My earliest work exploring the role of Clostridium perfringens epsilon toxin (ETX) in triggering new lesion formation in Multiple Sclerosis (MS).

2006 Stanford Undergraduate Honors Thesis entitled, "Blood-Brain Barrier Breakdown: A Cause or Consequence of Lesions in Multiple Sclerosis?"

Blood-Brain Barrier Breakdown: A Cause or Consequence of Lesions in Multiple Sclerosis?

An Honors Thesis Submitted to the Department of Biological Sciences in partial fulfillment of the Honors Program

STANFORD UNIVERSITY

by Kareem Rashid Rumah

May 2006

Approved for submittal to the Department of Biological Sciences

For consideration of granting graduation with honors:

Dr. Ben A. Barres____________________ Date __________

Dr. Liqun Luo ____________________ Date __________

Abstract

            Multiple Sclerosis (MS) is a debilitating, inflammatory disease of CNS white matter tracts that results in gross blood-brain barrier (BBB) breakdown, demyelination and axonal injury.  Despite extensive characterization of MS lesions, very little is known about the precise order in which these pathologies occur.  The prevailing thought has been that BBB breakdown occurs as a consequence of inflammatory demyelination.  Here I report that BBB breakdown may, instead, be a causative event, triggering demyelination and setting the stage for an MS lesion to unfold.  I found that transient BBB disruption resulted in significant demyelination 2.5 months after disruption.  With this finding in mind, I proceeded to explore the involvement of Clostridium Perfringens type D in the etiology of MS.  This bacterium has previously been suggested as a trigger for MS, as it secretes a potent Epsilon toxin that causes gross BBB breakdown like that observed in MS.  Serendipitously, I found that distinct CNS white matter tracts express Clostridium Perfringens Enterotoxin Receptor-2, making these regions susceptible to damage by an additional secreted toxin, Clostridium Perfringens Enterotoxin.

Introduction

Multiple Sclerosis (MS) is a debilitating neurodegenerative disease that affects about 300,000 individuals in the United States of America.  This disease kills approximately 3,000 of its victims annually, and diminishes the quality of life for its remaining survivors [1].  MS mainly affects Caucasian populations with females being twice as commonly affected as males.  Numbness, tremor, impaired vision and motor function, weakness, bladder dysfunction and psychological changes are the most common symptoms of this illness.  This disease can wax and wane for up to 30 years, but in perhaps half of all cases it steadily progresses to severe disability and premature death [1].

A classical MS lesion is characterized by the loss of CNS myelin, the fatty insulation of CNS axons which facilitates fast electrical conduction.  This pathological feature is called demyelination.  Immune infiltrates, such as activated T cells and phagocytic macrophages, heavily populate demyelinating MS lesions, leading to the assumption that MS is an autoimmune disease.  There is little doubt that the immune system contributes to the demyelination observed in MS, however, scientists have also entertained the idea that an environmental trigger, viral or perhaps bacterial, may cause primary insult to the CNS, triggering a cascade of secondary, immune mediated demyelination.

MS plaques vary in size and in number but are most conspicuous in areas with myelin enriched axon tracts, termed ‘white matter’.  These areas include the periventricular white matter, the visual system, deep white matter, brainstem, and spinal cord [11].  Active MS lesions are usually localized to white matter regions and are characterized by a mixture of lipid-laden macrophages, large reactive astrocytes, accompanied by varying perivascular inflammation [11].  Axons are relatively well preserved, although where damage is most severe, axons are also lost or fragmented and display irregular tortuous and clubbed profiles [11].  Besides focal demyelinated plaques, global diffuse injury of the so-called “normal”- appearing white matter (NAWM) is found in the brains of MS patients [11].  The pathology of NAWM in MS is characterized by diffuse, mainly CD8+, T-cell infiltrates, gliosis, microglial activation, diffuse axonal injury and nerve fiber degeneration [11].  Chronic inactive MS lesions are hypocellular with no evidence of active myelin breakdown.  Reactive gliosis is prominent and, although axonal density is relatively preserved when compared to the axonal loss observed in an infarction, it may still be markedly reduced [11].  Mature oligodendrocytes (the myelinating cells of the CNS) are markedly diminished or absent in chronic inactive lesions and inflammation, especially in perivascular regions, often remains [11].

The pathologic analysis of MS lesions reveals a profound heterogeneity between patients with respect to the inflammatory response, oligodendrocyte survival and patterns of demyelination [11].  Contrastingly, within a given patient a striking homogeneity among lesions is observed when considering pathologic features.  MS may, therefore, represent a common name for different pathologic entities that unify on the special vulnerability of CNS myelin to various immune and toxic mediators [11].  The heterogeneous phenotypes observed in MS patients have been classified into four distinctions:

Pattern I: Macrophage- associated demyelination

Pattern II: Antibody/complement-associated demyelination

Pattern III: Distal dying-back oligodendrogliopathy

Pattern IV: Primary oligodendrocyte degeneration

Although each of these patterns involve an inflammatory response composed mainly of T lymphocytes and macrophages, differences in plaque geography, extent and pattern of oligodendrocyte pathology, immunoglobulin deposition, complement activation, and myelin protein loss are observed [11].

