Multiple sclerosis is an inflammatory autoimmune disorder of the CNS that is characterized by demyelination, gliosis, and varying degrees of axonal pathology. Linkage and/or association studies have identified one of the most significant regions relevant to multiple sclerosis, chromosome 19q13, where the apolipoprotein E (APOE) gene locus resides (Haines et al., 1996; Pericak-Vance et al., 2001). Convergent lines of clinical evidence indicate that the ϵ4 allele of the APOE gene (APOE4) is associated with significantly more severe disability, earlier progression, and poorer recovery following relapse in multiple sclerosis patients (Evangelou et al., 1999; Chapman et al., 2001; Fazekas et al., 2001; Schmidt et al., 2002). In a mouse experimental autoimmune encephalomyelitis (EAE) model of human multiple sclerosis, APOE knockout mice showed worse disability and more inflammatory infiltrates in the CNS compared with wild-type controls (Karussis et al., 2003). In addition, APOE-knockout mice failed to recover from EAE, suggesting defective neuronal and/or glial repair mechanisms in the absence of apoE.

More widely known for its function in lipid and cholesterol metabolism, apoE has also proven to be an important contributor to proper functioning of the nervous system. In disease, the APOE4 gene encoding the apoE4 protein isoform has been identified as a risk factor for the susceptibility to, and/or severity of a number of disorders, including Alzheimer's disease, traumatic brain injury, atherosclerosis, Parkinson's disease, stroke, and cerebral hemorrhage (Corder et al., 1993; Sheng et al., 1998; Chapman et al., 2001; Lynch et al., 2002). We have demonstrated that biologically relevant concentrations of apoE3 attenuate the CNS response to injury by modulating glial activation and associated release of glutamate, NO, and proinflammatory cytokines such as TNF-α and IL-6 (Laskowitz et al., 1998, 2001; Colton et al., 2002, 2004; Lynch et al., 2003). These studies suggest that apoE represents a pharmacogenomically validated target for the development of novel therapeutics for multiple sclerosis, which may function by multiple mechanisms relevant to the pathogenesis of, and recovery from, multiple sclerosis.

Limited by the macromolecular nature of apoE and its inability to penetrate the blood-brain barrier (BBB), we developed COG133, a peptide derived from amino acids 133–149 in the receptor-binding region of the apoE holoprotein. We have demonstrated that COG133 retains the neuroprotective and anti-inflammatory properties of apoE holoprotein, both in vitro and in vivo. COG133 suppressed activation of microglia and macrophages; reduced release of TNF-α, IL-6, and NO; and protected primary culture neurons from glutamate excitotoxicity and oxidative stress (Laskowitz et al., 2001; Misra et al., 2001). In vivo, COG133 exhibited neuroprotective activity in a mouse model of traumatic brain injury and suppressed lipopolysaccharide (LPS)-induced peripheral and CNS inflammation (Lynch et al., 2003, 2005). We now report that COG133 significantly attenuated the severity of EAE, rendered a moderate recovery from the disease, inhibited cytokine release, inhibited free radical release, and reduced splenocyte proliferation. To facilitate transmembrane permeability of COG133, we fused a protein transduction domain (PTD) derived from the Drosophila antennapedia protein to COG133 and created COG112. Impressively, molecular fusion of this PTD dramatically enhances all the examined bioactivities of COG133, including inhibition of inflammatory cytokine release and reduction of lymphocyte proliferation in vitro, thereby providing more potent therapeutic effects on EAE in vivo and by extension, multiple sclerosis.

Materials and Methods

Mice. Female C57BL/6J mice (12 weeks old) were purchased from The Jackson Laboratory (Bar Harbor, ME) (and housed in the Experimental Animal Facility at Duke University, Durham, NC). Animal care and experimental procedures were approved by the Duke University Animal Care and Use Committee.

Reagents. Myelin oligodendrocyte glycoprotein (MOG) peptide is derived from residues 35 to 55 of the mouse MOG protein (MEVGWYRSPFSRVVHLYRNGK), and COG133 is derived from residues 133 to 149 of human apoE (acetyl-LRVRLASHLRKLRKRLL-amide). The reverse sequence of COG133 (rCOG133; acetyl-LLRKRLKRLHSALRVRL-amide) with an identical size and amino acid composition was used as a control. Antennapedia peptide (acetyl-YRQIKIWFQNRRMKWKKC) was fused with COG133 to give COG112 (acetyl-YRQIKIWFQNRRMKWKKCLRVRLASHLRKLRKRLL-amide). All the peptides were synthesized by the Peptide Synthesis Facility at University of North Carolina (Chapel Hill, NC) with a purity >95%. LPS and IFN-γ were purchased from Sigma-Aldrich (St. Louis, MO), and pertussis toxin was purchased from List Biological Laboratories Inc. (Campbell, CA). Quantitative ELISA kits for TNF-α and IL-6 were obtained from BioSource International (Camarillo, CA). Colorimetric BrdU ELISA kit for cell proliferation was purchased from Roche Applied Science (Penzberg, Germany). Antibodies for flow cytometry, including CD3-fluorescein isothiocyanate, CD4-PE Texas Red, CD8-PerCP-Cy5, and CD45-APC were obtained from BD Biosciences (San Jose, CA). Mouse CD45 MicroBeads and MACS separation columns/magnets were purchased from Miltenyi Biotec Inc. (Auburn, CA).

