L. Wang, Department of Molecular Biology, Medical College of Norman Bethune, Jilin University, Changchun, 130021, China. Email: email@example.com and
Y. Yu, Department of Immunology, Medical College of Norman Bethune, Jilin University, Changchun, 130021, China. Email: firstname.lastname@example.org. Senior author: Liying Wang
To explore the possibility that human mitochondrial genomic DNA-mimicking oligodeoxynucleotides could regulate the immune response, a series of mitochondrial DNA-based oligodeoxynucleotides (MTODNs) were designed and studied to determine their immunoregulatory effects on immune cells activated by toll-like receptor (TLR) stimulation. The results showed that a C-rich MTODN, designated MT01, was able to inhibit the proliferation of human peripheral blood mononuclear cells (PBMCs) induced by cytosine–phosphate–guanosine (CpG) oligodeoxynucleotides (ODNs) and the production of type I interferon (IFN) from human PBMCs stimulated by TLR agonists, including inactivated influenza virus, imiquimod, inactivated herpes simplex virus-1 (HSV-1) and CpG ODNs. In addition, MT01 inhibited the CpG ODN-enhanced antibody response and this inhibition could be related to the antagonism of TLR9-activation pathways in B cells. Notably, unlike the G-rich suppressive ODNs reported, MT01 is composed of ACCCCCTCT repeats. These data imply that MT01 represents a novel class of immunosuppressive ODNs that could be candidate biologicals with therapeutic use in TLR activation-associated diseases.
Functioning as pattern recognition receptors (PRRs), toll-like receptors (TLRs), which are broadly distributed on immune cells, initiate activation of the innate immune system in response to conserved microbial molecules, including lipopolysaccharide (LPS), flagellin, unmethylated cytosine–phosphate–guanosine (CpG) DNA and viral RNA.1,2 This results in the production of inflammatory cytokines, the activation of dendritic cells (DCs) and B cells, and the induction of specific, adaptive immunity to the pathogen. In homeostatic conditions, the activation is protective in a host. However, it is not surprising that if stimulation of the innate immune system occurs through dysregulation of the TLR receptor, the activation may be implicated in various genetic diseases and in a number of conditions, including susceptibility to infections, lung diseases, autoimmune diseases such as systemic lupus erythematosus (SLE), inflammatory bowel disease (IBD) and atherosclerosis3–10. Thus, the use of suppressive oligodeoxynucleotide (sODN) or novel small-molecule TLR inhibitors with a larger safety window and differentiated selectivity may potentially have significant clinical utility in those TLR activation-associated diseases (TAADs).
In the past decade, efforts have been made to identify novel TLR antagonists, especially antagonists to TLR9. Chloroquine, hydroxychloroquine and quinacrine, well-known anti-malaria drugs first used in 1940s, have been found to inhibit TLR9 activation by interfering with acidification of the cell endosome and have been used to treat rheumatoid arthritis (RA), SLE and Sjögren’s syndrome.11,12 In recent years, various DNA molecules have been tested for their ability to inhibit immune activation. Initially, a phosphorothioate deoxyguanosine oligomer [S-oligo (dG)20] was found to be capable of inhibiting murine interferon (IFN)-γ production induced by concanavalin A (Con A) and Escherichia coli DNA;13 calf thymus DNA, human placenta DNA and single-base phosphorothioate 30-mer oligodeoxynucleotides (ODNs) were subsequently demonstrated to inhibit interleukin (IL)-12 production from macrophages induced by bacterial DNA, LPS and a stimulatory phosphorothioate ODN.14 Later, suppressive motifs were identified in adenoviral DNA that selectively blocked CpG ODN-induced IL-12 production from mouse splenocytes and IFN-γ production from human peripheral blood mononuclear cells (PBMCs).15 ODNs containing the sequences of repetitive elements in mammalian telomeres were reported to down-regulate CpG ODN-induced IL-6, IL-12, IFN-γ and IL-10 production from mouse splenocytes, possibly by interfering with the maturation of endosomal vesicles and the co-localization of CpG DNA with TLR9 in these vesicles.16 Recently, ODNs synthesized based on the sequence of human microsatellite DNA have been found to inhibit the activation of human PBMCs when induced by stimulation in the presence or absence of TLR9 ligand.17 Various sODNs have been tested in animal models of SLE, RA, pulmonary inflammation, septic shock and multiple sclerosis.18–22 Overall, the reported sODNs were G-rich or rich in polyG and synthesized based on the sequence of either mammalian chromosome DNA or viral genome.