In patterns I and II, macrophages and T cells predominate in well demarcated plaques that surround veins and venules; only pattern II lesions demonstrate local preciptation of immunoglobulin and activated complement in regions of active myelin breakdown.  The expression of all myelin proteins is similarly reduced [11].  Pattern III lesions also contain an inflammatory infiltrate, composed of macrophages, activated microglia and T-cells; however, the lesions are ill defined and seldom surround vessels.  There is no evidence of immunoglobulin deposition or complement activation, and Myelin Associated Protein (MAG) is selectively lost compared to other myelin proteins.  MAG is a myelin protein localized to the most distal extension of the oligodendrocyte cell body (the periaxonal region), and early loss of this protein is believed to reflect a dying-back phenomenon in the oligodendrocyte, which precedes the apoptosis found to be prevalent in pattern III [11].  Pattern IV lesions also contain T-cells and macrophages but no preferential MAG loss, immunoglobulin deposition or complement activation is observed. Instead there is evidence of non-apoptotic oligodendrocyte death, the mechanisms of which are unclear [11].  In summary, inter-individual rather than intra-individual differences in lesion heterogeneity with respect to demyelination is a characteristic feature of MS [11].

Although demyelination is thought to be the major pathology occurring in MS, axonal damage has been consistently observed.  Only recently have researchers paid due attention to axonal pathology, as it is thought to result in the irreversible disability associated with MS [11].  Although MS includes inflammatory and demyelinating components, their relative influence on axonal loss is unclear.  Pathologic studies reveal that myelin and axonal pathology may occur independently [11].  The damage to axons does not seem to depend on the stage of demyelinating activity.  Furthermore, acute axonal injury is found in the NAWM of MS patients [11].

In addition to demyelination and axonal damage, an MS lesion also results in a gross breakdown of the neuroprotective blood-brain barrier (BBB).  The BBB is a protective barrier formed by the blood vessels of the brain. The capillary network of the central nervous system forms a barrier that limits the flow of solutes from the blood to the brain.  This barrier is extremely important for maintenance of brain homeostasis and suitable ionic concentrations for controlled neuronal excitability.  In addition to preventing the entry of many systemic toxins from entering brain tissue, the BBB also reduces the frequency of brain immune surveillance when compared to other organs.

To perform such specialized functions, brain endothelial cells (ECs), which form the lumen of blood capillaries, possess many unique properties when compared to endothelial cells in the periphery.  CNS ECs form high electrical resistance tight junctions between their adjoining cell walls [17].  Tight junctions prevent virtually all molecules from entering the brain by forming tight seals between adjacent EC membranes, creating fence like structures.  This adaptation prevents much of the paracellular transport that occurs in other organs.  CNS ECs also display lower rates of transcytosis and lack the fenenstra observed in peripheral ECs [18], adaptations which limit the transcellular transport of fluid phase molecules.  Fenestrated vessels are no more permeable to plasma proteins than are BBB vessels, but such vessels are much more permeable to water, ions, and small solute molecules [18].  Meanwhile, lipophilic molecules of low molecular weight can enter the brain through passive diffusion [17].  However, the brain possesses transporters such as P-glycoprotein, which generally transports these lipophilic molecules back into the blood [17].  While these features have evolved to maintain a stable CNS environment and restrict the entry of toxins, active transporters present on the surface of endothelial cells supply the brain with specific nutrients and survival factors. 

In MS, the prevailing idea has been that BBB breakdown results from the inflammatory autoimmune response and is thus a consequence of a MS lesion.  While scientists have historically assumed this to be the case, this view is now under question, as recent studies have suggested that hypoxic tissue pre-conditioning occurs in MS lesions [10,11, and 12].  In these lesions, cells have been reported to express hypoxia inducible protein alpha and heat-shock protein 70; both proteins are typically upregulated after hypoxic insult [10, 11, and 12]. This hypoxia has been postulated to occur after the BBB breakdown that results from microvascular disruption.  Despite these results, no studies have been performed to answer the question:  could BBB breakdown be a causative event, setting the stage for CNS demyelination, rather than a consequential phenomenon?

 Several different stimuli may cause BBB disruption:

1)  Environmental triggers such as heavy metal exposure [16].

2)  A primary autoimmune response directed against BBB components [7].

3)  Viral or bacterial pathogens [4].

In this paper I wish to explore the hypothesis that BBB breakdown is a preceding event that allows immune components to enter the brain to directly attack CNS myelin.  Additionally, I wish to explore the idea that BBB breakdown allows the entry of a systemic toxin that initiates or potentiates demyelination, which the immune system further propagates.  I will first present an experiment in which I test the hypothesis that transient breakdown of the BBB can result in demyelination.  The BBB of Sprague Dawley rats will be transiently disrupted, and demyelination and behavioral abnormalities will be assayed at various time points after BBB breakdown.  Secondly, I will explore the potential role of the BBB altering bacteria, Clostridium Perfringens types B and D, in the pathogenesis of MS.

