Immune Modulation by Antigenic Peptides and Antigenic Peptide Conjugates for Treatment of Multiple Sclerosis

The immune system defends our body by fighting infection from pathogens utilizing both the innate and adaptive immune responses. The innate immune response is generated rapidly as the first line of defense. It is followed by the adaptive immune response that selectively targets infected cells. The adaptive immune response is generated more slowly, but selectively, by targeting a wide range of foreign particles (i.e., viruses or bacteria) or molecules that enter the body, known as antigens. Autoimmune diseases are the results of immune system glitches, where the body’s adaptive system recognizes self-antigens as foreign. Thus, the host immune system attacks the self-tissues or organs with a high level of inflammation and causes debilitation in patients. Many current treatments for autoimmune diseases (i.e., multiple sclerosis (MS), rheumatoid arthritis (RA)) have been effective but lead to adverse side effects due to general immune system suppression, which makes patients vulnerable to opportunistic infections. To counter these negative effects, many different avenues of antigen specific treatments are being developed to selectively target the autoreactive immune cells for a specific self-antigen or set of self-antigens while not compromising the general immune system. These approaches include soluble antigenic peptides, bifunctional peptide inhibitors (BPI) including IDAC and Fc-BPI, polymer conjugates, and peptide-drug conjugates. Here, various antigen-specific methods of potential treatments, their efficacy, and limitations will be discussed along with the potential mechanisms of action.


Introduction
The body's immune system is a complex molecular and cellular system that protects the body from insult by a wide variety of pathogens that enter the host. Innate immune responses utilize physical barriers and surface pattern recognition of pathogens to generate immune responses using pattern recognition receptors (PRR) on the surfaces of immune cells. 1 The innate immune system reacts fast to infection as the first line of response by identifying pathogen associated molecular patterns (PAMPs), such as unusual mannose containing oligosaccharides on bacteria and fungi. Innate immune responses directed at PAMP containing pathogens include activation of the complement system, release of chemical mediators and cytokines to induce inflammation, and recruitment of white blood cells to sites of infection ( Figure 1). Cells that make up the innate immune system include dendritic cells, macrophages, neutrophils, basophils, eosinophils and natural killer cells ( Figure 1). 1 While prompt to respond to an infection, the innate immune system has no memory, meaning that it will launch the same response of engulfing or eliminating if it encounters the same pathogen later. A second and slower response, called the adaptive immune response, is more selective and targeted compared to the innate immune response, and is initiated through a process of antigen presentation and recognition of non-self antigens. 1 In the adaptive immune response, specific cells (i.e., B and T cells) are activated to target pathogens or cells infected with pathogen. For this response, antigen specific B and T cells must be activated and then proliferate to produce enough highly specific cells to respond against the pathogen. It should be noted that the maturation process of B and T cell receptors requires time. This means that while highly effective, it takes time for the adaptive immune system to mount an attack to fight off the pathogens. Though it takes longer, the adaptive immune system develops memory lymphocytes that can help launch a faster response if it encounters the same pathogen again.
Generally, when a pathogen attempts to invade the host, it will first encounter the innate immune system. 1 This includes physical barriers such as epithelial layers and/or mucosal layers, or cells with PRR that recognize non-host pathogen patterns and bind to them. The cells of the innate immune response will activate cells from the adaptive immune response, which are the antigen-specific B and T cells. 1,2 The adaptive immune response is broken down into humoral and cellular immune responses. The humoral response is initiated by activation of B cells for antibody production via plasma cells. The generated antibodies can target the pathogen for neutralization and opsonization followed by initiating the complement system to stop proliferation of the pathogen. 3 B cell activation is dependent on both antigen presentation to B cell receptors and the assistance of a helper T cell (Th), a type of CD4 + cell. Cell-mediated immunity activates CD8 + and CD4 + T cells that recognize specific antigens presented on the surface of cells via MHC-I and MHC-II complexes. 2 This recognition allows for the T cells to identify the infected cells followed by recruitment of other immune cells for eliminating the infected cells. 2 At the same time, both B and T cells create memory cells that allow for a faster adaptive immune response to that antigen if there is a reinfection. lymphocytes can potentially be specific to protein regions or peptide sequences native to the body, or self-antigens. Theoretically, lymphocytes that recognize self-antigens have been eliminated during development. In addition, there are multiple tolerance mechanisms present to remove autoreactive lymphocytes. However, autoreactive lymphocytes can evade the tolerance mechanisms. The onset of autoimmune disease can be triggered by the activation of these autoreactive lymphocytes; this activation has been observed in rheumatoid arthritis (RA), multiple sclerosis (MS), and type-1 diabetes (T1D). 4 Autoimmune diseases can be chronic and life-threatening; these patients need life-long therapy.
Disease-modifying therapies (DMTs), anti-inflammatories, and immunosuppressant drugs have been prescribed to patients for treatments of autoimmune diseases. 4,5 Antiinflammatories are aimed at reducing side effects of the disease for short-term treatment. 6 Because autoimmune diseases are a result of immune cells attacking self-tissues or organs, the immunosuppressant drugs are used long-term to lessen the activity of immune cells. However, these drugs are not cell specific; the use of these drugs can lead to global immunosuppression in the body. 5,7 As a result, patients are vulnerable to opportunistic infections and cancers. 5 DMTs treat autoimmune diseases by interfering with immune cell activity in general. 5 As an alternative, antigen-specific immunotherapies have been developed to target only rogue cells that cause the autoimmune disease and suppress their proliferation as well as disease progression. Therefore, various new developments of antigen-specific treatments and their potential mechanisms of action will be discussed here.

