno-stained for VEGF. Images A, B and C are serial
`sections from a non-EAE control at day 26. Images D, E and F are serial sections from an animal with moderate pathological scores (2–3) at day
`26. Sections show an infiltrated vessel without demyelination and high VEGF expression. Images G, H and I are serial sections at the edge of a
`demyelinating lesion (bottom left) from an animal at day 40. The area of demyelination and cellular infiltration shows a higher number of blood
`vessels compared with non-lesion white matter. A VEGF plaque is apparent at the border of the lesion. Images J, K and L are serial sections taken
`in a large demyelinating lesion in an animal with severe pathological scores (4) at day 70 post-immunization. Sections show extensive demyelination,
`myelin debris and cellular infiltration with numerous blood vessels and no VEGF expression. Arrows point to some examples of factor VIII stained
`blood vessels.
`
`(clinical score R2) that had a particular dichotomy
`of VEGF expression and blood vessel counts that
`illustrates
`this observation.
`In these animals,
`the
`
`lumbar–thoracic spinal cord sections showed a high
`number of blood vessels and pathological changes
`with little or no VEGF expression. The adjacent
`
`

`

`S.L. Kirk, S.J. Karlik / Journal of Autoimmunity 21 (2003) 353–363
`
`359
`
`Fig. 4. Temporal changes in number of blood vessels and VEGF. (A) Number of vessels increases with disease progression. A significant increase
`in factor VIII+ vessels was observed by day 15 post-immunization and remained high throughout the course of disease in CP-EAE animals (j) over
`controls (,). (B) Temporal sequence of VEGF expression. The area of VEGF immunoreactivity measured for the CP-EAE animals (j) increased
`through day 26 post-immunization before gradually returning to baseline by day 70. Control animals (,) were virtually devoid of VEGF
`immunoreactivity. Results are expressed as the meanSEM at each day of sacrifice. *Indicates a significant difference between non-EAE controls
`and EAE animals (P<0.05, Kruskal–Wallis ANOVA on ranks, Dunn’s).
`
`thoracic–cervical sections had low blood vessel counts
`accompanied by very high VEGF expression.
`
`4. Discussion
`
`A positive feedback relationship exists in which an
`inflammatory state can promote angiogenesis and angio-
`genesis
`can facilitate
`chronic
`inflammation. This
`relationship occurs through both the augmentation of
`cellular infiltration and proliferation and the overlap-
`ping roles of regulatory growth factors and cytokines
`[3,37]. In chronic inflammation, new blood vessels serve
`
`to transport inflammatory cells, nutrients and oxygen to
`the site of inflammation. Also, the resulting increased
`endothelial surface area enhances the production of
`cytokines, adhesion molecules and other inflammatory
`stimuli [3]. Inflammatory mediators act either directly or
`indirectly to promote angiogenesis;
`indirect action
`occurs through the induced expression of angiogenic
`factors such as VEGF [37]. Investigations into chronic
`inflammatory diseases such as rheumatoid arthritis and
`psoriasis have shown that angiogenesis plays a signifi-
`cant role in the progression of disease and does not
`simply act as a bystander, therefore presenting further
`therapeutic options [38].
`
`

`

`360
`
`S.L. Kirk, S.J. Karlik / Journal of Autoimmunity 21 (2003) 353–363
`
`Fig. 5. Spatial changes in blood vessel counts. ScR stained (A, B) (25 magnification) and factor VIII immunostained (C, D, E and F) (250
`magnification) lumbar spinal cord sections of control (day 90) (A, C, E) and CP-EAE (day 70, clinical score 2) (B, D, F). Box insets in A and B in
`the dorsal funiculi correspond to C, D and box insets in the lateral funiculi in A and B correspond to E, F. Chronic lesions (B) with infiltration and
`demyelination have an increased number of factor VIII stained vessels (F) compared to NAWM from the same section (D) or the corresponding area
`in control tissue (E). (G) The number of blood vessels in demyelinated lesions (j) is significantly higher than in NAWM of the same section (Q)
`(P<0.001, Paired t-test, n=18). Additionally, the number of vessels in the NAWM (Q) is higher than in non-EAE controls (,) (P<0.05, t-test, n=18
`EAE, n=17 non-EAE). Results are expressed as the meanSEM.
`
`

