The present study provides a comprehensive understanding of the immune system in patients with MDE across different stages of the disease, comparing it with HC. The most novel findings include increased monocytosis with an increment of intermediate and nonclassical monocyte subsets at the expense of classical monocytes, indicating a transitional activation of the monocytic population. We also observed a notable augmentation in the activation of CD4 T lymphocytes and elevated exhaustion markers in patients with active MDE compared to HC. Furthermore, there was a significant increase in the frequency of CD4+CD25+FOXP3+ Tregs in both active and remitted MDE patients compared to HC, which could be reflecting a compensatory anti-inflammatory immune system response. Finally, we observed increased levels of soluble markers of neuroinflammation, such as sSTREM2 and IL-17. Machine learning techniques identified a panel of biomarkers that can discriminate between patients with MDE and HC with an overall classification accuracy of 83.8%. Most of these biomarkers are related to immune cell activation. Finally, cluster analysis suggests three distinct clusters unrelated to the clinical expression of the disease.
Since the 1990s, it has been established that depression is associated with increased white blood cell count and monocytes 36, which led to the formulation of the monocytes and lymphocytes hypothesis in MDD 37. This hypothesis suggests that alterations in these immune cells play a role in the pathophysiology of depression. A recently published meta-analysis also supported this by demonstrating an overall increase in the total number of monocytes in depressed individuals (seven studies; SMD = 0.60; 95% CI, 0.19–1.01; P < 0.01; I2 = 66%)20. Consistent with these findings, our study observed a significant monocytosis in patients experiencing an active major depressive episode (MDE). These patients exhibited a clear elevation in the median percentage and absolute number of monocytes compared to HC. These results suggest an abnormal hematopoietic response in individuals with depression, specifically an enhanced production of monocytes. We further investigated the proportion of different subtypes of circulating monocytes (classical, intermediate, and nonclassical) using flow cytometry. The results indicate an expansion for the nonclassical and intermediate monocytes and a reduced percentage of classical monocytes in patients with MDE compared to HC. Even though we did not find statistical differences comparing active vs. remitted MDE, more pronounced changes were observed in active conditions. These results indicate that patients with MDE are characterized by a proinflammatory status and enhanced transition to intermediate and nonclassical subsets, as predicted by the monocyte transitional model of Patel et al. 38.
We have been one of the first groups to describe changes in the percentage and activation status of the three circulating monocyte subtypes in patients with severe MDD 17. Herein, we reinforce and expand the aforementioned findings by utilizing a distinct sample of patients, ensuring sex and age matching, and employing a simplified approach of directly measuring immune parameters in a small blood sample. This technical approach holds promise for potential translation into clinical practice, as it offers a convenient and feasible screening method.
Dysregulation of the adaptive immune system in MDE patients has been suggested, with decreased numbers of circulating T cells, an increase in the ratio of CD4+ relative to CD8+ T cells, and some immunosuppression features 15. Our finding also suggests a dysregulation of the T cell compartment in patients with mood disorders during MDE. While there were no significant differences in the absolute numbers of total lymphocytes, CD4 T cells, or the ratio of CD4 to CD8 T cells between MDE patients and HC, a notable increment in the activation status of CD4+CD69+, and exhausted CD4+PD1+ and CD4+LAG3+ T lymphocytes were observed in patients with active MDE compared to healthy controls. These results indicate that a CD4 lymphocyte activation process is ongoing in patients with MDE, associated with potential exhaustion of this compartment in the pathogenesis of depression. In this sense, it has been demonstrated that memory CD4+ T cells are abundant in adult humans, and its activation does not necessary depend on the encounter with the antigen 39, 40. The increased markers of cellular activation and exhaustion in CD4 T cells among patients with MDE is a novel concept that may explain the high comorbidity of these patients with non-psychiatric medical conditions 41, particularly autoimmune diseases 42. Furthermore, this concept aligns with recent studies that have demonstrated a higher degree of premature T cell aging 43.
Furthermore, there was a significant increase in the frequency of CD4+CD25+FOXP3+ Tregs in both active and remitted MDE patients compared to healthy controls. Tregs play a crucial role in maintaining immune homeostasis and suppressing excessive immune responses. The observed increase in Tregs suggests a compensatory mechanism by the immune system to counterbalance potential proinflammatory processes and regulate immune activity in MDE, as was suggested by the compensatory immune-regulatory reflex system (CIRS) concept 44. CIRS is involved in MDD and BD by regulating the primary immune-inflammatory response, thereby contributing to spontaneous and antidepressant-promoted recovery from the acute phase of illness 44. The simultaneously increased levels of both the pro-and antiinflammatory cytokines are reported in the brain of MDD patients; this indicates activity of both the IRS and CIRS in MDD. Speculation is rife that the disrupted IRS-CIRS elements might determine the onset, episodes, neuroprogressive processes, treatment response, and recovery of patients with MDD 45.
