Introduction

Autoimmune diseases affect around 5–10% of the world’s population [1], with autoimmune thyroid disease and type 1 diabetes (T1D) being the most common. There are nearly 100 distinct autoimmune diseases, which can be either organ-specific such as primary biliary cirrhosis, or systemic, involving multiple organs such as systemic lupus erythematosus (SLE) [2]. Incidence is influenced by demographic factors like age, gender, ethnicity, and others. They can occur at any age, but each disease has its characteristic age of onset, with females being usually more affected than males [2].

Due to the evolution of research, significant advances in the understanding of human autoimmunity have been achieved, allowing improvements in the classification, diagnosis, and, most importantly, management of these autoimmune diseases.

The cornerstone for the treatment of autoimmune diseases has long been immunosuppressants (e.g., corticosteroids), until the advent of immunomodulatory biologic drugs, which block inflammatory mediators, including proinflammatory cytokines. At the frontier of these biological drugs are tumor necrosis factor (TNF)-α blockers, used in rheumatoid arthritis (RA), psoriasis, ulcerative colitis, and Crohn’s disease for example [1]. They have been considered the “standard of care” due to their effectiveness in disease control and risk reduction [1].

Despite the success of these immunomodulatory therapies, there still exist significant unmet clinical demands. First, currently approved therapies are immunomodulatory, symptom-relievers, rather than direct curatives for loss of immune tolerance [1]. Second, due to their action on the reduction of systemic inflammation, they increase susceptibility to opportunistic infections or malignancies and have critical side effects [3–5]. Third, these agents are of no efficacy in the case of refractory autoimmune diseases (e.g., multiple sclerosis) and in patients who have grown unresponsive [6, 7].

The new treatment paradigm is based on targeting known immune tolerance mechanisms, such as increasing regulatory T cells (Tregs) and inducing anergy or elimination of auto-reactive immune cells. These comprise coinhibitory checkpoint agonists and co-stimulatory checkpoint antagonists to expand Tregs and/or suppress pathogenic effector cells [1]. Moreover, a review presented by Lopez-Santalla et al. [8] addressed a new type of treatment, which is cell therapy.

Autoimmunity is indeed a challenge; nevertheless, the prognosis for patients with these diseases has and will continue to remarkably improve as new approaches are being investigated.

Physiology of the immune system and key co-stimulatory molecules

DCs: key cells governing immune responses

DCs are professional APCs that efficiently link innate and adaptive immunity and modulate T cell responses. They constantly capture, process, and present Ags to T cells, influenced by the surrounding milieu. Subsequently, T cells either express stimulatory molecules to activate protective immunity against pathogens/cancer cells or inhibitory ones to tolerate self-Ags [12]. For this, understanding the role of distinct co-stimulatory pathways on DC function is indispensable.

DCs are said to be immature during homeostatic conditions, with minimal expression of co-stimulatory molecules and proinflammatory cytokines. Environmental cues such as pathogen- or danger-associated molecular pattern molecules direct their differentiation into mature DCs, with upregulated expression of co-stimulatory molecules and proinflammatory cytokines.

DC maturation leads to an upregulated expression of homing receptors [e.g., chemokine receptors (CCRs)], augmenting DC migration to lymph nodes, and the development of cellular extensions to expand the contact surface area with T cells. Additionally, there is an upregulation of MHC groups and an enhanced release of cytokines for more potent T cell recruitment [12]. On the other hand, interleukin-10 receptors (IL-10Rs) are downregulated to resist the immunosuppressive signals of interleukin-10 (IL-10) [13, 14].

Activation of naïve T cells by DCs relies on 3 signals (Figure 1): MHC presentation of the processed Ag to specific TCR, the action of multiple co-stimulatory molecules, and the role of the released cytokines in the milieu (e.g., IL-12, IL-6, TNF-α). The integration of all these signals directs T cell differentiation to an inflammatory versus tolerogenic profile [15].

Display full size

Figure 1. 

T cells activation by DCs. Th cell activation from DCs needs 3 signals which are 1) TCR recognition of the human leukocyte Ag (HLA) + Ag complex, 2) co-stimulatory molecule, B7, interaction with CD28, and 3) cytokines released from DCs to act on Th cells. Th cell activation involves proliferation through autocrine IL-2 signaling. This figure was created by Servier Medical Art (https://smart.servier.com/), licensed under a Creative Commons Attribution 4.0 Unported License

DCs also participate in mediating self-tolerance and thus the prevention of autoimmune diseases. This is due to their ability to eliminate autoreactive T cells and their support of the Treg population. In the absence of immune aggression, and thus the absence of proper co-stimulation and cytokines necessary for maturation of DCs and T cells, immature DCs presenting self-Ags will not yield any response and would rather result in autoreactive T cell apoptosis and anergy [12, 16]. Also, immunosuppressive molecules [e.g., IL-10, transforming growth factor-β (TGF-β)] act on DCs that present self-Ags on MHCII resulting in the expansion of the Treg pool [17].