Onset of Autoimmune Diseases
The trigger for the onset of autoimmune diseases is the recognition of self-antigens by B and T cells, followed by the activation and proliferation of T cells specific to the self-antigens. T cell activation is a product of naïve T cell interactions with dendritic cells and other antigen presenting cells (APCs). The antigen is engulfed and processed by APCs and subsequently antigen fragments are presented as antigen-MHC-II complexes (Ag-MHC-II) on the surface of APC. Then, T-cell antigen receptors (TCRs) bind with Ag-MHC-II as a cluster of Ag-MHC-II/TCR complexes (i.e., Signal-1). To activate T cells, Signal-1 is accompanied by Signal-2 generated by a cluster of complexes of costimulatory molecules at the APC/T cell interface ( Figure 2). 8 Besides Signal-1, T cells are also activated during the inflammatory response by Signal-2 through connectivity of co-stimulatory molecules ( Figure 3). 6,9 The interactions of both signals (Signal-1 and Signal-2) form the immunological synapse (IS), which involves positional translocation between Signal-1 and Signal-2 complexes. The Signal1/ Signal-2 translocation was demonstrated using peptide-MHC-II/TCR and ICAM-1/LFA-1 interactions, respectively, using T cells in vitro. [10][11][12] To study the IS, planar membranes were studded with fluorescently labeled peptide:MHC-II as a component of Signal-1 and ICAM-1 as a component of Signal-2 on the APC. Upon addition of the T cell, LFA-1 molecules on T cell bind to ICAM-1 molecules on planar membranes to form a cluster of Signal-2 in the center while a cluster of peptide:MHC-II/TCR complexes as Signal-1 form in the periphery ( Figure 3A). 10 After 5 min, the ICAM-1/LFA-1 complexes translocate from the center to peripheral to make peripheral supramolecular activation complex (pSMAC) through an intermediate ( Figure 3B), while peptide:MHC-II/TCR complexes translocate from the peripheral to the center to make central supramolecular activation complex (cSMAC). Finally, both clusters formed a "Bull's Eye" structure called the IS ( Figure 3C).

Multiple Sclerosis
Multiple sclerosis (MS) is an autoimmune disease characterized by neuroinflammation, which is caused by the infiltration of autoreactive T cells into the central nervous system (CNS) that attack and damage the myelin sheath on the axons of neurons. MS patients can suffer from several forms of the disease. Relapsing-remitting MS (RRMS) is found in 87% of MS patients where patients experience periods of disease progression followed by periods of improvement. 28,29 RRMS patients can progress to secondary progressive MS (SPMS) without remission periods and eventually reach a point of irreversible neuronal damage with signs of axonal degeneration. 28,29 A small number of patients suffer from primary progressive MS (PPMS) with constant neurological deterioration and no remittance periods. Finally, progressive-relapsing MS (PRMS) is the rarest MS form with constant neurological deterioration and heightened periods of attacks. 29 CNS infiltration by lymphocytes is one of the hallmarks of the disease in which autoreactive T cells recognize myelin sheath proteins as foreign-antigens. The myelin sheath consists of myelin oligodendrocyte glycoprotein (MOG), myelin basic protein (MBP), and proteolipid protein (PLP) that surround the axon. 6 In the CNS, microglia cells serve as APCs for presenting self-antigens to naïve CD4 + T cells that differentiate to Th1 and Th17 cells. 30 Th1 cells generate IFN-γ while Th17 cells produce IL-17 as inflammatory cytokines that stimulate immune cells to infiltrate the CNS. However, the mouse model studies indicate that IL-17-producing Th17 cells are more important than IFN-γ-producing Th1 cells in autoimmune diseases. 30 In addition, MS patients produce a high level of IL-17 which can be used as a marker for MS. 31 short-term to reduce of the inflammatory response. 6,32 Glatiramer acetate is an amino acid polymer that is believed to mimic MBP as a decoy that can slant the immune-response balance to a suppressor response. 33 DMTs are normally monoclonal antibodies (mAbs) that interfere with B and T cell activity to alter the disease progression. 5 The mAb, natalizumab, prevents CNS infiltration of immune cells by blocking interactions of α4 integrins on immune cells with vascular cell adhesion molecule-1 (VCAM-1) on the BBB endothelial cells. 5,34 Ocrelizumab, ofatumumab, and alemtuzumab were designed to target CD20 and CD52 and lyse immune cells. 5 Several of these mAbs (i.e., ocrelizumab, alemtuzumab, and natalizumab) have anti-inflammatory activities that reduce the rates of relapses. While mAb treatments help suppress MS, natalizumab has a side effect of general immunosuppression and it induces progressive multifocal leukoencephalopathy (PML) in a small number of patients. 35 This side effect can be dangerous to patients. Because of the PML side effect, there is a need to develop a specific treatment that is safer for MS patients.

Rheumatoid Arthritis
Rheumatoid Arthritis (RA) is another type of autoimmune disease that creates chronic inflammation of the joints due to T cell activation of self-antigen epitopes of the synovium. 36 As in MS, proliferation of Th17 cells and upregulation of the IL-17 cytokine are related to the chronic inflammation in RA. Disease-modifying antirheumatic drugs (DMARDS) help avert the onset immune responses to counteract the disease. Abatacept is a fusion protein drug made from a conjugate between the CD80/86-Fc-binding region of IgG1 and the extracellular domain of CTLA-4. This conjugate blocks interaction of CD80/86 to CD28 as costimulatory signals and suppresses the activation of inflammatory T cells ( Figure  2). 37 Several mAbs (i.e., infliximab, etanercept, adalimumab and certolizumab) have been used as DMARDS to reduce the levels of TNF-α in RA. Anti-CD20 mAb, rituximab, binds to CD20 to reduce the B cell population for suppressing disease symptoms and promoting remission in RA.

Type-1 Diabetes (T1D)
In type 1 diabetes (T1D), insulin-producing β cells are attacked by autoreactive CD4 + and CD8 + T cells that infiltrate the pancreatic islets in a process called insulitis. 38 Several self-antigens such as glutamic acid decarboxylase (GAD), insulinoma antigen-2 (IA-2), insulin, and zinc transporter 8 (ZNT8) stimulate autoreactive T cells in T1D. Suppression of GAD65 expression in beta cells could prevent diabetes in non-obese diabetic (NOD) mice. 39 Prediabetic patients produce antibodies specific to these antigens where insulin is the primary antigen before the appearance of sensitivity to other antigens. 38 These antigens are recognized by naïve T cells for their differentiation to inflammatory T cells (i.e., Th1 and Th17). It is interesting that soluble peptides from these antigens have been developed and successfully induced immunotolerance to T1D in animal models. by developing antigen-specific treatments characteristic to a specific autoimmune disease. Because autoreactive proinflammatory T cells are initiated from the activation of naïve T cells, the goal is to selectively suppress the autoreactive T cells in an antigen-specific manner while stimulating the proliferation of regulatory cells (i.e., Treg). During naïve T cell-APC interaction, T cell activation begins after the IS formation ( Figure 3); dendritic cells (DCs) normally have an important role in presenting the processed peptide antigens ( Figure 4). However, it has been shown that delivering soluble peptide epitopes of an antigenic protein can suppress a subpopulation(s) of reactive immune cells (i.e., T cells and B cells) that recognize the antigen ( Figure 4A).