Materials and Methods

Optic Nerve Crush Surgery.  All experimental procedures were in agreement with approved Stanford University animal care protocols. Adult Sprague Dawley rats and Swiss Webster mice were anaesthetized with an intraperitoneal injection of ketamine (40 mg/kg)/xylazine (10 mg/kg). The ON of the left eye was crushed intraorbitally in all experimental animals sparing the retinal artery. Briefly, after skin incision close to the superior orbital rim the orbit was exposed and part of the lachrymal gland and the dorsal eye muscles were dissected. The ON was exposed and compressed in jeweller's forceps until a visible nerve crush was performed at about 0.5 mm behind the eye cup.

Animals and SMI71 administration.  Juvenile Sprague Dawley rats (40g) were given intravenous (IV), tail vein injections in 1 ml per kg body weight commercially available antibody SMI71 (~40uL/kg IV) or saline control. A heating pad was used to dilate the animal’s tail vein and I proceeded to inject at a single site that is clean and free of debris.

Animals and toxin administration. Epsilon-prototoxin was purified from cultures of C. perfringens type D. It was activated by tryptic digestion and then diluted with a 1% Bacto peptone-saline solution to avoid overdigestion. Male or female BALB/c mice weighing 25–30 g were injected intravenously (tail vein) with 50% mouse LD₅₀  (approximately 70 ng kg⁻¹) or vehicle (1% Bacto peptone-saline solution containing trypsin). All experimental procedures were approved by the Animal Care and Use Committee of the California Animal Health and Food Safety Laboratory, University of California, Davis (permit 34).

Immunofluorescence. For immunostaining of optic nerves, adult rats were perfusion fixed with 4% paraformaldehyde.  The optic nerves were dissected and immediately fixed in 4% paraformaldehyde for an additional hour. The tissues were cryoprotected in 30% sucrose, and 7µm sections were cut and mounted on aminoalkylsilane-coated slides (Sigma). Alternatively, optic nerves and brains from both adult rats and mice were dissected from freshly sacrificed animals and immediately frozen in 100% OCT.  These tissues were sectioned at 7µm and post fixed by a 1hr submersion in 95% Ethanol followed by 1 min submersion 100% Acetone. All Sections were blocked and permeabilized with 50% goat serum and 0.4% Triton X-100 for 1hr. Section were incubated overnight at 4°C using rabbit anti-claudin 3 (1:200 dilution, Zymed), rabbit anti-OSP (1:200 dilution, Zymed),  monoclonal mouse anti-Neurofilament unphosphorlated (1:1000 dilution; Sternberger Inc.), monoclonal mouse anti-Neurofilament phosphorlated (1:1000 dilution; Sternberger Inc.),  monoclonal mouse anti-CD68(1:100 dilution; Serotec.),  monoclonal mouse anti-vimentin (1:500 dilution; Sigma Aldrich) to detect astrocytes, monoclonal mouse anti-Myelin Associated Glycoprotein (1:1000 dilution; Chemicon), BSL (1:200 dilution; Vector Laboritories ) to detect capillaries, Fluoromyelin (1:300 dilution; Invitrogen ) to detect myelin, followed by Alexa-conjugated anti-goat IgG (1:500; ) for 2 hr at room temperature. The coverslips were mounted in Vectashield with DAPI, sealed with nail polish, and examined in a Nikon Diaphot fluorescence microscope.

Western blotting. Western blotting was performed using typical protocols. Optic nerve and liver lysates were transferred to 0.45 µm nitrocellulose (Bio-Rad, Hercules, CA) and incubated for overnight with rabbit anti-claudin 3 (1:500 dilution) and 1 hr with horseradish peroxidase-conjugated secondary Ab (1:6000; Chemicon, Temecula, CA), both in 5% milk. Detection was performed using ECL (Amersham Biosciences, Arlington Heights, IL).

Results

Expt. 1.1 Optimization of a method for transiently breaking down the BBB.

To test whether the BBB could be successfully broken down, I took advantage of the commercially available antibody SMI71.  SMI71 is a monoclonal antibody that binds to an unknown antigen on the luminal surface of brain capillaries [2].  Upon binding, this antibody transiently breaks down the BBB for approximately 30 min. – 1 hr [2].

To determine if the BBB could be experimentally broken down, I intravenously injected juvenile Sprague Dawley rats with 40ul/kg SMI71.  To assay whether BBB breakdown had occurred, I next perfused the rats with a solution containing biotin by transcardiac injection, followed by perfusion with a 4% paraformaldehyde solution.  I prepared brain cryosections and stained them with Strepavidin-488 in order to detect, by flourescence microcopy, whether biotin had entered the brain tissue.  As shown in Figure 1, I found that biotin entered the CNS parenchyma of SMII71 injected animals, but remained confined to the blood vessels of animals injected with a control monoclonal antibody against CD31 (a luminal endothelial cell antigen with no known disruptive activity).  These data confirmed that the BBB had been successfully disrupted by this method, as previously reported.  Although SMI71 only transiently broke down the BBB, I found that antibody doses exceeding 40ul/kg resulted in death, demonstrating the importance of the BBB in animal viability.

Expt. 1.2 Does transient BBB breakdown cause demyelination?

