Type 1 diabetes is caused by adaptive immune responses, but innate immunity is important because monocytes infiltrate islets. Activated monocytes express cyclooxygenase (COX)-2, promoting prostaglandin-E2 (PGE2) secretion, whereas COX-1 expression is constitutive. We aimed to define monocyte COX expression in type 1 diabetes basally and after lipopolysaccharide (LPS) stimulation. Isolated CD14+ monocytes were analyzed for COX mRNA and protein expression from identical twins (discordant for type 1 diabetes) and control subjects. Basal monocyte COX mRNA, protein expression, and PGE2 secretion were normal in type 1 diabetic subjects. After LPS, twins and control subjects showed a COX mRNA isoform switch with decreased COX-1 mRNA (P < 0.01), increased COX-2 mRNA (P < 0.01), and increased COX-2 protein expression (P < 0.01). Compared with control subjects, both diabetic and nondiabetic twins showed greater LPS-induced downregulation of monocyte COX-1 mRNA (P = 0.02), reduced upregulation of COX-2 mRNA and protein (P < 0.03), and greater inhibition by the COX-2 inhibitor di-isopropylfluorophosphate (DFP) of monocyte PGE2 (P < 0.007). We demonstrate an alteration in monocyte COX mRNA expression as well as monocyte COX-2 and PGE2 production after LPS in type 1 diabetic patients and their nondiabetic twins. Because COX-2 response to LPS is proinflammatory, an inherited reduced response would predispose to chronic inflammatory diseases such as type 1 diabetes.

Type 1 diabetes is induced by environmental events affecting genetically susceptible individuals and resulting in the destruction of the insulin-secreting cells in the pancreatic islets of Langerhans. Consequently, even identical twins often remain discordant for the disease (1). This destructive immune process involves both the innate and adaptive immune response because monocytes, macrophages, and T-cells infiltrate the islets at onset of type 1 diabetes (2). The evidence is that adaptive immune effectors (including T- and B-cells) are important because circulating insulin autoantibodies, GAD, and islet cell–associated antigen, can predict type 1 diabetes (3). Recent studies suggest that innate effector cells, including natural killer T-cells and monocytes, also play a role in humans as well as in nonobese diabetic (NOD) mice, an animal model of autoimmune diabetes (49).

Activation of monocytes and macrophages induce the translocation of the transcription factor nuclear factor-κB, leading to expression of immune response genes with secretion of prostaglandins (PGs), nitric oxide, cytokines, and chemokines, which affect cellular immune responses (1011). Pathways contributing to PG output in antigen-presenting cells, such as macrophages and monocytes, are regulated by the enzyme cyclooxygenase (COX; PGG/H synthase). COX, a key enzyme in PG synthesis, has at least two isoforms (COX-1 and -2) encoded by separate genes and catalyzes a rate-limiting step in the conversion of arachidonic acid to PGH (1213). Whereas COX-1 is believed to be constitutively expressed, COX-2 expression is induced in response to cytokines, growth factors, tumor promoters, lipopolysaccharide (LPS), and bacterial endotoxin, resulting in increased proinflammatory PGs (e.g., PGE2) (1416). Aberrant expression of COX-2 in nonactivated monocytes has been reported in type 1 diabetes and subjects at increased risk of developing this disease, and the change was not dependent on disease duration, suggesting that it might be genetically determined (6). We addressed this possibility by studying COX-1 and -2 monocyte expression in patients with type 1 diabetes together with their nondiabetic identical cotwins, who were selected because they were unlikely to develop diabetes themselves and thus provided a genetically identical control subject.

Identical twin pairs were selected from the British Diabetic Twin Study (1). Twins from the registry are ascertained by referral through their physicians. Of 451 twin pairs, we selected 16 identical pairs discordant for type 1 diabetes (mean age 38 years [range 18–69], 8 male pairs) eligible according to the following criteria: 1) European origin, 2) affected twins had type 1 diabetes, 3) both twins of each pair were available for study, 4) neither twin was receiving drugs other than human insulin in the index case, and 5) the nondiabetic twin had a low disease risk, i.e., a risk <2% based on lack of diabetes-associated antibodies (1,17). Type 1 diabetes was defined according to standard criteria, and diabetes was excluded by glucose tolerance tests and random whole-blood glucose values (YSI, Yellow Springs, OH) <7.0 mmol/l (18). All diabetic twins were treated from the time of diagnosis with insulin and were taking highly purified human insulin at least twice daily; the mean duration of diabetes (mean ± SD) was 17 ± 11 years in the diabetic twin. Monozygosity was established in twin pairs using both clinical data and at least 22 blood groups, as previously described (1). Control subjects (n = 27, mean age 40 years [17–65], 13 male) were obtained from the local population; these subjects had no family history of diabetes and at the time of testing had no illness, were taking no drugs, had no clinical signs or symptoms of illness, and were on a normal diet. All of the subjects gave informed consent, and the study was approved by the ethics committee of the Royal Hospital Trusts.