In this study, based on the sequence of human mitochondrial DNA, we designed 13 ODNs, which were designated as mitochondrial DNA-based oligodeoxynucleotides (MTODNs). Among them, we described a candidate sODN (MT01). MT01 was subsequently analyzed in detail in terms of inhibitory potency, kinetics and mode of action. It was found that MT01 was able to down-regulate the production of TLR-dependent type I IFN from, and the proliferation of, cultured human PBMCs and mouse splenocytes that were induced by CpG ODNs, imiquimod, inactivated herpes simplex virus-1 (HSV-1), inactivated mouse influenza virus (FM1) and sera from SLE patients. Moreover, MT01 was examined for its functionality in vivo and was found to have the ability to suppress the antibody response in a hepatitis B surface antigen (HBsAg) + CpG ODN-immunized mouse model. These data imply that MT01 represents a novel class of immunosuppressive ODNs that could be of therapeutic use in TAADs.
Materials and methods
All nuclease-resistant phosphorothioate-modified ODNs used were synthesized by TaKaRa Biotechnology Company (Dalian, China). The following ODNs were used in this study: CpG 2216 (an A-type CpG ODN, 5′-GGgggacgatcgtcGGGGGg-3′), CpG 2006 (a B-type CpG ODN, 5′-TCGTCGTTTTGTCGTTTTGTCGTT-3′), CpG C274 (a C-type CpG ODN, 5′-TCGTCGAACGTTCGAGATGAT-3′), CpG BW006 (a B-type CpG ODN developed by our laboratory, 5′-TCGACGTTCGTCGTTCGTCGTTC-3′) and MS19 (a negative control for sODN, developed in our laboratory, 5′-AAAGAAAGAAAGAAAGAAAGAAAG-3′). Uppercase and lowercase letters represent phosphorothioate and phosphodiester linkages, respectively. The ODNs were stored in phosphate-buffered saline (PBS) and had < 5 endotoxin U/mg of ODN, determined using the Limulus amebocyte lysate assay (Associates of Cape Cod, Inc., East Falmouth, MA).
Cells and cell culture conditions
Human PBMCs were isolated from buffy coats of healthy blood donors (The Blood Center of Jilin Province, Changchun, China) using Ficoll–Hypaque density-gradient centrifugation (Pharmacia, Peapack, NJ). The donors were chosen randomly and had an even gender distribution and an age range of 18–55 years. The viability of the PBMCs was 95–99%, as determined by Trypan Blue exclusion. Six-week-old mice were killed by cervical dislocation following anaesthesia. Their spleens were removed and a single-cell preparation was prepared by gently teasing the spleens against sterile glass slides and removing red blood cells by lysis in Tris-buffered NH4Cl;23 the cells were then resuspended at 5 × 106 cells/ml after being washed twice in Iscove’s modified Dulbecco’s medium (IMDM). Vero E6 cells (African green monkey kidney cell line) and L929 mouse fibroblasts were from the ATCC (American Type Culture Collection, Manassas, VA). The cells were cultured at 37°C in a 5% CO2 humidified incubator and maintained in IMDM supplemented with 10% (v/v) heat-inactivated fetal bovine serum (Gibco, Langley, OK) and antibiotics (100 IU of penicillin/ml and 100 IU of streptomycin/ml).