I next investigated whether BBB breakdown is sufficient to cause demyelination.  I injected fourteen test and seven control P20 (postnatal day 20) rats with SMI71 and anti-CD31 monoclonal antibodies, respectively.  At various time points after antibody injection, I sacrificed some of the rats and prepared brain and optic nerve cryosections in order to determine whether demyelination could be detected by immunohistochemical staining.  I stained the brain and optic nerve cryosections with a rabbit polyclonal antibody against Oligodendrocyte Specific Protein (OSP is a compact myelin protein) and with a mouse monoclonal antibody against MAG (a non-compact myelin protein).  The optic nerve was the CNS white matter region chosen for inspection, as it is abundant in myelin and is one of the first and most commonly affected central nervous system regions in MS [19]. 

In my first experiment, I analyzed four SMI71 animals and two CD31 animals 2.5 months after injection.  I detected demyelination in one of the optic nerves in three of the four SMI71 injected animals but did not observe demyelination in any the optic nerves harvested from the two control CD31 animals (Figure 2).  Although there was evidence of demyelination in the test animals, I did not detect immune infiltrates in the demyelinated plaques.  The lesion sites were not hypercellular. This lack of hypercellularity was reminiscent of the chronic inactive lesions described above.  I hypothesize that the demyelination was caused by immune cells that entered the optic nerve early after BBB breakdown, but then migrated away.  An alternative possibility is that the demyelination was not cell mediated, but induced by humoral factors present in the serum.  The hypothesis that immune cells may be present at earlier time points is currently being tested by assaying immune infiltration and demyelination within days and weeks of BBB breakdown.

To determine whether the demyelination occurring after BBB breakdown affected neurological function, I next examined the behavioral proficiency of a separate group of animals eight and fifteen months after injection.  To test behavioral proficiency, I examined lower limb and tail strength. There seemed to be no behavioral defects in test or control animals (data not shown).  Demyelination was also assessed after fifteen months, but no lesions were observed.  Interestingly, recent studies have suggested that demyelinating lesions in the optic nerve are less capable of repair than in other CNS regions [29].  Thus I hypothesize that many CNS demyelinating lesions may have occurred at early timepoints after BBB breakdown that spontaneously remyelinated, in contrast to the lesions I observed in the optic nerve.  Unexpectedly, however, two SMI71 but no CD31 animals injected rats died fourteen months after inoculation, yet these animals showed no overt behavioral defects eight months after injection.  Unfortunately, I was unable to recover the bodies for autopsy.  I am currently observing the remaining animals for behavioral defects, and will soon assay demyelination by immunohistochemistry.

Although I need to do many more experiments, which I plan to continue during the next year, so far my experiments provide evidence that BBB breakdown, by itself, is sufficient to cause demyelination.  This finding challenges the prevailing idea that BBB breakdown is merely a consequence of a demyelinating lesion.  The discovery that BBB breakdown is sufficient to induce demyelination is important because it raises new hypotheses about the initial trigger of Multiple Sclerosis, one of which I have tested (see below).  At least some of the demyelination in these experiments seems reparable and does not seem to result in full-blown MS.  What additional stimuli may cause an episodic form of demyelination and/or more extensive demyelination?  An intriguing possibility is that the demyelination that occurs after BBB breakdown is exacerbated by the entry of a toxin to CNS white matter that would otherwise be excluded from the brain parenchyma by an intact BBB.  One provocative candidate for this hypothesis is Clostridium Perfringens types B and D enterotoxaemia. 

Expt. 2.1  Does CNS injury result in white matter susceptibility to Clostridium Perfringens Enterotoxin?

Clostridium Perfringens Epsilon toxin (ETX) is a potent toxin secreted from type B and D strains.  This toxin binds to an unknown receptor on the luminal surface of brain endothelial cells and causes microvascular injury, which results in BBB breakdown and cerebral edema [21].  Clostridium Perfringens types B and D also secrete an Enterotoxin (CPE). CPE is a cytolytic toxin, which binds to claudin 3 positive cells and results in cellular damage by increasing intracellular calcium concentrations [3]. Assaying claudin 3 immunoreactivity after surgical optic nerve injury led to the serendipitous finding that this CPE receptor is upregulated in the degenerating nerve, but is undetectable in the contralateral, uninjured nerve (Figure 3).  This led to the hypothesis that CNS injury, such as the edema that occurs after BBB breakdown, may also result in claudin 3 upregulation and susceptibility to CPE, as it can now enter the brain.

One week after optic nerve crush, the myelin closest to the axon begins to degenerate.  This degenerating debris takes the form of ovoid structures, in which the periaxonal myelin protein, MAG is found. The formation of these periaxonal myelin ovoids has previously been described after spinal cord crush [3]. Surprisingly, within these degenerating ovoids, claudin 3 (Clostridium Perfringens Enterotoxin Receptor – 2) accompanies MAG (Figure 4). I later confirmed Claudin 3 expression by Western Blot analysis (Figure 5).

This serendipitous result led to the hypothesis that a non-surgical injury, such as the prolonged edema caused by ETX mediated BBB breakdown, may be sufficient to induce claudin 3 expression and white matter susceptibility to CPE.  Not only could ETX cause demyelination by merely breaking down the BBB, but perhaps it could also cause the injury needed to facilitate more extensive demyelination by inducing claudin 3 expression and white matter susceptibility to CPE.

