CD4+ T cell activation leads to increased IGF-1R, but not IR, expression
Previous publications demonstrating that CD4+ T cells have increased binding of fluorescently labeled insulin following activation have led to the assumption that T cell activation increases IR expression [21,26,27]. However, insulin can bind both IR and IGF-1R, albeit with a lower affinity for IGF-1R [28]. Thus, we examined both Insr and Igf1r gene expression following CD4+ T cell activation. CD4+ T cells were isolated from splenocytes of wildtype C57BL/6J mice and activated for up to 48 hours with anti-CD3 and anti-CD28 antibodies. Interestingly, we observed that Insr gene expression was only mildly changed over the course of activation; however, Igf1r gene expression was dynamically upregulated, with mRNA expression peaking at 6 hours post activation (Fig. 1a-b). To support these results, we also examined protein expression of IR and IGF-1R following CD4+ T cell activation using immunoblot and found that IR protein levels were unchanged following CD4+ T cell activation, whereas IGF-1R protein levels were indeed upregulated following activation and accumulated over time (Fig. 1c).
We next interrogated the gene expression levels of Insr and Igf1r on CD4+ T helper cell subsets (Th1, Th2, Th17, Treg), which we generated in vitro by activating splenic CD4+ T cells in the presence of cytokines and antibodies that promote subset differentiation, as previously described [29]. Each functional subset showed unique gene expression of Insr and Igf1r (Fig. 1d-e). Notably, there was a striking increase in Igf1r expression in Th17 cells compared to the other CD4+ T helper subsets. These results were consistent with protein expression, as determined by immunoblot, which confirmed that Th17 cells have the highest expression of IGF-1R protein among the CD4+ T helper subsets (Fig. 1f).
IGF-1 treatment increases IL-17 production by CD4+ T cells
Given the increased expression of IGF-1R on Th17 cells, we next asked if IGF-1 would increase T cell production of IL-17 in vitro. To test this, splenic CD4+ T cells were isolated from wildtype C57BL/6J mice and activated with plate-bound antibodies to CD3 and CD28 in serum free media supplemented with either Insulin Free Media Supplement or bovine serum albumin (BSA) to prevent any confounding effect of growth factors or hormones found in serum. Following 24 hours of activation, cells were treated with or without 50 ng/mL IGF-1, and IL-17 production was examined at 48 hours. We found that IGF-1 increased the percentage of CD4+ T cells producing IL-17, as measured by flow cytometry (Fig. 2a), as well as the production of IL-17 by CD4+ T, as measured by ELISA (Fig. 2b-c), whereas insulin treatment did not significantly increase the production of IL-17 (Fig. 2b-c). Furthermore, neither insulin nor IGF-1 altered IFN-g production by activated CD4+ T cells (Fig. 2d). We also observed no changes in viability, cell size, activation marker expression, or proliferation in CD4+ T cells activated in the presence of insulin or IGF-1 (Supplementary Fig. S1).
To understand whether this effect of IGF-1 on IL-17 production is dependent on IR or IGF-1R signaling, we activated splenic CD4+ T cells from IR conditional knockout (cKO) mice (CD4Cre+IRfl/fl) or IGF-1R cKO mice (CD4Cre+IGF1Rfl/fl) and littermate controls in full serum conditions and measured IL-17 production by ELISA. Activated CD4+ T cells from both IR cKO and IGF-1R cKO mice showed reduced IL-17 production, suggesting that both IR and IGF-1R signaling are required for activated CD4+ T cells to produce IL-17 (Fig. 2e).
