figure a

Introduction

Pancreatic beta cell compensation for obesity-associated insulin resistance is crucial for the maintenance of glucose homeostasis. Type 2 diabetes develops only in individuals who are unable to sustain the beta cell compensatory response. Beta cell failure in humans and rodents with type 2 diabetes is characterised by the disruption of beta cell function and gene expression [1, 2]. The latter features the downregulation of beta cell enriched genes, including key transcription factors, beta cell function genes and adaptive unfolded protein response (UPR) genes [1,2,3,4]. However, the triggering stimuli are poorly understood.

Inflammatory stress has been suggested to play a role in the pathogenesis of beta cell failure and type 2 diabetes [5,6,7]. The number of islet macrophages is increased in type 2 diabetes and they have been proposed as key contributors to beta cell inflammation [5,6,7]. However, the current knowledge of the molecular differences between islet macrophages in conditions of beta cell compensation and failure is poor. Here we used well-established obese mouse models with opposing disposition to diabetes development (diabetes-resistant ob/ob mice and diabetes-prone db/db mice) [4] to examine the gene expression changes in flow-sorted islet macrophages and the potential involvement of islet macrophages in beta cell (mal)adaptation.

Methods

Mice

Sixteen-week-old male and female C57BL/KsJ db/db and C57BL/6J ob/ob mice, and their littermate lean controls (C57BL/KsJ or C57BL/6J, respectively) were taken from the Garvan Institute breeding colonies (Australian BioResources, Moss Vale, NSW, Australia). Animals were kept under conventional conditions with free access to food and water. Procedures were approved by the Garvan Institute/St Vincent’s Hospital Animal Experimentation Ethics Committee, following guidelines issued by the National Health and Medical Research Council of Australia. Islets were isolated by liberase digestion (Roche Diagnostics, Castle Hill, Australia), gradient centrifugation (Ficoll-Paque PLUS gradient; GE Healthcare Bio-Sciences, Uppsala, Sweden) and handpicking under a stereomicroscope [4].

Flow cytometry and cell sorting

Flow cytometry and cell sorting of macrophages were performed in a BD FACSAria III cell sorter (Becton Dickinson, Franklin Lakes, NJ, USA) for cells stained positive with F4/80-APC (1:250 dilution; catalogue no. 123115) and CD11b-FITC (1:200 dilution; catalogue no. 101205) antibodies (Biolegend, San Diego, CA, USA) (see Fig. 1a–d and electronic supplementary material [ESM] Methods for further details). DAPI (0.1 μg/ml) was added to samples prior to sorting. Initial gating was applied to capture all cells, followed by single-cell gating. DAPI-negative cells were gated as live cells followed by macrophage gating with positive CD11b and F4/80 staining. All gating boundaries were determined using unstained samples. Where possible, sorted cells were re-analysed to ensure collected samples contained only cells that were stained positive for both markers.

Fig. 1
figure 1

 Changes in macrophage content and gene expression levels in islets from male and female db/db and ob/ob mice. Isolated islets were dissociated into single cell suspensions using trypsin, blocked, and then stained for macrophage-specific markers using F4/80-APC and CD11b-FITC antibodies. (a–d) Subsequently, cells were sorted using a BD FACSAria III cell sorter. An example of the flow cytometry gating is shown. Four levels of gating (labelled ‘Cells’ and ‘P2’–‘P4’) were used to separate single cells from the cell suspension (a). Live cells were then gated from the single cell population with DAPI staining (b). From live single cells, macrophages were gated on F4/80 and CD11b (c). A summary of the results of the gating shown in (a–c) is given in the table (d). The percentage of gated cells from the parent population is given under the column heading ‘%Parent’. Macrophage content was expressed as a percentage of total live islet cells (e). Total RNA was extracted from macrophages, reverse-transcribed and analysed by real-time RT-PCR. Shown are mRNA levels in macrophages of db/db (white bars) and ob/ob (grey bars) mouse islets expressed as fold change of the levels in respective controls (represented by the dashed line) (f). All results are mean ± SEM. *p<0.05, **p<0.01, ***p<0.001 vs control. Mac, macrophages

Full size image

Macrophage depletion in islets

Islets were cultured in liposomes containing either PBS or clodronate (1 mg/ml) (Liposoma, Amsterdam, the Netherlands) for 2 days. Islets were randomly handpicked for treatment. RNA was extracted for gene expression analysis. Insulin and proinsulin secretion was assessed in groups of five islets, as previously described [4]. Insulin was measured by a homogeneous time-resolved fluorescence (HTRF) assay kit (Cisbio; Genesearch, Arundel, QLD, Australia). Proinsulin was measured by ELISA (Mercodia AB, Uppsala, Sweden).