To locate the cellular origin of claudin 3 expression, I co-labeled claudin 3 with components of other cell types in the CNS parenchyma such as axonal neurofilament , astrocytic vimentin , and the activated macrophage marker, CD68 (Figure 6).  There was no apparent overlap, between claudin 3 and these markers.  Furthermore, no CNS component showed as close an association with claudin 3 as Myelin Associated Glycoprotein.  Despite these results, it remained difficult to confirm the cellular origin of claudin 3, as the nerve was not intact, but in the process of degenerating.  Claudin 3 clearly associated with degenerating debris, and was eventually cleared from the CNS.  These data suggested that either neurons or oligodendrocytes expressed claudin 3, as these cells are known to degenerate after optic nerve crush.

Expt. 2.2 Can prolonged BBB breakdown by Clostridium Perfringens Epsilon toxin result in white matter susceptibility to Clostridium Perfringens Enterotoxin?

In collaboration with Dr. Francisco Uzal of the University of California at Davis, I explored this hypothesis.  The Uzal lab inoculated four adult BALB/ c mice with ETX at half L₅₀ and four adult mice with a similar volume of vehicle control.  They sacrificed one ETX injected animal at three, five and seven days after inoculation and proceeded to inject the fourth ETX animal once more with ETX, a week after the first inoculation.  Three days after the second injection, they sacrificed the fourth animal.  Dr. Uzal inoculated one vehicle control animal following the same schedule as each ETX animal.  I snap froze, sectioned and fixed with Ethanol/Acetone or 4% paraformaldehyde, each test and control brain, and analyzed Claudin 3 immunoreactivity by fluorescence microscopy.  There was no detectable claudin 3 induction in myelin rich regions of the CNS such as the corpus callosum or cerebellum of any test and control animal (Figure 7). However, I did observe claudin 3 immunoreactivity in the brainstem of a control animal (Figure 8).  Upon further investigation, these claudin 3 cells predominantly localized to conserved white matter tracts in the brainstems of both control and ETX injected animals (Figure 9).  Surprisingly, unlike the optic nerve, the white matter tracts of the brain stem expressed claudin 3 without injury.

Observing claudin 3 immunofluorescence in preserved white matter tracts, rather than in degenerating optic nerves, allowed for better judgment of the exact localization of the receptor.   Claudin 3 is a transmembrane receptor and the intimate relationship between the apposing axonal and oligodendrocyte plasma membranes created great difficulty in determining the cellular origin of this molecule by immunohistochemistry.  To accurately pinpoint which cell type was immunoreactive for claudin 3, I co-labeled claudin 3 with the axonal markers, phosphorylated and non-phosphorylated neurofilament (Figure 10), and closely compared this micrograph to claudin 3 co-labeling with fluoromyelin, a marker that brightly stains the lipid enriched regions of the myelin sheath (Figure 11).  Claudin 3 seemed to colocalize well with neurofilament while it closely associated with, but did not co-localize with fluoromyelin, suggesting that the axonal plasma membrane, rather than the inner leaflet of the myelin sheath, expressed claudin 3.

This finding led to the reinvestigation of the cellular origin of claudin 3 in the degenerating optic nerve, as claudin 3 seemed to preferentially localize with the myelin protein MAG, rather than neurofilament.  I performed an optic nerve crush on an adult Sprague Dawley rat and harvested the optic nerves and retinas three days after crush.  The goal of this experiment was to determine if claudin 3 co-localized with neuronal soma in the retinal ganglion cells (RGC) layer of the retina or oligodendrocyte soma in the optic nerve.  Retinal ganglion cells (RGCs) project their axons along the optic nerve and are the primary cells to degenerate after optic nerve crush.  RGC cell bodies do not reside in the optic nerve as do their axons, but are instead found in a distinct layer in the retina [22].  I harvested all tissues three days after crush rather than a full week, to ensure that degenerating RGC cell bodies did not completely degenerate and could be observed for claudin 3 immunoreactivity.  In the optic nerve, I co-labeled claudin 3 with the oligodendrocyte cell body marker CC1.  These two markers failed to co-localize (Figure 12a).  I proceeded to co-label claudin 3 with the nuclear marker, DAPI, in the RGC cell body layer of the retina.  I observed claudin 3 immunoreactivity around the nuclei of RGCs of the crushed optic nerve, but not in the contralateral uncrushed optic nerve (Figure 12b and 12c).  These results suggest that the neurons in white matter tracts, rather than the oligodendrocytes, express claudin 3.

Discussion

The above data suggest that BBB breakdown can indeed result in demyelination.  This simple finding may allow us to more accurately define the pathological sequence that occurs during an MS relapse.  Revealing this piece of the puzzle may allow us to refocus our ideas as to what may cause MS.  The preeminent focus of MS research has been to examine the immune reaction against immunogenic myelin antigens; the assumption being that BBB breakdown is secondary to demyelination.  It may be more fruitful to explore ways that BBB disruption may occur that closely mimic the frequency and severity of the BBB disruptions observed in MS patients.