Hence, it can be conferred that DCs may perform different actions under different conditions. Eliciting a tolerogenic vs. an immunogenic T cell response relies heavily, yet not solely, upon the nature of the available co-stimulation.

DC abnormalities in autoimmune diseases

In the normal physiological steady state and in the absence of danger or pathogen signals, DCs are naturally tolerogenic. This means that they do not activate T cells to mount an immune response. Instead, they promote the development of Tregs. Naturally, tolDCs have low MHC and co-stimulatory molecules on their surface [16]. They also produce IL-10 and TGF-β, which are well-known for tolerance induction and immunomodulation [16]. All these properties, among other complex intricate interactions between DCs and T cells, lead to self-tolerance. On the other hand, the presence of proinflammatory molecules that get upregulated chronically leads to an imbalance in the DC profile, which in turn promotes the mounting of an immune response to self-Ags [72]. DC-T cell interactions through surface molecules, as well as secreted cytokines, play an important role in achieving tolerance versus disease [73]. It thus comes with no surprise that these cellular interactions are extremely targeted in therapies aiming to treat autoimmune diseases.

DC-based therapies

Many new potential therapies for autoimmune diseases are based on DCs (Table 1). Using this approach, DCs are cultured ex vivo and manipulated in various ways to generate DCs that have tolerogenic properties. These DCs are then administered to the patients with the goal of reaching T cell anergy or Treg differentiation. TolDCs are characterized by low presentation of co-stimulatory molecules (CD80/CD86), high production of the anti-inflammatory cytokine, IL-10, and low secretion of pro-inflammatory cytokine [74].

There are several techniques to generate this immature tolerogenic profile of DCs. In most cases, monocytes are first extracted from the patients, transformed into DCs by using granulocyte-macrophage colony-stimulating factor (GM-CSF) and IL-4, and then modulated in a certain way for them to be tolerogenic [75]. Moreover, using low concentrations of GM-CSF by itself was proven to generate ATDCs [76]. The generated DCs may be pulsed with the auto-Ag to be targeted in order to produce a specific tolerogenic response for this particular Ag, or they may be injected in specific regions of the body where the disease mainly manifests.

A widely used technique for generating tolDCs is based on Dexa addition during DC differentiation from monocytes as it prevents the formation of fully mature DC [77]. This approach has been used in many clinical trials for RA, inflammatory bowel disease (IBD), Crohn’s disease, and multiple sclerosis. For RA, in a phase 1 clinical trial, ATDCs treated with Dexa were injected intra-articularly in the knees. After 3 months, synovitis in the knee joint has decreased, and this treatment proved to be safe and well-tolerated [78]. Furthermore, in a phase 1 clinical trial for Crohn’s disease, semi-mature DCs generated by adding IL-1β, IL-6, TNF-α, and PGE2, in addition to Dexa and vitA, were administered intraperitoneally. A clinical improvement was observed in one-third of patients, and the treatment was also generally well-tolerated. However, a significant increase in Tregs was observed in the circulation after 12 weeks with T cell hypo-responsiveness to polyclonal activation [79].

Moreover, 1,25-dihydroxyvitD3 [1,25(OH)₂D₃], the active form of VitD3, generates tolDCs when added to DCs in the differentiation phase by causing low Ag presentation, low co-stimulatory molecule presentation, and inhibition of IL-12 [80]. Thus, it comes as no surprise that this molecule has been implemented in clinical trials. In T1D, a phase 1 clinical trial using intradermal vitD3-treated ATDCs, pulsed with proinsulin peptide C19-A3, proved to be safe without producing systemic immunosuppression [81]. An interesting study used DC treated with Dexa applied on day 3 of DC differentiation and both VitD3 and Dexa applied on day 6 [82]. These DCs were then stimulated with monophosphoryl lipid A (MPLA), to become semi-mature, and pulsed with autologous Ags collected from the synovial fluid of inflamed joints. After intra-articular injection of the resulting tolDCs, patients who received the highest dose had improved clinical symptoms without side effects or systemic immunomodulatory consequences [82]. VitD3-treated DCs are also currently being studied for multiple sclerosis treatment [83].