`

`S.L. Kirk, S.J. Karlik / Journal of Autoimmunity 21 (2003) 353–363
`
`361
`
`Fig. 6. Vessel counts correlate with tissue pathology. Scatter plots for vessel counts in CP-EAE (C) and controls (B) with blinded scoring of
`meningeal inflammation (A), perivascular infiltrates (B), myelitis (C) and demyelination (D). Spearman ♵-values were 0.841, 0.850, 0.818, and 0.772,
`respectively (all P<0.01, Spearman correlation, n=64).
`
`Circumstantial evidence for angiogenesis common to
`MS and EAE include cellular infiltration and prolifer-
`ation, the roles of MMPs, growth factors and cytokines,
`activities of ET-1 and NO and MRI findings. Herein we
`have shown that there is a significant increase in the
`number of blood vessels in chronic, demyelinating EAE
`and that VEGF, a key angiogenic signal for the prolifer-
`ation and migration of endothelial cells, appeared in the
`disease process with a peak expression on day 26, and
`preceded the observed vascular changes which persisted
`to day 90. Although the vessel counts for the acute phase
`(<day 20) were about 25 vessels per section consistent
`with the findings of Sobel et al. [39], we observed very
`high numbers of vessels in the chronic phase reaching a
`maximum of 169 vessels per section. The highest vessel
`counts were concentrated in demyelinating lesions.
`There was a small but significant increase in the number
`of blood vessels in the NAWM of CP-EAE animals
`compared to non-EAE controls. Previously, our labora-
`tory reported a change in the magnetization transfer
`
`ratio in the NAWM of CP-EAE guinea pigs [40]. Our
`results suggest that, as in MS, the NAWM in CP-EAE is
`also abnormal [41].
`The timing of the appearance of maximum VEGF
`expression suggests that VEGF signals the proliferation
`of vessels in growing lesions and then decreases as the
`lesions mature. Although we did not determine the cell
`type expressing VEGF, a previous study of MS and
`EAE demonstrated VEGF expression in astrocytes,
`macrophages and some endothelial cells [23].
`As VEGF is known as one of the earliest factors in
`angiogenesis [42], the presence of a high number of
`vessels without VEGF in proximity to tissue with lower
`vessel counts and high VEGF expression may indicate
`the evolution of the pathological process. This obser-
`vation is consistent with the ascending paralysis and
`progression of pathology seen this EAE model. In the
`early stages of the inflammatory response, the produc-
`tion of VEGF in normal appearing tissue may signal the
`formation of new blood vessels that would contribute to
`
`

