Blog

By Dejan Mamula, Shervin Khosousi, Yachao He, Vesna Lazarevic, and Per Svenningsson

Excerpt from the article published in NPJ Parkinsons Dis. 8, 171, 10 December 2022. DOI: https://doi.org/10.1038/s41531-022-00438-0

Editor’s Highlights

• Growing evidence suggests that adaptive immunity-related neuroinflammation contributes to Parkinson’s disease (PD).
• Lymphocyte migration is shown to be regulated by mitochondria positioning at the posterior part of the cell, known as the trailing edge or uropod.
• PD-associated transcriptomic signatures, for example, oxidative stress and mitophagy, was found in CD4⁺ memory T cells suggesting their association with disease pathogenesis.
• CD4⁺ T cells derived from PD patients have altered migration potential along with differentially increased p11 levels in CD4⁺ T cell subsets across Th subsets, impaired mitochondrial positioning within the cell, and reduced mitochondrial functionality.
• Migration and mitochondrial alterations we found are not affected by dopaminergic medication, nor change over the course of the disease. This is in accordance with other functional T cell studies that did not find any effect of dopaminergic medication on T cell function in PD.

Abstract

Dysfunctions in the immune system appear implicated in both disease onset and progression of Parkinson’s disease (PD). Neurodegeneration observed in the brain of PD patients has been associated with neuroinflammation that is linked to alterations in peripheral adaptive immunity, where CD4⁺ T cells are key players. In the present study, we elucidated the immunological aspect of PD by employing a wide range of cellular assays, immunocytochemistry and flow cytometry to examine CD4⁺ T cells. We particularly investigated the role of CD4⁺ T cell migration in the proper functioning of the adaptive immune system. Our data reveal the altered migration potential of CD4⁺ T cells derived from PD patients, along with impaired mitochondrial positioning within the cell and reduced mitochondrial functionality. In addition, a cross-sectional study of p11 levels in CD4⁺ T cell subsets showed a differentially increased level of p11 in Th1, Th2 and Th17 populations. Taken together, these results demonstrate major impairments in the functionality of peripheral CD4⁺T cells in PD.

Introduction

Parkinson’s disease (PD) is the second most common neurodegenerative disorder, with up to 1% of prevalence among people older than 60¹. PD is clinically diagnosed based on the presence of bradykinesia, rigidity and tremor. PD patients also suffer from a multitude of non-motor symptoms, including sleep disorders, depression and cognitive impairments, largely contributing to poor quality of life². The major neuropathological hallmarks of PD are the loss of dopaminergic neurons in substantia nigra and the formation of neuropathological inclusions, named Lewy bodies and Lewy neurites³. The therapeutic approaches towards PD are symptomatic replacement of dopaminergic neurotransmission, but they do not slow down PD progression⁴.

Important progress towards understanding etiology and progression of PD, and other neurodegenerative disorders, has been made by widening the focus of research to the critical involvement of the immune system⁵. For example, PD patients have been shown to have increased microgliosis, elevated levels of peripheral proinflammatory cytokines, higher risk for auto-immune diseases, and aberrant adaptive immune responses⁶. Also, it has been shown that particular major histocompatibility complex (MHC) alleles linked with PD, display α- synuclein derived peptide epitopes, thereby inducing T cell-mediated immunity in PD patients⁷. The adaptive immune system includes different cell types with the distinctive role of CD4⁺ T cells which is usually observed through the interaction with other immune cells in order to cope with environmental challenges^8,9,10. PD-associated transcriptomic signatures, for example, oxidative stress and mitophagy, was found in CD4⁺ memory T cells suggesting their association with disease pathogenesis¹¹.

During random migration, CD4⁺ T cells need to optimally scan their environment to properly exert their function and this process is critically influenced by velocity and directionality¹². The velocity and directionality of CD4⁺ T cells are regulated by specific molecular signaling to GTPases and cofilin¹³. Lymphocyte migration is highly dependent on the proper positioning of mitochondria at the uropod with their localized ATP production^14,15,16. Mitochondria are indispensable for facilitating chemotactic lymphocyte movement and mitochondrial reactive oxygen species (mROS) play an important role in proper T cell activation and proliferation¹⁷. It has been shown that mROS amplifies T- cell receptor (TCR) proximal signaling and that transient formation of mROS is important for the activation of certain transcription factors^18,19,20,21. In the case of mitochondrial DNA deficit, mROS are reduced, subsequently leading to a decrease in the expression of the cytokines IL-2 and IL-4²⁰. It has been shown that ROS promotes polarization of anti-CD3/CD28 primed human T cells into Th2 phenotype²². Interestingly, PD patients demonstrate Th1 bias²³. However, the involvement of ROS and underlying cellular mechanisms of CD4⁺ T cells in disease progression are poorly understood.

