Abnormal apoptotic clearance can trigger TLRs and nucleic acid sensors on immune and non-immune cells and produce an immune response with cytokine production (7). Rare hereditary genetic mutations e.g. in DNASE1L3 and PRKCD that lead to abnormal apoptotic pathways provide crucial insight into the role of apoptotic breakdown and debris clearance in SLE (73, 74). DNase I activity degrades chromatin in the apoptotic process and mice with a mutation in this enzyme had increased levels of anti-DNA antibody production (75). Smoking induces cellular damage and promotes cytokine production, and UV light enhances apoptotic turnover, and thus may increase self-antigen burden in susceptible individuals (76, 77)

Nucleic acid sensors are important surveyors of the environment and are specifically able to recognize viral infections and induce type I IFN production. Toll-Like receptors 3, 7, 8, and 9 shape the immune response by sensing cellular debris (78). In a pristane-induced lupus mouse model, TLR7, which senses single stranded RNA, was required for RNA-reactive autoantibodies (8). TLR9 senses unmethylated CpG sequence motifs. SLE patients with active disease have higher level of TLR9+ B cells and monocytes than healthy controls, and TLR9 levels correlated with antibodies to dsDNA (79, 80).

Type I and Type II IFN contribute a large role to the pathogenesis of SLE and become elevated prior to development of autoantibodies (81). Rare single gene disorders, grouped together as Aicardi-Goutiere’s syndrome, display gene defects that cause an overproduction of type I IFN (82). These patients display similar phenotypes to classic SLE, including autoantibodies.

There is a marked imbalance of T cell cytokines in SLE, with low levels of IL-2 accompanied by elevated IL-17 and IL-6 (83). IL-2 is a key cytokine in Treg development, survival and maintenance. It restricts Th17 cell development (84, 85). Elevated levels of IL-17 are thought to induce tissue inflammation and recruitment of immune cells. B cell activating factor (BAFF or BLyS), expressed by stromal and immune cells, promotes B cell activation in SLE and its levels positively correlate with antibody levels (86, 87).

The process of autoimmune disease development can be roughly categorized into three stages: 1) a priming phase that includes an inciting event or accumulation of events in individuals at genetic and environmental risk; 2) the onset of clinical symptoms marked by organ-specific inflammation; and 3) a chronic inflammatory tissue-destructive phase (88). During the transition to clinically significant symptoms, regulatory processes, including Treg cells, fail to control pathological autoreactive B and T cells. This imbalance perpetuates the processes of bystander activation, epitope spreading and uncontrolled cytokine and antibody production. Epitope spreading involves the diversification of epitope specificity from the initial dominant epitope-specific immune response (89). The specificity of the autoimmune response spreads to include additional self-epitopes besides the initiating self-antigens. Chronic inflammation promotes tissue damage and cascading self-antigen presentation, expanding autoreactive T-cell specificities, including cryptic or sequestered epitopes (90). For example, late-stage SLE is characterized by an explosion of autoantibodies, apparently the result of chronic inflammation and epitope spreading (19). Bystander activation occurs with stress, infection or trauma-induced activation of tissue APCs, activating T cells of additional specificities, which further promote inflammation and tissue damage. Bystander T cells can provide help to B cells for autoantibody production, or to cross-presenting DCs presenting tissue-derived self-antigen (91). Treg cells may control bystander T cells and epitope spreading through interaction with cross-presenting DCs. In a rheumatoid arthritis mouse model Treg cell depletion promoted the expansion of pathogenic autoreactive T cells, an increase in inflammatory cytokines, and B‐cell epitope spreading (92).

SLE is marked by abnormal B and T cell interactions and spontaneous germinal centers in secondary lymphoid organs (93–95). In SLE there is loss of functional Treg and induction of effector T cells that produce proinflammatory cytokines and BAFF, which is not normally observed in healthy people (96, 97) (98). Multiple lines of evidence demonstrate the importance of Treg in lupus pre-clinical models. In the NZB/NZWF1 spontaneous model, Treg cell adoptive transfer delayed SLE progression, reduced renal pathology, and improved survival (61), while Treg depletion accelerated disease development (99). In human SLE, most but not all studies demonstrate a reduced frequency of Treg cells (100, 101). Targeted depletion of pDCs decreased SLE-associated glomerulonephritis in mice (102, 103). In human SLE, while pDC are decreased in the blood, they are increased in lupus-affected organs, suggesting their chemo-attraction and possible expansion at these sites (104–106).