Vesicular stomatitis virus protection bioassay
Vesicular stomatitis virus (VSV) was grown in Vero E6 cells. After titration, the virus was stored in aliquots at −70°C until use. The antiviral activities in the supernatants of CpG ODN- and/or MTODN-treated human PBMCs were measured as described previously.24 Briefly, the supernatants of human PBMCs (5 × 106 cells/ml) that had been stimulated, for 48 hr, with CpG ODN alone or in the presence of MTODNs, were collected and stored in aliquots at −70°C until use. The Vero E6 cells were seeded onto 96-well flat-bottom plates (3 × 105 cells/ml) and cultured for 24 hr to confluence. The cells were incubated with 100 μl of the diluted supernatants for 24 hr and then challenged with 10 × 50% tissue culture infectious doses (TCID50) of VSV for another 48 hr. After staining with 0·5% crystal violet, the cytopathic effects were examined using a Multi-well Microtiter Plate Reader (Thermo Fisher Scientific Inc. Waltham, MA, USA) at 578 nm and expressed as mean optical density value ± standard deviation (SD). When the supernatant of mouse splenocytes stimulated with CpG ODNs and/or MTODNs was detected, the Vero cells were changed for L929 cells. Imiquimod used in the VSV protection bioassay was purchased from Lvye Chemical Co., Ltd (Yancheng, China). HSV-1 was inactivated for 10 min at 70°C and used at a multiplicity of infection (MOI) of 200. Influenza A virus A/Fort Monmouth/1/1947 H1N1 strain (Flu virus FM1) was inactivated for 30 min at 56°C in a water bath and used at an MOI of 2. Sera were from antinuclear antibody (ANA)-positive and anti-double-strand DNA-positive SLE patients.
Proliferation of human PBMCs or of mouse splenocytes was determined using the [3H]thymidine incorporation assay, as described previously.25 Briefly, human PBMCs or mouse splenocytes (5 × 106 cells/ml) were plated in 96-well U-bottom plates (CoStar, Cambridge, MA) and cultured with CpG ODNs or LPS (Sigma-Aldrich Corp., St Louis, MO) and/or MTODN for 48 hr, followed by pulsing with [3H]thymidine (0·5 μCi/well) (New England Nuclear, Boston, MA) for an additional 16 hr. The cells were harvested on glass fiber filters and detected in a scintillation counter. Cell proliferation was expressed as mean counts per minute (c.p.m.) ± SD.
Immunization of mice with HBsAg
All experiments involving animals were conducted in accordance with the national guidelines for the care and the use of laboratory animals. Vaccines were prepared with physiological saline. Female BALB/c mice, 6–8-weeks of age (Vital River, Beijing, China), were immunized, by injection into the tibialis posterior muscle, with HBV vaccine (Walvax Biotech Co., Ltd., Kunming, Yunnan Province, China) containing 1 μg of recombinant HBsAg and/or 5 μg of CpG BW006 and/or 50 μg of MT01. Blood was drawn from the caudal vein, and serum was separated from the blood by centrifugation of the samples at 1250 × g for 10 min; the serum samples thus obtained were stored at −20°C until required for testing by enzyme-linked immunosorbent assay (ELISA). The pre-immune sera were collected from the mice 2 days before immunization.
Mouse splenocytes were incubated with or without ODNs for 12 hr, washed twice with fluorescence-activated cell sorter (FACS) staining buffer (PBS supplemented with 0·5% bovine serum albumin, 0·01% NaN3 and 100 mm EDTA) and then analyzed for the expression of cell-surface molecules by dual-colour staining with fluorescein isothiocyanate (FITC)-labelled mouse CD19 monoclonal antibodies (mAbs) and phycoerythrin (PE)-labelled mouse CD69 mAbs followed by analysis on a FACSCalibur. For detection of CD80, mouse splenocytes were incubated for 72 hr. The mAbs were purchased from Becton Dickinson (Franklin Lakes, NJ).
Data are shown as means ± SD. Statistical significance of differences was evaluated using the paired two-tailed Student’s t-test. Differences were considered statistically significant for P <0·05.