We obtained endotoxin-free Ficoll-Hypaque from Amersham Pharmacia (Bucks, U.K.); PBS, RPMI 1640, Dulbecco’s modified Eagle’s (DMEM), l-glutamin, penicillin/streptozotocin, and fetal bovine serum (FBS) from Life Technologies (Paisley, U.K.); CD14+ magnetic beads for a positive selection column from Miltenyi Biotec (Surrey, U.K.); ELISA kits for detection of PGE2 from R&D Systems (Oxfordshire, U.K.); human COX-1 and -2 antibodies for Western detection from Cayman Chemical (Boldon, U.K.); LPS from Sigma (Dorset, U.K.); bicinchoninic acid kit (Pierce, U.K.); nitro-blue tetrazolium chloride (NBT)/5-bromo-4-chloro-3′-indolyphosphate p-toluidine salt (BCIP) solutions from Bio-Rad (Hemel Hempstead, U.K.); a RNeasy mini kit from Qiagen (Crawley, U.K.); a RiboGreen RNA quantitation kit from Invitrogen (Paisley, U.K.); human and mouse COX-1/2 amplicon, COX-1/2 primers (R and F), and probes for real-time quantitative RT-PCR from MWG-Biotech (Ebersberg, Germany); the COX-2–specific inhibitor di-isopropylfluorophosphate (DFP; a kind gift from Prof. T. Warner); anti–COX-2 fluorescein isothiocyanate (FITC) and COX-2 blocking peptide from Cayman Chemical (Boldon, U.K.); and mouse IgG1 isotype control and anti-CD14 phycoerythrin from SeroTec (Oxford, U.K.).

Human lymphocyte separation.

Peripheral blood mononuclear cells (PBMCs) from human subjects (twins and control subjects) were prepared from heparinized blood using standard Ficoll-Hypaque separation.

CD14+ monocytes isolation and magnetic separation.

Isolated PBMCs from human blood were washed in RPMI 1640 and then resuspended in complete medium (RPMI containing 10% FBS, 100 μg/ml penicillin/streptozotocin, and 2 mmol/l l-glutamine). PBMCs were washed with 15 ml of magnetic cell sorting buffer (PBS, 2 mmol/l EDTA, and 5% FBS), and CD14+ cells were separated according to the manufacturer’s instructions using a positive selection column. For COX mRNA and protein expression, CD14+ monocytes were plated onto 48-well plates (at 5 × 105 per well). Human cells were then cultured as appropriate with either medium alone or medium supplemented with 1 μg/ml LPS in the presence or absence of the COX-2–selective inhibitor DFP and then incubated at 37°C in 5% CO2 for 18–24 h.

Detection of monocyte intracellular COX-2 by fluorescence-activated cell sorting.

Briefly, isolated human PBMCs (3 × 105 cells) were immediately transferred to azide-containing buffer (PBS, 0.5% BSA, and 2 mmol/l EDTA) for immunostaining. For stimulation, cells were resuspended in RPMI 1640 plus 5% autologous serum and cultured in polypropylene tubes in 1.5-ml aliquots of 1 × 106 cells/ml. Cultured aliquots were stimulated for 18 h with LPS (0.1, 10, and 1,000 ng/ml) or left in medium alone.

Unstimulated or stimulated cells were washed in fluorescence-activated cell sorting (FACS) buffer (PBS, 1% FBS [0.2-μm filtered], 5 mmol/l EDTA, and 0.1% sodium azide) and surface stained with phycoerythrin-conjugated mouse anti-human CD14 for 20 min on ice. Thereafter, cells were washed twice in labeling buffer and fixed for 10 min on ice with 3% (wt/vol) paraformaldehyde or formaldehyde (0.2-μm filtered) in PBS. Fixed cells were washed twice with saponin buffer (FACS buffer plus 0.2% [wt/vol] saponin) and incubated with FITC-conjugated mouse anti-human COX-2 (0.5 μg/sample) for 30 min on ice.

Control treatments were carried out on parallel samples as follows: the ligand blocking control for specificity was anti-human COX-2 FITC (0.5 μg/sample) preincubated with 10 μg/ml blocking peptide for 1 h at room temperature and then added to the sample. The isotype control was mouse IgG1 FITC (0.5 μg/sample). Samples were acquired on a cytometer (LSR; Becton Dickinson), and data were analyzed using WinMDI software (Joseph Trotter, Scripps Institute, La Jolla, CA). The intensity of COX-2 expression was calculated after gating on CD14+ monocytes and by subtracting isotype control as background.

Detection of monocyte COX-1 and -2 protein expression by Western blotting.

After overnight stimulation with and without LPS, CD14+ monocytes (5 × 105 cells per well) were washed with cold PBS and lysed with 100 μl of lysis buffer (1% Triton X-100, 50 mmol/l HEPES, 10% glycerol, 2 mmol/l EDTA, 10 mmol/l sodium fluoride, and 10 mmol/l sodium pyrophosphate) supplemented with 2 mmol/l Na3VO4, 1 mmol/l phenylmethylsulfonyl fluoride, 10 μg/ml leupeptin, and 10 μg/ml aprotonin. Total protein concentration was determined using a bicinchoninic acid protein assay kit. Proteins were loaded in the range of 25–30 μg protein per lane, and cell lysates were separated by 10% SDS-PAGE and protein transferred onto nitrocellulose membrane. After blocking nonspecific binding, blots were incubated with COX-1 and -2 antibodies at 1:1,000 and then developed using NBT/BCIP solution. To semiquantify the band, a Spot Denso AlphaImager was used to estimate protein bands (in arbitrary units) and the mean COX-2 fold increase after LPS stimulation (19). We focused on COX-2 protein expression and therefore had only six cell lysates available from both the twins and control subjects to analyze COX-1 protein levels; there were insufficient numbers to perform a statistical analysis.

Detection of monocyte COX-1 and -2 mRNA expression by quantitative RT-PCR.

After a 3-h LPS stimulation of isolated CD14+ monocytes, RNA was extracted using an RNeasy mini kit (Qiagen). RNA was quantified in triplicate using RiboGreen quantification kits (Molecular Probes, Leiden, the Netherlands) with minor amendments to the original protocol (20).