RNA analysis

Total RNA from macrophages and islets was extracted using the RNeasy Micro Kit and RNeasy Mini Kit, respectively (Qiagen, Doncaster, VIC, Australia). cDNA was synthesised using the QuantiTect Reverse Transcription Kit (Qiagen). Real-time PCR with macrophage and islet cDNA was performed on a 7900HT Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) using TaqMan assays (Applied Biosystems) and Power SYBR Green PCR Master Mix (Applied Biosystems), respectively. Primer details are provided in ESM Table 1. The value obtained for each specific product was normalised to a control gene (cyclophilin A) and expressed as a fold change of the value in control extracts.

Statistical analysis

All results are presented as means ± SEM. Student’s unpaired t test was used to compare differences between two groups. Differences between more than two groups were calculated using ANOVA with Fisher’s least significant difference (LSD) post hoc test. The researchers were not blinded to experimental groups.

Results

Differential gene expression changes in macrophages of islets from db/db and ob/ob mice

The ob/ob mouse maintains normoglycaemia because of successful beta cell compensation, whereas the db/db mouse progresses to diabetes due to the failure of beta cell compensation (ESM Fig. 1b). Here, an increase in macrophage levels was observed in islets from diabetic db/db mice compared with lean control mice, but not in islets from ob/ob mice (Fig. 1e). This suggests that macrophage accumulation in islets is not solely dependent on obesity or increased fatty acids because these variables are similar in db/db and ob/ob mice (ESM Fig. 1a and [4]). We next assessed changes in gene expression in flow-sorted macrophages using RT-PCR. In macrophages from db/db islets, the mRNA levels of the proinflammatory macrophage marker interferon regulatory factor 5 (Irf5) were increased, as compared with control (Fig. 1f). In contrast, Irf5 mRNA levels were unchanged in macrophages from ob/ob islets (Fig. 1f). Moreover, the mRNA levels of proinflammatory IL-1β (Il1b) were maintained in macrophages from db/db islets vs control, whereas they were markedly suppressed in macrophages from ob/ob islets (Fig. 1f). IL-1 receptor antagonist (Il1rn) is triggered by the same stimuli as IL-1β. Il1rn mRNA levels were markedly increased in macrophages from db/db islets, whereas they were unchanged in macrophages from ob/ob islets (Fig. 1f). The effects of TGF-β1 on macrophages are mainly anti-inflammatory. Interestingly, Tgfb1 expression levels were significantly increased in macrophages from ob/ob islets, whereas they were unchanged in macrophages from db/db islets (Fig. 1f). The mannose receptor C-type 1 (MRC1) is expressed on the surface of macrophages and is regarded as a marker of alternative activation. The expression levels of Mrc1 were upregulated in macrophages from db/db islets, but not in macrophages from ob/ob islets (Fig. 1f). Overall, these findings demonstrate for the first time that markedly different patterns of macrophage gene expression accompany beta cell compensation and failure.

Effects of depleting macrophages on gene expression in islets from db/db and ob/ob mice

