Classification of T lymphocyte motility behaviors using a machine learning approach.

T lymphocytes migrate into organs and interact with local cells to perform their functions. How human T lymphocytes communicate with organ-specific cells and participate in pathobiological processes remains unresolved. Brain infiltration of T lymphocytes is associated with multiple neurological disorders. Thus, to characterize the behavior of human T lymphocytes reaching the human brain, we performed time-lapse microscopy on human CD8+ T lymphocytes co-cultured with either primary human astrocytes or neurons. Using traditional manual and visual assessment of microscopy data, we identified distinct CD8+ T lymphocyte motility behaviors. However, such characterization is time and labor-intensive. In this work, we trained and validated a machine-learning model for the automated classification of behaviors of CD8+ T lymphocytes interacting with astrocytes and neurons. A balanced random forest was trained for the binary classification of established classes of cell behaviors (synapse vs. kinapse) as well as visually identified behaviors (scanning, dancing, and poking). Feature selection was performed during 3-fold cross-validation using the minimum redundancy maximum relevance algorithm. Results show promising performances when tested on a held-out dataset of CD8+ T lymphocytes interacting with astrocytes with a new experimenter and a held-out independent dataset of CD8+ T lymphocytes interacting with neurons. When tested on the independent CD8+ T cell-neuron dataset, the final model achieved a binary classification accuracy of 0.82 and a 3-class accuracy of 0.79. This novel automated classification approach could significantly reduce the time required to label cell motility behaviors while facilitating the identification of interactions of T lymphocytes with multiple cell types.

GM-CSF distinctly impacts human monocytes and macrophages via ERK1/2-dependent pathways.

Human monocytes and macrophages are two major myeloid cell subsets with similar and distinct functions in tissue homeostasis and immune responses. GM-CSF plays a fundamental role in myeloid cell differentiation and activation. Hence, we compared the effects of GM-CSF on the expression of several immune mediators by human monocytes and monocyte-derived macrophages obtained from healthy donors. We report that GM-CSF similarly elevated the expression of CD80 and ICAM-1 and reduced HLA-DR levels on both myeloid cell subsets. However, GM-CSF increased the percentage of macrophages expressing surface IL-15 but reduced the proportion of monocytes carrying surface IL-15. Moreover, GM-CSF significantly increased the secretion of IL-4, IL-6, TNF, CXCL10, and IL-27 by macrophages while reducing the secretion of IL-4 and CXCL10 by monocytes. We show that GM-CSF triggered ERK1/2, STAT3, STAT5, and SAPK/JNK pathways in both myeloid subsets. Using a pharmacological inhibitor (U0126) preventing ERK phosphorylation, we demonstrated that this pathway was involved in both the GM-CSF-induced increase and decrease of the percentage of IL-15+ macrophages and monocytes, respectively. Moreover, ERK1/2 contributed to GM-CSF-triggered secretion of IL-4, IL-6, TNF, IL-27 and CXCL10 by macrophages. However, the ERK1/2 pathway exhibited different roles in monocytes and macrophages for the GM-CSF-mediated impact on surface makers (CD80, HLA-DR, and ICAM-1). Our data demonstrate that GM-CSF stimulation induces differential responses by human monocytes and monocyte-derived macrophages and that some but not all of these effects are ERK-dependent.

Specific alterations in NKG2D+ T lymphocytes in relapsing-remitting and progressive multiple sclerosis patients.

BACKGROUND: T lymphocytes exhibit numerous alterations in relapsing-remitting (RRMS), secondary progressive (SPMS), and primary progressive multiple sclerosis (PPMS). The NKG2D pathway has been involved in MS pathology. NKG2D is a co-activating receptor on subsets of CD4+ and most CD8+ T lymphocytes. The ligands of NKG2D are expressed at low levels in normal tissues but are elevated in MS postmortem brain tissues compared with controls. Whether the NKG2D pathway shows specific changes in different forms of MS remains unclear. METHODS: We performed unsupervised and supervised flow cytometry analysis to characterize peripheral blood T lymphocytes from RRMS, SPMS, and PPMS patients and healthy controls (HC). We used an in vitro microscopy approach to assess the role of NKG2D in the interactions between human CD8+T lymphocytes and human astrocytes. RESULTS: Specific CD8+, CD4+, and CD4-CD8- T cell populations exhibited altered frequency in MS patients' subgroups. The proportion of NKG2D+ T lymphocytes declined with age in PPMS patients but not in RRMS and HC. This reduced percentage of NKG2D+ cells was due to lower abundance of γδ and αß CD4-CD8- T lymphocytes in PPMS patients. NKG2D+ T lymphocytes were significantly less abundant in RRMS than in HC; this was caused by a decreased frequency of CD4-CD8- and CD8+ T lymphocytes and was not linked to age. Blocking NKG2D increased the motility of CD8+ T lymphocytes co-cultured with astrocytes expressing NKG2D ligand. Moreover, preventing NKG2D from interacting with its ligands increased the proportion of CD8+ T lymphocytes exhibiting a kinapse-like behavior characterized by short-term interaction while reducing those displaying a long-lasting synapse-like behavior. These results support that NKG2D participates in the establishment of long-term interactions between activated CD8+ T lymphocytes and astrocytes. CONCLUSION: Our data demonstrate specific alterations in NKG2D+ T lymphocytes in MS patients' subgroups and suggest that NKG2D contributes to the interactions between human CD8+ T lymphocytes and human astrocytes.