Sequence-specific primers and probes.

Intron-spanning primers and probes specific for the human COX-1 and -2 genes were designed using Primer Express (ABI, Warrington, Cheshire, U.K.) (Table 1). Human and mouse COX-1 and -2 probes had a FAM reporter dye at the 5′ end and a quencher dye (6-carboxy-tetramethyl-rhodamine, TAMRA) on the 3′ end. The levels of human and mouse COX-1 and -2 mRNA were quantitated relative to amplicon-specific standard curves by real-time RT-PCR using 50 ng total RNA in duplicate and analyzed on an ABI Prism 7700 sequence detector. Standard curves were generated by serial dilution of single-stranded sense oligonucleotides specifying COX-1 or -2 amplicons (Table 2), as previously described (21). All serial dilutions were carried out in duplicate. Standard curves in duplicate were repeated three times, with duplicate no-template controls included with every RT-PCR run. mRNA levels were expressed as COX-1 or -2 mRNA copy numbers per microgram total RNA.

The RT-PCR was carried out in a 25-μl reaction mixture containing 1 × TaqMan EZ buffer, 3 mmol/l Mn(OAc)2, 300 mmol/l dA/dC/dG/dUTP, 2.5 units rTth (recombinant Thermus thermophilus) DNA polymerase, 10 pmol primers (forward and reverse primer sets) (Table 1), 5 pmol TaqMan probe, 0.5 units AmpErase UNG (uracil-N-glycosylase), and 100 ng total RNA at 50°C for 2 min, 60°C for 30 min, and 92°C for 5 min followed by 40 cycles of 20 s at 92°C and 1 min at 62°C. All samples were run in duplicate with water as a no-template control and a positive control containing RNA specifying both COX-1 and -2 RNA levels from the human colon cancer cell line HT-29

Functional activity of COX-2 expression.

Conversion of PGH2 to PGE2 was used to assess the functional activity of COX-2 expression. The accumulated levels of PGE2 from culture supernatants of CD14+ monocytes pre- and poststimulation were measured by a competitive enzyme immunoassay (Amersham Pharmacia). All samples were batched, and PGE2 assays were performed blind with internal controls in duplicate. The limit of detection for PGE2 assay was 36.2 pg/ml. PGE2 concentration in the supernatants was standardized to picograms per milliliter per 5 × 105 cells for reporting.

Murine macrophage cell line J774.

The murine monocyte/macrophage cell line (from American Type Culture Collection) was grown in Dulbecco’s modified Eagle’s medium supplemented with 100 units of penicillin/streptomycin, 2 mmol/l l-glutamine, and 10% FBS (Invitrogen). Confluent cells were cultured in 1 μg/ml LPS overnight and lysed for RNA extraction. Protein lysate was also prepared for Western blotting, and culture supernatants were analyzed for PGE2 levels as a marker of COX activity.

Statistical analysis.

Data were assembled using Microsoft Excel software and are the means ± SD. Data were analyzed using the Prism statistical package (version 3; GraphPad Prism). All statistics were performed with either Student’s t test (two tailed unless otherwise stated) or Mann-Whitney U test (two tailed unless stated otherwise), whichever was appropriate. Differences between variables were considered significant at P < 0.05. The LPS dose-response relationship to COX-2 protein expression was assessed by flow cytometry as the mean fluorescent intensity of COX-2–positive CD14+ monocytes (Fig. 1). For each subject, the curve-fitting precision was estimated by R2, the coefficient of determination. We anticipated that normal subjects would maintain a perfect logarithmic dose-response relationship between LPS dose and COX-2 expression such that R2 would approximate to 1.0. Thus, loss of relationship of LPS dose to COX-2 expression results in a decreased coefficient of determination (22). Basal COX-1 mRNA expression in CD14+ monocyte showed marked variation between individuals; 5% had undetectable levels, so we analyzed COX-1 mRNA expression only in cells with detectable basal levels.

Basal monocyte COX-1 and -2 expression is normal in type 1 diabetes.

Basal monocyte COX-1 mRNA expression was not consistently detected by quantitative RT-PCR, and when it was detected, levels did not differ between diabetic and nondiabetic twins or control subjects (Fig. 1A). COX-1 protein was detected in basal samples in which sufficient cell lysates were available. Diabetic and nondiabetic twins and control subjects did not differ in terms of basal COX-2 mRNA expression (Fig. 1B) and basal monocyte COX-2 protein (by FACS and Western blotting) (Fig. 1C and D, respectively), as well as PGE2 production (Table 3), a marker of COX-2 functional activity.

Downregulation of monocyte COX-1 mRNA after LPS stimulation.

When CD14+ monocytes were stimulated with LPS, those subjects with detectable basal monocyte COX-1 mRNA showed a significant downregulation of COX-1 mRNA expression in all three groups (diabetic twins: P = 0.003, n = 12; nondiabetic twins: P = 0.0009, n = 10; and control subjects: P = 0.003, n = 13) (Fig. 1A). We extended this observation by studying COX-1 mRNA expression after LPS stimulation in normal human subjects using whole PBMCs (n = 20) (data not shown). COX-1 mRNA expression level was also significantly reduced after LPS stimulation in PBMCs (P < 0.007) (data not shown). COX-1 protein was present after LPS stimulation with no clear pattern of response.

Upregulation of monocyte COX-2 mRNA and protein expression after LPS stimulation.