Costimulatory Signals in Autoimmune Diseases
The important role of the costimulatory signal (Signal-2) has been elucidated by showing the suppression of inflammatory immune cell activation by blocking this signal in animal models of autoimmune diseases. Inhibition of ICAM-1/LFA-1 interactions (Signal-2) by antibodies, peptides, and small molecules modulates T-cell activation to suppress T1D, 40,41 psoriasis,42,43 RA,44,45 and allograft rejection 46,47 in animal models ( Figure  3D). Administration of anti-ICAM-1 and anti-LFA-1 mAbs together inhibits allograph rejection of heart transplantation between two different strains of mice. 46 The results indicate that blocking the ICAM-1/LFA-1 signal (Signal-2) interaction prevents activation of inflammatory T cells that attack the foreign organ from a different strain of mice ( Figure  3D). 46 In this case, blocking Signal-2 in the presence of Signal-1 induces T cell anergy by preventing the complete formation of IS ( Figure 3D). It has been shown that psoriasis can be treated with an anti-CD11a mAb (efalizumab) that blocks the ICAM-1/LFA-1 signal (Signal-2). [48][49][50] Unfortunately, the side effect of efalizumab was progressive multifocal leukoencephalopathy (PML), which is often fatal, in a population of patients; in 2009, the use of this drug was terminated. 51,52 Natalizumab (Tysabri®) was developed for treatment of MS. This mAb binds the α4-subunit and blocks the α4β1 and α4β7 integrins interactions on leukocytes to VCAM-1 molecules of the vascular endothelial cells of the blood-brain barrier (BBB) to inhibit leukocyte adhesion to the BBB endothelial cells. 53,54 As a result, it prevents the infiltration of immune cells into the brain and halts disease progression in MS. The use of natalizumab was halted for a short period of time because it caused a small population of MS patients to develop PML. It was speculated that blocking integrin activity could lower the activity of the general immune response to fight pathogenic infections. Therefore, an alternative solution in treating MS patients without compromising their general immune systems will benefit patients. In this case, a new and novel method is needed to selectively suppress only a subpopulation of autoreactive T cells that react to a specific self-antigen in a specific autoimmune disease. As a result, such treatments will avoid suppression of the general immune responses.
As mentioned previously, positive and negative costimulatory signals alter T cell activation. 13,15,19,55 Interactions between CD28 and CD80/86 are one of the most important positive costimulatory signal that leads to sustain activation of proinflammatory T cells. 15,[55][56][57] It was proposed that blocking this signal could be a method for specific immunomodulation. Alternatively, CD80/86 can also interact with CTLA-4 as a negative costimulatory signal and this signal has been shown to suppress the activation of inflammatory T cells in cancers. A fusion between CTLA-4 and the Fc region of IgG1 has been shown to induce T cell anergy in vitro; however, it did not modulate T cell activation in vivo. 15

Natural Antigenic Peptides:
Administrations of soluble antigenic peptides in animal models have been shown to suppress various autoimmune diseases; they act in an antigen-specific manner and avoid suppression of the general immune response. The antigenic peptides mimic T cell epitopes with linear sequences (<20 amino acids) that induce tolerogenic effects via immune deviation. [58][59][60] Their small size allows them to evade immunogenicity and they are not able to cross-link for IgE activation. 60 It is critical to note that the antigenic peptide must be soluble because insoluble peptide could induce local inflammation. 60 For example, the GAD 208-217 peptide ( Table 1) from GAD65 epitope suppressed T1D in NOD mice and the peptide complexed with empty I-A g7 MHC-II and HLA-DQ8 MHC-II on the surface of APC. 61,62 The soluble antigenic peptides prevent inflammatory T cell activation and insulitis to protect insulin-producing β cells in T1D animal models. Similarly, antigenic peptides from myelin sheath proteins can suppress the progression of multiple sclerosis in animal models. For example, PLP [139][140][141][142][143][144][145][146][147][148][149][150][151] peptide from an epitope of myelin sheath PLP suppresses experimental autoimmune encephalomyelitis (EAE) disease symptoms in mice as an animal model for MS.
Designing antigenic peptides for immunotolerance is more complex than administering a fragment of the protein antigen. 63 Antigenic peptides can be administered to stimulate upregulation of different T cell phenotypes without needing the whole protein antigen. These antigenic peptides can induce tolerance towards the antigen by increasing the number of Tregs rather than inflammatory T cells. Antigenic peptides from epitopes of the antigen protein with 13-17 amino acids have the best fit to MHC-II binding pocket. At times, a newly designed peptide from the antigenic protein may not have immunological activities as the naturally processed antigenic peptides; this is because the newly designed peptide has a different conformation than the peptide epitope on an intact protein antigen. 64 Thus, the conformation of antigenic peptide (apitope) must match that of the naturally processed epitope that stimulates T cells without antigen processing. 63 Different antigenic peptides in MS were thought to be linked to different forms of MS. 65 Different types of EAE as different models of MS can be stimulated by PLP, MBP, or MOG peptide in complete Freund's adjuvant (CFA). Activations of PLP-and MOG-specific autoreactive T cells have been proposed to be correlated with RRMS and SPMS, respectively. In MS, MBP-specific autoreactive T cells are also stimulated as a result of epitope spreading. 65 Administration of PLP peptides suppressed in animal models for RRMS while MOG peptides suppressed animal models for SPMS. Thus, different antigenic peptides could be targeted to various stages of MS. Due to antigen specificity, the MOG peptide can only suppress MOG-stimulated EAE, not PLP-stimulated EAE or MBPstimulated EAE. 66 Due to the antigenic specificity of this approach, it may be necessary to administer a mixture of multiple antigenic peptides for a more robust regulatory response to suppress MS. 58 The mechanism of antigenic peptide is due to its binding to empty unprocessed MHC-II on the surface of immature dendritic cells (iDCs) to form peptide/MHC-II complex ( Figure  4A). 59,60 TCRs on naïve T cells then bind to the peptide/MHC-II complexes on the iDCs in the absence of costimulatory molecules (i.e., CD80/CD86). The peptides bind to empty MHCs without regular processing through the cytoplasmic domain; therefore, the naïve T cells become regulatory T cells (Treg) that produce regulatory cytokines (i.e., IL-10, IL-35, TGF-β) to provide a tolerogenic response ( Figure 4A). The upregulation of Tregs produces regulatory cytokines that suppress the activation of Th1 and Th17 cells in an antigenic-specific manner. 58 In autoimmune diseases, the disease onset is due to the skewed balance of inflammatory (i.e., Th1, Th17) over regulatory (i.e., Treg) immune cells. Therefore, an important feature when trying to suppress autoimmune diseases is to stimulate the proliferation of Tregs. Yu et al. showed that suppression of autoimmune diseases cannot be done without the presence of Tregs. 67 Furthermore, the upregulation of Tregs was able to suppress EAE in an antigen-restricted manner. 67 Treg cell proliferation has been associated with the upregulation of regulatory cytokines such as IL-10, IL-35, and TGF-β as well as FoxP3 and CD25 markers on T cells. 68,69 IL-10 reduces the BBB's disruption and leakiness in MS and decreases CNS infiltration of proinflammatory molecules and immune cells through the BBB. Thus, the administration of antigenic peptides leads to a specific immunotolerance without suppressing the overall immune system. 63 The EAE disease has been developed in rats and mice as animal models for MS to evaluate potential therapeutics for MS. These animal models have been used to find different active peptide epitopes of the PLP antigen protein, including PLP [130][131][132][133][134][135][136][137][138][139][140][141][142][143][144][145][146][147] and PLP [103][104][105][106][107][108][109][110][111][112][113][114][115][116] . Immunization of animals with the different PLP peptides in CFA can induce EAE. 70 The nasal administrations of soluble MBP peptide in rats suppressed EAE. 58 Peptide epitopes from MBP and MOG were also found to be active. These antigenic peptides could be delivered via intravenous (i.v.), subcutaneous (s.c.) or oral route to suppress EAE. For the oral route, the antigenic peptides need to be administered as a vaccine, indicating that vaccine delivery of peptide is necessary to prime naïve T cells to become regulatory T cells (Treg) for suppressing the disease. A co-administration of four different MBP peptides as ATX-MS1467 (MBP [30][31][32][33][34][35][36][37][38][39][40][41][42][43][44] ) had reached clinical for SPMS; during a phase-I study, patients treated with these peptides showed improvement from the disease symptoms. 71,72 Soluble antigenic peptides derived from collagen II (CII) such as CII 256-270 , CII 707-721 , and CII 1237-1249 can suppress RA in animal models by suppressing inflammatory T cell activity. 36 Collagen II is found in the synovial joints and can elicit an inflammatory T cell response in RA. [73][74][75] Both CII 256-270 and CII 707-721 suppressed RA scores compared to the control in the CIA model while the CII 1237-1249 peptide did not suppress the disease symptoms. 36 The activities of CII 256-270 and CII 707-721 were presumably due to their binding to I-A g7 MHC-II molecules on APCs (i.e., dendritic cells). [75][76][77] Although molecular modeling with the X-ray structure of DR4 MHC-II has shown that the CII 1237-1249 peptide can form a peptide MHC-II complex, it does not have activity in vivo. 78 Therefore, peptide binding to MHC-II doesn't necessarily modulate the immune response. 78 CII 256-270 lowered the IL-6 levels in the serum indicating suppression of Th17 cells. 36 It should be noted that Th17 cells are also important in the inflammatory response in CIA. 79,80