Inhibitory effect of MTODNs on the production of type I IFN from human PBMCs induced by CpG ODNs
To develop immunosuppressive ODNs with novel sequences as potential candidates for treating immune response-associated diseases, such as autoimmune diseases, a serials of ODNs were designed and screened in our laboratory. Considering the high incidence of autoimmune diseases in women and a tendency for familial inheritance, maternal inherited mitochondrial DNA was an attractive option for investigation. According to the basic sequence feature of suppressive ODNs as reported, we attempted to identify similar sequences in human mitochondrial DNA;26 from these, 13 MTODN sequences were designed. (Table 1). Among the 13 MTODN sequences identified, the repeat units ranged from 2 to 9 nucleotides. The inhibitory effect of each MTODN was tested in a VSV protection bioassay for determining the production of type I IFN from human PBMCs induced by CpG 2216, a typical A-type CpG ODN.25 Freshly isolated human PBMCs from four healthy donors were, respectively, incubated for 48 hr with CpG 2216 (1 μg/ml) alone or in the presence of MTODNs (8 μg/ml). In a parallel experiment, the human PBMCs were incubated with MTODNs alone. The supernatants were collected to determine whether they could protect Vero cells from VSV attack. As shown in Fig. 1a, the supernatant of human PBMCs induced by CpG 2216 could protect Vero E6 cells from VSV challenge. Moreover, MT01 displayed significant inhibition of the production of type I IFN from human PBMCs induced by A-type CpG ODNs, whereas the other MTODNs tested did not. When used alone, none of the MTODNs were able to stimulate human PBMCs to produce IFN (Fig. 1b). MT01 was the only MTODN to possess a C-rich sequence, and, of note, G-rich MTODNs, such as MT02, had no inhibitory effect on the production of type I IFN in this study.
Table 1. The sequences of mitochondrial DNA-based oligodeoxynucleotides (MTODNs)
MTODN sequence (5′→3′)
Dose effect and kinetics of MT01 on the production of type I IFN and the proliferation of human PBMCs induced by CpG ODNs
Upon the first round of screening, MT01 was selected as a sODN candidate. To further investigate the inhibitory role of MT01 on immune responses, we chose three TLR9 agonists – CpG 2216, CpG 2006 (a B-type CpG ODN characterized by stimulating the proliferation of primary human B cells)27 and CpG BW006 (a B-type CpG ODN developed in our laboratory)28– as the stimuli. Human PBMCs were incubated with CpG ODNs (1 μg/ml) alone or in the presence of various concentrations of MT01 (0·25–32 μg/ml). As shown in Fig. 2, MT01 was found to inhibit the CpG ODN-induced production of type I IFN and proliferation of human PBMCs in a dose-dependent manner. MT01 could inhibit CpG 2216-induced type I IFN production (Fig. 2a) in a VSV protection bioassay and CpG 2006-induced (Fig. 2b) or CpG BW006-induced (Fig. 2c) proliferation in a [3H]thymidine incorporation assay, and the concentrations of MT01 at which suppression first occurred were 1, 8 and 1 μg/ml, respectively; when the dose of MT01 increased, the inhibitory effect was enhanced. At 4 μg/ml, the suppressive effect of MT01 reached a plateau when human PBMCs were stimulated with CpG 2216 and CpG BW006. Based on the titrations of MT01 in combination with CpG 2216, CpG 2006 or BW0006, and our unpublished data on the consensus dose responses of human PBMCs from different donors to MT01, 8 μg/ml was selected as the optimal dose of MT01 in latter in vitro experiments. Considering the inhibitory role and possible chemotoxicity of MT01, we selected a negative-control ODN (MS19) that exhibited no obvious suppressive effect on human PBMCs, as described previously.17 As shown in Fig. 2, as the parallel control, MS19 was unable to inhibit CpG ODN-induced IFN production and proliferation of human PBMCs, indicating that MT01 certainly possessed an inhibitory role on immune responses without any toxicity. Additionally, when used alone, MT01 and MS19 had no stimulatory or suppressive effects. Overall, these results showed that MT01 could inhibit IFN production and proliferation induced by TLR9 activation.