Several groups have developed methods to expand Treg cells ex vivo for reintroduction as an autologous cell therapy product. Treg cells can be isolated from peripheral blood or umbilical cord blood, but must be expanded due to their low frequency. In vitro expansion strategies include anti-CD3/CD28-coated beads, with addition of IL-2 and/or TGF-β and rapamycin (121). Proof of concept experiments in lupus-prone mice showed that ex vivo-expanded Treg cells suppressed glomerulonephritis and prolonged survival (61, 122). Ex vivo-expansion of Treg cells in the presence of immunosuppressive drugs or Treg transfer into patients on immunosuppressants can be challenging, as the drugs may hinder expansion or change function (112). Furthermore, the process requires a good manufacturing practice (GMP) environment, which is challenging and expensive. A clinical trial using ex vivo-expanded autologous polyclonal Treg cells in patients with autoimmune disease was terminated in November 2019 due to screen failures and low enrolment. In a case report, the treatment was shown to be safe and clinical disease activity to be stable in a single SLE patient. Infused labeled Treg cells were transiently observed in PB then in diseased SLE skin, accompanied by skewing from Th1 to Th17 immunity locally (123). Treg are highly plastic and may differentiate to Th17 in inflammatory settings and where IL-2 is limiting (124). Larger studies are needed to understand the impact of Treg therapy on disease severity.

Hematopoietic and/or mesenchymal stem cell transfer (HSCT and MSCT, respectively) have been trialed in patients with severe autoimmune diseases, including SLE, who have failed standard therapy. In SLE patients, HSCT has successfully induced long-term remission (125). In 15 patients with severe SLE evaluated up to 8 years after HSCT, CD4⁺CD25^highFoxp3⁺ Treg and LAP^highTGF-β⁺CD8⁺Foxp3⁺ cells were restored to levels and function similar to healthy subjects (117). These promising results suggest that HSCT may reestablish immune tolerance by replenishing multiple types of Treg cells. However, as HSCT is associated with significant risks, treatment complications and cost, it is currently reserved for treatment-refractory patients. A 4-year follow-up of an open-label trial of MSCT in 87 treatment-refractory SLE patients found a 28% remission rate post-infusion (118). While double-blind placebo-controlled trials are needed to understand the true benefits of MSCT, these trials provide evidence that tolerance may be successfully re-established in SLE.

IL-2 levels and CD25 expression by Treg are reduced in SLE patients and murine lupus models (126–128). IL-2 plays a pleiomorphic role in the immune system. One of its functions is to expand and promote survival of Treg cells (129, 130). Reduced IL-2 favors the differentiation of IFN-γ-producing Th1 and IL-17 producing Th17 cells and their accumulation in skin and kidneys (131, 132), and is associated with inflammation. In lupus-prone mice, IL-2 treatment increased levels of Treg cells in lymphoid and peripheral organs and protected them from SLE-related organ damage (99, 133). There have been several trials in lupus showing safety and Treg expansion (128, 134). In a recent double-blind placebo-controlled clinical trial in patients with suboptimally controlled SLE, LD IL-2 for 12 weeks (s.c. alternate days for three 2-week cycles), the SLE Responder Index (SRI)-4 response rates at week 12 were 55.17% and 30.00% in LD IL-2 and placebo groups respectively (p=0.052). Although the primary end point was not met, the significantly greater lupus nephritis complete remission rate in the LD IL-2 arm was notable. Immunologically, IL-2 supplementation significantly increased Tregs and NK cells but did not change total CD4+ or CD8+ T cells and there was no increase in viral load of pre-existing viruses (3). While promising, LD IL-2 dosing may be complicated by concomitant expansion of regulatory and cytotoxic cells. Furthermore, development of neutralizing autoantibodies with continued treatment is a potential risk (135). Targeted IL-2 therapies may allow more precise manipulation of the immune response and longer duration of action. For example, anti-CD4 and anti-CD2-coated poly(lactic-co-glycolic) acid (PLGA) nanoparticles loaded with IL-2 and TGFβ expanded Treg cells in vitro and in vivo in the BDF1 lupus pre-clinical model (136). In a recent phase 1b clinical trial of a polyethylene glycol (PEG) conjugate of IL-2 (NKTR-358) in patients with mild to moderate SLE, dose-dependent increases in Tregs (up to 11 fold) were observed, which returned to baseline 20-30 days post-dose (137). Anti-IL-2 antibodies were not reported.

DCs play a critical role in maintaining self-tolerance. Indeed, targeting steady-state skin migratory DC with antigen coupled to DC-selective antibodies induced antigen-specific tolerance (138). Tolerogenic DCs can also be generated in vitro from monocytes or murine bone marrow precursors in the presence of NF-κB inhibitors 1,25 (OH)₂ vitamin D3 (calcitriol), rapamycin or glucocorticoids. After proof-of-concept studies in experimental animal models (139, 140), several groups translated antigen-specific immunotherapy using modified or tolerogenic autologous DCs and autoantigenic peptides to clinical trials for MS (113) and RA (114, 115). These trials demonstrate the feasibility and safety of this approach, with preliminary evidence of an immunomodulatory effect in RA. In two pre-clinical lupus models, histone antigen-loaded tolerogenic DCs improved clinical scores, increased Treg in affected skin and reduced anti-histone autoantibodies (141). Tolerogenic DCs exposed to apoptotic cells were generated from PB monocytes derived from lupus patients (142). Other approaches have been developed to target DCs directly in situ, including a PLGA nanogel to deliver the immunomodulator mycophenolic acid (MPA) to DCs (119, 120). DCs took up the PLGA-lipid-MPA nanogel more efficiently and with better DC suppression than a PLGA nanogel. In a murine lupus model, PLGA-MPA nanogel increased median survival by 3 months when given prophylactically and by 2 months when given to mice with advanced renal damage. Consistent with the local effects of MPA on DCs, treated mice had a substantial reduction in DC-derived inflammatory cytokines such as IFN-γ and IL-12. Although not strictly immune tolerance, this approach achieves sustained delivery of MPA to induce a prolonged anti-inflammatory effect.