Altered Peptide Ligands (APLs):
Altered peptide ligands have been developed as an alternative to soluble natural antigenic peptides to improve regulatory immune responses compared to natural antigenic peptides. The initial idea behind dosing APLs is that they can interrupt the interaction between the TCR and MHC-II to prevent the progression of EAE and other autoimmune diseases. 81 Altering the sequence of the native antigenic peptide to make the APL can potentially modulate TCR/Ag-MHC-II interactions with the hope that the APL is more effective than the native antigenic peptide. 81 In APLs, the amino acids of the parent peptide are mutated to improve binding to MHC-II and/or enhance the binding of Ag-MHC-II complex to TCR. 82 The best known APL is the glatiramer acetate polymer, which is currently marketed as Copaxone® for treating MS. 83 The polymer was designed to mimic the MBP antigenic peptide with randomly polymerized L-alanine, L-glutamic acid, L-lysine, and L-tyrosine. 83,84 The potential side effect of repeated administrations of APL such as glatiramer acetate is that a small population of patients developed hypersensitivity reactions, specifically anaphylaxis. 85 Naturally occurring APLs, that are generated by a single amino acid modification via a naturally occurring reaction(s), can also be responsible for the onset of autoimmune diseases. 86 This natural APL causes altered interactions between TCR and Ag-MHC-II to form TCR/Ag-MHC-II complexes. 86 These modifications that can form APLs include citrullination, deamidation, and deimidation. [86][87][88][89] The citrullination reaction specifically changes the charge of the overall peptide due to the conversion of the positively charged arginine residue to a neutral citrulline residue; this alters the APL binding properties to MHC. 86 Citrullination is caused by peptidyl arginine deiminases (PAD), which are normally overexpressed during inflammation by inflammatory cells. 86,90 Post-translational modifications have been associated with an increase in peptide binding affinity to MHC-II as well as the onset of autoimmune diseases (i.e., RA and MS). 86,91,92 Deamidation of gliadin peptides by transglutaminase 2 (TG2) enzyme could alter the peptide conformation and increase peptide affinity to MHC-II that causes the induction of a Th1 response, known as Celiac disease. 86,[93][94][95][96] It was previously believed that blocking MHC-II with an APL was enough to prevent autoimmune disease; however, blocking of MHC alone is nonspecific and could lead to suppression of other immune responses (i.e., innate and humoral responses). 81,97 Because TCR genes are highly conserved, even between different rodent species, TCRs could be a better target for more specific autoimmune therapies. [98][99][100] In addition, another important characteristic of T cells is that they can recognize multiple ligands. Thus, it allows soluble antigenic peptides to be mutated so that they can still be recognized by the TCR while inducing a different immune response. Therefore, generating TCR antagonists using APLs can be developed for preventing inflammatory T cell activation. 81 Antagonist peptide sequences are determined by replacing a single amino acid on the antigenic peptide to see which peptide still binds to the TCR.
Similarly, APLs from the allergen Par J1 were developed by substituting its residues with alanine and valine amino acids to alter binding to MHC-II, and also interfere with TCR interactions. 104 Thus, the APLs behaved as TCR antagonists rather than solely MHC antagonists. APLs specific to TCR do not all work the same way because they can be antagonists, partial agonists, or super agonists depending on the sequence. 105 Although APLs as partial agonists cannot induce T cell proliferation, they can induce cytokine release and generate a Th2 response rather than a Th1 response. 86,106 APLs as antagonists decrease T cell proliferation by acting as competitive inhibitors to induce a negative feedback loop. 105,107,108 APL antagonists were selected because of their binding properties to MHC-II while inhibiting activation of inflammatory T cells. Some of the antagonist APLs inhibited the proliferation of T cells to stop the induction of EAE in animal models. 81,82 Administration of APLs also led to the upregulation of Treg cells reflected by production of IL-10. 109,110 In a short clinical trial, MBP [83][84][85][86][87][88][89][90][91][92][93][94][95][96][97][98][99] and an APL derivative called NBI-5788 were evaluated to understand their effects on immediate as well as long-term immune deviation. The treatments reduced the amount of lesions in the brain and shifted the immune response from Th1 to Th2 in a group of MS patients. 111,112 MS patients who received placebo showed low levels of APL-reactive T cell with high levels of the inflammatory cytokine, IFN-γ. 112 These same observations were made for MS patients that received MBP [83][84][85][86][87][88][89][90][91][92][93][94][95][96][97][98][99] .
To achieve the targeting of a sub-population of immune cells, various types of bifunctional peptide inhibitors (BPIs) were designed by conjugating an antigenic peptide and a Signal-2 peptide or protein (i.e., the LABL peptide and I-domain of LFA-1)( Figure 5). The idea is that a BPI will bind to the surface of an APC by interacting with empty MHC-II via the antigenic peptide and simultaneously bind to ICAM-1 via the LFA-1 peptide known as LABL, derived from CD11a (237-247) ( Figure 5A). In BPIs, antigenic peptides are separated from the LABL peptide via a long linker using several aminocaproic acid or polyethylene glycol based linker ( Figure 5A). The required linker length for simultaneous binding was estimated from molecular modeling using X-ray structures by docking of the antigenic peptide to MHC-II and LABL peptide to the D1 domain of ICAM-1 where these two proteins were both anchored to cell membranes. Thus, the BPI will disrupt the cluster formation of pSMAC as well as cSMAC to inhibit the IS formation ( Figure 3E). The inhibition of the IS formation suppresses the proliferation of Th17 and Th1 cells while stimulating antigen-specific Treg upregulation ( Figures 4B-C). The presence of the antigenic peptide on BPIs ensures they selectively target a subpopulation of disease specific T cells as well as avoiding suppression of general immune responses to other antigens.