The results of Expt. 1.2 suggest that BBB breakdown may be a causative rather than a consequential phenomenon in relation to CNS demyelination, however, animals that survive transient BBB breakdown seem to be capable of repairing these lesions.  Why doesn’t this demyelination develop into the episodic demyelination observed in MS? One explanation worth consideration is that it may take years or even decades for these animals to develop a relapsing form of demyelination.  MS is a progressive disease, which worsens with time [6].  Secondly, MS is thought to have a genetic component.  Animals of different genetic backgrounds may yield varying demyelinating phenotypes, thus an animal model using a strain of rat more genetically susceptible to demyelination may provide a more severe phenotype.  Interestingly, in EAE animal models, differing strains produce differing degrees of demyelination [20].  Thus genetic background may be a crucial determinant of whether an initial triggering event, which transiently breaks down the BBB, eventually results in a relapsing and remitting illness.  Finding chromosomal loci linked to increased severity or susceptibility to demyelination may provide the information necessary to produce transgenic animals. Knockdown of these target genes may attenuate the severity of demyelination observed after BBB breakdown.

Additionally, the lesions in Expt. 1.2 are devoid of immune infiltrates.  Have these immune cells migrated away from the lesion site, leaving a chronic inactive lesion behind?  Or has the myelin sheath degraded without cell mediated demyelination?  A provocative publication, which gives some credence to the latter hypothesis, is the 2004 Barnett and Prineas study.  Amazingly, Barnett and Prineas were able to isolate and observe lesions as they formed.  This publication strongly challenges the prevailing views on MS pathology, as until this study, all pathological observations had been conducted on previously existing lesions [15].  These researchers found that newly forming lesions occurred in the absence of immune infiltrates, suggesting that immune cells arrive after a lesion has already ensued.  This finding contradicts the prevailing idea that immune cells mediate the demyelination that occurs during MS relapse [15].  Could this be what we have observed in the lesions of Expt.1.2?  Further investigation is currently underway to determine whether immune infiltrates populate these lesions at earlier time points.

As BBB breakdown can cause demyelination, it may be useful to ask what stimuli can cause episodic BBB disruption.  An interesting hypothesis is that a sporadic autoimmune response to components of the BBB may result in intermittent periods of demyelination.  To test this hypothesis, one could replicate the Experimental Autoimmune Encephalomyelitis (EAE) inoculation protocol, but replace myelin peptides with the homogenates of BBB components such as astrocyte endfeet and CNS pericytes.  CNS astrocytes and pericytes are thought to provide the microenvironment necessary for brain endothelial cells to acquire such unique, barrier forming properties [17].  Could CNS astrocytes and pericytes be prime targets in an autoimmune cascade resulting in periodic BBB breakdown?  It is unlikely that the body’s immune system will mount an autoimmune response to the CNS endothelium because immune cells are constantly in contact with CNS endothelial cells.  Autoimmune attack directed against CNS vasculature would lead to constitutive BBB breakdown, probably resulting in death.  Supporting this idea, an overdose of SMI71 results in death, presumably because of an overwhelming breakdown of the BBB.  Since immune cells survey the CNS parenchyma with a very low frequency [23], one could imagine a stochastic autoimmune response to the CNS astrocyte and/or pericyte BBB components, as they reside behind the barrier, and are seldom encountered by the immune system.  As CNS surveillance is episodic and low frequency, BBB disruption and consequent demyelination would follow suit.

Interestingly, antibodies to Aquaporin 4, a protein found in the endfeet of the astrocytes that ensheath bloodvessels, have been discovered in patients suffering from Devic’s disease, a demyelinating disease very similar to MS [7].  The 2006 publication “Neuromyelitis optica,” suggests that compromise of the BBB can precede demyelination.  The results from Expt. 1.2 support such a hypothesis.  Could this sequence of events be conserved in the demyelination observed in MS?

An alternate hypothesis is that a persistent infection may trigger cyclic BBB disruptions.  Several viral and bacterial pathogens are known to disrupt BBB function as they invade the brain.  However, to date, no bacterial or viral strain has been found to specifically populate the CNS parenchyma of MS patients.  Clostridium Perfringens types B and D distinguish themselves from such pathogens, as they can disrupt the BBB without invading the brain. Interestingly, Clostridium Perfringens releases the majority of its toxins during cyclical sporulation that can be triggered by various environmental stressors [21, 24].  Additionally, CPE is only produced during the sporulation phase of the bacterium’s life cycle [21].  As the sporulating bacteria lyse, they release not only their spores but also their enteric toxins [8]. The ETX that is secreted from these bacteria can disrupt the BBB by merely entering the blood stream.

Surprisingly, Clostridium Perfringens type D has been proposed as a trigger for MS in the past.  Dr. Timothy Murrell’s 1986 publication of A Review of The Sheep-Multiple Sclerosis Connection, reviews the notion that MS is associated with the concentration of global sheep populations.  He identifies historically aberrant MS outbreaks, some of which with astounding probabilities that make random chance unlikely [4].  Murrell proposes that these epidemics stem from human contact with sheep and goes on to offer possible pathogens that may be transmitted from sheep to humans, resulting in the development of MS.  Intriguingly, he offers Clostridium Perfringens type D as a possible infectious agent, as sheep are the natural reservoir for this bacterium.   According to Dr. Murrell, the toxins of Clostridium Perfringens type D cause MS like symptoms upon entering the animal’s blood stream [4].  He notes that afflicted sheep experience MS like symptoms such as blindness, central nervous system derangement due to malacia in brain areas with a distribution similar to MS [4].  Dr. Murrell also mentions that steroids are used to treat these animals and dually serve as standard treatments for MS patients.  Furthermore, he proposes that the beneficial effects of hyperbaric oxygen therapy in MS may be explained by reducing anaerobic pathogens such as Clostridium Perfringens.  Murrell ends his discussion of Clostridium Perfringens type D by suggesting that BBB breakdown by ETX may open the CNS to a neurotropic virus, which may cause the demyelination observed in MS.  However, data from Expt. 1.2 suggest that BBB breakdown alone may to trigger demyelination.