Granulocyte-macrophage colony-stimulating factor-stimulated human macrophages demonstrate enhanced functions contributing to T-cell activation.

Granulocyte-macrophage colony-stimulating factor (GM-CSF) has been implicated in numerous chronic inflammatory diseases, including multiple sclerosis (MS). GM-CSF impacts multiple properties and functions of myeloid cells via species-specific mechanisms. Therefore, we assessed the effect of GM-CSF on different human myeloid cell populations found in MS lesions: monocyte-derived macrophages (MDMs) and microglia. We previously reported a greater number of interleukin (IL)-15+ myeloid cells in the brain of patients with MS than in controls. Therefore, we investigated whether GM-CSF exerts its deleterious effects in MS by increasing IL-15 expression on myeloid cells. We found that GM-CSF increased the proportion of IL-15+ cells and/or IL-15 levels on nonpolarized, M1-polarized and M2-polarized MDMs from healthy donors and patients with MS. GM-CSF also increased IL-15 levels on human adult microglia. When cocultured with GM-CSF-stimulated MDMs, activated autologous CD8+ T lymphocytes secreted and expressed significantly higher levels of effector molecules (e.g. interferon-γ and GM-CSF) compared with cocultures with unstimulated MDMs. However, neutralizing IL-15 did not attenuate enhanced effector molecule expression on CD8+ T lymphocytes triggered by GM-CSF-stimulated MDMs. We showed that GM-CSF stimulation of MDMs increased their expression of CD80 and ICAM-1 and their secretion of IL-6, IL-27 and tumor necrosis factor. These molecules could participate in boosting the effector properties of CD8+ T lymphocytes independently of IL-15. By contrast, GM-CSF did not alter CD80, IL-27, tumor necrosis factor and chemokine (C-X-C motif) ligand 10 expression/secretion by human microglia. Therefore, our results underline the distinct impact of GM-CSF on human myeloid cells abundantly present in MS lesions.

IL-27 shapes the immune properties of human astrocytes and their impact on encountered human T lymphocytes.

BACKGROUND: Interleukin-27 (IL-27) can trigger both pro- and anti-inflammatory responses. This cytokine is elevated in the central nervous system (CNS) of multiple sclerosis (MS) patients, but how it influences neuroinflammatory processes remains unclear. As astrocytes express the receptor for IL-27, we sought to determine how these glial cells respond to this cytokine and whether such exposure alters their interactions with infiltrating activated T lymphocytes. To determine whether inflammation shapes the impact of IL-27, we compared the effects of this cytokine in non-inflamed and inflamed conditions induced by an IL-1ß exposure. MAIN BODY: Transcriptomic analysis of IL-27-exposed human astrocytes showed an upregulation of multiple immune genes. Human astrocytes increased the secretion of chemokines (CXCL9, CXCL10, and CXCL11) and the surface expression of proteins (PD-L1, HLA-E, and ICAM-1) following IL-27 exposure. To assess whether exposure of astrocytes to IL-27 influences the profile of activated T lymphocytes infiltrating the CNS, we used an astrocyte/T lymphocyte co-culture model. Activated human CD4+ or CD8+ T lymphocytes were co-cultured with astrocytes that have been either untreated or pre-exposed to IL­27 or IL-1ß. After 24 h, we analyzed T lymphocytes by flow cytometry for transcription factors and immune molecules. The contact with IL-27-exposed astrocytes increased the percentages of T-bet, Eomes, CD95, IL-18Rα, ICAM-1, and PD-L1 expressing CD4+ and CD8+ T lymphocytes and reduced the proportion of CXCR3-positive CD8+ T lymphocytes. Human CD8+ T lymphocytes co-cultured with human IL-27-treated astrocytes exhibited higher motility than when in contact with untreated astrocytes. These results suggested a preponderance of kinapse-like over synapse-like interactions between CD8+ T lymphocytes and IL-27-treated astrocytes. Finally, CD8+ T lymphocytes from MS patients showed higher motility in contact with IL-27-exposed astrocytes compared to healthy donors' cells. CONCLUSION: Our results establish that IL-27 alters the immune functions of human astrocytes and shapes the profile and motility of encountered T lymphocytes, especially CD8+ T lymphocytes from MS patients.