GAD-BPI for T1D
The concept of the BPI was tested using the GAD-BPI molecule by conjugating the GAD65 208-217 peptide with the LABL peptide (CD11a 237-246 ) via a linker that is made from a combination of aminocaproic acid and glycine (Table 1). 140 The GAD 208-217 peptide binds to I-A g7 MHC-II on APCs. On the surface of APC, GAD-BPI molecules block the IS formation while GAD peptide/MHC-II complexes are still recognized by TCRs on a specific population of naïve T cells. Therefore, other subpopulations of T cells with TCRs that recognize other antigens (i.e., pathogens) will not be suppressed and can still be activated to fight an infecting pathogens. As a result, GAD-BPI and other BPI molecules do not suppress general immune responses. To test the activity of GAD-BPI, a T1D mouse model was developed by administration of the GAD 208-217 peptide (40 nmol) in complete Freund's adjuvant (CFA) via subcutaneous (s.c.) route. Administration of GAD-BPI (80 nmol, 100 μL) on days 0 and 7 significantly suppressed insulitis in T1D mice compared to control mice injected with phosphate buffer saline (PBS).
GAD-BPI-treated mice had low insulitis with 83% of normal islets on day 8 while PBStreated mice had high insulitis with only 35% of normal islets. 39,140 Next, CD4 + T cells isolated from GAD-BPI-treated and PBS-treated NOD mice were transferred (adoptive transfer) to evaluate their effects in NOD severe combined immunodeficiency (NOD.Scid) mice. About 72% NOD.Scid mice that were received CD4 + T cells from GAD-BPI-treated mice did not have hyperglycemia (i.e., blood glucose ≥ 250 mg/dl) after 7 weeks while only 17% of NOD.Scid mice did not have hyperglycemia when treated with CD4 + T cells from PBS-treated mice. The mice receiving CD4 + T cells from GAD-BPI-treated mice had significantly lower insulitis compared to those that received CD4 + T cells from PBStreated mice. 140 Incubation of isolated B cells from NOD mice with GAD-BPI had high spots of colocalization between I-A g7 MHC-II and ICAM-1 when evaluated by confocal microscopy studies. 140 This suggests that the GAD-BPI can simultaneously bind to MHC-II and ICAM-1 on the surface of B cells or APCs and can disrupt the IS formation.
Because autoimmune disease patients will normally show signs of disease before receiving treatments, BPI administrations for three consecutive days were delivered after the observed disease symptoms began. Thus, the animals were injected with BPI when initial disease symptoms were observed and stopped when remission was started. 141 Treatments with three BPIs (i.e., Ac-PLP-BPI-PEG3, Ac-PLP-BPI-PEG6, Ac-PLP-BPI-NH 2 -2; Table 1) for three consecutive days showed quick recovery from the disease before reaching the peak of the disease and before the normal remission time on day 20. 141 Recovery was observed immediately after the first injection of all three BPIs, and when the 3 injections were completed, the mice had very low disease scores or were close to that of heathy animals within 2-3 days. 141 As expected, PBS-treated mice had disease exacerbation without recovery from the disease peak until the normal remission days. Overall, BPI treatments reversed disease exacerbation when treated after the observed disease symptoms began.