Other than the previously mentioned studies about RA, there are two more worth mentioning. CreaVax-RA consists of autologous semi-mature DCs pulsed with 4 Ags related to RA [recombinant peptidylarginine deiminase 4 (PAD4), RA33, citrullinated-filaggrin, vimentin]. When CreaVax-RA is administered repetitively as a subcutaneous injection, a significant decrease in auto-Abs and IFN-γ-secreting T cells was noted, in addition to a good-to-moderate EULAR response. These clinical responses were achieved with minimal side effects [84]. Rheumavax, which is similar to CreaVax-RA in regard to pulsing with putative RA Ags and intradermal injection, uses however DCs generated in the presence of NF-kB inhibitor (BAY 11-7082). After one month from the injection, Rheumavax decreased DAS-28, with a reduction in pathogenic T cells and inflammatory chemokines secretion [85].

More techniques are also implemented, such as using DC treated with antisense oligonucleotides targeting co-stimulatory molecules (CD40, CD80, and CD86) decreasing their expression, or with rapamycin, which is an antibiotic with immunosuppressant properties [86, 87]. Furthermore, adding IL-10 during DC differentiation and maturation would result in a tolerogenic phenotype [88, 89]. In addition, experimental work in mouse models proposed that DC expressing Fas ligand would cause apoptosis of active T cells that express Fas, and thus, when these DC were injected in mouse models of collagen-induced arthritis (CIA), they resulted in disease amelioration [90]. In another murine model of T1D, autoimmune regulator (AIRE)-overexpressing DCs that express T1D Ags induced Ag-specific immune tolerance with lower blood glucose levels and less pancreatic inflammation. AIRE expression seems to downregulate the expression of co-stimulatory molecules (CD40, CD80/86) which would therefore cause tolerance to auto-Ags specifically [91].

In vivo manipulation of DCs

As generating DCs ex vivo, making them tolerogenic, and injecting them into the patient is quite complex and expensive, researchers thought of manipulating DCs in vivo by using nano-medicines and Ag-delivering Abs. These approaches, however, are quite new with limited translation into clinical studies.

Nanoparticles can be engineered to deliver certain drugs to the inside of a target cell-based on the nanoparticles’ size and chemical interactions [92]. This technique has been exploited with the aim of treating autoimmune diseases by developing nanoparticles that would target DCs. In pre-diabetic mice, T1D onset was prevented in 40% of mice treated with nanoparticles containing insulin B, GM-CSF, TGF-β1, and VitD3, which have been stated above to be tolerogenic [93]. This outcome was achieved because DCs would take up the potential auto-Ag, along with the tolerogenic molecules [93]. Thus the Ag would be presented in a tolerogenic environment, and cause T cell tolerance with an increase in Tregs [93]. In another approach, nanoparticles containing antisense oligonucleotides, targeting CD40, CD80, and CD86 co-stimulatory molecules, injected in an area that is drained by pancreatic lymph nodes, caused a reversal of diabetes in new-onset non-obese diabetic (NOD) mice [94]. The concept in this study is that even if the auto-Ag is not present within the nanoparticle, injecting nanoparticles with tolerogenic molecules into an area where DCs would naturally present the auto-Ag would also yield the same result [94].

Furthermore, mice with experimental autoimmune encephalomyelitis (EAE), which are models of multiple sclerosis, were also used in preclinical studies for nanoparticle therapy. Similar to the previously mentioned experiment with prediabetic mice, nanoparticles encapsulating the auto-Ag of EAE, proteolipid protein, with a tolerogenic agent, rapamycin, prevented EAE relapse [95]. IL-10 was also used as a tolerogenic molecule in nanoparticles for EAE [96]. Nanoparticles encapsulating the target Ag only, such as in the case of TAK-101 nanoparticles that encapsulate gliadin, also show promising results. In a phase 2 clinical trial, intravenous administration of TAK-101 nanoparticles in celiac disease patients on a gluten-free diet was associated with reduced gluten challenge-induced intestinal damage [97]. When TAK-101 nanoparticles deliver gliadin into the inside of DCs, this Ag will be presented in a tolerogenic environment as no inflammation will be provoked through this delivery mechanism [98].

Other than nanoparticles, Ag-delivering Abs also prove to be an efficient way of delivering molecules into target cells [99]. In theory, when Ags of interest are conjugated to Abs that would bind the target cell, these Ags would get internalized into this cell [99].

For example, naturally pro-tolDCs in the human body express DEC-205; thus, delivering anti-DEC-205 Abs conjugated to auto-Ags of interest would make DEC-205⁺ DCs present this Ag and promote tolerance. Anti-DEC-205 chimeric Abs fused with proteolipid protein helped reduce the severity of EAE in mouse models by inducing anergy in effector T cells, thus decreasing the number of pathogenic T cells in the CNS [100]. Several other studies have been carried out in a similar fashion for NOD, proteoglycan-induced arthritis, and experimental colitis mouse models for T1D, RA, and IBD respectively [101].