The results of experiments 2.1 and 2.2 demonstrate that claudin 3 (Clostridium Perfringens Enterotoxin Receptor-2) is expressed by myelinated axons of the rodent brain stem, and by the degenerating RGCs of the crushed optic nerve rather than by CNS myelin.  With this in mind, it is clear that CPE cannot cause direct damage to CNS myelin.  However, it is unclear as to whether CPE mediated insult to myelinated axons and/or the mere presence of foreign bacterial products in the brain may serve as adjuvant, further priming the immune system against immunogenic myelin peptides.  Such a cascade is reminiscent of EAE, and could exacerbate the demyelination that occurs after BBB breakdown.

The potential role of CPE in exacerbating demyelination is speculative.  However, CPE’s ability to injure white matter axons, which incur unexplained injury in MS [11], is certainly not.  Axonal injury in MS seems to be unrelated to the stage of demyelination [11].  Could CPE mediate the axonal damage observed in MS?  Luccinetti et al. discuss how axonal damage is likely to occur in MS.  According to this group, unknown agents trigger a cascade that leads to a disturbance in axoplasmic membrane permeability and subsequent energy failure.  This leads to uncontrolled sodium influx into the axoplasm, which reverses the sodium/calcium exchanger and results in excess intraxonal calcium, which activates calcium dependent proteases and degrades cytoskeletal proteins, further impairing axonal transport. Voltage-gated calcium channels (VGCC) accumulate at sites of disturbed axonal transport, leading to further calcium influx, eventual dissolution of the axonal cytoskeleton and axonal disintegration [11].

While Luccinetti’s proposal is certainly a plausible mechanism for how axonal injury may occur in MS, CPE mediated damage offers a calcium dependent mechanism that is far simpler.  CPE damages cells by forming a complex with the extracellular loop of claudin 3, forming a pore in the cell membrane.  This pore formation results in increased calcium influx into the cell and consequent activation calcium dependent proteases that degrade the cytoskeleton [13].  Moreover, calcium is required for CPE mediated toxicity, as claudin 3 positive cells are refractory to CPE in calcium free media [13].  In summary, the results of experiments 2.1 and 2.2 demonstrate that a fascinating relationship may exist between the toxins of Clostridium Perfringens types B and D and the mammalian CNS.  Could Clostridium Perfringens types B and/or D be the pathogenic/environmental triggers that MS researchers have postulated to exist?

If the toxins of Clostridium Perfringens types B and D are not found to be involved in MS, the expression of claudin 3 in degenerating axons may still be useful in elucidating the pathology of this disease.  Assaying the degree of axonal degeneration that occurs in MS lesions has proven to be a difficult task [28].  Using claudin 3 as a marker for degenerating axons may be a simple way of addressing this problem, providing that there is no basal claudin 3 expression in these cells, as is observed in the rodent brainstem.

In conclusion, it may be of interest to search for triggers of BBB breakdown to further our knowledge of MS pathology.  BBB disruptions may occur in many ways and autoimmune attacks directed against barrier components and pathogens with such unique barrier altering capacities as Clostridium Perfringens types B and D are certainly just a few possibilities.  It may be interesting to search for antibodies to barrier components and barrier altering pathogens, rather than just antibodies against myelin peptides as has been past method.  If clinical investigation shows no signs of such antibodies in MS patients, then perhaps an animal model for episodic BBB disruption may provide a more accurate model for MS than the current EAE model, which fails to accurately recapitulate the pathologies observed in Multiple Sclerosis [9].

References

1. Kidd PM. Multiple sclerosis, an autoimmune inflammatory disease: prospects for its integrative management. Altern Med Rev. 2001 Dec;6(6):540-66. Review.

2. Ghabriel MN, Zhu C, Hermanis G, Allt G. Immunological targeting of the endothelial barrier antigen (EBA) in vivo leads to opening of the blood-brain barrier. Brain Res. 2000 Sep 29;878(1-2):127-35.

3. Chakrabarti G, McClane BA. The importance of calcium influx, calpain and calmodulin for the activation of CaCo-2 cell death pathways by Clostridium perfringens enterotoxin. Cell Microbiol. 2005 Jan;7(1):129-46.

4. Murrell TG, O'Donoghue PJ, Ellis T. A review of the sheep-multiple sclerosis connection. Med Hypotheses. 1986 Jan;19(1):27-39.

5. Lucchinetti CF, Parisi J, Bruck W. The pathology of multiple sclerosis.
   Neurol Clin. 2005 Feb;23(1):77-105, vi. Review.

6. Ramsaransing GS, De Keyser J. Related Articles, Benign course in multiple sclerosis: a review. Acta Neurol Scand. 2006 Jun;113(6):359-69.