A long latency between the first pathological changes and clinical diagnosis of PD requires finding reliable molecular and cellular biomarkers. We previously showed that p11 has an elevated expression in distinctive leukocyte populations in PD patients²⁴. p11 (S100A10, annexin II light chain) together with Annexin A2 constitute a heterotetrameric protein complex localized preferentially at cell membranes which coordinates multiple cellular processes²⁵. Annexin A2 regulates actin bundling and couples cytoskeleton with plasma membrane²⁶. The neuronal function of p11 has been implicated in PD and L-dopa responses²⁷ but also in the pathophysiology of depression and antidepressant therapies²⁸. Increased levels of p11 in specific subsets of peripheral leukocytes isolated from PD patients, introduced p11 as a potential biomarker, primarily in PD patients with associated depression²⁴.

The aim of the present study was to further elucidate immunological aspects of PD, focusing primarily on CD4⁺ T cells. To this end, we evaluated the migratory phenotype of PD-derived CD4⁺ T cells and underlying cellular events. Additionally, we carried out a cross-sectional study of p11 expression in different CD4⁺ Th subsets in attempt to come closer to the application of p11 as a potential inflammation-related PD biomarker.

Results

Subjects

The study included 30 healthy controls and 49 PD patients from neurological clinics in Region Stockholm (Table 1 and Supplementary Tables 1, 2). All research participants gave their consents according to the Declaration of Helsinki further approved by the regional ethical committees. Clinical diagnostic criteria for PD²⁹ were fulfilled by all participants. PD severity was scored using both Unified Parkinson’s Disease Rating Scale (UPDRS) and Hoehn and Yahr (H&Y) scale. Montreal Cognitive Assessment (MoCA) scores was used for cognitive assessment, Montgomery Åsberg Depression Rating Scale (MADRS) was used to assess depression, and levodopa equivalent daily dose (LEDD) for a total amount of dopaminergic medication³⁰.

BMCs
Group	Sex (M/F) (%)	Age mean ± SD	Disease duration (years) median (range)	H&Y score median (range)	MDS- UPDRS III median (IQR)	MoCA median (IQR)	MADRS median (IQR)	LEDD median (IQR)
Control (n = 30)	43/57	64,4 ± 9,5	n/a	n/a	n/a	n/a	n/a	n/a
PD (n = 49)	59/41	67,3 ± 10,2	4,3 (0,0–21,1)	2 (1–5)	36 (25,5–43,5)	26 (24,5–28)	8 (4–13)	610 (160–937,5)

Control subjects were selected among healthy volunteers and were age- and sex-matched to PD patients (Table 1). Subjects with systemic inflammatory disorders or anti-inflammatory therapies were excluded.

Migratory deficit of PD-derived CD4 + T cells

Several lines of evidence propose that immune system dysregulations, not least in Th cells, play an important role in PD pathogenesis^31,32. Taking into account that efficient immune response highly depends on the directional migration of lymphocytes, we focused our study on the assessment of the migration potential of CD4⁺ T cells. Hence, we employed a transwell assay to measure the ability of CD3/CD28-activated CD4⁺ T cells to migrate in in vitro conditions. For this experiment, cells from 11 PD and 11 age and sex-matched controls were used.

As shown in Fig. 1, both PD and healthy control (hCTRL) derived activated CD4 + T cells readily migrated to the lower chamber with and without the addition of a chemoattractant (stromal cell-derived factor-1α, SDF-1α). However, the percentage of input of PD-derived CD4⁺ T cells that migrated to the lower chamber was significantly lower compared to hCTRL in both conditions suggesting a migratory deficit in these cells (Interaction, F(1, 52) = 0.4887 p = 0.4876; SDF-1α, F = 19, 32 p < 0.0001; PD, F = 34, 13 p < 0.0001).