Both insulin and IGF-1 modulate CD4+ T cell metabolism
Both insulin and IGF-1 alter the metabolism of metabolic cells and tissues, with insulin particularly well known for increasing glucose uptake [1,9,13]; thus, we next examined and compared the effects of insulin and IGF-1 on CD4+ T cell metabolism. Treatment of activated splenic CD4+ T cells with a physiological dose of IGF-1 increased glucose uptake, as measured by uptake of tritiated 2-deoxyglucose, while treatment with a physiological dose of insulin did not (Fig. 3a). However, insulin significantly increased basal oxygen consumption rate (OCR; a surrogate for mitochondrial metabolism), extracellular acidification rate (ECAR; a surrogate for glycolytic metabolism), the ratio of oxidative to glycolytic metabolism (OCR/ECAR), maximal respiration, spare respiratory capacity (SRC), ATP production, and proton leak in activated CD4+ T cells, as measured using extracellular flux analysis and using the Seahorse Mito Stress test (Fig. 3b-h). Similar to insulin, IGF-1 treatment of activated CD4+ T cells significantly increased basal OCR, ECAR, the OCR/ECAR ratio, maximal respiration, ATP production, and proton leak, but SRC. (Fig. 3b-h). Overall, however, the metabolic changes in activated CD4+ T cells treated with 50 ng/mL IGF-1 were less potent than the metabolic changes seen in activated CD4+ T cells treated with 10 ng/mL insulin.
IGF-1 uniquely impacts the mitochondrial membrane potential of activated CD4+ T cells
IGF-1 has been implicated in several mitochondrial processes such as mitochondrial biogenesis, mitophagy, and mitochondrial function [23-25]. We therefore measured the mitochondrial mass and mitochondrial membrane potential of activated CD4+ T cells treated with insulin or IGF-1. First, activated CD4+ T cells were stained with MitoTracker Green (MTG) to measure mitochondrial mass. Neither insulin nor IGF-1 caused any significant change in mitochondrial mass compared to untreated cells (Fig. 3i). In order to generate ATP in the mitochondria, mitochondrial membrane potential is generated by pumping protons into the intermembrane space which creates a gradient and polarizes the inner mitochondrial membrane. This proton gradient is then used to generate ATP via ATP synthase. To measure the degree of inner mitochondrial membrane polarization, we stained cells with tetramethylrhodamine (TMRE), which is a cell permeant positively charged dye that accumulates in active mitochondria due to their relative negative charge. Inactive or depolarized mitochondria with decreased membrane potential fail to sequester TMRE. IGF-1 treatment of activated CD4+ T cells significantly decreased TMRE staining, while insulin treatment had no significant effect (Fig. 3j), suggesting that IGF-1, but not insulin, decreases mitochondrial membrane potential in activated CD4+ T cells.
IGF-1 treatment of Th17 cells promotes metabolism and function
Given our findings that IGF-1R is significantly upregulated on Th17 cells compared to other CD4+ T helper subsets (Fig. 1e-f), and treatment with IGF-1 causes an increase in IL-17 production by activated bulk CD4+ T cells (Fig. 2a-c), we next investigated the effect of IGF-1 treatment on the function and metabolism of Th17 cells. Th17 cells were differentiated from CD4+ T cells in vitro, as previously described [29], and then treated with or without IGF-1 for an additional 48 hours. IGF-1 treatment of differentiated Th17 cells did not increase the percentage of IL-17 positive cells but did increase IL-17 production as measured by mean fluorescent intensity flow cytometrically (Fig. 4a-b). IGF-1 also altered the metabolism of differentiated Th17 cells, with increased basal OCR, ECAR, maximal respiration, spare respiratory capacity, ATP production, and proton leak (Fig. 4c-h), mirroring the results seen in bulk activated CD4+ T cells treated with IGF-1. In contrast, differentiated Th1 and Treg cells treated with IGF-1 did not show any change in IFN-g or Foxp3 expression, respectively, nor increased cellular metabolism (Supplementary Fig. S2).
IGF-1 treatment of Th17 cells reduces mitochondrial membrane potential and mROS
IGF-1 treated Th17 cells also had decreased mitochondrial membrane potential compared to untreated Th17 cells, as measured by TMRE staining (Fig. 4i), but showed no significant difference in mitochondrial mass (Fig. 4j). Decreased membrane polarization can be an indicator that the mitochondria are uncoupled, meaning that protons are translocating across the inner mitochondrial membrane through a mechanism other than ATP synthase, and are thus not used to make ATP. This mechanism of uncoupling can be used by cells to protect against oxidative stress since increased mitochondrial polarization causes an increase in the production of ROS. To test whether IGF-1 treatment of Th17 cells decreased mROS along with decreased mitochondrial membrane potential, we measured mROS using the MitoSOX stain. Indeed, we saw that Th17 cells treated with IGF-1 had reduced MitoSOX staining (Fig. 4k). Furthermore, cellular ROS production, as measured by CellROX staining, also showed decreased ROS production in IGF-1 treated Th17 cells (Fig. 4l), suggesting that IGF-1 exerts a cytoprotective effect in addition to a metabolic effect.