To exclude the possibility that the inhibition displayed by MT01 was caused by its extracellular interference with CpG ODNs, the kinetics of the inhibition of MT01 were studied using CpG 2216 as the stimulatory CpG ODN. MT01 was added to human PBMCs (from two healthy donors) at various time-points (0·25, 0·5, 1, 2, 3, 6, 12 or 24 hr) before or after the addition of CpG 2216. As shown in Fig. 3, significant inhibition was observed when MT01 was co-administered with CpG 2216 or added before CpG 2216 (Fig. 3b) (P <0·01). MT01 was noticeably inhibitory, even when added less then 2 hr after the addition of CpG 2216 (P <0·01) and could also inhibit CpG 2216-induced IFN production when added 3 and 6 hr after CpG 2216 (Fig. 3a) (P <0·05). It was demonstrated that CpG 2216-induced production of type I IFN could be inhibited by MT01 6 hr after delivery of the stimulatory signal by CpG 2216, indicating that the inhibition of MT01 on TLR9 activation was consistent and did not occur as a result of extracellular interference with CpG ODNs.
The repeat numbers and some base changes of the motif in MT01 influenced the suppression on TLR9 activation
Structurally, MT01 was composed entirely of three repeats of the motif ‘ACCCCCTCT’. To determine whether the number of motifs contributed to the suppressive effect, ODNs that contained one or two of motifs were synthesized, and named MT01a and MT01b respectively. We used the VSV protection bioassay and proliferation assay to compare their inhibitory effects. CpG 2216 and CpG BW006 were chosen as the stimuli, and the doses of MT01, MT01a and MT01b were 8 μg/ml. MT01a, a 9-mer ODN composed of a single ‘ACCCCCTCT’ motif, failed to inhibit CpG 2216-induced type I IFN production (Fig. 4a) and CpG BW006-induced proliferation of human PBMCs (Fig. 4b). MT01b, composed of two ‘ACCCCCTCT’ motifs, was able to inhibit CpG 2216-induced type I IFN production (Fig. 4a) (P <0·01) and CpG BW006-induced proliferation (Fig. 4b) (P <0·05) of human PBMCs. MT01, composed of three ‘ACCCCCTCT’ motifs, could inhibit both CpG 2216-induced IFN production (Fig. 4a) (P <0·01) and CpG BW006-induced proliferation of human PBMCs (Fig. 4b) (P <0·01). The inhibitory effect of MT01 on CpG BW006-induced proliferation was stronger than that of MT01b (P <0·01). The inhibitory effect of MT01 on CpG ODN-induced type I IFN production and proliferation was maximal, so we did not observe an inhibitory effect of ODNs containing four or more ‘ACCCCCTCT’ motifs.
We then changed some bases in the ‘ACCCCCTCT’ motif to pinpoint the key base(s) in the motif responsible for the inhibitory effect. As seen in Table 2, four 27-mer ODNs, named MT01c, MT01d, MT01e and MT01f, were synthesized to contain three repeats of the ‘ACCCCCTCT’ motif with changes to some of the bases. The changes to the motifs in MT01c, MT01d, MT01e and MT01f were, respectively: ‘T’ to ‘A’ in base seven and ‘T’ to ‘A’ in base nine; ‘C’ to ‘A’ in bases two, three and four; ‘C’ to ‘A’ in bases two, three, four and five; and ‘A’ to ‘C’ in the first base. As shown in Fig. 4c, compared with the production of IFN by human PBMCs induced by CpG 2216, MT01, MT01c and MT01f were able to inhibit the activity (P <0·01) when used together with CpG 2216, whereas the inhibitory effects of MT01d (P <0·05) and MT01e (P <0·05) were weaker than the inhibitory effect of MT01 on CpG 2216-induced IFN production. When used alone, MT01c, MT01d, MT01e and MT01f were unable to stimulate human PBMCs to produce IFN.