To further explore the observed phenotype we proceeded with a 2D migration assay on ICAM1 substrate (Fig. 2a). This allowed us to characterize not only migration potential of cells but also migratory parameters such as velocity and directionality. As shown in Fig. 2, our data revealed both reduced velocity (hCTRL 11.69 ± 0.23 µm/min vs PD 9.94 ± 0,22 µm/min, p < 0,0001; Fig. 2b) and directionality (hCTRL 0.458 ± 0.012 vs PD 0.355 ± 0.010 p < 0,0001; Fig. 2c) of PD-derived CD4⁺ T cells. Consequently, that also led to a reduction of displacement (hCTRL 230.1 ± 4.54 µm vs PD 194.9 ± 4.36 µm, p < 0.0001; Fig. 2d), confirming altered migratory properties of PD-derived CD4⁺ T cells.

We finally investigated whether the velocity and directionality of the PD-derived CD4 + cells were affected by disease stage, disease duration or levodopa medication, but found no significant correlations (Supplementary Fig. 1, Spearman’s rho p > 0.05).

Found migration defects can be a consequence of disturbed actin dynamics. We used a specific fluorescent dye to label F-actin filaments of both PD and control CD3/CD28-activated CD4⁺ T cells. Confocal imaging showed no difference in actin turnover between PD and the control group, suggesting other explanations for the observed migration defects (Supplementary Fig. 2).

Alteration in mitochondrial functionality

Lymphocyte migration is shown to be regulated by mitochondria positioning at the posterior part of the cell, known as the trailing edge or uropod¹⁷. In order to test mitochondrial alignment with uropod in both PD and healthy control CD4⁺ T cells, we performed ICC using CD44 as a specific marker for uropod and MitoTracker green as a marker for mitochondria. As shown in Fig. 3, impaired alignment of mitochondria and uropod was observed in cells derived from PD patients.

To further assess whether spatial alteration of mitochondria impacts their functional properties, we used flow cytometry to analyze mitochondrial membrane potential and reactive oxygen species (ROS) production with and without CD3/CD28 activation in 11 PD and 11 controls. The mitochondrial membrane potential was assessed utilizing tetramethylrhodamine ethyl esters (TMRE), the total cellular ROS and mitochondrial superoxide (SOX), one of the most important ROS from which other ROS are derived³³ and which has a significant role for stress signaling³⁴. Our data revealed that all three parameters were significantly reduced in CD3/CD28-activated PD CD4⁺ T cells vs healthy control: TMRE (Interaction, F(1, 40) = 4.397 p = 0.0424; activation, F = 7098 p < 0.0001; PD, F = 8, 680 p = 0.0053); ROS (Interaction, F(1, 39) = 1.786 p = 0.1892; activation, F = 15,16 p = 0.0004; PD, F = 4457 p = 0.0412) and SOX (Interaction, F(1, 40) = 3456 p = 0.0704; activation, F = 2861 p < 0.0001; PD, F = 6319 p = 0.0161) (Fig. 4a–c).

To investigate the relationship between oxidative stress markers and clinical parameters, we correlated SOX, ROS, and TMRE levels in PD with disease stage, disease duration or levodopa medication, but found no significant correlations (Supplementary Fig. 1, spearman’s rho p > 0.05). We did, however, observe significant correlations between SOX, ROS, and TMRE in PD CD4⁺ T cells, which suggests these markers are interlinked. We finally analyzed the data separately for drug naïve and levodopa-treated PD patients, and noticed comparable changes of SOX and TMRE showing the same trend, while ROS was significantly reduced in activated CD4⁺ T cells derived from drug naïve PD patients (Interaction, F(1, 20) = 5107 p = 0.0351; activation, F = 7920 p = 0.0107; PD, F = 2737 p = 0.1137) (Supplementary Fig. 3).

We also quantified the total mitochondria content by analyzing MitoTracker with flow cytometry. There was no difference in mitochondrial quantification between PD-derived cells and control subjects with or without CD3/CD28 activation (Fig. 4d). Taken together, we can conclude that both improper mitochondria-positioning within the cells and their functional alteration may account for the migratory deficit observed in PD-derived CD4⁺ lymphocytes.

In order to restore observed defects, we have treated CD4⁺ T cells derived both from PD patients and age/sex-matched healthy controls with Coenzyme Q10 and performed the transwell assay with and without SDF-1α. Both groups increased the percentage of migrated cells to the lower chamber in the presence of chemokine, but there was no difference in percentages of migrated cells between the groups both with and without SDF-1α (Interaction, F(1, 8) = 0.0428 p = 0.8413; SDF-1α, F = 2702 p = 0.0008; PD, F = 0.9399 p = 0.3607). Coenzyme Q10 treatment of cells with described mitochondrial defect, provided a rescue (Fig. 4e).

