Desipramine decreases expression of human and murine indoleamine-2,3-dioxygenases

Abstract

Abundant evidence connects depression symptomology with immune system activation, stress and subsequently elevated levels of kynurenine. Anti-depressants, such as the tricyclic norepinephrine/serotonin reuptake inhibitor desipramine (Desip), were developed under the premise that increasing extracellular neurotransmitter level was the sole mechanism by which they alleviate depressive symptomologies. However, evidence suggests that anti-depressants have additional actions that contribute to their therapeutic potential. The Kynurenine Pathway produces tryptophan metabolites that modulate neurotransmitter activity. This recognition identified another putative pathway for anti-depressant targeting. Considering a recognized role of the Kynurenine Pathway in depression, we investigated the potential for Desip to alter expression of rate-limiting enzymes of this pathway: indoleamine-2,3-dioxygenases (Ido1 and Ido2). Mice were administered lipopolysaccharide (LPS) or synthetic glucocorticoid dexamethasone (Dex) with Desip to determine if Desip alters indoleamine-dioxygenase (DO) expression in vivo following a modeled immune and stress response. This work was followed by treating murine and human peripheral blood mononuclear cells (PBMCs) with interferon-gamma (IFNγ) and Desip. In vivo: Desip blocked LPS-induced Ido1 expression in hippocampi, astrocytes, microglia and PBMCs and Ido2 expression by PBMCs. Ex vivo: Desip decreased IFNγ-induced Ido1 and Ido2 expression in murine PBMCs. This effect was directly translatable to the human system as Desip decreased IDO1 and IDO2 expression by human PBMCs. These data demonstrate for the first time that an anti-depressant alters expression of Ido1 and Ido2, identifying a possible new mechanism behind anti-depressant activity. Furthermore, we propose the assessment of PBMCs for anti-depressant responsiveness using IDO expression as a biomarker.

Keywords: Depression; Desipramine; Ido1; Inflammation; Kynurenine pathway; PBMCs.

Figure 1. Desip decreased Ido1-FL expression

After i.p. treatment of mice with either LPS or Dex (in vivo, A–D) or cells/brain tissue with IFNγ (ex vivo, E–H) with and without Desip, expression of Ido1-FL was quantified in A) Hippocampi, B) Astrocytes (ACSA-2⁺ enriched cells), C) Microglia (CD11b⁺ enriched cells), D) PBMCs, E) PBMC^−T, F) T cells (CD90⁺ enriched cells), G) OHSCs ± Dex and H) OHSC ± Gal-9. * p ≤ 0.05 for Ido1-FL induction by LPS or IFNγ; ^δ p ≤ 0.05 for the inhibitory effect of Desip within LPS or IFNγ; ^ψ p ≤ 0.05 for the ability of Dex or Gal-9 to synergistically enhance IFNγ-stimulated gene expression.

Figure 2. Attenuation of Ido1-v1 expression by Desip

After treatment of mice i.p. with either LPS or Dex (in vivo, A–D) or cells/tissue with IFNγ (ex vivo, E–H) with and without Desip, expression of Ido1-v1 was quantified in A) Hippocampi, B) Astrocytes (ACSA-2⁺ enriched cells), C) Microglia (CD11b⁺ enriched cells), D) PBMCs, E) PBMC^−T, F) T cells (CD90⁺ enriched cells), G) OHSCs ± Dex and H) OHSC ± Gal-9. * p ≤ 0.05 for Ido1-v1 induction by LPS, Dex or IFNγ; ^δ p ≤ 0.05 for the inhibitory effect of Desip within LPS or IFNγ; ^ψ p ≤ 0.05 for the ability of Dex or Gal-9 to synergistically enhance IFNγ-stimulated gene expression; ^ϕ p ≤ 0.05 main effect of Desip.