Induction and Clinical Evaluation of Acute Monophasic EAE. EAE was induced in mice following the method of Linker et al. (2002) with minor modifications. First, 250 μg of pMOG_35-55 peptide in 0.1 ml of PBS was emulsified with an equal volume of complete Freund's adjuvant containing 500 μg of Mycobacterium tuberculosis H37RA (Difco, Detroit, MI). On the day of immunization (referred to as day 0), mice were injected s.c. with 100 μl of this emulsion on two adjacent sites at the base of the tail, followed by immediate injection of 200 ng of pertussis toxin via the tail vein (i.v.) and then a second pertussis toxin dose on day 2 postimmunization.

Following this encephalitogenic challenge, mice were monitored and weighed daily, and a clinical score (CS) was assigned by an investigator blinded to group assignments. Criteria for clinical score (CS) scoring were as follows: 0, no disease; 1, limp tail; 2, flaccid tail and abnormal gait (ataxia and/or paresis of hind limbs); 3, severe hind limb paresis; 4, complete paralysis with hind body; and 5, moribund or dead, using increments of 0.5 points for intermediate clinical findings. The onset of EAE was defined as the day a mouse showed a clinical score ≥0.5.

Peptide Treatment Regimens. To test the possible protective effect of apoE-mimetic peptides for EAE, mice were randomly assigned to four groups with 10 mice per group following immunization with MOG: 1) vehicle control; 2) COG133, 1 mg/kg; 3) COG112, 1 mg/kg; and 4) rCOG133, 1 mg/kg. All peptides were reconstituted in sterile saline and injected i.p. every other day starting on day 2 and ending on day 30 postimmunization.

To evaluate the possible therapeutic effects of these peptides, a post-treatment paradigm was adopted to treat the animals after onset of symptoms. In brief, on the day that the animals showed a clinical score of 2 or above, they were randomly designated to one of three groups that were intraperitoneally injected on a daily basis with either 1) saline vehicle, 2) COG133 at 1 mg/kg, or 3) COG112 at 1 mg/kg and evaluated each day until the end of the experiment.

Histopathology. On day 35 postimmunization, mice were anesthetized and perfused with 25 ml of PBS followed by 25 ml of 4% formaldehyde in PBS. Brains and spinal cords were dissected, post-fixed in 4% formaldehyde for 24 h, and embedded in paraffin. Five-micrometer-thick sections were cut from three levels of spinal cord (cervical, thoracic, and lumbar). Sections were stained with Luxol fast blue and eosin to reveal demyelination and peripheral leukocyte infiltration. Photographs of stained sections were taken with a Nikon Eclipse TE200 microscope equipped with a Nikon DXM1200 digital camera.

Quantification of T-Cell Infiltration into Spinal Cord by Flow Cytometry. Three mice from control or COG112-treated groups were randomly chosen on day 21 postimmunization. Single-cell suspension was achieved from freshly dissected spinal cords and then labeled with MACS mouse CD45 magnetic beads. CD45⁺ leukocytes were separated through a MACS separation column. After blocking with CD16/CD32 Fc Block (BD Biosciences), cells were stained for surface markers with preconjugated antibodies in fluorescence-activated cell sorting buffer. The cells were measured after a final washing step on a BDLSRII Instrument (BD Biosciences) in the Duke University Human Vaccine Institute Flow Cytometry Facility and analyzed using FlowJo software (TreeStar, San Carlos, CA).

Peritoneal Macrophage Culture, Treatments, and Nitrite Assays. Macrophages were obtained on day 35 postimmunization by peritoneal lavage and seeded to 96-well microplate at a density of 1 × 10⁵ cells/well in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum, 2 mM l-glutamine, 1% HEPES, 100 U/ml penicillin, and 100 μg/ml streptomycin at 37°C in a humidified incubator with 5% CO₂. Twenty-four hours later, the serum-containing medium was removed, and cells were washed once with serum-free medium. Designated concentrations of these peptides were applied 30 min before LPS + IFN-γ treatment. Conditioned medium was collected at 45 h after LPS and IFN-γ exposure for analysis of TNF-α and IL-6 or at 72 h for measurement of nitrite, a stable metabolite and marker of NO. In brief, nitrite levels in conditioned medium were determined by injecting 50 μl of sample into a 280 NOA analyzer (Sievers, Boulder, CO). NO levels are expressed as micromolar NO₂ per milligram of total protein. Total protein (micrograms per well) was measured using the bicinchoninic acid method (Pierce Chemical, Rockford, IL) according to manufacturer's instructions, with bovine serum albumin as a standard.