Reduced ROS has been shown to lead to Th1-mediated immunity³⁵ and a recent publication reported that PD-derived CD4⁺ T cells are Th1 biased²³. In order to investigate the Th1 phenotype in our model system, we quantified the level of IFNγ in supernatants of CD4⁺ T cells derived from both PD and healthy individuals. Our data showed an increase in IFNγ level in cells originated from PD patients (hCTRL 4340 ± 4529 vs PD 6276 ± 6055 p = 0,0148) (Supplementary Fig. 4) corroborating a Th1 profile in PD.

Furthermore, we also compared CD3/CD28 activation level in both PD-derived cells and their healthy counterparts by performing surface staining for CD25 and CD69. Flow cytometry analysis showed no significant differences in CD69⁺CD25⁻, CD69⁺CD25⁺, and CD69⁻CD25⁺subpopulations with or without activation, between PD and healthy controls, although there was a significant increase in percent of each subpopulation upon activation within each group (Supplementary Fig. 5).

Cross-sectional study of p11 expression levels in CD4 + Th1/Th2/T17 cells

p11 is a neuronal protein known to associate with PD and L-dopa responses^24,27 as well as depression and the action of antidepressant drugs^28,36. Our previous publication showed that, apart from the brain, p11 is also expressed in peripheral blood cells, and its level has been increased in various populations of peripheral leukocytes derived from PD patients²⁴. Protein p11 is an interacting partner of Annexin A2 and is found in the same complex²⁵. Since Annexin A2 regulates actin bundling²⁶, p11 may also have a role affecting the actin cytoskeleton. It has also been shown that the p11 gene and protein expression in human epithelial cells are induced by IFNγ³⁷. In the present study, the intention was to characterize the expression level and distribution of p11 in different subsets of CD4⁺ T cells in both PD and control individuals.

Purified CD4⁺ T cells were activated by priming with CD3/CD28 and subsequently stained with Annexin A2 antibody. There was 25% reduction in Annexin A2 protein level in activated PD cells as observed by confocal imaging (hCTRL 1000 ± 603 vs PD 7562 ± 441 p = 0.0014) (Fig. 5a, b). Purified CD4⁺ T cells, activated with CD3/CD28 were stained with p11 antibody. Confocal microscopy showed more than 80% increase in p11 expression level in activated PD-derived CD4⁺ T cells (hCTRL 1000 ± 523 vs PD 1854 ± 1008 p < 0.0001) (Fig. 5c, d) although there was no difference in p11 mRNA level between analyzed groups (Fig. 5e).

In order to determine the expression level of p11 protein in CD4⁺ T cell Th1, Th2 and Th17 cell populations, PBMCs derived from 32 PD and 16 control individuals (Suppl. Table 2) were stained and analyzed by flow cytometry (Supplementary Fig. 6). As shown in Fig. 6, our data revealed that p11 was significantly increased in Th1 (hCTRL 1000 ± 955 vs PD 2073 ± 2363 p = 0,0021), Th2 (hCTRL 1000 ± 1972 vs PD 17,740 ± 2005 p = 0.0155) and Th17 cells (hCTRL 1000 ± 758 vs PD 1888 ± 2909 p = 0,0040) derived from PD patients when compared to control cells, but this increase seems to be most prominent in the Th1 subset (Fig. 6a, c, d). However, there was no difference in p11 expression within the Th1/17 subset between PD and control groups (Fig. 6b).

Discussion

Growing evidence suggests that adaptive immunity-related neuroinflammation contributes to PD, although the precise modes of action are yet to be unraveled⁶. It has been shown that human dopaminergic substantia nigra and stem cell-derived dopamine neurons express MHC-I, thereby becoming possible targets for T cell-mediated immunity in case of specific epitope display³⁸. CD4⁺ T cells play a pivotal role in adaptive immunity^8,10 therefore, any of their functional impairments may contribute to the inflammatory profile seen in PD. It has also been shown that PD patients have functional Th1 bias with specific cytokine profile³⁹. Another study found increased frequency of IFNγ⁺ CD8⁺ T cells, IL-4⁺ Th2 cells and IL-17A⁺ Th17 cells in PD patients⁴⁰. Detrimental effect of IL-17 resulting from neuronal death is suggested as one of mechanisms underlying PD pathogenesis characterized by increased cell counts of Th17 cells in the blood of patients and T lymphocytes in postmortem PD brains⁴¹. Nevertheless, to our knowledge, there are no data exploring the migration phenotype of CD4⁺T cells, which may critically be involved in immune system dysfunction observed in PD patients. Proper migration of CD4⁺ T cells is important for competent immunological scanning of the environment and maintaining homeostasis. Murine studies demonstrated that impairment in migration phenotype introduces serious developmental defects⁴² that may lead to an immunocompromised profile.