When CD14+ monocytes were stimulated with LPS overnight, monocyte COX-2 expression—whether determined by FACS, Western blotting, or quantitative RT-PCR—increased, as anticipated, in all three groups, i.e., in diabetic and nondiabetic twins as well as control subjects (paired one-tailed P < 0.003 for each) (Fig. 1). Concomitantly, PGE2 secretion of these cells also increased significantly with LPS stimulation (P < 0.0001 in all groups) (data not shown). Time course experiments confirmed that COX-2 mRNA, COX-2 protein, and PGE2 production were significantly increased by LPS at the selected time of study (data not shown).

Monocyte COX-1 and -2 response to LPS is altered in type 1 diabetic patients and their twins.

The magnitude of the COX-2 response to LPS was significantly reduced in both diabetic and nondiabetic twins compared with control subjects, as assessed by FACS and Western blotting (P < 0.015 for both) (Fig. 1C and D). Dose-response studies using FACS indicated that the response to LPS in control subjects increased as little as 0.1 ng/ml with a stepwise increase in responses above the basal level at a dose range of 10–1,000 ng/ml (Fig. 1C) and a good dose-response relationship (R2) approximating to 1.0 (R2 = 0.89 ± 0.08). The FACS profiles confirmed that the intracellular COX-2 protein response to LPS was significantly reduced at the higher LPS dose in both diabetic and nondiabetic twins (Fig. 1C). In line with a familial, possibly genetic alteration causing these changes, there was no difference between the diabetic and nondiabetic identical twins in their COX-2 response to LPS by FACS analysis, and that response was correlated between twins of each pair for COX-2 (r = 0.6, P < 0.01) and for PGE2 secretion (r = 0.62, P = 0.02) (data not shown). However, the dose-response relationship using FACS reflected by R2 did not differ between twins (diabetic R2 = 0.90 ± 0.07, nondiabetic R2 = 0.86 ± 0.09) and control subjects, indicating a similar sensitivity to LPS.

The observed downregulation of COX-1 mRNA expression after LPS stimulation was significantly greater in both diabetic and nondiabetic twins compared with control subjects (P = 0.02 for both) (Fig. 1A). Moreover, consistent with the altered responses not being caused by metabolic changes, there was no difference between diabetic and nondiabetic identical twins in both the decrease in COX-1 and the increase in COX-2 mRNA levels after stimulation with LPS.

Monocyte functional response to LPS is altered in type 1 diabetes.

We investigated whether the response to LPS was functional by determining the PGE2 production by CD14+ monocytes in response to overnight exposure to LPS. PGE2 secretion rose substantially after LPS in diabetic and nondiabetic twins as well as in control subjects (P < 0.0001) with no difference in the level of PGE2 secretion between the groups (data not shown). The production of PGE2 by monocytes after LPS was markedly reduced, as expected, by DFP, a specific COX-2 inhibitor (P < 0.0005 for each group) (data not shown). Consistent with a COX enzyme activity defect in the twins’ monocyte responses to LPS, there was a greater inhibition by DFP of monocyte PGE2 production after LPS in both diabetic and nondiabetic twins compared with control subjects (P < 0.007 and P < 0.002, respectively) (Fig. 1E). Levels of PGE2 after LPS were strongly correlated between diabetic and nondiabetic twins of each pair (r = 0.70, P = 0.03) and remained correlated after DFP induced inhibition (r = 0.61, P < 0.02). These results are consistent with a familial alteration in monocyte function in twins genetically susceptible to type 1 diabetes.

Murine macrophages show a COX mRNA isoform switch after LPS stimulation.

The observed COX mRNA isoform switch in human peripheral blood monocytes after LPS stimulation was confirmed in the established murine macrophage cell line J774. LPS stimulation of murine macrophage cell line J774 resulted in a significant increase in COX-2 mRNA with a concomitant decrease in COX-1 mRNA levels (P < 0.02 and P < 0.01, respectively) (Fig. 2A). Moreover, the observed induction of COX-2 mRNA (Fig. 2A) and protein levels in these cells was confirmed by an increase in PGE2 production after LPS stimulation (Fig. 2B), consistent with COX expression being functional in J774 cells (Fig. 2B).

The observations presented in this report demonstrate for the first time that there is a COX mRNA isoform switch in monocytes stimulated with the nonspecific antigen LPS, resulting in decreased COX-1 and increased COX-2 mRNA expression levels; this COX mRNA switch was altered in type 1 diabetes. Furthermore, monocytes from diabetic twins compared with control subjects after LPS showed decreased COX-2 protein expression levels and a greater inhibition of LPS-induced PGE2 secretion by a COX-2–specific inhibitor. We demonstrated that these alterations in monocyte COX mRNA, COX-2 protein, and PGE2 secretion to LPS in type 1 diabetes is familial, and most likely inherited, because they can also be detected in nondiabetic twins genetically at risk for type 1 diabetes but selected to be unlikely to develop diabetes. Because COX-2 response to LPS is proinflammatory, the reduced response we observed would predispose to chronic inflammatory disease, such as type 1 diabetes.

Basal monocyte COX-1 and -2 expression levels were normal in our type 1 diabetes patients and their nondiabetic identical twins using three techniques (Western blotting, quantitative RT-PCR, and FACS). In addition, we did not detect any difference in basal monocyte COX-2 functional activity as determined by PGE2 secretion. In contrast, Litherland et al. (6) found an increase in monocyte basal COX-2 expression using FACS analysis alone. Several differences between their study and that described here might account for this apparent discrepancy. Importantly, the subjects used in the two studies differed because we did not study individuals at high risk of developing diabetes, in contrast to Litherland et al. (6), because we were interested in genetically determined changes. It remains possible that increased basal monocyte COX-2 expression is a feature of the pre-diabetic phase of the disease studied by Litherland et al. (6) reverting to normal levels several years after diagnosis.