Patients with MDE exhibit the typical features of an ongoing inflammatory response, including increased expression of proinflammatory cytokines and their receptors, elevated acute phase reactive proteins levels, and adhesion molecules in peripheral blood, cerebrospinal fluid, and brain 14, 46. Nonetheless, soluble proinflammatory biomarkers can be increased in depression as well as numerous inflammatory pathologies; there are clear differences in the magnitude or concentration levels of these factors comparing acute inflammation in response to infections compared to the low-grade systemic inflammation as reflected by chronic conditions. High sensitivity CRP (hs-CRP) is a marker of acute phase response, but it has been used extensively as a measure of low-grade inflammation in psychiatric 47 and physical conditions 48, 49. Meta-analyses of cross-sectional studies confirm that mean concentrations of circulating hs-CRP and inflammatory cytokines such as interleukin 6 (IL-6) are higher in patients with acute depression than controls 10, 12, 50, 51. Our findings also show increased hs-CRP levels in patients with active MDE compared to HC, but interestingly, we found that individuals with MDE in remission still exhibited elevated hs-CRP levels compared to HC, suggesting that residual inflammation may persist even after symptom improvement. Similarly, our study revealed elevated levels of IL-6 in individuals with active MDE compared to those with MDE in remission. This indicates that IL-6 may serve as a marker of ongoing inflammation during the active phase of the disease. Altogether, these findings underscore the role of inflammation in MDE.
In addition to the two well-described nonspecific markers of inflammation, our study found a potential novel, more specific biomarker of neuroinflammation for MDE, the soluble triggering receptor expressed on myeloid cell 2 (sTREM2). This is a protein receptor largely expressed in microglial cells in the brain. It plays a crucial role in regulating microglial function and modulating the immune response in the central nervous system (CNS). sTREM2 refers to the soluble form of this receptor, which can be measured in the cerebrospinal fluid (CSF) or peripheral blood. Changes in sTREM2 levels have been associated with the activation of microglia in neurodegenerative and neuroinflammatory diseases. A recent metanalysis of 22 observational studies, which included 5716 participants, comparing individuals with Alzheimer’s vs. controls, showed a significant increase in CSF of sTREM2 level (standardized mean difference [SMD]: 0.41, 95% confidence intervals [CI]: 0.24, 0.58, p < 0.001) 52. This marker also increased in conditions of neuroinflammation, such as angiitis of the CNS 53 and amyotrophic lateral sclerosis (ALS) 54. The role of sTREM2 in MDE has not been studied, and the present study is the first one reporting increased sTREM2 levels in the plasma of patients with active MDE compared with HC. Further research is needed to fully understand the role of sTREM2 in MDE and its role as a diagnostic or therapeutic target.
We also identified an increase in IL-17 levels, constituting an exciting soluble marker due to its association with autoimmune pathologies 55. IL-17 is considered a signature cytokine of CD4+ T helper 17 (Th17) cells; however, it can also be produced by different cell types, including CD8+ T cells, natural Th17 cells, innate lymphoid cells (ILCs), γδ T cells, natural killer (NKT) cells, and neutrophils. Animal studies have indicated that inflammatory Th17 cells contribute to depression-like behavior 56. Interestingly, studies have demonstrated that the administration of anti-interleukin-17A (IL-17A) antibodies can lead to a reduction in depressive symptoms in mice 57. In humans, there have been few studies that have examined the role of IL-17 in depression. One of the most recent studies found an increase in IL-6 and IL-17 levels in patients with a first depressive episode compared to controls 58. In the same study, treatment with antidepressants decreased plasma levels of IL-6 and IL-17, although the latter remained elevated compared to controls 58.
Furthermore, this study also revealed that the HAMD score exhibited a moderate correlation with IL-6 and a strong correlation with IL-17 58. This study suggests that autoimmunity may play a role in the etiology or pathogenesis of depression. Our results support this idea, as we have observed significant elevated levels of IL-17 in patients with remitted MDE and a trend (p = 0.06) in active MDE compared to HC. The fact that this cytokine is elevated in remitted patients may indicate that inflammation can have a chronic role in depression beyond periods of active illness, which could be highly relevant for those cases of mood disorders characterized by neuroprogression.
The two Boruta analyses revealed different markers for discriminating between MDE patients and HC. First, comparing between MDE (active and remitted) and HC, the selected markers, including lymphocytes percentage, monocytes absolute count, classical monocytes, non-classical monocytes, intermediate monocytes, hs-CRP, MCP1, CD4+PD1+, CD4+LAG3+, and CD4+CD25+FOXP3+Tregs, demonstrated a significant discriminative potential. When the Random Forest model was applied to an independent test dataset, an impressive overall classification accuracy of 83.8% was achieved. Secondly, when considering the classification of active MDE, remitted MDE, and healthy controls, the Boruta algorithm identified important markers, including classical monocytes, non-classical monocytes, intermediate monocytes, monocytes absolute count, ESR, hs-CRP, CD4+CD69+, CD4+LAG3+, and CD4+CD25+FOXP3+ Tregs. Subsequently, the Random Forest model, trained using these selected variables, demonstrated an overall classification accuracy of 70%. Altogether, these findings highlight the potential of utilizing immune cell biomarkers to differentiate MDE patients from HC, as well as distinguish between different states of the disorder. The discriminatory power exhibited by the selected markers in the Random Forest model suggests their relevance in understanding the underlying mechanisms and aiding in the diagnostic process of MDE. Further research and validation studies are warranted to explore these markers' clinical utility and generalizability in larger and more diverse patient populations.
