Cyclic Adenosine Monophosphate (cAMP)-Dependent Phosphodiesterase Inhibition Promotes Bone Anabolism Through CD8+ T Cell Wnt-10b Production in Mice

Abstract

Cyclic adenosine monophosphate (cAMP)-dependent phosphodiesterase (PDE) inhibitors such as pentoxifylline (PTX) suppress cAMP degradation and promote cAMP-dependent signal transduction. PDE inhibitors increase bone formation and bone mass in preclinical models and are used clinically to treat psoriatic arthritis by targeting inflammatory mediators including activated T cells. T cell activation requires two signals: antigen-dependent CD3-activation, which stimulates cAMP production; and CD28 co-stimulation, which downregulates cAMP-signaling, through PDE activation. PDE-inhibitors consequently suppress T cell activation by disrupting CD28 co-stimulation. Interestingly, we have reported that when CD8⁺ T cells are activated in the absence of CD28 co-stimulation, they secrete Wnt-10b, a bone anabolic Wnt ligand that promotes bone formation. In the present study, we investigated whether the bone anabolic activity of the PDE-inhibitor PTX, has an immunocentric basis, involving Wnt-10b production by CD8⁺ T cells. When wild-type (WT) mice were administered PTX, biochemical markers of both bone resorption and formation were significantly increased, with net bone gain in the axial skeleton, as quantified by micro-computed tomography (μCT). By contrast, PTX increased only bone resorption in T cell knockout (KO) mice, causing net bone loss. Reconstituting T cell-deficient mice with WT, but not Wnt-10b knockout (KO) CD8⁺ T cells, rescued bone formation and prevented bone loss. To study the role of cAMP signaling in Wnt-10b expression, reverse-transcription polymerase chain reaction (RT-PCR) and luciferase-reporter assays were performed using primary T cells. PDE inhibitors intensified Wnt-10b promoter activity and messenger RNA (mRNA) accumulation in CD3 and CD28 activated CD8⁺ T cells. In contrast, inhibiting the cAMP pathway mediators protein kinase A (PKA) and cAMP response element-binding protein (CREB), suppressed Wnt-10b expression by T cells activated in the absence of CD28 co-stimulation. In conclusion, the data demonstrate a key role for Wnt-10b production by CD8⁺ T cells in the bone anabolic response to PDE-inhibitors and reveal competing T cell-independent pro-resorptive properties of PTX, which dominate under T cell-deficient conditions. Selective targeting of CD8⁺ T cells by PDE inhibitors may be a beneficial approach for promoting bone regeneration in osteoporotic conditions. © 2022 The Authors. JBMR Plus published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research.

Keywords: ANABOLICS; CELLS OF BONE; OSTEOBLASTS; OSTEOIMMUNOLOGY; SYSTEMS BIOLOGY–BONE INTERACTORS; THERAPEUTICS.

Fig. 1

PTX induces both bone formation and resorption in WT mice leading to increased lumbar spine trabecular and cortical bone mass. Two‐month‐old WT C57BL6/J female mice were treated with PTX or vehicle (PBS) 5 times/week, for 12 weeks. Vertebral (L₃) trabecular and cortical structure was analyzed by μCT and serum bone turnover markers by ELISA. μCT trabecular indices: (A) BV/TV; (B) TV; (C) BV; (D) Tb.N; (E) Tb.Th; (F) Tb.Sp; (G) Conn.D; and (H) TV.D. μCT cortical indices: (I) Ct.Th and (J) Ct.Ar. (K) Representative 6‐μm μCT reconstructions of vertebrae for vehicle‐treated and PTX‐treated mice. Red scale bar = 500 μm. (L) Bone resorption marker CTx and (M) bone formation marker P1NP. (N) RT‐PCR analysis of Wnt‐10b gene expression in bone marrow. Mean ± SEM. *p < 0.05; **p < 0.01, ***p < 0.001, ****p < 0.001, or p = ns (not significant) by Student's t test or Mann‐Whitney test (P1NP) following assessment for normal distribution by Shapiro‐Wilk normality test. n = 11 mice/group. One vehicle RT‐PCR sample was undetectable in N. BV = bone volume; BV/TV = trabecular bone volume fraction; Conn.D = connectivity density; Ct.Ar = cortical area; Ct.Th = cortical thickness; CTx = C‐terminal telopeptide of collagen type I; P1NP = N‐terminal propeptide of type‐I procollagen; Tb.N = trabecular number; Tb.Sp = trabecular separation; Tb.Th = trabecular thickness; TV = tissue volume; TV.D = volumetric bone density.

Fig. 2

PTX does not increase femoral trabecular and cortical bone mass in WT mice. Two‐month‐old healthy intact WT C57BL6/J female mice were treated with PTX or vehicle (PBS) 5 times/week, for 12 weeks and femoral bone analyzed by μCT. Trabecular bone indices: (A) BV/TV; (B) TV; (C) BV; (D) Tb.N; (E) Tb.Th; (F) Tb.Sp; (G) Conn.D; and (H) TV.D. Cortical indices: (I) Ct.Th and (J) Ct.Ar. (K) Representative 6‐μm μCT reconstructions of the femur are shown for vehicle‐treated and PTX‐treated mice with trabecular bone on top and cortical bone below. Red scale bar = 500 μm. Mean ± SEM. *p < 0.05; **p < 0.01, ****p < 0.001, or p = ns (not significant) by Student's t test or Mann‐Whitney test (Tb.N, Ct.Th, and Ct.Ar) following assessment for normal distribution by Shapiro‐Wilk normality test. n = 11 mice/group. One femur in the vehicle group was damaged and could not be analyzed. BV = bone volume; BV/TV = trabecular bone volume fraction; Conn.D = connectivity density; Ct.Ar = cortical area; Ct.Th = cortical thickness; CTx = C‐terminal telopeptide of collagen type I; P1NP = N‐terminal propeptide of type‐I procollagen; Tb.N = trabecular number; Tb.Sp = trabecular separation; Tb.Th = trabecular thickness; TV = tissue volume; TV.D = volumetric bone density.