Lupuzor (rigerimod or IPP-201101) is a 21aa peptide representing residues 131–151 of the 70K spliceosomal protein within the U1 small nuclear RNP, phosphorylated at Ser140. This promiscuous peptide sequence was identified using ex vivo peptide screening techniques (143). This epitope is recognized by IgG antibodies and CD4+ T cells from H‐2^k MRL/lpr and H‐2^d/z (NZB × NZW)F1 lupus‐prone mice (143, 144). With i.v. delivery, the peptide inhibits chaperone-mediated autophagy and reduces B cell MHC class II expression (145). Two trials of IPP-201101 immunotherapy in SLE demonstrated safety and potential efficacy (146, 147). However, IPP-201101 failed to meet its primary end point of superiority over standard care in phase III clinical trials (148). The peptide seemed to have non antigen-specific immunomodulatory properties, rather than inducing antigen-specific regulation, and this may be why it was not superior to standard care. Standard of care high dose glucocorticoids and immunosuppressive drugs are likely more bioavailable than an immunosuppressive peptide.

These treatment strategies are antigen non-specific and use nanoparticles (NP) to deliver biologics or immunosuppressive drugs. In the following sections we consider antigen-specific tolerizing approaches using NP in SLE.

Liposome formulations loaded with peptide or protein antigens and various NF-κB inhibitors, including curcumin, quecertin and BAY11-7082 induced antigen-specific tolerance in mice with antigen-induced arthritis (155). We also developed and undertook pre-clinical studies of liposomes co-encapsulating calcitriol and peptide. Calcitriol/peptide liposomes promoted the differentiation of antigen-specific Foxp3⁺ Treg, anergy of Tmem, and IL-10 production upon restimulation with antigen ex vivo (156). Notably, liposomes were preferentially taken up by activated PD-L1⁺ migratory DCs, and regulation was PD-L1-dependent. We translated this to a phase 1b clinical trial in RA. Other groups have co-encapsulated antigens in NPs with either rapamycin (157) or aryl hydrocarbon receptor (AhR) ligands (158) for in vivo uptake by DC. With substitution of suitable lupus antigenic peptides, these liposome or NP approaches could be adapted to lupus patients.

Other research groups have developed NPs that resemble apoptotic bodies, to promote a tolerogenic response to encapsulated antigen. Specifically, i.v. administration of 500nm PLGA particles encapsulating antigen induced antigen specific tolerance (159, 160). These relatively large, negatively-charged particles are preferentially taken up by DCs and macrophages expressing MARCO, and induce antigen-specific suppression in the absence of an immunomodulatory drug (161). Another strategy to mimic signals from apoptotic bodies uses phosphatidylserine (PS) liposomes. During apoptosis, the PS phospholipid translocates from the inner leaflet to the outer leaflet of the lipid bilayer of the dying cell. PS liposomes suppressed pre-clinical models of T1D and acute EAE in a non-antigen-specific manner (162, 163). It is unclear whether this technique would succeed in SLE, which is characterized by impaired clearance of apoptotic cells.

The TCR may also be directly targeted with NPs coated with peptide loaded onto MHC class I or II, without co-stimulation. After i.v. delivery of iron oxide nanoparticles coated with peptide-MHC class I complexes (pMHC-I) they suppressed autoreactive CD8+ memory T cells and converted them to a regulatory, anergic phenotype (110). Nanoparticles coated with pMHC-II differentiated cognate autoreactive CD4 memory T cells into Tr1 cells producing IL-10 (111, 164). Nanoparticles coated with pMHC-II suppressed autoimmune symptoms in several pre-clinical models in an antigen-specific manner, without compromising systemic immunity (111). To date, this approach has not been translated to clinical trials.

Thus, a wide array of nanoparticle technologies has been developed. Figure 1 describes some of the technologies incorporating autoantigens, immunomodulatory drugs, or targeting strategies, or a combination of strategies. In summary, approaches that promote the expansion of antigen-specific Treg cells, particularly Tr1 cells derived from autoreactive memory T cells, will be required to control bystander cytokine production and epitope spreading in multi-system autoimmune diseases, such as SLE.