Controlled-Release Formulation using Microparticles:
Because of the need for repeated i.v. administrations and the potential rapid clearance of Ac-PLP-BPI-NH 2 -2 or other BPIs (Table 1), alternative methods for delivery of BPI molecules were investigated. In this case, Ac-PLP-BPI-NH 2 -2 was loaded into PLGA microparticles in different doses and they were administered via s.c. route for a controlled-release formulation. The BPI particles preparation and characterization are described in Zhao et al. 142 The activity of the BPI (or Ac-PLP-BPI-NH 2 -2) in solution and in loaded microparticles were evaluated in six groups of EAE mice (ten/group). 142 As a control (Group 1), PBS-treated EAE mice had high disease scores and 20% decrease in body weight on days 10-12. The second group was injected with 4 × 100 nmol of Ac-PLP-BPI-NH 2 -2 (BPI) in solution via an i.v. route on days 4,7,10,and 14,respectively (Group 2). This treatment suppressed the EAE symptoms very effectively with no clinical signs of disease and no appreciable weight loss. Group 3 was injected with 4 × 100 nmol of BPI via s.c. route on days 4, 7, 10, and 14 and the treatments resulted in no symptoms and weight loss.
In Group 4, the animals were injected four times (i.e., days 4, 7, 10, and 14) with 100 nmol of BPI in solution + blank particles using the s.c. route. The animals in Group 4 had similar profiles as in Groups 2 and 3. The results from Group 4 suggested that blank particles has no activity. The Group 5 mice were injected with one injection of solution BPI molecule (100 nmol) on day 4 followed by one injection of nanoparticles containing 300 nmol of BPI on day 7. Clinical scores and body weight of Group 5 mice were clearly better than untreated controls but slightly worse than those receiving Ac-PLP-BPI-NH 2 -2 in solution. Mice in Group 6 were injected on day 4 with nanoparticles loaded with 400 nmol of BPI via s.c. route and this one treatment suppressed the disease completely, resulting in healthy animals (clinical score = 0). In summary, BPI can be delivered via a single s.c. injection using a controlled-release method and this method is more convenient than multiple injections of soluble BPI via i.v. or s.c. route.
Another controlled-release method to deliver Ac-PLP-BPI-NH 2 -2 was developed using two types of nanoparticles made from a mixture of poly(D,L-lactic-co-glycolic acid) (PLGA) coated with alginate and chitosan to form colloidal gel systems. The BPI was loaded by physical mixing into the gel. 143 The colloidal gel containing BPI was administered 11, 8, and 5 days before induction of EAE on day 0 to test the BPI efficacy as a vaccine for EAE. 143 Compared to the PBS control and blank colloidal gel, the BPI-colloidal gel showed a high effectiveness in suppressing EAE. The disease suppression levels were further evaluated by observing the level of inflammatory and regulatory cytokines after the treatment period. The mice in the BPI-colloidal gel group had significantly lower levels of inflammatory cytokines (i.e., IL-6, IL-17, and IFN-γ) compared to the control. 143 This was further confirmed by the increased levels of the regulatory cytokine IL-2. These results indicate that the colloidal gel method is an efficient method to deliver BPI molecules as a vaccine-like treatment as well as prophylactic or regular controlled-release treatment.

The Effects of BPI on the BBB Integrity and Production of Cytokines in EAE Mice:
The pathologies of the CNS in EAE mice are analogous to those of MS patients in terms of the immune cells (i.e., T cells, macrophages, monocytes) present in the brain lesions because of the leakiness in the blood-brain barrier (BBB) in both diseases. 21,144 The BBB leakiness can be detected by magnetic resonance imaging (MRI) in which a high brain deposition of gadolinium diethylene-triamine-pentaacetate (Gd-DTPA) is observed after its i.v. delivery. 145 The effects of the BPI in preventing the BBB leakiness was also evaluated using MRI. In this study, healthy mice (Group 1) were administered with Gd-DTPA; a low amount of Gd-DTPA was deposited in the brain. 145 The second group of mice had EAE, and after administration of Gd-DTPA, a significantly higher deposition of Gd-DTPA was found in the brains compared to those in control healthy mice. 145 This is due to the leakiness of the BBB in EAE mice. In the third group, the BPI-treated EAE mice treated were administered with Gd-DTPA and they had significantly low amounts of Gd-DTPA in the brains compared to those in untreated EAE mice. 145 These results indicated that BPI treatments in EAE mice suppressed the disease and prevented the BBB leakiness. At the same time, BPI molecules prevented axon demyelination and CNS infiltration of immune cells. 32 Increased levels of IL-17 have been correlated with the increase in disease activity in MS patients, where a high level of IL-17 was found in the cerebrospinal fluid and at the lesions. 68,146 The levels of IL-17 in the blood of Ac-PLP-BPI-NH 2 -2-treated mice were lower than in PBS-treated mice on day 35 after stimulation of the disease. 141 PLP-BPI also enhanced the levels of TGF-β, IL-2, IL-4, IL-5, and IL-10 cytokines, indicating the stimulation of Tregs and Th2 cells. 103,145 BPI delivered in solution or by a controlledrelease method lowered the production of inflammatory cytokines (i.e.,  implying that BPI molecules prevented the proliferation of Th1 and Th17 cells. 143,145 In summary, BPI molecules can alter and modulate the immune cell balance between inflammatory T cells (i.e., Th17 and Th1) and regulatory T cells (i.e., Treg) to control the disease.