Co-stimulatory molecules-targeted therapies

Co-stimulatory molecules present on DCs and on other immune cells are being targeted by drugs for the aim of treating autoimmune diseases. This approach is not only used for autoimmune diseases as the new immune checkpoint inhibitors targeting these molecules showed robust results in cancer therapy as well [102]. Therapies developed for autoimmune diseases mainly target CD80/86 complex with CD28, CTLA-4, CD40, and CD40L, and some other less known molecules such as ICOS, CD137, OX40, etc.

CD80/86, CD28/CTLA-4

CD80/86 is an important co-stimulatory molecule, as aforementioned, since it provides “signal II” needed for launching an immune response. Without this signal, tolerance and anergy would rather develop; hence, it becomes clear why many therapies focus on blocking this molecule (Table 2 and Figure 2). Abatacept (CTLA-4-Ig), an Food and Drug Administration (FDA)-approved drug for the treatment of RA and psoriatic arthritis, is basically a fusion protein holding the extracellular portion of CTLA-4 fused with the modified fragment crystallizable (Fc) portion of human IgG1. CTLA-4 binds CD80/86 and blocks its interaction with CD28, mimicking a tolerogenic state in which CD80/86 molecules are less expressed. This drug proved to be efficacious and relatively safe in the long term [103]. The downsides to this medication are that it is costly and that it may predispose to infections as it targets all CD80/86 molecules and not only those related to the disease [103, 104]. A very similar drug is belatacept, which binds approximately 4-fold more strongly to CD86 and 2-fold more strongly to CD80 than abatacept [105]. Belatacept is FDA-approved for post-renal transplant immunosuppression as it does not cause the nephrotoxicity associated with calcineurin inhibitors, namely cyclosporine. Moreover, treatment with belatacept showed higher patient and graft survival after 7 years compared to cyclosporine. However, this treatment was associated with more early acute rejection [106]. Two additional fusion proteins XPro952348 and MEDI5256 bind CD80/86 molecules even more avidly than the previously mentioned drugs, with MEDI5256 having a preferentially way more stronger binding to CD80. These molecules also have a longer half-life due to changes in the Fc region of the fusion protein. Both newer drugs proved to be more efficacious and more potent than abatacept and belatacept in suppressing humoral responses following keyhole limpet hemocyanin (KLH) immunization [107, 108].

Table 2.

Summary of Abs targeting CD80/86 interaction with CD28

Ab	Properties	Clinical application	Desired effects	Downsides	References
Abatacept	Fusion protein of the extracellular domain of CTLA-4 is linked to modified Fc portion	FDA-approved for: RA Psoriatic arthritis	Blocks CD80/86 interaction with CD28 Efficacious and relatively safe	Predisposes to infections Costly	[103, 104]
Belatacept	Similar to abatacept in structure 4-fold higher affinity to CD86 2-fold higher affinity to CD80	FDA-approved for: Immunosuppression after renal transplant	Higher graft survival compared to cyclosporine Does not cause nephrotoxicity associated with cyclosporine	Higher early acute rejection	[105, 106]
XPro952349 & MEDI5256	Similar to abatacept in structure Higher avidity to CD80/86 than abatacept and belatacept	-	Decreased humoral response to KLH immunization	-	[107, 108]
TAB08	Agonistic anti-CD28 Ab	Phase 2 clinical trial for RA	At low doses, minimal activation of CD28 preferentially expands Tregs and increases serum IL-10	Causes cytokine release syndrome if administered with a higher dose	[109–111]
FR104	Pegylated Fab with antagonistic CD28 activity	-	Decreased humoral response to KLH immunization in healthy human volunteers Improved GVHD-free survival in primate models	Increased infectious complications	[112, 113]

Display full size

This table shows the various molecules generated to target co-stimulatory molecules (CD80/86) interaction with CD28. Inhibiting this interaction leads to T cell anergy and an inhibited immune response. -: no data. Fab: fragment Ag-binding; GVHD: graft-versus-host disease; TAB08: theralizumab

Display full size

Figure 2. 