Measurement of TNF-α and IL-6 by ELISA. Levels of TNF-α and IL-6 in conditioned medium were quantified with ELISA according to manufacturer's instructions. First, 96-well ELISA plates were precoated with rabbit anti-mouse polyclonal to TNF-α or rat anti-mouse monoclonal to IL-6 for 12 to 18 h at 2–8°C and then blocked for 2 h at room temperature. Following incubation with samples or mouse TNF-α standards for 2 h at room temperature, wells were washed, and biotinylated secondary antibody and horseradish peroxidase conjugated-avidin were added for 1 h. After washing, tetramethylbenzidine and peroxide were added, and color was developed for 30 min and then stopped by acidification with 2 N H₂SO₄. Absorbance at λ = 450 nm was measured with a microplate reader (Molecular Devices, Sunnyvale, CA), and the concentrations were calculated with respect to the standard cure and expressed as the mean concentration (picograms per milliliter) ± S.E.M.

Real-Time Polymerase Chain Reaction Analysis for NOS2 Expression. To examine the effects of COG112 and COG133 on NOS2 mRNA expression, macrophages were harvested from 4 EAE mice from the saline vehicle-treated group on day 35 postimmunization at the end of the experiment and cultured in six-well plates. After pretreatment for 30 min with designated peptides at concentration of 5 μM, LPS was added to a final concentration of 100 ng/ml to stimulate the cells for 10 h. In brief, total RNA in macrophages was extracted using an RNeasy Mini kit (QIAGEN, Valencia, CA) and reverse-transcribed to cDNA using a High-Capacity cDNA Archive kit (Applied Biosystems, Foster City, CA). The resulting cDNA templates were then mixed with primer/probe sets for murine NOS2 and Taqman Universal PCR Master mix (Applied Biosystems). The PCR amplification was performed using an Applied Biosystems ABI 7000HT sequence detection system. Relative quantification of mRNA expression was calculated by the comparative C_T method (the cycle number at which the amount of amplified target gene reaches a fixed threshold) with the amount of target = 2^–ΔΔCT compared with the endogenous control (18S) (Livak and Schmittgen, 2001). Data are presented as the average fold change for triplicate samples compared with untreated-cells.

Splenocyte Proliferation Assay. Four C57BL6/J mice were primed with MOG using the same protocol as EAE induction. On day 6 postimmunization, a single-cell suspension was prepared from freshly dissected spleens and suspended at 2 × 10⁶/ml in RPMI 1640 medium containing 10% heat-inactivated fetal calf serum, 2 mM l-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin. Two hundred microliters of cell suspensions was added to 96-well microtiter plates containing MOG alone, MOG + COG112, or MOG + COG133 and incubated for 72 h at 37°C with 5% CO₂. Cell proliferation was determined by Colorimetric BrdU Cell Proliferation kit (Roche Applied Science) following manufacturer's instructions. In brief, 20 μl of BrdU labeling solution diluted 1:100 in culture medium was added to each well for the last 18 h of culture. After removing the medium, cells were fixed, and anti-BrdU peroxidase working solution was added to each well and incubated for 90 min at room temperature. Following several washes, substrate solution was added, color was developed, and absorbance was measured with a microplate reader at 370 nm.

Statistical Analysis. For each group, the average daily clinical score was calculated from day 10 postimmunization (when the clinical score for all mice was 0) to day 35 postimmunization and data were analyzed by the nonparametric Friedman ANOVA for repeated measures followed by Dunn's multiple comparison test. Maximal clinical score and cumulative clinical score were analyzed by the nonparametric Kruskal-Wallis ANOVA followed by Dunn's multiple comparison test, and differences between pairs of groups were tested by Mann-Whitney U-test. Analysis of percentage of mice in remission was conducted by the nonparametric Friedman ANOVA for repeated measures. The difference between treatment groups in the subset of mice in remission on a given day was conducted by Fisher's exact test. All biochemical experiments were done in at least triplicate and statistical significance was assessed by ANOVA followed by Dunnett's multiple comparison test. Prism 4.0 software (GraphPad Software Inc., San Diego, CA) was used for statistical analysis. All t-tests were two-tailed, and a level of p < 0.05 was considered significant.