Fig. 3

PTX induces vertebral bone loss due to increased resorption, but not formation, in T cell–deficient mice, but is rescued by reconstitution of CD8⁺ T cells from WT, but not Wnt‐10b KO mice. C57BL6/J female TCRβ KO T cell–deficient mice, and TCRβ KO mice reconstituted with WT or CD8⁺ Wnt‐10b KO T cells, were treated with PTX or vehicle (PBS) 5 times/week, for 12 weeks. Vertebral structural indices were analyzed by μCT for: (A) BV/TV; (B) TV; (C) BV; (D) Tb.N; (E) Tb.Th; (F) Tb.Sp; (G) Conn.D; and (H) TV.D, and for cortical indices: (I) Ct.Th and (J) Ct.Ar. (K) Representative 6‐μm μCT reconstructions. Red scale bar = 500 μm. (L) Bone resorption marker C‐terminal telopeptide of CTx and (M) bone formation marker P1NP. (N) RT‐PCR analysis of Wnt‐10b expression in bone marrow. Mean ± SEM. *p < 0.05; **p < 0.01, ***p < 0.001, ****p < 0.001, or p = ns (not significant) by one‐way ANOVA or Kruskal‐Wallis test (BV/TV, Tb.N, Tb.Sp, Conn.D, TV.D, and Ct.Th). n = 10 mice/group. One extreme outlier in the Wnt‐10b KO T cells + PTX group, with an exceptionally high BV/TV (=17%) and other parameters was excluded from the μCT data. BV = bone volume; BV/TV = trabecular bone volume fraction; Conn.D = connectivity density; Ct.Ar = cortical area; Ct.Th = cortical thickness; CTx = C‐terminal telopeptide of collagen type I; P1NP = N‐terminal propeptide of type‐I procollagen; Tb.N = trabecular number; Tb.Sp = trabecular separation; Tb.Th = trabecular thickness; TV = tissue volume; TV.D = volumetric bone density.

Fig. 4

In vitro analysis of the role of the cAMP pathway in Wnt‐10b expression by CD8⁺ T cells. The effect of cAMP pathway activators and inhibitors on Wnt‐10b gene expression in purified CD8⁺ T cells was assessed by (A) RT‐PCR and (B) by luciferase assay using a transfected Wnt‐10b promoter reporter and normalized to Renilla. In both (A) and (B) purified T cells were stimulated with plate bound anti‐CD3e (10 μg/mL) and/or soluble anti‐CD28 activating antibodies (25 μg/mL). Cells were treated with the inducers of cAMP accumulation (PTX [100μM] and apremilast [10μM]) or with H89 (25μM) an inhibitor of PKA or 666‐15 (1μM) a CREB inhibitor. After 6 hours (A) mRNA was extracted for RT‐PCR quantification of Wnt‐10b expression with normalization to 18S, or (B) quantification of luciferase activity using a dual‐luciferase reporter assay with normalization for Renilla expression. Mean ± SEM. *p < 0.05; **p < 0.01, ***p < 0.001, ****p < 0.001, or p = ns (not significant) by one‐way ANOVA following assessment for normal distribution by Shapiro‐Wilk normality test. n = 4 replicates/group and the data are representative of two independent experiments.

Fig. 5

Model for PDE‐induced bone anabolism by CD8⁺ T cells. (A) Under immunosufficient conditions, (1) APCs present antigen to the TCR activating the associated CD3 complex (signal 1) and leading to activation of adenylate cyclase with synthesis of cAMP. (2) APC‐expressed CD80/CD86 ligands activate T cell CD28 receptors (signal 2) promoting catabolism of cAMP to signaling inert AMP, through PDE4. PDE inhibitors such as PTX, counteract cAMP degradation allowing accumulation of cAMP. cAMP activates PKA, which in turn activates CREB. CREB translocates to the nucleus and associates with CREs in gene promoters, inducing transcription of target genes including Wnt‐10b. (3) Wnt‐10b protein binds to LRP5/6 receptors leading to osteoblastogenesis/osteoblast activation and driving bone formation. (4) Direct effects of PDE‐inhibitor on osteoclasts and/or osteoblasts, induce resorption, causing a high bone turnover state, but with net gain of bone mass in the axial skeleton. (B) Under immunodeficient conditions the catabolic effects of PDE dominate, leading to bone loss. APC = antigen presenting cell; CRE = CREB responsive element; CREB = cAMP responsive element binding protein; LRP5/6 = low‐density lipoprotein receptor‐related protein 5/6; PDE4 = phosphodiesterase‐4; PKA = protein kinase A; TCR = T cell receptor.