We previously reported that the mRNA levels of genes important for the maintenance of the beta cell phenotype are reduced in db/db islets, whereas they are preserved in ob/ob islets [4]. Here, we investigated the effects of depleting macrophages on gene expression in islets from db/db and ob/ob mice. The studies were performed ex vivo to assess the direct effects of depleting macrophages in islets. Clodronate-liposome treatment depleted macrophages in islets from control, db/db and ob/ob mice (vs PBS-liposome treatment), as evidenced by reduced mRNA expression of macrophage markers, Cd11b (also known as Itgam) and F4/80 (also known as Adgre1) (ESM Fig. 2a,b and Fig. 2a,b). The mRNA levels of transcription factors important for islet development and the maintenance of beta cell differentiation were increased in clodronate-liposome-treated db/db islets compared with PBS-liposome-treated db/db islets. Mafa and Pdx1 mRNA levels were significantly increased in db/db islets after macrophage depletion (Fig. 2c,d). In contrast, no significant difference was observed in the mRNA levels of Mafa and Pdx1 in ob/ob islets following depletion of macrophages (Fig. 2c,d). Genes involved in beta cell function were evaluated. The mRNA levels of the glucose transporter GLUT2 (Glut2 [also known as Slc2a2]), the pore-forming subunit of the ATP-sensitive K+ channel (Kcnj11) and the incretin receptor (Gipr) were increased in macrophage-depleted db/db islets compared with PBS-liposome-treated db/db islets, but they were not significantly affected following macrophage depletion in ob/ob islets (Fig. 2f,h,i). These data suggest that macrophage depletion in db/db islets is associated with partial restoration of beta cell gene expression. Furthermore, the mRNA levels of adaptive UPR genes, Xbp1, Hspa5, Pdia4 and Fkbp11, were increased in macrophage-depleted db/db islets vs db/db islets treated with PBS-containing liposomes, whereas expression levels of the deleterious UPR effector, Ddit3, were reduced (Fig. 2k–o). In contrast, differences in UPR gene expression were not detected in macrophage-depleted ob/ob islets (Fig. 2k–o). These data indicate that macrophages contribute to the alterations in the pattern of UPR gene expression in db/db mice by downregulating Xbp1 and the adaptive UPR, while promoting deleterious UPR gene expression. Moreover, macrophage depletion reduced the expression of genes involved in inflammatory (Il1b, Fas and Nfkbia) (ESM Fig. 3) and oxidative stress (catalase [Cat] and heme oxygenase-1 [Hmox1]) in db/db islets (ESM Fig. 4), whereas only catalase was reduced in ob/ob islets. Overall, our studies suggest that depleting macrophages would be beneficial for reducing beta cell stress and preserving beta cell identity.

Fig. 2
figure 2

 Effects of macrophage depletion on gene expression in islets from male and female db/db and ob/ob mice. Isolated islets were cultured with liposomes containing PBS or clodronate for 2 days. Total RNA was extracted from islets, reverse-transcribed and analysed by real-time RT-PCR. Shown are mRNA levels in db/db and ob/ob mouse islets cultured in the presence of PBS-liposome (white bars) or clodronate-liposome (grey bars) expressed as fold change of the levels in respective controls (represented by the dashed line). All results are mean ± SEM. *p<0.05, **p<0.01, ***p<0.001, clodronate-liposome treatment vs PBS-liposome treatment per group

Full size image

The therapeutic utility of targeting macrophages for type 2 diabetes is unclear. Recent findings from the Canakinumab Anti-inflammatory Thrombosis Outcome Study (CANTOS) showed that IL-1β inhibition did not have long-lasting effects on HbA1c or fasting plasma glucose levels among patients with diabetes [8]. In our study, insulin secretory function remained impaired in db/db islets after macrophage depletion as compared with PBS-liposome treatment (ESM Fig. 5). Another feature of beta cell dysfunction in type 2 diabetes is elevated proinsulin secretion. Interestingly, proinsulin secretion was reduced following macrophage depletion in db/db islets as compared with PBS-liposome treatment (ESM Fig. 6a,b). This occurs in concert with restoration of adaptive UPR gene expression (Fig. 2k–n), which is crucial for proinsulin processing [9].

Discussion

This study shows for the first time that: (1) differential shifts in macrophage genotypes accompany beta cell compensation and failure in obese mice; (2) islet macrophages contribute to beta cell stress and dedifferentiation in a rodent model of diabetes; and (3) islet macrophages are crucial regulators of the balance between adaptive and deleterious UPRs during obesity-associated endoplasmic reticulum (ER) stress. These observations reveal the potential importance of macrophage plasticity in the regulation of beta cell identity during obesity.

Recent studies have shown that the islet-resident macrophage has a homeostatic role [10]. It is not clear whether the increased number of macrophages in db/db islets is due to monocyte recruitment. However, if so, one may postulate that infiltrating macrophages may have a less homeostatic function than resident macrophages, which could help explain the major differences in the phenotypes between db/db and ob/ob mice. Another possibility is that alternatively activated macrophages could be causing beta cell dedifferentiation in an attempt to regenerate damaged tissue.

Our studies in rodents support a model in which distinct alterations in islet macrophages during obesity are critically important for beta cell abnormalities. We propose that the accumulation of alternatively activated macrophages in islets contributes to beta cell stress and the loss of beta cell identity in diabetes.