Stress Signal ULBP4, an NKG2D Ligand, Is Upregulated in Multiple Sclerosis and Shapes CD8+ T-Cell Behaviors.

BACKGROUND AND OBJECTIVES: We posit the involvement of the natural killer group 2D (NKG2D) pathway in multiple sclerosis (MS) pathology via the presence of specific NKG2D ligands (NKG2DLs). We aim to evaluate the expression of NKG2DLs in the CNS and CSF of patients with MS and to identify cellular stressors inducing the expression of UL16-binding protein 4 (ULBP4), the only detectable NKG2DL. Finally, we evaluate the impact of ULBP4 on functions such as cytokine production and motility by CD8+ T lymphocytes, a subset largely expressing NKG2D, the cognate receptor. METHODS: Human postmortem brain samples and CSF from patients with MS and controls were used to evaluate NKG2DL expression. In vitro assays using primary cultures of human astrocytes and neurons were performed to identify stressors inducing ULBP4 expression. Human CD8+ T lymphocytes from MS donors and age/sex-matched healthy controls were isolated to evaluate the functional impact of soluble ULBP4. RESULTS: We detected mRNA coding for the 8 identified human NKG2DLs in brain samples from patients with MS and controls, but only ULBP4 protein expression was detectable by Western blot. ULBP4 levels were greater in patients with MS, particularly in active and chronic active lesions and normal-appearing white matter, compared with normal-appearing gray matter from MS donors and white and gray matter from controls. Soluble ULBP4 was also detected in CSF of patients with MS and controls, but a smaller shed/soluble form of 25 kDa was significantly elevated in CSF from female patients with MS compared with controls and male patients with MS. Our data indicate that soluble ULBP4 affects various functions of CD8+ T lymphocytes. First, it enhanced the production of the proinflammatory cytokines GM-CSF and interferon-γ (IFNγ). Second, it increased CD8+ T lymphocyte motility and favored a kinapse-like behavior when cultured in the presence of human astrocytes. CD8+ T lymphocytes from patients with MS were especially altered by the presence of soluble ULBP4 compared with healthy controls. DISCUSSION: Our study provides new evidence for the involvement of NKG2D and its ligand ULBP4 in MS pathology. Our results point to ULBP4 as a viable target to specifically block 1 component of the NKG2D pathway without altering immune surveillance involving other NKG2DL.

Capturing T Lymphocytes' Dynamic Interactions With Human Neural Cells Using Time-Lapse Microscopy.

To fully perform their functions, T lymphocytes migrate within organs' parenchyma and interact with local cells. Infiltration of T lymphocytes within the central nervous system (CNS) is associated with numerous neurodegenerative disorders. Nevertheless, how these immune cells communicate and respond to neural cells remains unresolved. To investigate the behavior of T lymphocytes that reach the CNS, we have established an in vitro co-culture model and analyzed the spatiotemporal interactions between human activated CD8+ T lymphocytes and primary human astrocytes and neurons using time-lapse microscopy. By combining multiple variables extracted from individual CD8+ T cell tracking, we show that CD8+ T lymphocytes adopt a more motile and exploratory behavior upon interacting with astrocytes than with neurons. Pretreatment of astrocytes or neurons with IL-1ß to mimic in vivo inflammation significantly increases CD8+ T lymphocyte motility. Using visual interpretation and analysis of numerical variables extracted from CD8+ T cell tracking, we identified four distinct CD8+ T lymphocyte behaviors: scanning, dancing, poking and round. IL-1ß-pretreatment significantly increases the proportion of scanning CD8+ T lymphocytes, which are characterized by active exploration, and reduces the proportion of round CD8+ T lymphocytes, which are less active. Blocking MHC class I on astrocytes significantly diminishes the proportion of poking CD8+ T lymphocytes, which exhibit synapse-like interactions. Lastly, our co-culture time-lapse model is easily adaptable and sufficiently sensitive and powerful to characterize and quantify spatiotemporal interactions between human T lymphocytes and primary human cells in different conditions while preserving viability of fragile cells such as neurons and astrocytes.