The clustering analysis revealed the presence of three distinct clusters based on immunological profiles among patients with MDE. First, these results suggest that patients with MDE, regardless of whether they are experiencing or have remitted from a MDE, exhibit signs of an inflammatory state. Cluster 1 is characterized by the highest number of leukocytes, mainly given by the increment in lymphocyte count. Nonetheless, this cluster showed the lowest proinflammatory cytokines levels, probably due to a different state of the inflammation process. Cluster 3 displayed the most robust inflammatory pattern, with high levels of TNFα, CX3CL-1, IL-12p70, IL-17A, IL-23, and IL-33, associated with the highest level of IL-10, as well as increased medians for b-NGF and the lowest level for BDNF. This profile is also associated with the highest absolute number and percentage of circulating monocytes as well as the lowest absolute number and percentage of circulating lymphocytes, denoting an active inflammatory process. Noteworthy, a lower percentage of individuals in Cluster 3 were receiving pharmacological treatment, indicating a potential association between the immunological profile and treatment status. Cluster 2 has some cardinal signs of more acute inflammation as the elevated levels of MCP1, which precede the monocytosis, but also increased levels of some proinflammatory cytokines such as IL-1β, IFN-γ, and IL-8. Similarly, the absolute number of monocytes is closer to a HC value, as well as the percentage of lymphocytes, suggesting as possible initiation of the inflammatory process.
Based on these results, the following questions emerge: Do the observed clusters represent distinct stages of the same underlying process, or do they indicate different inflammatory pathways that converge to produce a common phenotype? The lack of significant differences in the distribution of active or remitted MDE and the severity of depressive symptoms across the clusters suggests that they may not signify distinct stages of the same illness. However, it cannot be ruled out that these clusters represent different trajectories of the same disease, considering the limitations of our cross-sectional study design. Definitive answers to these questions will require future studies with a longitudinal design, which will provide further insights into this matter.
Our study demonstrates several noteworthy strengths. Firstly, we assessed changes in cellular levels of the monocyte compartment and T cells, considering a specific plasma cytokine milieu. This allowed us to establish a distinct profile for MDE patients, defining subtypes of the condition. This approach fills a critical gap in the literature, as this area has received inadequate attention thus far. Secondly, our standardized methodology employed three cocktails of antibodies with a minimal blood sample volume of only 100 µl each. This approach allowed us to accurately measure the proportion and activation of monocytes, the proportion of CD4 to CD8 lymphocytes, and Tregs, as well as the activation and exhaustion of T cells, utilizing direct blood staining. Such an approach holds promise for rapid translation into clinical practice. Thirdly, our study is a multicenter investigation that carefully matched participants based on age and sex, two variables known to significantly influence the immune system. By controlling for these factors, we strengthened the validity and generalizability of our results. Finally, our rigorous patient selection process excluded individuals with known causes of inflammation or immune system activation, ensuring the focus remained on the specific MDE condition. Additionally, we included patients at various stages of the disease, a novel aspect not previously explored in MDE immunotyping studies.
Some limitations of the study need to be acknowledged. Firstly, we measured the immune cell profile in peripheral blood, which may not fully reflect the immune activity in the central nervous system (CNS). However, evidence suggests that inflammatory factors originating in the blood can reach the CNS through various pathways, including passive or active transport across the blood-brain barrier, immune cell transmigration, and vagal nerve signaling. While peripheral blood analysis provides valuable insights, it is important to recognize the potential disparities between peripheral and CNS immune responses. Another significant limitation is that most participants received psychopharmacological treatment at the time of inclusion. It is well-known that many psychotropic medications can impact the immune system, potentially confounding the interpretation of immunological findings. Furthermore, our study focused exclusively on individuals with MDE, which could be a limitation as MDD and BP may exhibit distinct immune profiles. However, in clinical practice, MDE is the most commonly encountered presentation, and there are currently no precise indicators that reliably classify between these two groups. Therefore, we included MDD and BP patients to explore whether immunological markers could provide insights into their shared pathophysiology.
Despite these limitations, our study provides valuable insights into the immunological aspects of MDE providing a global view of the phenomenon, analyzing both the humoral and the innate and adaptive cellular components. Further research is needed to fully understand the implications of these immune alterations in a longitudinal process to pave the way for potential advancements in clinical practice.