Results

COG112 and COG133 Attenuated the Severity and Promoted Remission from Active EAE in a Pretreatment Paradigm. Because the lack of apoE enhances symptoms of EAE in mice (Karussis et al., 2003), we tested whether the apoE-mimetic peptides COG112 and COG133 could modify the clinical course of EAE in wild-type mice that express endogenous mouse apoE. We administered COG112, COG133 (both at 1 mg/kg), or saline vehicle by i.p. injection starting on day 2 postimmunization and continuing every other day through day 30 of the experiment. Clinical observations continued through 35 days after initial induction of EAE. As depicted in Fig. 1, the onset and severity of EAE were significantly improved, albeit to varying degrees, following alternate day injections of COG112 or COG133 compared with saline-treated mice. The day of onset was significantly delayed following treatment with COG133 compared with saline (15.5 ± 0.63 versus 13.5 ± 0.35, respectively; p < 0.05), whereas COG112 had no effect on this parameter (13.5 ± 0.6). The magnitude of clinical score on the day of onset was equivalent for all groups (clinical score 0.95–1.0) and was followed by a progressive course of neurological impairment. Saline-treated mice exhibited a maximal clinical score of 3.3 ± 0.16, which was decreased in COG112-treated mice (clinical score 2.0 ± 0.35; p < 0.001 versus saline), and to a lesser degree in COG133-treated mice (clinical score 2.4 ± 0.54; p = 0.054 versus saline), a treatment that barely missed statistical significance due in to the death of a COG133-treated mouse that resulted in a clinical score of 5 for that one mouse. Analysis of cumulative disease burden, based on the sum of clinical scores throughout the duration of the study (CS_cum), indicated that disease severity was significantly decreased in the presence of COG112 (CS_cum = 14.6 ± 3.6; p < 0.001) and COG133-treated mice (CS_cum = 29 ± 10.5; p < 0.05) compared with saline-treated mice (CS_cum = 50.7 ± 3.9). The difference in CS_cum scores between COG112 and COG133 was not significant. These data demonstrate that COG112 and COG133, when administered every other day from day 2 through day 30 postimmunization, significantly decreased the clinical outcome severity scores in these EAE mice.

Another index of disease that profoundly changed after treatment with COG112 and COG133 is the percentage of mice exhibiting remission of disease, which was defined as mice that previously exhibited clinical signs of EAE greater than 1 and then returned to a clinical score of 0. An ANOVA of the percentage of mice in remission starting on day 19 postimmunization, when no mice were in remission, through day 35 indicated a statistically significant treatment effect (p < 0.0001; Fig. 1B), with post hoc analyses indicating that over the course of the study, significantly more mice treated with COG112 (p < 0.001) or with COG133 (p < 0.05) were in remission compared with saline-treated mice, whereas the difference between COG112- and COG133-treated mice was not significant (p > 0.05). Fisher's exact analysis on days 25, 30, and 35 postimmunization, when the EAE mice start the remission phase, showed that treatment with COG112 or COG133 resulted in a higher percentage of mice in remission compared with saline treatment. Together, these data indicate that COG133 conjugated to a PTD (i.e., COG112) or COG133 alone not only decreased the severity of the disease but also robustly promoted remission from active symptoms of EAE disease.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1.

Treatment with apoE-mimetic peptides significantly improved the clinical outcome of MOG-induced EAE in C57BL/6J female mice. MOG-immunized mice were randomly assigned to three groups receiving i.p. injection of saline (○), COG112 (♦), or COG133 (▴) at the dose of 1 mg/kg every other day starting on day 2 postimmunization and ending on day 30 postimmunization. A, mean of the daily clinical score for each group of mice is shown. ANOVA of clinical scores from day 10 postimmunization (when no mice were exhibiting clinical score) through day 35 postimmunization indicated a statistically significant treatment effect, p < 0.0001, and both COG112 and COG133 were significantly different from saline (p < 0.001). B, treatment with COG112 and COG133 significantly promoted recovery from MOG-induced EAE in C57BL/6J female mice. Remission was defined as a return to clinical score of 0 in mice that previously exhibited clinical signs of EAE. The percentage of mice in remission is shown for each treatment group. ANOVA indicated a statistically significant treatment effect, p < 0.001, with significant differences between saline versus COG112 (p < 0.001) and saline versus COG133 (p < 0.05).

To precisely control the readout of these studies, both saline vehicle and reverse-COG133 peptide (with the reverse sequence of COG133 and the same molecular weight) were used. We found that clinical scores of saline- and reverse COG133-treated mice were not statistically different for any of the EAE disease parameters that were investigated (data not shown). As a control for COG112, a peptide corresponding to its antennapedia prefix domain was tested in a separate pilot experiment. No difference was found between antennapedia peptide and saline; therefore, only data from saline-treated mice are presented.

COG112 and COG133 Promote Recovery from EAE in a Post-Treatment Paradigm. The success of apoE mimetics to ameliorate EAE using a preventive administration paradigm led us to examine whether apoE-mimetic peptides could demonstrate a therapeutic effect in animals with established disease. C57BL6/J mice were inoculated with MOG to induce EAE using the same protocol as in the previous experiment. On the day when the animals showed a clinical score ≥2, they were randomly assigned to groups treated with COG112, COG133, or saline vehicle by daily i.p. injections. As shown in Fig. 2, A and B, COG112 and COG133 significantly slowed the deleterious progress of the disease and promoted recovery to a normal clinical behavior. Even though this treatment starts much later (from approximately day 14 postimmunization on average) than that of the pre-treatment paradigm (day 2 postimmunization), COG112 and COG133 still significantly reduced the maximal severity of disability. The most profound improvement was found with post-treatment of COG112 where significantly more animals displayed complete recovery or remission from the disease compared with saline controls (Fig. 2C). On day 35 postimmunization, 71% of COG112-treated and 50% of COG133-treated mice exhibited complete remission; however, only 22% of mice in the saline control group completely recovered. In addition, COG112, but not COG133 or saline controls, also significantly reduced the days of sickness, counting from the first day that animals showed clinical signs to the day when they fully recovered from the disability of EAE (data not shown).