Table 2. The sequences of MT01c–f
Effect of MT01 on the production of type I IFN from human PBMCs induced by other TLR agonists
As reported previously, HSV and FM1 were able to activate TLR9, TLR3 and TLR7;29–31 imiquimod activated immune cells to secrete large amount of IFN-α32 by ligating TLR7/8;33 and DNA-containing immune complexes (ICs) in lupus serum stimulated plasmacytoid dendritic cells (pDCs) to produce IFN and other cytokines.34 To evaluate the inhibitory effect of MT01 on IFN production induced by TLR agonists other than CpG ODNs, human PBMCs were incubated for 48 hr with MT01 in the presence or absence of heat-inactivated HSV, heat-inactivated FM1 virus, imiquimod or the serum from SLE patients, respectively. The supernatants were collected and assayed to determine their levels of type I IFN in a VSV protection bioassay. The results showed that MT01 could significantly inhibit the production of type I IFN induced by heat-inactivated HSV (Fig. 5a) (P <0·05), heat-inactivated FM1 (Fig. 5b) (P <0·01), imiquimod (Fig. 5c) (P <0·01) and the sera from a patient with SLE (Fig. 5d) (P <0·01), indicating that the inhibition was also related to the antagonism of TLR7/8, possibly TLR3.
Effect of MT01 on CpG ODN-induced production of type I IFN and proliferation of mouse splenocytes in vitro and on the production of antibodies in mice
It is generally accepted that CpG motif sequence specificity is granted in a species-specific manner. To determine whether MT01 was also species dependent and the feasibility of evaluating its in vivo effect in mice, splenocytes from two BALB/c mice were incubated respectively with CpG 2216, CpG BW006 and/or MT01 for 48 hr and then investigated for their production of type I IFN using a VSV protection bioassay and for proliferation using a [3H]thymidine-incorporation assay. The results from the two experiments and the mean value are shown in Fig. 6. MT01 was able to inhibit CpG 2216-induced type I IFN production from mouse splenocytes (Fig. 6a) (P <0·01) and CpG BW006-induced proliferation of mouse splenocytes (Fig. 6b) (P <0·01). When used alone, MT01 was neither stimulatory nor inhibitory on the IFN production and proliferation of mouse splenocytes. The data indicated that MT01 is less species dependent and able to act as an sODN in mice.
Based on the results of the above experiments, mice immunized with HBsAg or HBsAg + CpG BW006 were used as an animal model to test whether MT01 could inhibit antibody production in vivo. Female BALB/c mice (n =6) were immunized with hepatitis B virus (HBV) vaccine (containing 1 μg of HBsAg/mouse) or with HBV vaccine + CpG BW006 (5 μg/mouse) on Day 1, 29 by injecting the agents into the tibialis posterior muscle, and were injected with MT01 (50 μg/mouse) on days 0, 7, 14, 21, 28, 35, 42 and 49 to investigate whether MT01 could inhibit the production of anti-HBs IgG in mice. The mice were bled on days 2, 13, 20, 27, 34, 41, 48 and 55 to collect sera for detecting anti-HBs IgG by ELISA. The experiment was conducted following a schedule described in Fig. 7a. The pre-immune sera of the mice were used as negative controls. The result (Fig. 7) demonstrated that HBV vaccine elicits detectable anti-HBs IgG after the first immunization and elevated production of anti-HBs IgG on day 41 (2 weeks after the second immunization), and CpG BW006 significantly promoted the production of anti-HBs IgG. MT01, in mice immunized with HBV vaccine + CpG BW006, inhibited the production of anti-HBs IgG, as indicated by the level of antibodies in the sera collected on days 34, 41, 48 and 55 (Fig. 7b) (P <0·01). However, in mice immunized with HBV vaccine alone, MT01 showed no inhibition of antibody production (Fig. 7c), indicating that MT01 inhibited only the CpG ODN-enhanced antibody response.
To determine the cells on which MT01 exerted its inhibition on antibody production induced by antigens, splenocytes from BALB/c mice were stimulated with CpG BW006, a TLR9 agonist, or with LPS, a TLR4 agonist, in the presence of MT01 or MS19 (as a negative-control ODN) for 12 or 72 hr. The CpG BW006-stimulated cells were then double-stained with FITC-labelled anti-CD19 mAb and PE-labelled anti-CD69 mAb or PE-labelled anti-CD80, respectively, followed by analysis on a FACSCalibur. [3H]Thymidine incorporation was used to determine the proliferation of LPS-stimulated cells. The results showed that CpG BW006 obviously up-regulated the expression of CD69 as well as of CD80 on CD19+ B cells, and that MT01 was able to abolish, almost completely, the CD69 up-regulation (Fig. 8a) and mediate inhibition, of approximately 50%, of the up-regulation of CD80 (Fig. 8b). Notably, MS19 inhibited the up-regulation of CD80 (although more weakly compared with MT01) and did not inhibit the up-regulation of CD69. By contrast, neither MT01 nor MS19 were able to inhibit LPS-induced proliferation of the cells (Fig. 9). These results demonstrate that the inhibition of MT01 on antibody production could be related to the antagonism of the TLR9, but not the TLR4, activation pathway in B cells.