Figure 3. Attenuation of Ido2-v1 expression by Desip

After treatment of mice i.p. with either LPS or Dex (in vivo, A–D) or cells/tissue with IFNγ (ex vivo, E–H) with and without Desip, expression of Ido2-v1 was quantified in A) Hippocampi, B) Astrocytes (ACSA-2⁺ cells), C) Microglia (CD11b⁺ enriched cells), D) PBMCs, E) PBMC^−T, F) T enriched cells (CD90⁺ enriched cells), G) OHSCs ± Dex and H) OHSC ± Gal-9. * p ≤ 0.05 for Ido2-v1 induction by LPS or IFNγ; ^δ p ≤ 0.05 for the effect of Desip within LPS or IFNγ; ^ψ p ≤ 0.05 for the ability of Dex or Gal-9 to synergistically enhance IFNγ-stimulated gene expression.

Figure 4. Attenuation of Ido2-v3 expression by Desip

After treatment of mice i.p. with either LPS or Dex (in vivo, A–D) or cells/tissue with IFNγ (ex vivo, E–H) with and without Desip, expression of Ido2-v3 was quantified in A) Hippocampi, B) Astrocytes (ACSA-2⁺ enriched cells), C) Microglia (CD11b⁺ enriched cells), D) PBMCs, E) PBMC^−T, F) T cells (CD90⁺ enriched cells), G) OHSCs ± Dex and H) OHSC ± Gal-9. * p ≤ 0.05 for Ido2-v3 induction by LPS, Dex or IFNγ; ^δ p ≤ 0.05 for the inhibitory effect of Desip within LPS or IFNγ; ^ϕ p ≤ 0.05 main effect of Desip.

Figure 5. Attenuation of IFNγ expression by Desip

After treatment of mice i.p. with either LPS or Dex (in vivo, A, B and C) or cells with IFNγ (ex vivo, D and E) with and without Desip, expression of IFNγ was quantified in A) Hippocampi, B) Microglia (CD11b⁺ enriched cells), C) PBMCs, D) PBMC^−T, and E) T cells (CD90⁺ enriched cells). * p ≤ 0.05 for IFNγ induction by LPS; ^δ p ≤ 0.05 for the inhibitory effect of Desip within LPS; ^ϕ p ≤ 0.05 main effect of Desip.

Figure 6. Desip decreased IDO1 and IDO2 expression by human PBMCs

After human PBMCs were treated with IFNγ with or without Desip, expression of A) IDO1-001, B) IDO1-005 and C) IDO2-201 was quantified. * p ≤ 0.05 for induction by IFNγ; ^δ p ≤ 0.05 for the inhibitory effect of Desip within IFNγ.

Figure 7. Proposed model for Desip blocking Kynurenine Pathway activation in brain and periphery

Multiple studies have linked inflammation and stress with major depression in humans and depressive-like behaviors in rodent models. Inflammation or stress increase peripheral pro-inflammatory cytokine release (ex. IFNγ) into the circulation, which increases Ido expression by PBMCs and most tissues. Either peripheral IFNγ or infiltrating immune cells capable of producing IFNγ enter the brain to increase Ido expression by both microglia and astrocytes. We’ve extended this area of research by confirming that Desip can act in vivo by 1) blocking peripheral IFNγ expression as a mechanism to decrease Ido expression by PBMCs, microglia and astrocytes. 2) Ex vivo, IFNγ increases Ido expression by PBMCs (both human and mouse) and synergizes with multiple factors (i.e. Dex or Gal-9) to increase the DOs within the brain. As such, we now show that Desip is able to directly block IFNγ-dependent Ido induction in both peripheral cells (human and murine PBMCs) and within OHSCs. The later findings suggest that 3) Desip would also act to directly block DO expression in the periphery and brain by attenuating IFNγ activity. Kynurenine Pathway activation (via the DOs) results in production of additional downstream kynurenine metabolites (kynurenines: quinolinic acid: QuinA and kynurenic acid: KynA) that directly alter behavior and immune activity. Thus, Desip is able to modulate DO-dependent behavior and immune responses indirectly by regulating IFNγ expression and directly by blocking IFNγ activity, both in the brain and periphery.