COG112 and COG133 Reduced Demyelination and T-Cell Infiltrates in the Spinal Cord. One of the pathological hallmarks of multiple sclerosis is extensive demyelination of the myelin sheath surrounding neurons accompanied by massive infiltration of peripheral leukocytes into the brain and the spinal cord, which are associated with a diversity of neurological disabilities in the disease state. We further examined the effects of apoE-mimetic treatments on demyelination and leukocyte influx into the spinal cords by Luxol fast blue and eosin staining. Saline-treated mice displayed severe demyelination in all segments of the spinal cord as depicted in Fig. 3, A to C, corresponding to the cervical, thoracic and lumbar segments, respectively. As shown in Fig. 3G (magnified picture of the inset in Fig. 3A), two demyelination plaques are clearly visible as a loss of blue staining of the white matter on the left side, and there are increased numbers of mononuclear cells infiltrating these demyelinated areas. This is consistent with the hypothesis that infiltrating leukocytes are the major contributor to pathogenesis and demyelination in both multiple sclerosis and EAE. Interestingly, COG112 and COG133 treatments greatly reduced this demyelination as depicted by an absence of observable demyelination in these regions in the representative section of a COG112-treated mouse in Fig. 3 (D, E, F, and H corresponding to inset in D) (pictures from COG133-treated animals not shown). These histopathological findings are consistent with the decreased clinical severity of EAE observed in COG112- and COG133-treated mice.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

COG112 and COG133 promote the recovery of animals with clinical disability of EAE when applied after the onset of disease. C57BL/6J female mice were first immunized with MOG to create an EAE model. On the day that an animal first showed a clinical score ≥2, it was randomly assigned to a group treated daily with COG112 (1 mg/kg i.p.) or COG133 (1 mg/kg i.p.) or normal saline. A, changes in the mean daily clinical score after treatment with COG112 or COG133. Statistical analysis was conducted by ANOVA comparing the clinical score after day 18 postimmunization, which was defined as the recovery phase. Other disease indices, i.e., average of maximal clinical score (B) and percentages of animals in remission (C), are also shown.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

Representative histopathology of EAE in mice treated with COG112 or saline control. Five-micrometer-thick sections were stained with Luxol fast blue and eosin to reveal demyelination and peripheral infiltrates. The left panels show the pathological changes in spinal cord from cervical (A), thoracic (B), and lumbar (C) segments of saline-treated control animal. Demyelination area is shown as loss of blue staining in white matter, and peripheral infiltrates are stained with purple particles. D, E, and F show the corresponding segments of COG112-treated animal where no obvious demyelination was observed. The boxed insets of A and D are magnified 10× in G and H, respectively.

Although peripheral influx of mononuclear cells into the spinal cord and demyelination has been demonstrated by histological staining in Fig. 3, reliable quantification and profiling of T-lymphocyte infiltrates into the entire spinal cord is needed to evaluate the effect of apoE mimetics on T-cell migration. Because COG112 displays a more potent therapeutic effect (as shown in Figs. 1 and 2), we chose COG112 for this infiltration study. Animals received saline vehicle or COG112 at 1 mg/kg/day by daily i.p. injection starting from day 2 postimmunization. Single-cell suspensions from spinal cords were prepared on day 21 postimmunization when COG112-treated animals had started to recover, whereas saline-treated animals still displayed active clinical disease. As depicted in Fig. 4, A and C, COG112 reduced the frequency of CD3⁺ T cells compared with saline vehicle controls. Figure 4E clearly demonstrated COG112 treatment significantly reduced the absolute numbers of total CD3⁺ T cells and specifically, CD4⁺ T cell infiltration into the spinal cord. In contrast, no significant difference was found in the number of CD8⁺ T cells between COG112 and saline groups. These findings are consistent with previous observations and our hypothesis that CD4⁺ T cells play a significant role in the pathogenesis of multiple sclerosis and EAE.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4.

Profiling and quantification of T-cell infiltrates in the spinal cords of animals on day 21 postimmunization that were treated with or without COG112. Single-cell suspensions of spinal cord were stained with anti-CD3, CD4, and CD8 monoclonal antibodies, as described under Materials and Methods. Frequency of CD3⁺CD45⁺ cytometry (T cell) is represented control (A) and COG112 (C). Frequency of CD4⁺ and CD8⁺ T cells is represented in saline control (B) and COG112 (D)-treated mice. Total cell number for each of the specific cell surface markers is calculated and expressed as number per spinal cord (E).