In this study we demonstrated that MT01 could negatively regulate immune activation. MT01, a 27-mer ODN, was synthesized with reference to a motif with a sequence of 5′-ACCCCCTCTACCCCCTCT-3′ in human mitochondrial DNA. Characteristically, the sequence consists of a 9-mer motif with the sequence 5′-ACCCCCTCT-3′. Based on this sequence, ODNs with one motif (MT01a), two tandem motif repeats (MT01b) and three tandem motif repeats (MT01) were synthesized and tested for their ability to inhibit the production of type I IFN and to stimulate the proliferation of human PBMCs. MT01a showed the weakest activity, MT01b manifested significant inhibition and MT01 showed maximum activity. Based on these data we propose that ODNs containing repeat sequences in the genomes of mammals, such as humans, could display inhibitory activities on immune cells. This speculation is supported by accumulating evidence that repeat elements in mammalian telomeres16 and human microsatellite DNA mimicking ODNs17 have a negative regulatory effect on the innate immune response triggered by TLR activation. The phenomenon could have biological significance, especially in pathophysiological settings. Specifically, in mammalian cells, each mitochondrion generally contains several identical copies of a conserved mitochondrial DNA. When cells undergo excessive necrosis or apoptosis during an over-reactive immune response, abundant mitochondrial DNA or its fragments could be released as a type of endogenous danger signal that may warn the immune cells to decrease their activities, thus avoiding undesirable injuries to the immune system. Broadly, mammalian telomere DNA16 and human microsatellite DNA,17 when released in excessive quantities, could also display immune-inhibitory activities similar to those of mitochondrial DNA.
Structurally, there are six ‘C’ bases in each of the three 9-mer motifs in MT01 (5′-ACCCCCTCTACCCCCTC-TACCCCCTCT-3′); this is quite different from the sequences of the suppressive ODNs reported by other groups.16 The host telomere-derived molecules that down-regulate the innate immune response contain a G-rich motif, such as TTAGGG. Adenovirus-derived DNA, which blocks the induction of bacterial DNA by cytokines, is a GC-rich molecule.15 In addition, a 20-mer ODN that inhibits IFN-γ production from mouse splenocytes is a poly G oligomer.35 The MT01 is a C-rich ODN and contains five successive ‘C’ bases in each of its 9-mer motifs. Seemingly, five successive ‘C’ bases are required for the inhibitory activities of MT01. The substitution of three successive ‘C’ bases with three successive ‘A’ bases significantly reduced the MT01-mediated inhibition of IFN production. If the first four successive ‘C’ bases were substituted with four successive ‘A’ bases, the inhibitory activity of the altered MT01 was reduced more significantly. The five successive ‘C’ bases in the context of MT01 seem optimal, which is supported by evidence that the substitution of ‘A’ bases in the motifs with an extra ‘C’ base cannot make the altered MT01 more inhibitory.