IGF-1 treatment of Th17 cells decreases mitochondrial membrane potential and mROS production in an IGF-1R and IR dependent manner
To determine if these IGF-1-mediated changes on Th17 cell mitochondrial membrane potential and mROS require signaling by IR, IGF-1R, or both, we differentiated Th17 cells in vitro from IGF-1R cKO mice or IR cKO mice, and littermate controls. Following differentiation, Th17 cells were treated with or without IGF-1 for an additional 48 hours. We observed that IGF-1 treatment of Th17 cells from IGF-1R cKO mice did not decrease TMRE, as it did in Th17 cells from control mice (Fig. 5a). Consistent with our earlier results, mitochondrial mass was not significantly changed in either control or IGF-1R cKO Th17 cells treated with IGF-1 (Fig. 5b). Similar to TMRE, MitoSOX staining was decreased in control Th17 cells but not in Th17 cells generated from IGF-1RcKO mice (Fig. 5c). These data suggest that the effect of IGF-1 on Th17 cell mitochondrial membrane polarization and mROS production is IGF-1R dependent.
Since IGF-1R and IR are highly homologous and are both able to bind IGF-1, we also investigated the requirement for IR in mediating the effect of IGF-1 on mitochondrial membrane potential and mROS production in Th17 cells. IGF-1 treatment of Th17 cells from IR cKO mice did not decrease mitochondrial membrane potential as it did in Th17 cells from control mice (Fig. 5d). Mitochondrial mass was not significantly changed following IGF-1 treatment of Th17 cells from either control or IR cKO mice (Fig. 5e). Lastly, IGF-1 treatment of Th17 cells from IR cKO mice did not decrease mROS as it did in Th17 cells from control mice (Fig. 5f). Interestingly, these results suggest that the effect of IGF-1 on Th17 cell mitochondrial membrane polarization and mROS production require both IR and IGF-1R, and therefore indicate a potential role for the IR/IGF-1R hybrid in mediating effects of IGF-1 on Th17 cell mitochondria.
Uncoupling protein 2 mediates the cytoprotective effect of IGF-1 on Th17 cells
Since IGF-1 treatment of Th17 cells was shown to reduce mitochondrial membrane potential and mROS production, we next set out to understand how IGF-1 signaling could be mediating this effect. One mechanism by which the mitochondrial membrane potential can be decreased is through uncoupling proteins, which move protons back across the inner mitochondrial membrane, bypassing ATP synthase [30]. Uncoupling protein 2 (UCP2) is a mitochondrial transmembrane protein which functions to allow protons to flow across the inner mitochondrial membrane, thus uncoupling ATP synthesis [31]. UCP2 is closely related to uncoupling protein 1, which is involved in brown fat thermogenesis. UCP2 is known to be expressed on immune cells and is upregulated in CD4+ and CD8+ T cells during activation [32]. In CD8+ T cells, UCP2 was shown to inhibit glycolysis and fatty acid synthesis, while reducing ROS [33]. Moreover, total body knockout of UCP2 led to more severe disease in the EAE model, due to increased ROS produced by innate immune cells [34]. Interestingly, UCP2 has been shown to be upregulated by IGF-1 treatment in cancer cell lines [35].
We therefore investigated whether IGF-1 treatment of Th17 cells increased UCP2 expression. Indeed, we found that IGF-1 upregulated Ucp2 in both activated CD4+ T cells and in differentiated Th17 cells (Fig. 6a-b). We therefore used an inhibitor of UCP2, genipin [36], to probe whether UCP2 activity was required for IGF-1 effects on mitochondrial membrane metabolism. When Th17 cells were treated with IGF-1 in the presence of the UCP2 inhibitor, IGF-1-induced changes in mitochondrial membrane potential and mROS production were reversed, without affecting mitochondrial mass (Fig. 6c-e). UCP2 inhibition also reversed IGF-1 effects on oxygen consumption, proton leak, and ATP production (Fig. 6f-h). These data suggest that the cytoprotective effects of IGF-1 are mediated, at least in part, through UCP2 activity.