In Vivo Treatment with COG133 Significantly Dampened the Release of NO, TNF-α, and IL-6 in Macrophages Subsequently Challenged with Immunogen in Vitro. To examine the roles played by inflammatory cytokines such as TNF-α and IL-6, and the free radical NO, in the pathogenesis of EAE, macrophages were collected from a subset of saline- and COG133-treated mice on day 35 postimmunization and then challenged in vitro with the immune stimulators LPS + IFN-γ. As depicted in Fig. 5, none of the macrophages exhibited significant levels of NO, TNF-α, or IL-6 in the absence of immune stimulation. However, upon treatment with LPS and IFN-γ, macrophages derived from saline-treated mice exhibited a robust release of NO, whereas macrophages derived from COG133-treated mice released significantly lower levels of NO in response to stimulation (p < 0.001; Fig. 5A). Likewise, macrophages derived from saline-treated mice responded with a robust release of TNF-α and IL-6 when treated in vitro with LPS + IFN-γ, which is in contrast to the responses of macrophages derived from COG133-treated mice (p < 0.001; Fig. 5, B and C). In summary, macrophages obtained from saline-treated mice exhibited clinically severe EAE and a much higher responsiveness to in vitro challenge compared with macrophages obtained from their COG133-treated counterparts. It is noteworthy that in vivo treatment with COG133 afforded protection against subsequent in vitro triggering of macrophages.

In Vitro Treatment with COG112 and COG133 Inhibited LPS + IFN-γ-Induced Release of NO, TNF-α, and IL-6. We have already demonstrated that macrophages from saline-treated mice exhibited robust release of NO, TNF-α, and IL-6 following in vitro challenge with LPS and IFN-γ (Fig. 5). We then examined the ability of COG112 and COG133 to inhibit the in vitro release of NO, TNF-α, and IL-6 using macrophages obtained from saline-treated mice on day 35 postimmunization. In LPS + IFN-γ-treated macrophages, COG112, at the concentration of 5 μM, completely inhibited NO release (p < 0.01 versus control), whereas the same concentration of COG133 inhibited NO release by approximately 37% (p < 0.05 versus control; Fig. 5A). Consistent with the previous experiment, LPS + IFN-γ-induced macrophages exhibited robust release of TNF-α and IL-6 (compare Fig. 4, A, B, and C, with Fig. 5, A, B, and C). COG112 treatment resulted in a dose-dependent suppression of TNF-α and IL-6 levels, with complete inhibition occurring at approximately 5 μM. Indeed, the levels of NO, TNF-α, and IL-6 produced by macrophages treated with 5 μM COG112 were not significantly different from the levels measured in unstimulated macrophages. However, 5 μM COG133 exhibited only moderate, yet statistically significant suppression of both TNF-α and IL-6. Therefore, fusion of the protein/peptide transduction domain to COG133 created COG112 with significantly enhanced anti-inflammatory effects.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 5.

In vivo exposure to apoE-mimetic peptide COG133 suppressed in vitro production of inflammatory effectors by peritoneal macrophages. Nitrite in culture medium was measured with a Sievers NO analyzer, and TNF-α and IL-6 were measured by ELISA as described under Materials and Methods. Statistical analysis was performed by two-tailed Student's t test, and significance is shown as **, p < 0.01 and ***, p < 0.001.

COG112 and COG133 Inhibited LPS + IFN-γ-Induced NOS2 mRNA Expression by Quantitative Real-Time Reverse Transcription-Polymerase Chain Reaction. To further substantiate the finding that COG112 and COG133 inhibited NO release, we examined their effects on NOS2 mRNA expression in macrophages harvested from saline-treated EAE mice on day 35 postimmunization. Macrophages were pretreated with saline vehicle, 5 μM COG112, or 5 μM COG133 for 30 min and subsequently stimulated with LPS + IFN-γ for 10 h. As depicted in Fig. 7, LPS + IFN-γ stimulation in the absence of COG peptides increased NOS2 expression by more than 14,000-fold versus unstimulated peritoneal macrophages. Pretreatment of cultures with COG112 or COG133 resulted in significant reductions of NOS2 mRNA. Similar to its effect on the reduction of NO product levels, 5 μM COG112 treatment reduced NOS2 mRNA levels by 92%, whereas 5 μM COG133 treatment inhibited levels by only 24%. These data suggested that the inhibition of NO release in COG112- and COG133-treated macrophages might be due to their inhibitory effects on NOS2 mRNA expression.

COG112 and COG133 Inhibited MOG-Induced Splenocyte Proliferation. T-cell proliferation in response to myelin antigen is a key part of the pathogenesis of both EAE and multiple sclerosis; therefore, we examined the ability of COG112 and COG133 to block splenocyte proliferation in response to stimulation with MOG peptide ex vivo. At day 6 postimmunization with MOG (before exhibiting clinical signs of EAE), spleens were dissected from four saline-treated mice and then dispersed into single-cell suspension for culturing in 96-well microtiter plates. As shown in Fig. 8, MOG stimulation induced splenocyte proliferation in a concentration-dependent manner that was significantly inhibited in the presence of both doses of COG112 (0.5 and 1 μM) and the highest dose of COG133 (5 μM). Given that 0.5 μM COG112 inhibited splenocyte proliferation to the same level as 5 μM COG133, we conclude that fusing the antennapedia-PTD to COG133 to create COG112 increased the potency of COG133 by ∼10-fold. In the absence of MOG stimulation, both COG112 and COG133 (5 μM) did not show a direct impact on splenocyte proliferation (data not shown). These data suggested that suppression of T-cell proliferation might be another mechanism contributing to the ability of COG112 and COG133 to decrease the systemic severity of EAE.