I-Domain Antigen Conjugate (IDAC) Molecules
A different approach to the BPI molecules is by investigating the I-domain antigen conjugate (IDAC) for the treatment of EAE. In this case, the insert-domain of the LFA-1 α-subunit (or I-domain of CD11a) was linked to several antigenic peptides to make IDAC molecules ( Figure 5C; Table 2). [147][148][149] The I-domain can bind to D1 of ICAM-1 on APCs. 150 IDAC molecules were proposed to have a similar mechanism as the BPI molecules on APC ( Figure  3E). Thus, this binding process prevents the formation of the IS and alters naïve T cell commitment from inflammatory to regulatory T cells ( Figures 3E & 4D). [147][148][149] One of the benefits of the IDAC molecules is that multiple antigenic peptides can be conjugated to the 20 lysine residues within the I-domain. Therefore, a single I-domain molecule can have several homogenic or heterogenic antigenic peptides derived from PLP, MBP, and MOG to target various EAE animal models. In other words, IDAC molecules can be conjugated with diverse antigenic peptides from PLP, MOG, and MBP for treatments of patients who are sensitive to different antigens and to prevent antigenic spreading.
IDAC molecules (i.e., IDAC-1, -2, and -3) have been shown to suppress EAE animal models as vaccines, prophylactics, or therapeutic agents. [147][148][149] The lysine residues in the I-domain were modified with the maleimide groups followed by conjugation of the antigenic peptides (i.e., PLP-Cys-OH, PLP-Cys-NH 2 , and Ac-PLP-Cys-NH 2 ) via a thiol group on the Cys residue. Using mass spectrometry, it was determined that each I-domain contained an average of 2.5 antigenic peptides on the IDAC molecule. 147 The IDAC and I-domain molecules have similar circular dichroism (CD) spectra, indication that there is no secondary structural change upon conjugation of antigenic peptides to the I-domain. IDAC molecules (IDAC-1 and IDAC-3, Table 2) were evaluated in PLP-stimulated EAE by i.v. administrations (26 nmol/injection) on days 4 and 7. 147 Another BPI molecule, called Ac-PLP-cIBR1-NH 2 ( Table 1), was administered via i.v. injections (50 nmol/injection) on days 4, 7, and 10 as a positive control. The EAE disease symptoms were significantly suppressed in animals treated with IDAC-1, IDAC-3, and Ac-PLP-cIBR1-NH 2 compared to those treated with PBS or I-domain. IDAC-3 suppressed EAE symptoms better than IDAC-1. Two injections of IDAC-3 (26 nmol/injection) had similar efficacy as three injections of Ac-PLP-cIBR-NH 2 (50 nmol). These results also suggest that N-and C-termini capping of the PLP peptide improves the efficacy of the IDAC-3 molecule compared to that of IDAC-1. The IDAC molecules are more potent than BPI molecules presumably due to simultaneous delivery of multiple antigenic peptides by IDAC molecules.
Similarly, MOG-IDAC (IDAC-4, Table 2) was also administered on days 4 and 7 via s.c. (26 nmol/injection) to MOG-stimulated mice. As controls, EAE mice were injected with MOG 38-50 peptide (Table 1; 100 nmol/injection) and PBS on days 4, 7 and 10. IDAC-4 suppressed both disease progression and incidence better than those injected with MOG 38-50 peptide and PBS. Finally, IDAC-5 (Table 2, Figure 6) was synthesized to contain 3 different antigenic peptides derived from PLP, MOG, and MBP to test whether IDAC-5 could suppress both PLP-and MOG-stimulated EAE. n = 6) or PBS was administered once via s.c. on day 4 to PLP-and MOG-stimulated EAE mice. The results showed that IDAC-5 suppressed EAE symptoms in both PLP-and MOG-stimulated EAE mice better than PBS control (Figure 6; p<0.01 between days 12-17). These results suggest that IDAC molecules can be used to simultaneously deliver multiple heterogenic antigens to potentially prevent antigen spreading in EAE and MS.
IDAC-3 was administered in vaccine-like and regular treatment schedules on PLP-stimulated EAE mice. 149 The first and second groups of healthy SJL/J mice were injected via s.c. with IDAC-3 (26 or 10 nmol/injection, respectively) on days −11, −8, and −5 prior to EAE disease induction on day 0 as a vaccine-like administration. Similarly, the third and fourth groups of SJL/J mice were treated on days −11, −8, −5 with Ac-PLP-BPI-NH 2 -2 (100 nmol/injection) as a positive control and PBS as a negative control. In the fifth group, IDAC-3 (26 nmol/injection) was injected via s.c. on days 4 and 7 after disease stimulation on day 0. When administered using a vaccine-like schedule, both IDAC-3 and Ac-PLP-BPI-NH 2 -2 significantly suppressed the EAE disease symptoms compared to the PBS-treated animals. 149 The IDAC-3 effectiveness upon a vaccine-like administration was dose-dependent, where injections of 26 nmol IDAC-3 were more effective than 10 nmol injections. IDAC-3 administrations were more effective than Ac-PLP-BPI-NH 2 -2 and vaccine-like treatment may alter the immune cells prior to the disease induction. Therefore, further studies are needed to elucidate the mechanisms of IDAC-3 and Ac-PLP-BPI-NH 2 -2 in altering the immune cells prior to disease stimulation.
Compared to PBS treatment, IDAC-3 (26 nmol/injection) treatments significantly enhanced the IL-2 production during the peak of the disease on day 13. During the remission on day 35, the levels of IL-2 were very low in IDAC-3-and PBS-treated mice. 149 Antigen-specific activated T cells produced IL-2 upon antigen recognition; in contrast, Treg cells suppress IL-2-producing cells to prevent T cells from self-antigen recognition. 151,152 Animals treated with IDAC-3 produced significantly higher levels of IL-5 but not IL-4 when compared to PBS-treated mice during the remission on day 35. Normally, Th2 and mast cells produce IL-5 for inducing B cell growth and activating eosinophils; however, overproduction of IL-5 may induce allergic reactions. 68,83 Higher levels of IL-10 were observed in IDAC-3-treated mice compared to those of treated with PBS. IL-10 cytokine is normally generated by T-regs, Th2, and monocytes to down-regulate inflammatory response from Th17 and Th1 cells. 33,84 This result was supported by suppression of IL-17, TNF-α and IFN-γ cytokines via the upregulation of Treg cells due to IDAC-3 treatments.

Fc-BPI Molecules
The Fc-BPI ( Figure 5C) was designed to improve the delivery of antigenic peptides compared to BPI and IDAC molecules. The Fc-BPI is composed of the Fc-domain of an IgG1 mAb that is linked to the PLP peptide on the C-terminus while the N-terminal end is linked to the LABL peptide ( Figure 5C). 7 Since the IgG1 Fc is a homodimer, there are two N-termini and two C-termini for conjugating LABL and antigenic peptides respectively onto a single Fc molecule. The Fc region is stable on its own and remains stable with the addition of the antigenic and signal-2 blocking peptides. The advantage of the Fc-BPI compared to current IDACs is that, rather than a random mixture, the Fc-BPI is a single molecule which is easily controlled during synthesis. The molecule is more tractable analytically during preclinical and clinical studies compared to IDAC molecules. The presence of the Fc domain increases the solubility and may increase plasma stability of Fc-BPI because the Fc-domain is recognized by the FcRn receptor to improve the Fc-BPI's circulation time in the blood stream. Because of its size, Fc-BPI avoids glomerular filtration compared to BPI molecules. Intravenous injection of the Fc-BPI containing PLP peptides on days 4 and 7 (25 nmol/injection) significantly suppressed EAE in mice compared to PBS, as determined by disease scores and body weights on days 12-17. 7 Fc-BPI also prevented demyelination of axons in the EAE mouse brains. 7

Antigenic Peptide-Drug Conjugates
Peptide-drug conjugates are another class of drugs that aim to specifically target immunomodulatory agents against diseased cells in autoimmune diseases. 162 The PLP antigenic peptide was conjugated to dexamethasone (Dex) to make a PLP-Dex conjugate that was evaluated for its ability to suppress EAE in mice. 162 Dexamethasone effectively suppressed proinflammatory cytokines in the EAE mice to counteract the costimulatory signal of B7/CD28. The conjugate induced the upregulation of CTLA-4 to reduce a proinflammatory response. 162 While effective, dexamethasone can lead to general immune suppression; therefore, the addition of the antigenic peptide helps its targeted delivery to a subpopulation of immune cells. PLP-Dex can effectively suppress EAE in mice compared to those treated with dexamethasone alone or mannitol. 162 The results suggest that a combination of antigenic peptide and drug together can suppress the disease in an antigen-specific manner.