After LPS stimulation in diabetic and nondiabetic twins, as well as in control subjects, COX-1 mRNA levels decreased, whereas COX-2 mRNA levels increased. This COX mRNA isoform switch after LPS stimulation was confirmed in human PBMCs as well as in the murine macrophage cell line J774. A COX mRNA isoform switch has never been quantitatively demonstrated before, although previous studies have documented such a switch without using quantitative methods (23). Accumulating evidence suggests that COX-1 expression can be modulated and expressed differentially by different cells under various conditions, including nerve cells and human brain injury, phorbol ester stimulation of THP-1 cells, phorbol ester (TPA [12-O-tetradecanoylphorbol-13-acetate]) stimulation of human megakaryoblastic-like cells, and tobacco carcinogen treatment of a human macrophage cell line (2325). Another system in which LPS has been shown to differentially regulate constitutive and inducible transcription factor activity in rats includes the downregulation of DNA-binding activities of the constitutive transcription factors SP-1 and AP-2 (26). These observations, taken together with our own, indicate that COX-1 mRNA can be regulated in both monocytes and macrophages and that after stimulation by the nonspecific antigen LPS, there is an isoform switch with COX-1 mRNA decreasing as COX-2 mRNA increases.

We found that this monocyte COX mRNA isoform switch is altered in type 1 diabetes. The diabetic twins showed a greater decrease in COX-1 mRNA than control subjects, whereas the inducible COX-2 mRNA response was lower than in control subjects. Moreover, the same changes were found in the nondiabetic twins, who were selected to be at low disease risk, so we can assume that the altered COX expression does not presage disease, nor does it result from the metabolic disturbance of diabetes, consistent with the altered isoform switch being genetically determined. In support of a familial effect determining this reduced monocyte response to LPS, both the COX-2 response and the dose-response relationship after LPS stimulation between twins were strongly correlated. As expected, LPS stimulation promoted monocyte production of PGE2, and we showed that this response was functional in that it could be inhibited by the COX-2 inhibitor DFP. Consistent with a functional defect in monocytes from both diabetic and nondiabetic twins, both groups showed a greater inhibition of LPS-induced PGE2 secretion compared with control subjects, and, in line with a functional genetic defect, this PGE2 response was strongly correlated between twins of each pair.

LPS is a microbial product that acts by stimulating receptors on the monocyte cell surface, including Toll-like receptors. However, our results cannot be taken as implicating either microbial products or abnormalities in Toll-like receptors in the etiology of type 1 diabetes because it is possible that monocyte stimulants other than LPS might have the same effect. Further studies are required to define the nature of the proposed abnormality and confirm that the abnormality is indeed genetic. In a recent study of monocyte-derived macrophages from a limited number of HLA-heterozygous type 1 diabetic patients, the secretion of PGE2 and cytokines in response to LPS was increased compared with HLA-matched relatives (27). Those authors did not study the macrophage response to LPS after COX-2 inhibitors, as we have done. Nevertheless, as in our current study, the changes noted in macrophage cell responses after LPS were consistent with an abnormality in these innate effector cells in type 1 diabetes (27). Because the COX-2 response to LPS is essentially proinflammatory, the reduced response we detected would predispose to inflammation and might thereby predispose to chronic inflammatory diseases such as type 1 diabetes.

Nongenetic, probably environmental, factors play a major role in causing type 1 diabetes and operate through the induction of an autoimmune response (28). The detailed mechanism resulting in the activation of that immune response remains unclear. Although the induction of autoimmunity probably involves the adaptive immune system, innate effector cells are important in priming or promoting these responses. These innate effectors include a few relatively inflexible cell populations such as monocytes/macrophages, dendritic cells, natural killer cells, natural killer T-cells, and γδ T-cells (29). Alteration in these innate effectors could predispose to type 1 diabetes: for example, the frequency of Vα24JαQ T-cells was higher in nondiabetic twins than in their diabetic identical twins or triplets, and the latter show a dramatic reduction in interleukin-4 secretion capacity compared with their nondiabetic twins (5).

In conclusion, our data indicates that there is a familial, most likely inherited, functional abnormality in monocyte COX responses to LPS in type 1 diabetic patients, which may predispose to the disease. These observations confirm and extend evidence that monocyte function is altered in type 1 diabetes (30). If these monocyte changes were to predispose to type 1 diabetes, then modulation of the innate immune system could be of therapeutic value in preventing it.

FIG. 1.

Human COX-1 and -2 mRNA, protein expression, and functional activity. Abnormal COX isoform show a switch in identical twins discordant for type 1 diabetes. Results are shown for 27 control subjects and 17 diabetic monozygotic twins and their 17 nondiabetic twins. A: Monocyte COX-1 mRNA expression using quantitative RT-PCR. Data are basal means ± SE. COX-1 mRNA expression was not significantly different between the three groups. Post-LPS, both sets of twins as well as control subjects showed a decrease in monocyte COX-1 mRNA expression (P < 0.01 for all groups). B: Monocyte COX-2 mRNA expression using quantitative RT-PCR. Data are basal means ± SE. COX-2 mRNA expression was not significantly different between groups. Post-LPS, both sets of twins and control subjects showed an increase in monocyte COX-2 mRNA expression (P < 0.01 for all groups). C: LPS dose-response for intracellular monocyte COX-2 protein expression by FACS analysis. Data are the fold increase above basal levels at each LPS dose. Mean response in control subjects increased with stepwise increases in LPS dose (P < 0.05 for all comparisons). At each LPS dose, COX-2 response in both sets of twins was lower compared with control subjects at the same dose. *P < 0.04; **P < 0.01; ***P < 0.003. D: Monocyte COX-2 protein expression by Western blotting. Data are the fold increase above basal levels post-LPS. COX-2 protein expression was significantly greater post-LPS in control subjects compared with diabetic and nondiabetic twins. E: Monocyte PGE2 secretion. Monocyte PGE2 secretion post-LPS stimulation in the presence of the COX-2–specific inhibitor DFP. Data shows DFP inhibition of PGE2 production post-LPS. There was no difference in basal or LPS-induced PGE2 production between the groups, but DFP inhibition of PGE2 production post-LPS was significantly greater in both diabetic and nondiabetic twins compared with control subjects. ▪, control subjects; , diabetic twins; , nondiabetic twins.