Discussion

We have demonstrated that a robust, debilitating form of EAE was significantly ameliorated by treatment with apoE-mimetic peptides COG112 and COG133, when administered either before or after onset of disease. The histopathological data indicate that this reduction in clinical severity was accompanied by suppression of inflammation, demyelination, and infiltration of cells into the spinal cord. The data suggest that apoE-mimetic peptides, particularly COG112, may have promising utility in the treatment of EAE and the human disease it models, multiple sclerosis.

Experiments designed to elucidate the mechanism underlying these clinical and histopathological observations suggest that apoE-mimetic peptides function via several mechanisms relevant to the pathogenesis of EAE and multiple sclerosis, including activation of macrophages with subsequent modulation of NO and inflammatory cytokine release and suppression of lymphocyte proliferation. First, peritoneal macrophages derived on day 35 postimmunization from COG133-treated mice released significantly less NO, TNF-α, and IL-6 than macrophages from saline-treated mice in response to in vitro stimulation with LPS + IFN-γ (Fig. 5). This finding suggests that continuous treatment with COG133 in vivo and in vitro maintained the macrophages in a relatively anergic state. This is extremely important with respect to multiple sclerosis, because activated macrophages are thought to be a major player in mediating events that result in demyelination. Specifically, the majority of inflammatory infiltrates in multiple sclerosis are macrophages, and they have been identified as the proximate mediators of tissue injury in multiple sclerosis due to their direct involvement in the phagocytosis of the myelin sheath and their amplifying role as antigen-presenting cells that activate antigen-specific T lymphocytes (Prineas and Connell, 1978; Smith and Lassmann, 2002). Interestingly, Tenger et al. recently reported that holo-apoE protein inhibited T-cell proliferation by acting on the antigen-presenting ability of macrophages (Tenger and Zhou, 2003). This is matched with our observation that apoE-derived peptides suppressed MOG-induced proliferation of splenocytes in cultures that may contain macrophages (Fig. 8), suggesting apoE peptides possibly mimic the bioactivity of holo-apoE protein in this respect. In addition, initiating production of NO and cytokines by macrophages is thought to mediate the disruption of BBB with subsequent invasion of antigen-specific T lymphocytes (Smith and Lassmann, 2002). Therefore, suppression of initial macrophage activation may greatly contribute to block the deleterious inflammatory cascade and further axonal damage.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 6.

In vitro effects of apoE-mimetic peptides on LPS- and IFN-γ-induced production of inflammatory mediators in cultured peritoneal macrophages isolated from EAE mice. Macrophages were treated with COG112 or COG133 and then stimulated with LPS (100 ng/ml) and IFN-γ (10 U/ml). Conditioned medium was collected 45 h after immunogen exposure for analysis of TNF-α (A) and IL-6 (B) or 72 h after inflammatory stimulation for measurement of nitrite (C). The level of nitrite was normalized to protein concentration of cell lysates measured at 72 h. Statistical analysis was performed by ANOVA followed by Dunnett's post hoc test compared with respective LPS + IFN-γ control only. Significance is shown as *, p < 0.05 and **, p < 0.01.

In support of this notion, we found that the pharmaceutical effects of apoE mimetics are highly correlated with their potency to inhibit macrophage activation, so that COG112, with enhanced efficiency in inhibiting activation of macrophages compared with COG133 (Fig. 6), presented an improved therapeutic effect on the recovery from the disability of EAE (Figs. 1 and 2). These data are consistent with our published reports characterizing APOE genotype-specific differences in immune responses both in vitro and in vivo. In these studies, humans and mice expressing the human apoE4 isoform exhibited significantly more robust production of NO, TNF-α, and/or IL-6 compared with those expressing the more common human apoE3 isoform (Brown et al., 2002; Colton et al., 2002, 2004; Lynch et al., 2003). Together, these findings suggest that the apoE4 isoform is functionally deficient, and like the results in mice lacking apoE, possesses an insufficient immune modulating ability to face an inflammatory insult. Therefore, apoE peptides may potentially compensate for these APOE4-associated deficits and dampen the overwhelming inflammation occurring in these neurological disorders (Laskowitz et al., 2001). For example, we have demonstrated that i.v. administration of COG133 suppressed LPS-induced systemic and CNS inflammatory responses and recently published reports in which COG133 reduced brain inflammation and improved outcome in a mouse model of traumatic brain injury (Lynch et al., 2003, 2005). Therefore, this body of work demonstrates that COG133 may function like an endogenous functional apoE protein (human apoE3 isoform or normally expressed mouse apoE) to suppress microglial and macrophage activation and subsequent release of potentially harmful effectors, including inflammatory cytokines and NO.