7. Wingerchuk D. Related Articles, Neuromyelitis optica. Int MS J. 2006 May;13(2):42-5.

8. Duncan CL. Related Articles, Time of enterotoxin formation and release during sporulation of Clostridium perfringens type A. J Bacteriol. 1973 Feb;113(2):932-6.

9. Sriram S, Steiner I. Experimental allergic encephalomyelitis: a misleading model of multiple sclerosis. Ann Neurol. 2005 Dec;58(6):939-45. Review.

10. Lassmann H, Reindl M, Rauschka H, Berger J, Aboul-Enein F, Berger T, Zurbriggen A, Lutterotti A, Bruck W, Weber JR, Ullrich R, Schmidbauer M, Jellinger K, Vandevelde M. A new paraclinical CSF marker for hypoxia-like tissue damage in multiple sclerosis lesions. Brain. 2003 Jun;126(Pt 6):1347-57.

11. Lucchinetti CF, Parisi J, Bruck W. The pathology of multiple sclerosis. Neurol Clin. 2005 Feb;23(1):77-105, vi.

12. Stadelmann C, Ludwin S, Tabira T, Guseo A, Lucchinetti CF, Leel-Ossy L, Ordinario AT, Bruck W, Lassmann H. Tissue preconditioning may explain concentric lesions in Balo's type of multiple sclerosis. Brain. 2005 May;128(Pt 5):979-87. Epub 2005 Mar 17.

13. Chakrabarti G, McClane BA.  The importance of calcium influx, calpain and calmodulin for the activation of CaCo-2 cell death pathways by Clostridium perfringens enterotoxin. Cell Microbiol. 2005 Jan;7(1):129-46.

14. Chakrabarti G, Zhou X, McClane BA.  Death pathways activated in CaCo-2 cells by Clostridium perfringens enterotoxin. Infect Immun. 2003 Aug;71(8):4260-70.

15. Barnett MH, Prineas JW. Relapsing and remitting multiple sclerosis: pathology of the newly forming lesion. Ann Neurol. 2004 Apr;55(4):458-68.

16. Struzynska L, Walski M, Gadamski R, Dabrowska-Bouta B, Rafalowska U. Lead-induced abnormalities in blood-brain barrier permeability in experimental chronic toxicity.  Mol Chem Neuropathol. 1997 Aug;31(3):207-24.

17. Rubin LL, Staddon JM.  The cell biology of the blood-brain barrier.  Annu Rev Neurosci. 1999;22:11-28. Review.

18. Kaya M, Chang L, Truong A, Brightman MW.  Chemical induction of fenestrae in vessels of the blood-brain barrier.  Exp Neurol. 1996 Nov;142(1):6-13.

19. Cerovski B, Vidovic T, Petricek I, Popovic-Suic S, Kordic R, Bojic L, Cerovski J, Kovacevic S. Multiple sclerosis and neuro-ophthalmologic manifestations.  Coll Antropol. 2005;29 Suppl 1:153-8. Review.
    
    

20. Dal Canto MC, Melvold RW, Kim BS, Miller SD.   Two models of multiple sclerosis: experimental allergic encephalomyelitis (EAE) and Theiler's murine encephalomyelitis virus (TMEV) infection. A pathological and immunological comparison.  Microsc Res Tech. 1995 Oct 15;32(3):215-29. Review.

21. McDonel JL.  Clostridium perfringens toxins (type A, B, C, D, E).  Pharmacol Ther. 1980;10(3):617-55. Review.

22. Holder GE. Electrophysiological assessment of optic nerve disease.  Eye. 2004 Nov;18(11):1133-43. Review.

23. Bailey SL, Carpentier PA, McMahon EJ, Begolka WS, Miller SD. Innate and adaptive immune responses of the central nervous system. Crit Rev Immunol. 2006;26(2):149-88.

24. Murrell TG, Ingham BG, Moss JR, Taylor WB.  A hypothesis concerning Clostridium perfringens type A enterotoxin (CPE) and sudden infant death syndrome (SIDS).  Med Hypotheses. 1987 Apr;22(4):401-13.

25. Murrell TG, Harbige LS, Robinson IC. A review of the aetiology of multiple sclerosis: an ecological approach.  Ann Hum Biol. 1991 Mar-Apr;18(2):95-112. Review.

26. Wolfgram F.  What if multiple sclerosis isn't an immunological or a viral disease? The case for a circulating toxin.  Neurochem Res. 1979 Feb;4(1):1-14. Review.

27. Sternberger NH, Sternberger LA. Blood-brain barrier protein recognized by monoclonal antibody.  Proc Natl Acad Sci U S A. 1987 Nov;84(22):8169-73.

28. Bruck W. Inflammatory demyelination is not central to the pathogenesis of multiple sclerosis.  J Neurol. 2005 Nov;252 Suppl 5:v10-5. Review.

29. Lachapelle F, Bachelin C, Moissonnier P, Nait-Oumesmar B, Hidalgo A, Fontaine D, Baron-Van Evercooren A. Failure of remyelination in the nonhuman primate optic nerve.  Brain Pathol. 2005 Jul;15(3):198-207.