FIG. 1.

Human COX-1 and -2 mRNA, protein expression, and functional activity. Abnormal COX isoform show a switch in identical twins discordant for type 1 diabetes. Results are shown for 27 control subjects and 17 diabetic monozygotic twins and their 17 nondiabetic twins. A: Monocyte COX-1 mRNA expression using quantitative RT-PCR. Data are basal means ± SE. COX-1 mRNA expression was not significantly different between the three groups. Post-LPS, both sets of twins as well as control subjects showed a decrease in monocyte COX-1 mRNA expression (P < 0.01 for all groups). B: Monocyte COX-2 mRNA expression using quantitative RT-PCR. Data are basal means ± SE. COX-2 mRNA expression was not significantly different between groups. Post-LPS, both sets of twins and control subjects showed an increase in monocyte COX-2 mRNA expression (P < 0.01 for all groups). C: LPS dose-response for intracellular monocyte COX-2 protein expression by FACS analysis. Data are the fold increase above basal levels at each LPS dose. Mean response in control subjects increased with stepwise increases in LPS dose (P < 0.05 for all comparisons). At each LPS dose, COX-2 response in both sets of twins was lower compared with control subjects at the same dose. *P < 0.04; **P < 0.01; ***P < 0.003. D: Monocyte COX-2 protein expression by Western blotting. Data are the fold increase above basal levels post-LPS. COX-2 protein expression was significantly greater post-LPS in control subjects compared with diabetic and nondiabetic twins. E: Monocyte PGE2 secretion. Monocyte PGE2 secretion post-LPS stimulation in the presence of the COX-2–specific inhibitor DFP. Data shows DFP inhibition of PGE2 production post-LPS. There was no difference in basal or LPS-induced PGE2 production between the groups, but DFP inhibition of PGE2 production post-LPS was significantly greater in both diabetic and nondiabetic twins compared with control subjects. ▪, control subjects; , diabetic twins; , nondiabetic twins.

FIG. 2.

Murine macrophage cell line (J774) COX-1 and -2 mRNA expression and PGE2 secretion. Data are the means ± SE from five experiments. A: COX mRNA expression levels pre- and post-LPS in a murine macrophage cell line. After LPS stimulation, COX-1 mRNA expression levels decreased, and COX-2 mRNA expression levels increased. B: PGE2 secretion pre- and post-LPS in a murine macrophage cell line. After LPS, the level of PGE2 increased, as assessed by a competitive enzyme immunoassay.

FIG. 2.

Murine macrophage cell line (J774) COX-1 and -2 mRNA expression and PGE2 secretion. Data are the means ± SE from five experiments. A: COX mRNA expression levels pre- and post-LPS in a murine macrophage cell line. After LPS stimulation, COX-1 mRNA expression levels decreased, and COX-2 mRNA expression levels increased. B: PGE2 secretion pre- and post-LPS in a murine macrophage cell line. After LPS, the level of PGE2 increased, as assessed by a competitive enzyme immunoassay.

TABLE 1

Human and mouse COX-1 and -2 primers and probes

GenesAccession no.Forward primerReverse primerTaqMan probe
hCOX-2 GI:4506264 5′GAATCATTCACCAGGCAAATTG-3′ 5′TTTCTGTACTGCGGGTGGAAC-3′ 5′TTCCTACCACCAGCA ACCCTGCCA-3′ 
hCOX-1 GI:38045923 5′GGATGCCTTCTCTCGCCAG-3′ 5′ATGTGGTGGTCCATGTTCCTG-3′ 5′CCACCGATCCGGCAGCAA-3′ 
mCOX-1 GI:13542734 5′GGGAATTTGTGAATGCCACC-3′ 5′GGGATAAGGTTGGACCGCA-3′ 5′-TCCGAGAAGTACTCATGCGCCTGGTACT-3′ 
mCOX-2 GI:31981524 5′AGCGAGGACCTGGGTTCAC-3′ 5′AAGGCGCAGTTTATGTTGTCTGT-3′ 5′-AAGTCCACTCCATGGCCAGTCCTCG-3′ 
GenesAccession no.Forward primerReverse primerTaqMan probe
hCOX-2 GI:4506264 5′GAATCATTCACCAGGCAAATTG-3′ 5′TTTCTGTACTGCGGGTGGAAC-3′ 5′TTCCTACCACCAGCA ACCCTGCCA-3′ 
hCOX-1 GI:38045923 5′GGATGCCTTCTCTCGCCAG-3′ 5′ATGTGGTGGTCCATGTTCCTG-3′ 5′CCACCGATCCGGCAGCAA-3′ 
mCOX-1 GI:13542734 5′GGGAATTTGTGAATGCCACC-3′ 5′GGGATAAGGTTGGACCGCA-3′ 5′-TCCGAGAAGTACTCATGCGCCTGGTACT-3′ 
mCOX-2 GI:31981524 5′AGCGAGGACCTGGGTTCAC-3′ 5′AAGGCGCAGTTTATGTTGTCTGT-3′ 5′-AAGTCCACTCCATGGCCAGTCCTCG-3′ 

Table shows both human and mouse COX-1 and -2 primers and probes used for real-time quantitative RT-PCR. h, human; m, mouse.