In an attempt to improve the permeability of COG133 through the blood-brain barrier and the plasma membrane, we designed a new peptide, COG112, by fusion of a 17-amino acid PTD-fragment of antennapedia to the amino terminus of COG133. In all of the in vitro experiments (Figs. 5, 6, 7, 8), COG112 was consistently more potent in suppressing LPS and IFN-γ-induced NO, IL-6, TNF-α, and NOS2 expression as well as in inhibiting lymphocyte proliferation. More importantly, COG112 more potently reduces the severity of clinical signs throughout the duration of the study, including reduction of mean maximal clinical score, cumulative clinical score, and an increase in the percentage of mice in remission compared with COG133. We have yet to empirically determine whether this is due to enhanced penetration of the BBB, as has been reported for other PTD-drug conjugates (Rousselle et al., 2003). The exact mechanism underlying this difference is a matter of active investigation, and our unpublished data suggest that COG133 interferes with certain intracellular signal transduction pathways (N. Ohkubo, C. A. Colton, F.-Q. Li, and M. P. Vitek, unpublished data). In this study, we also found that COG112 and COG133 potently inhibited intracellular expression of NOS2 mRNA, suggesting that these peptides may influence certain signal transduction factors upstream of iNOS expression, such as nuclear factor-κB, which has also been associated with multiple sclerosis and EAE (Hilliard et al., 1999).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 7.

ApoE-mimetic peptides significantly inhibit LPS and IFN-γ-induced NOS2 mRNA expression in peritoneal macrophages. Level of NOS2 expression was quantified by quantitative real-time polymerase chain reaction as described under Materials and Methods. NOS2 expression was first normalized by the endogenous control (18S rRNA) and then expressed as -fold change compared with control group (LPS + IFN-γ-treated only). Data were analyzed by ANOVA followed by Dunnett's post hoc test compared with LPS + IFN-γ-treated only group. **, p < 0.01.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 8.

ApoE-mimetic peptides inhibited pMOG_35-55-induced splenocyte proliferation. Mice were immunized with pMOG_35-55, and splenocytes were prepared from the spleens 6 days later. Dispersed splenocytes were cultured in 96-well plates in the presence of increasing doses of pMOG_35-55 peptide. Separate wells of splenocytes were cultured with 20 μg/ml pMOG_35-55 and the indicated concentrations of COG112 or COG133. After 54 h, BrdU was added to the cultures, which were then harvested at 72 h and assayed for BrdU incorporation. pMOG_35-55 induced proliferation in a concentration-dependent manner (**, p < 0.01). Statistical significance was analyzed by ANOVA followed by Dunnett's test as expressed by ##, p < 0.01 and ###, p < 0.001 compared with MOG-treated only. Data are expressed at mean optical density value ± S.E.M for each treatment (n = 6).

The current investigation focused on the anti-inflammatory aspect of apoE peptides, leaving the likelihood open that other mechanisms might also be involved. For example, our laboratory and other laboratories have found that both apoE protein and apoE-derived peptides exert neuroprotective effects against glutamate-mediated excitotoxicity and neurotrophic effect (Misra et al., 2001; Aono et al., 2002; Linker et al., 2002; Gay et al., 2006). It has been well documented that excessive release of glutamate by macrophages and brain microglia during inflammation in both multiple sclerosis and EAE contributes to the pathogenesis of the disease and that blockade of glutamate receptors can ameliorate EAE symptoms (Pitt et al., 2000). Along these lines, Aono et al. (2003) reported that COG133 effectively reduced N-methyl-d-aspartate-mediated killing of primary neuronal cultures as did the well-known N-methyl-d-aspartate receptor antagonist MK-801. Axonal degeneration in the chronic progressive phase of disease is a crucial factor responsible for the irreversible long-term disability of patients with multiple sclerosis (Losseff et al., 1996).

Neurotrophic factors may also help to prevent axonal degeneration and encourage repair of damaged axons. Linker et al. (2002) recently reported that the ciliary neurotrophic factor, which was originally identified as a survival factor for isolated neurons, was a major protective factor in demyelinating disease through its function to rescue oligodendrocytes. Interestingly, apoE has been found to bind to and potentiate the biological activity of ciliary neurotrophic factor (Gutman et al., 1997). Furthermore, we found recently that COG133 can also potentiate the neurotrophic effect of nerve growth factor by promoting the dendrite outgrowth in PC12 cells (N. Ohkubo, C. A. Colton, F.-Q. Li, and M. P. Vitek, unpublished data). In the present study, the effect of apoE peptides on promoting the recovery of EAE seems more robust than that of inhibiting the onset of disease, suggesting the peptides may significantly affect the neuronal repair process in which neurotrophic effects may dominate.

In summary, we now report that apoE-mimetic peptides significantly attenuated the severity of EAE and promoted recovery from the disease, indicating that COG112 and COG133 may represent multidimensional therapeutics capable of inhibiting the inflammatory cascade, protecting against oxidative stress, and/or simultaneously protecting neurons and oligodendrocytes from damage. Further studies may validate whether these compounds have utility as therapeutic agents for the human autoimmune disease multiple sclerosis.