Anti-B7 molecules mechanism of action. Inhibiting the interaction between the co-stimulatory molecule, B7, with CD28 through an anti-B7 molecule (e.g., abatacept, belatacept) leads to T cell anergy and an inhibited immune response. This figure was created by Servier Medical Art (https://smart.servier.com/), licensed under a Creative Commons Attribution 4.0 Unported License

Other than CTLA-4-Ig, agonistic anti-CD28 Ab (TAB08) serves an intriguing finding, where at low doses of this Ab, minimal activation of CD28 would preferentially expand Tregs and thus help with tolerance [109]. However, careful dose titration is required since a previous trial of the same Ab but with a higher dose resulted in cytokine release syndrome [110]. With the appropriate dosing though, TAB08 increases serum IL-10 without causing cytokine storm. This Ab has entered phase 2 clinical trials in RA as it showed a clinical response with an acceptable level of adverse effects [111]. On the other hand, antagonistic anti-CD28 Ab also has a role in treating autoimmune diseases. FR104, a pegylated Fab with antagonistic CD28 properties, hampered humoral response to KLH challenge in human healthy volunteers [112]. Moreover, this Ab served better than CTLA-4-Ig in improving GVHD-free survival in non-human primate models, yet it has infection-related complications [113].

Genetic engineering of DCs

In addition to the previously mentioned approaches, gene therapy has also gained popularity in generating tolDCs. There are many techniques involved in gene therapy, such as using viruses and liposomes for gene transfer or using RNA interference (RNAi) to silence particular genes.

DCs producing high levels of IL-10 have been genetically engineered using a lentiviral vector encoding the IL-10 gene. These cells, called DC-IL-10, have decreased allogeneic T cell responses in humanized mouse models and seem promising for use in clinical practice in the context of allogeneic transplantation [137]. This technique has also been exploited in generating Ag-specific tolerance in T cell-mediated diseases. Transferring immunodominant Ag-derived peptides along with the IL-10 gene to DC using lentiviral vectors helps in producing Ag-specific Treg [138]. These DC-IL-10/Ag cells prevented the development of T1D in preclinical disease models.

Other than IL-10, researchers have also tried to transfer other genes to DCs. Since NF-kB is very important for DC maturation, a gene coding for kinase-defective dominant negative form of IKK2 (dnIKK2), which blocks NF-kB activation, was transferred to DCs using an adenoviral vector. This way DCs will not be able to mature properly and would produce a regulatory profile of T cells. These produced Tregs also inhibit naive T cell responses to antigenic stimuli with mature DCs [139]. Moreover, IL-4 coding adenoviral vectors were used in the transduction of DCs producing DC/soluble IL-4 (sIL-4). These genetically modified DCs alter the immune response to the Th2 profile rather than Th1 and were proven to reduce the number of prediabetic NOD mice that developed diabetes [140].

Furthermore, RNAi is a widely used genetic engineering method for DCs. As previously mentioned, antisense oligonucleotides targeting CD40, CD80, and CD86 molecules produce tolDCs [86, 94]. In addition, silencing IL-12 expression with siRNA stimulates the production of Th2 cytokines, as well as IL-10. This therapy would help reduce Th1 dominant alloimmune or autoimmune responses [141]. Finally, immunoliposomes are also used in genetic engineering as non-viral vectors. For example, immunoliposomes can deliver to DCs a gene encoding for CTLA-4 protein fused to an endoplasmic reticulum retention signal. This fusion protein prevents CD80/86 molecules from reaching the cell surface by keeping them in the endoplasmic reticulum, thus promoting tolerance [142].

Conclusions

As previously mentioned, novel therapies with advanced strategies are being implemented for autoimmune diseases. These therapies are being improved based on intensive research studies ranging from preclinical and experimental works to clinical trials. Each therapy is designed to be efficient in inhibiting or reversing a specific disease process, thus targeting one single disease or a group of diseases with a similar pathophysiology. Moreover, the most appropriate dosing, routes of administration, and other pharmacokinetic properties are being established to have the best clinical results possible. These therapies, however, need time to be implemented into clinical practice as they are still new and need to be tested thoroughly before they get approved for the market. Nonetheless, it’s expected that in the near future, the management of autoimmune disease will evolve, and these molecular and cell-based therapies will be a major milestone in immunotherapy.

Furthermore, the novel immunotherapy agents match the modern shift in medicine towards precision and personalized treatments. As everyone has a specific set of genetic characteristics, the disease process in each patient will be slightly different from that in others. Thus, therapy should be tailored to the specific needs of each patient. Certain genetic mutations and pathological markers can be evaluated, and appropriate therapy will be administered. Novel immunotherapy can achieve this and oppose the disease process in each individual patient.

Many other treatment modalities not mentioned in this review are also being studied, such as mesenchymal stem/stromal cell-based therapies among others [10]. The hope is that autoimmune diseases and similar disease processes, such as allergies and transplant rejection, can be cured or prevented without the need for long term immunosuppressive therapy with devastating side effects.