TABLE 2

Human and mouse COX-1 and -2 amplicons

GenesAmplicon sequence
hCOX-1 5′-GGATGCCTTCTCTCGCCAGATTGCTGGCCGGATCGGTGGGGGCAGGAACATGGACCACCACAT-3′ 
hCOX-2 5′-GAATCATTCACCAGGCAAATTGCTGGCAGGGTTGCTGGTGGTAGGAATG TTCCACCCGCAGTACAGAAA-3′ 
mCOX-1 5′-GGGAATTTGTGAATGCCACCTTCATCCGAGAAGTACTCATGCGCCTGGTACT CACAGTGCGGTCCAACCTTATCCC-3′ 
mCOX-2 5′-AGCGAGGACCTGGGTTCACCCGAGGACTGGGCCATGGAGTGGACTTAAATCACA TTTATGGTGAAACTCTGGACAGACAACATAAACTGCGCCTT-3′ 
GenesAmplicon sequence
hCOX-1 5′-GGATGCCTTCTCTCGCCAGATTGCTGGCCGGATCGGTGGGGGCAGGAACATGGACCACCACAT-3′ 
hCOX-2 5′-GAATCATTCACCAGGCAAATTGCTGGCAGGGTTGCTGGTGGTAGGAATG TTCCACCCGCAGTACAGAAA-3′ 
mCOX-1 5′-GGGAATTTGTGAATGCCACCTTCATCCGAGAAGTACTCATGCGCCTGGTACT CACAGTGCGGTCCAACCTTATCCC-3′ 
mCOX-2 5′-AGCGAGGACCTGGGTTCACCCGAGGACTGGGCCATGGAGTGGACTTAAATCACA TTTATGGTGAAACTCTGGACAGACAACATAAACTGCGCCTT-3′ 

Table shows both human and mouse COX-1 and -2 amplicons used for real-time quantitative RT-PCR. h, human; m, mouse.

TABLE 3

Basal monocytes COX-2 protein expression and PGE2 secretion

COX-2nFACS (MFI)Western blot (intensity)PGE2 (pg/ml)
Control subjects 27 128.1 ± 47.9 149.8 ± 136.9 4,786 ± 6,427 
Twins     
    Diabetes 17 172.1 ± 139.4 155 ± 38.6 4,442 ± 5,341 
    Nondiabetic 17 167.6 ± 101 161.9 ± 38.1 6,204 ± 8,656 
COX-2nFACS (MFI)Western blot (intensity)PGE2 (pg/ml)
Control subjects 27 128.1 ± 47.9 149.8 ± 136.9 4,786 ± 6,427 
Twins     
    Diabetes 17 172.1 ± 139.4 155 ± 38.6 4,442 ± 5,341 
    Nondiabetic 17 167.6 ± 101 161.9 ± 38.1 6,204 ± 8,656 

Data are the means ± SD for basal COX-2 levels in PBMCs and CD14+ monocytes assessed by FACS, Western blotting, and PGE2 secretion. Results are the mean fluorescence intensity for FACS and mean color intensity of protein bands for Western blotting. Both PBMCs for FACS and CD14+ monocytes for protein and PGE2 secretion were cultured as described in research design and methods. There was no difference between twins and control subjects for basal COX-2 or PGE2. MFI, mean fluorescence intensity.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

This study was supported by the British Diabetic Twin Research Trust (to R.D.G.L.), Diabetes UK (to M.L. and R.D.G.L.), and the Joint Research Board at St. Bartholomew’s Hospital (to H.B. and R.D.G.L.). H.B. was supported by the Juvenile Diabetes Research Foundation International.

We thank the twins, their families, research fellows, and research nurses for their assistance throughout this project. We also thank the late Professor Derek Willoughby for advice and Professor Tim Warner for advice and the kind gift of DFP.