The results of this study demonstrate that MT01 was able to inhibit the production of type I IFN from human PBMCs induced by CpG 2216, heat-inactivated HSV-1, imiquimod, heat-inactivated FM1 virus and sera from SLE patients. CpG 2216, a TLR9 agonist, induces high amounts of IFN-α and IFN-γ in PBMCs.25 Acting as a TLR7/8 agonist, imiquimod stimulates the production of a large number of cytokines, including IFN-α or IFN-β.32,33 It is well established that natural IFN-producing cells specialize in the production of high levels of type I IFN in response to DNA and RNA viruses. HSV-1, as a TLR9 agonist29,36 and influenza virus (FM1),31 as a TLR3 and TLR7 agonist, are capable of inducing large amounts of type I IFN. Mammalian DNA and RNA in the serum of lupus patients, in the form of immune complexes, stimulate pDCs to produce IFNs and ILs via the activation of TLR9 and TLR7, respectively.34,37 The present data imply that MT01, acting as a TLR antagonist, may be developed as a therapeutic agent for the treatment of diseases associated with the over-production of IFNs. Ample evidence strongly suggests a link between IFNs and disease pathogenesis. IFN has been found to be involved in the pathogenesis of psoriasis. The mRNAs of type I IFNs and type I IFN-inducible genes and proteins were all consistently and significantly overexpressed in the lesional skin of psoriatic patients.38 The treatment of non-psoriatic disease with IFN or with imiquimod, a TLR7/8 agonist that induces the production of type I IFNs, can induce or exacerbate psoriasis.39–41 The data suggest that type I IFNs may be potential therapeutic targets in psoriasis treatment. The inappropriate activation of type I interferon also has a role in SLE pathogenesis.9 Serum levels of IFN-α are elevated in SLE patients, and gene-expression profiling of peripheral blood cells shows that most lupus cases demonstrate an up-regulation of IFN-responsive genes.42 In SLE patients, IFN production by TLR7/9 activation is caused, at least in part, by immune complexes formed by auto-antibodies with nucleoprotein particles released from dead and dying cells.43–45 Therapeutic administration of recombinant IFN-α was associated with the induction of SLE-like symptoms in patients.46,47 In addition, the pathogenesis of type I diabetes, RA, myositis and Sjögren’s syndrome are also involved in the over-production of IFN.48–51 The involvement of IFN suggests that blockade of IFN production of MT01 by inhibiting TLRs could be used as a promising approach to treat psoriasis, SLE and other IFN-mediated autoimmune diseases.
Notably, our unpublished data indicated that MT01 could also inhibit the production of TNF-α induced by TLR9 agonists. It has been documented that over-production of TNF-α contributes to the development of RA and other autoimmune diseases.52 To target TNF-α, several biological compounds including infliximab (a mouse–human chimeric mAb against TNF-α), etanercept (a TNF receptor–IgG fusion protein) and adalimumab (a fully human mAb) are being used for the treatment of RA. Thus, it might be possible to develop MT01 as a candidate biological therapy to prevent the over-production of TNF-α.
In this study, using mice immunized with HBV vaccine, either alone or with CpG BW006, we observed the in vivo effect of MT01 on the antibody response induced by antigens. The results showed that MT01 could inhibit anti-HBs IgG production in mice immunized with HBV vaccine + CpG ODN, but not in mice immunized with HBV vaccine alone, indicating that MT01 was able to inhibit the CpG ODN-enhanced antibody response induced by the antigens. The data imply that MT01 could be used as an agent to inhibit the overproduction of antibodies induced by TLR9 activation. TLR9, present in pDCs and B cells in humans, is a receptor for microbial CpG DNA,53 and contributes to innate immunity. Recent studies have shown that some abnormality of TLR9 activation is involved in the development of autoimmune diseases related to an increased number of activated B cells and auto-antibody production. The involvement is confirmed by the evidence that anti-DNA auto-antibody production is impaired in TLR9 gene-knockout lupus-prone mice54 and that the level of TLR9 expression on peripheral blood B cells from patients with active SLE is correlated with its clinical parameters and pathogenic auto-antibody production.55 The results demonstrated that MT01 could inhibit CpG ODN-enhanced humoral immunity. The accumulated data led to a novel hypothesis regarding therapeutic approaches that might interfere with the development and progression of SLE.56 Likewise, we can extrapolate that MT01 may be used to treat SLE and other autoimmune diseases associated with the overproduction of auto-antibodies.
We gratefully acknowledge the Blood Center of Jilin Province for providing human buffy coats. The CAL-1 cell line was kindly provided by Dr Takahiro Maeda, Nagasaki University Graduate School of Biomedical Science, Japan. This study is supported by National Nature Scientific Foundation of China (30972671) and CHANGCHUN HUAPU BIO-TECH Co., Ltd., China.
The authors declare having no financial or commercial conflicts of interest.