`

`362
`
`S.L. Kirk, S.J. Karlik / Journal of Autoimmunity 21 (2003) 353–363
`
`persistent and invading inflammation. We cannot verify
`that VEGF expression precedes vessel proliferation in an
`individual tissue section because sampling was at one
`time point only. However, the combination of the tem-
`poral sequence for the entire cohort of animals (seen in
`Figs. 2 and 4) and the pattern seen in the three animals
`with the dichotomy of VEGF and vessels suggests that
`this scenario is possible.
`It has been suggested that therapeutic intervention in
`angiogenesis has the potential to alleviate chronic in-
`flammation [37]. Restricting nutrient and new inflam-
`atory cell delivery to sites of active inflammation,
`endothelial cell activation, proliferation and vascular
`remodeling are decreased. Consequently, endothelial
`cell derived factors such as MMPs and cytokines are
`also reduced [43]. Our results suggest that the process of
`chronic neuroinflammation has vascular features in
`common with other chronic inflammatory processes and
`that similar treatment strategies may be effective.
`
`References
`
`[1] Folkman J. Angiogenesis in cancer, vascular, rheumatoid and
`other disease. Nat Med 1995;1(1):27–31.
`[2] Stromblad S, Cheresh DA. Integrins, angiogenesis and vascular
`cell survival. Chem Biol 1996;3(11):881–5.
`[3] Jackson JR, Seed MP, Kircher CH, Willoughby DA, Winkler JD.
`The codependence of angiogenesis and chronic inflammation.
`FASEB J 1997;11:457–65.
`[4] Putnam TJ. Studies in multiple sclerosis: “encephalitis” and
`sclerotic plaques produced by venular obstruction. Arch Neurol
`Psychiat (Chic) 1935;33:929–40.
`[5] Putnam TJ. Evidence of vascular occlusion in multiple sclerosis
`and “encephalomyelitis”. Arch Neurol Psychiat (Chic) 1937;
`37:1298–321.
`[6] Noseworthy JH, Lucchinetti C, Rodriguez M, Weinshenker BG.
`Multiple sclerosis. N Engl J Med 2000;28:938–52.
`[7] Lucchinetti C, Bruck W, Parisi J, Scheithauer B, Rodriguez M,
`Lassmann H. Heterogeneity of multiple
`sclerosis
`lesions:
`implications for the pathogenesis of demyelination. Ann Neurol
`2000;47(6):707–17.
`[8] Ludwin SK. The neuropathology of multiple sclerosis. Neuro-
`imaging Clin N Am 2000;10:625–48.
`[9] Rindfleisch E. Pathological histology: an introduction to the
`study of pathological anatomy. Translated from the German by
`Kloman WC and Miles FT. London: Trubner and Co.; 1872,
`p. 652–658.
`[10] Kristensson K, Wisniewski HM. Chronic relapsing experimental
`allergic encephalomyelitis. Studies
`in vascular permeability
`changes. Acta Neuropathol (Berl) 1977;39(3):189–94.
`[11] Kwon EE, Prineas JW. Blood-brain barrier abnormalities in
`longstanding multiple sclerosis lesions. An immunohistochemical
`study. J Neuropathol Exp Neurol 1994;53(6):625–36.
`[12] Adams C. Vascular aspects of multiple sclerosis A color atlas of
`multiple sclerosis & other myelin disorders. UK: Wolfe Medical
`Publications, Ltd; 1989;184–201.
`[13] Tan IL, van Schijndel RA, Pouwels PJ, van Walderveen MA,
`Reichenbach JR, Manoliu RA et al. MR Venography of multiple
`sclerosis. AJNR Am J Neuroradiol 2000;21(6):1039–42.
`[14] Benveniste EN. Role of macrophages/microglia in multiple scler-
`osis and experimental allergic encephalomyelitis. J Mol Med
`1997;75:165–73.
`
`[15] Dines KC, Powell HC. Mast cell interactions with the nervous
`system: relationship to mechanisms of disease. J Neuropathol Exp
`Neurol 1997;56:627–40.
`[16] Van Meir EG. Cytokines and tumors of the central nervous
`system. Glia 1995;15:264–88.
`[17] Brod SA, Kerman RH, Nelson LD, Marshall GD Jr, Henninger
`EM, Khan M et al. Ingested IFN-alpha has biological effects in
`humans with relapsing-remitting multiple sclerosis. Mult Scler
`1997;3:1–7.
`[18] Giovannoni G, Miller DH, Losseff NA, Sailer M, Lewellyn-
`Smith N, Thompson AJ et al. Serum inflammatory markers and
`clinical/MRI markers of disease progression in multiple sclerosis.
`J Neurol 2001;248:487–95.
`[19] Ziche M, Morbidelli L. Nitric oxide and angiogenesis. J
`Neurooncol 2000;50:139–48.
`[20] Haufschild T, Shaw SG, Kesselring J, Flammer J. Increased
`endothelin-1 plasma levels in patients with multiple sclerosis.
`J Neuroophthalmol 2001;21:37–8.
`[21] Salani D, Taraboletti G, Rosano L, Di Castro V, Borsotti P,
`Giavazzi R et al. Endothelin-1 induces an angiogenic phenotype
`in cultured endothelial cells and stimulates neovascularization
`in vivo. Am J Pathol 2000;57:1703–11.
`[22] Shin T, Kang B, Tanuma N, Matsumoto Y, Wie M, Ahn M et al.
`Intrathecal administration of endothelin-1 receptor antagonist
`ameliorates autoimmune
`encephalomyelitis
`in Lewis
`rats.
`Neuroreport 2001;12:1465–8.
`[23] Proescholdt MA, Jacobson S, Tresser N, Oldfield EH, Merrill
`MJ. Vascular endothelial growth factor is expressed in multiple
`sclerosis plaques and can induce inflammatory lesions in exper-
`imental allergic encephalomyelitis rats. J Neuropathol Exp
`Neurol 2002;61(10):914–25.
`[24] Hiehle JF, Grossman RI, Ramer KN, Gonzalez-Scarano F,
`Cohen JA. Magnetization transfer effects in MR-detected mul-
`tiple sclerosis lesions: comparison with gadolinium-enhanced
`spin-echo images and nonenhanced T1-weighted images. A J N R
`1995;16:69–77.
`[25] He J, Grossman RI, Ge Y, Mannon LJ. Enhancing patterns in
`multiple sclerosis: evolution and persistence. A J N R 2001;
`22:649–64.
`[26] Gill M, Miller SL, Evans D, Scatliff JH, Meyerand ME, Powers
`SK et al. Magnetic resonance imaging and spectroscopy of small
`ring-enhancing lesions using a rat glioma model. Invest Radiol
`1994;29:301–6.
`[27] Rashid W, Parkes LM, Ingle GT, Chard DT, Symms M, Tofts PS
`et al. Comparative investigation of cerebral perfusion in multiple
`sclerosis using a novel technique. Baltimore (USA): ECTRIMS,
`2002.
`[28] Hyduk SJ, Karlik SJ. Apoptotic cells are present in the CNS
`throughout acute and chronic-progressive EAE in the absence of
`clinical recovery. J Neuropathol Exp Neurol 1998;57:602–14.
`[29] Sehested M, Hou-Jensen K. Factor VII related antigen as
`an endothelial cell marker in benign and malignant diseases.
`Virchows Arch A Pathol Anat Histol 1981;391:2