Mechanisms of BPI and BPI Related Molecules
In all previous studies, BPI molecules are more potent than the respective antigenic peptides alone. It is proposed that BPI molecules induce production of Treg cells and suppress inflammatory T cell proliferation through three different mechanisms while the antigenic peptides work via one mechanism ( Figure 4A). In the first mechanism, BPIs work in the same manner as soluble antigenic peptides ( Figure 4A). In this case, the BPI binds to the surface of the immature dendritic cell (iDC) via empty MHC-II through docking of the antigenic peptide portion of the BPI. 60,63,66,163 Because iDC does not express CD80/ CD86, binding of naïve T cell to iDC via TCR/Ag-MHC-II complex in the absence of CD28 binding to CD80/CD86 results in naïve T cell differentiation to Treg cells. In the second mechanism, the antigenic peptide fragment of the BPI binds to the empty MHC-II while LABL binds to ICAM-1 on the surface of the iDC ( Figure 4B). Because of the additional binding of LABL peptide on BPI to ICAM-1, the population of BPI is higher on iDC compared to antigenic peptide alone. Furthermore, binding of LABL fragment blocks the LFA-1/ICAM-1 costimulatory signal as well as the absence of CD28/CD80/86 costimulatory signal. Therefore, the production of Treg cells is more pronounced in mice treated with BPI compared to those treated with antigenic peptide alone. In inflammatory conditions, naïve T cells can also binds to mature DC (mDC) by forming immunological synapse via TCR/MHC-II and LFA-1/ICAM-1 or CD28/CD80/86 to generate inflammatory Th17 and Th1 cells ( Figure 4C). Thus, the third potential mechanism of BPI is that it interacts with mDC in which BPI molecule simultaneously binds to MHC-II and ICAM-1 on the surface of mDC ( Figure 4D). Upon mDC interaction with T cell, there is a connection between TCR and Ag-MHC-II complex; however, because LABL fragment binds to ICAM-1, it blocks of the binding between LFA-1 and ICAM-1 as costimulatory signal. Thus, the IS formation is inhibited to generate anergic effects on T cell activation. 6 Therefore, naïve T cells do not undergo clonal expansion to produce inflammatory Th1 and Th17 cells ( Figure 4D). In summary, it is proposed that the superior efficacy of BPI molecules over antigenic peptides is due three different mechanisms of BPI activity. Overall, treatments with BPI molecules and antigenic peptides tip the balance of immune cells from inflammatory to regulatory response followed by the restoration of the immune system balance to normal conditions.

Conclusion
Autoimmune diseases are difficult to treat due to their autoreactive nature. Many therapies help reduce the intensity of the disease, but often can suppress the general immune system. This shortcoming makes patients susceptible to opportunistic infections. To account for this limitation, antigen-specific immunotherapies are being developed to selectively target the diseased cell population while leaving the general immune system uncompromised. Soluble antigenic peptides and peptide conjugates (BPI, IDAC, Fc-BPI, polymer conjugates, and peptide-drug conjugates) have shown good-to-excellent efficacies in suppressing the EAE disease. BPIs and antigenic peptides are able to suppress the disease by binding to APCs to modulate the differentiation and proliferation of a specific subpopulation of T cells to a regulatory response rather than an inflammatory response. Some encouraging results have shown that antigenic peptides and BPI molecules can be used as prophylactics, treatments, and vaccines to suppress autoimmune diseases and alter the balance of immune cells to a regulatory phenotype. Although these treatments have been effective in animal models of MS, the mechanisms of action and potential side effects of these antigen-specific therapeutics still need further investigation. In the future, their mechanisms of action will be used to improve their efficacy and safety in treating MS and other autoimmune diseases patients.  The immune system has innate and adaptive immunity. The innate immune response is a fast response generated by macrophages, dendritic cells, neutrophils, eosinophils, basophils and natural killer cells. The adaptive immune response is a slow immune response generated by B cells and T cells. B cells are responsible to produce antibodies that recognize the pathogen for the humoral response. Antigen-specific T cells such as CD4 + and CD8 + T cells are generated as the cell-mediated response to pathogen infection. T cells and natural killer cells are at the interface of innate and the adaptive immunity.  There are various signaling proteins on the surface of antigen presenting cells (APCs) and T cells. The first signal, Signal-1, is generated by interactions between Ag-MHC-II complex and TCR to produce TCR/Ag-MHC-II complex. The costimulatory signal or Signal-2 is divided into positive and negative signals. The positive costimulatory signal can be generated by OX40L/OX40, CD40L/CD40, CD28/CD80, CD28/CD86, LFA-1/ICAM-1, or ICOS/ICOSL. The negative signal can be stimulated by CTLA-4/CD80, CTLA-4/CD86, or PD1/PDL1. The mechanism of immunological synapse (IS) formation at the interface of T cell and APC to activate the proliferation of antigen-specific inflammatory T cells (i.e., Th1 and Th17).

CIA
(A) The initial interaction between the T cell and APC is mediated by LFA-1/ICAM-1 complexes that form a cluster at the center (Red) and TCR/Ag-MHC-II complexes to form a cluster at the peripheral region (Green). LFA-1/ICAM-1 complexes translocate from the center to the periphery and at the same time TCR/Ag-MHC-II complexes translocate into the center. II and ICAM-1 on APC. Interaction of the T cell with the APC generates TCR/Ag-MHC-II complexes but binding of cell adhesion fragment (LABL peptide) to ICAM-1 blocks the LFA-1/ICAM-1 interactions. As a result, the BPI inhibits the formation of the IS to stop proliferation of inflammatory Th1 and Th17 cells. The potential mechanisms of action of antigenic peptides and BPI molecules to suppress autoimmune disease by controlling the autoreactive inflammatory T cells. (A) The antigenic peptide binds to empty MHC-II on the surface of immature DC (iDC). Binding of the naïve T cell to the iDC, using TCR to Ag-MHC-II in the absence of binding between CD28 to CD80/86, causes the differentiation of naïve T cells to Treg cells. (B) The BPI binds to both empty MHC-II and ICAM-1 on the surface of iDC. The binding of the naïve T cell to the iDC via the TCR/Ag-MHC-II complex in the absence of costimulatory signals from LFA-1/ICAM-1 and/or CD28/CD80 or CD28/CD86 results in production of Treg cells.