1.
Redondo MU, Yu L, Hawa M: Heterogeneity of type 1 diabetes: analysis of monozygotic twins in Great Britain and the United States.
Diabetologia
44
:
354
–362,
2001
2.
Leslie RD, Delli Castelli M: Age-dependent influences on the origins of autoimmune diabetes: evidence and implications.
Diabetes
53
:
3033
–3040,
2004
3.
Leslie RD, Lipsky P, Notkins AL: Autoantibodies as predictors of disease.
J Clin Invest
108
:
1417
–1422,
2001
4.
Carroll M: Innate immunity in aetiopathology of autoimmunity.
Nat Immunol
2
:
1089
–1090,
2001
5.
Wilson SB, Kent SC, Patton KT, Orban T, Jackson RA, Exley M, Porcelli S, Schatz DA, Atkinson MA, Balk SP, Strominger JL, Hafler DA: Extreme Th1 bias of invariant Valpha24JalphaQ T cells in type 1 diabetes.
Nature
391
:
177
–181,
1998
6.
Litherland SA, Xie XT, Hutson AD, Wasserfall C, Whittaker DS, She JX, Hofig A, Dennis MA, Fuller K, Cook R, Schatz D, Moldawer LL, Clare-Salzler MJ: Aberrant prostaglandin synthase 2 expression defines an antigen-presenting cell defect for insulin-dependent diabetes mellitus.
J Clin Invest
104
:
515
–523,
1999
7.
Adler T, Akiyama H, Herder C, Kolb H, Burkart V: Heat shock protein 60 elicits abnormal response in macrophages of diabetes-prone non-obese diabetic mice.
Biochem Biophys Res Commun
294
:
592
–596,
2002
8.
Fernandez NC, Lozier A, Flament C, Ricciardi-Castagnoli P, Bellet D, Suter M, Perricaudet M, Tursz T, Maraskovsky E, Zitvogel L: Dendritic cells directly trigger NK cell functions: cross-talk relevant in innate anti-tumor immune responses in vivo.
Nat Med
5
:
405
–411,
1999
9.
Naumov YN, Bahjat KS, Gausling R, Abraham R, Exley MA, Koezuka Y, Balk SB, Strominger JL, Clare-Salzer M, Wilson SB: Activation of CD1d-restricted T cells protects NOD mice from developing diabetes by regulating dendritic cell subsets.
Proc Natl Acad Sci U S A
98
:
13838
–13843,
2001
10.
DeWitt D, Smith WL: Yes, but do they still get headaches?
Cell
83
:
345
–348,
1995
[erratum Cell 84:following 650,
1996
]
11.
Sweet MJ, Hume DA: Endotoxin signal transduction in macrophages.
J Leukoc Biol
60
:
8
–26,
1996
12.
Feng L, Sun W, Xia Y, Tang WW, Chanmugam P, Soyoola E, Wilson CB, Hwang D: Cloning two isoforms of rat cyclooxygenase: differential regulation of their expression.
Arch Biochem Biophys
307
:
361
–368,
1993
13.
Van JR, Botting RM: New insights into the mode of action of anti-inflammatory drugs.
Inflamm Res
44
:
1
–10,
1995
14.
Hemple SL, Kujubo DA, Herschman HR, Hunniinghake GW: Lipopolysaccharide induces prostaglandin synthase-2 protein and mRNA in human alveolar macrophages and blood monocytes.
J Clin Invest
93
:
391
–396,
1993
15.
Smith WL, Garavito RM, De Witt DL: Prostaglandin endoperoxide H synthases (cyclooxygenases)-1 and -2.
J Biol Chem
271
:
33157
–60,
1996
16.
Vane JR, Bakhle YS, Botting RM: Cyclooxygenases 1 and 2.
Annu Rev Pharmacol Toxicol
38
:
97
–120,
1998
17.
Hawa M, Rowe R, Lan MS, Notkins AL, Pozzilli P, Christie MR, Leslie RD: Value of antibodies to islet protein tyrosine phosphatase-like molecule in predicting type 1 diabetes.
Diabetes
46
:
1270
–1275,
1997
18.
National Diabetes Data Group: Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance.
Diabetes.
28
:
1039
–1057,
1980
19.
Faour WH, He Y, He QW, Ladurataye M, Quintero M, Mancine A, Battista D: Prostaglandin E2 regulates the level and stability cyclooxygenase-2 mRNA through activation p38 mitogen activated protein kinase in interleukin-1 B-treated human synovial fibroblasts.
J Biol Chem
276
:
31720
–31731,
2001
20.
Bustin SA: Quantification of mRNA using real-time reverse transcription PCR (RT-PCR): trends and problems.
J Mol Endocrinol
29
:
23
–39,
2002
21.
Bustin SA: Absolute quantification of mRNA using real-time reverse transcription polymerase chain reaction assays.
J Mol Endocrinol
25
:
169
–193,
2000
22.
Lo SSS, Hawa M, Beer SF, Pyke DA, Leslie RDG: Altered islet beta cell function before the onset of type 1 (insulin-dependent) diabetes mellitus.
Diabetologia
35
:
277
–282,
1992
23.
Smith CJ, Morrow JD, Roberts LJ, Marnett LJ: Differentiation of monocytoid THP-1 cells with phorbol ester induces expression prostaglandin endoperoxide synthase-1 (COX-1)
Biochem Biophys Res Commun
192
:
787
–793,
1993
24.
Ueda N, Yamashita R, Yamamoto S, Ishimura K: Induction of cyclooxygenase-1 in a human megakaryoblastic cell line (CMK) differentiated by phorbol ester.
Biochem Biophys Acta
1344
:
103
–110,
1997
25.
Rioux N, Castonguay A: The induction of COX-1 by tobacco carcinogen U973 human macrophages is correlated to the activation of NFkB.
Carcinogenesis
21
:
1745
–1751,
2000
26.
Ye X, Lui SF: LPS regulates constitutive and inducible transcription factor activities differentially in vivo in the rat.
Biochem Biophys Res Commun
288
:
927
–932,
2001
27.
Plesner A, Greenbaum CJ, Gaur LK, Ernst RK, Lernmark A: Macrophages from high-risk HLA-DQB1*0201/*0302 type 1 diabetes mellitus patients are hypersensitive to lipopolysaccharide stimulation.
Scand J Immunol
56
:
522
–529,
2002
28.
Wucherpfennig KW, Eisenbarth GS: Type 1 diabetes.
Nat Immunol
2
:
767
–768,
2001
29.
Beyan H, Buckley LR, Yousaf N, Londei M, Leslie RD: A role for innate immunity in type 1 diabetes?
Diabetes Metab Res Rev
19
:
89
–100,
2003
30.
Devaraj S, Glaser N, Griffen S, Wang-Polagruto J, Miguelino E, Jialal I: Increased monocytic activity and biomarkers of inflammation in patients with type 1 diabetes.
Diabetes
55
:
774
–779,
2006