In the present study, we showed that PD-derived cells display migratory deficits characterized by velocity, directionality and displacement reduction, suggesting improper functionality of T cell-mediated immunity. An earlier study has revealed that mitochondrial relocation at the uropod is critical for lymphocyte migration as their concentration at the uropod regulates the motor of migrating cells¹⁴. In our study, we report imaging data of impaired alignment of mitochondria with uropod in PD-derived cells, suggesting that improper mitochondria redistribution within these cells may represent the part of the underlying cellular mechanism of an observed migratory phenotype. This is in line with mitochondrial microtubule-dependent trafficking which has been implied in PD. Namely, disturbed mitochondrial function in PD does not sufficiently support the transport along microtubules, resulting both in impaired transport of autophagosomes to lysosomes for degradation and inefficient recruitment of mitochondria where needed in the cell^43,44,45. We further proceeded with the functional investigation and found that PD-derived CD4⁺T cells show significant reduction of mitochondrial functional parameters such as ROS, SOX and mitochondrial membrane potential. These alterations may contribute to less competent adaptive immunity and increased infection incidence in PD patients. Indeed, it has been shown that, together with impaired swallowing and urine bladder emptying dysfunctions, PD patients have an increased risk of respiratory and urinary tract infections⁴⁶.

Observed mitochondrial defects were restored by Coenzyme Q10 treatment resulting in normal migratory capacity of PD-derived CD4⁺ T cells. The migration level was equal to control regardless of chemokine conditions, further confirming that mitochondrial defect has an impact on the migration of PD cells.

In addition, the precise control of ROS production is crucial for regulating the appropriate signal transduction and T cell activation^17,47. It is known that oxidative stress promotes the differentiation of human T cells toward Th2 phenotype²². Reduced homeostatic ROS production observed in our study drives Th1 bias in affected PD cells. By assessing IFNγ in supernatants of activated CD4⁺ T cells from PD patients and control subjects, we showed an increase in IFNγ production in cells originated from PD patients, which is in line with the previous publications²³.

The PD and control subjects in each experiment were carefully age and sex-matched. Other covariates within the PD group, such as disease duration, disease stage, motor symptoms, and levodopa medication were explored, but no significant correlations were found between CD4⁺ T cell migration/mitochondrial parameters and clinical parameters. These findings suggest that migration and mitochondrial alterations we found are not affected by dopaminergic medication, nor change over the course of the disease. This is in accordance with other functional T cell studies that did not find any effect of dopaminergic medication on T cell function in PD²³.

Besides observed mitochondrial defects, we investigated actin dynamics. Confocal imaging showed no difference in F-actin turnover between PD and control samples suggesting other actin regulation. Annexin A2 is a regulator of actin bundling coupling cytoskeleton with the plasma membrane and it is found either as a monomer or in a heterotetrameric complex with p11²⁶. Also the role of p11 as a possible PD biomarker was suggested in our previous study that showed a differential expression pattern of p11 in various populations of peripheral leukocytes²⁴. Here we investigated the expression profile of p11 in activated CD4⁺ T cells more deeply. Our data revealed an overall increase in p11 level in PD cells vs controls suggesting that either its synthesis or degradation may be altered. Since p11 mRNA expression was unchanged between PD and control cells, that may suggest that degradation mechanism is hampered in PD cells. Considering that mitochondrial dysfunction is shown to impair both proteasome assembly and activity⁴⁸, the observed increase of p11 levels in PD might be a consequence of disturbed mitochondria. Moreover, our findings of elevated levels of both IFNγ and p11 in CD4⁺ T cells derived from PD patients are in line with previously published data showing that IFNγ increases p11 protein level in a time and dose-dependent manner³⁷. Taken together, a simultaneous decrease in Annexin A2 and an increase in p11 protein levels can change complex stoichiometry, disturb actin regulation and contribute to the migration phenotype of PD-derived CD4⁺ T cells.

In order to analyse if p11 expression is more specifically distributed across various Th populations, we performed a cross-sectional study in different Th subsets. Previously we showed that there was no significant difference in the levels of p11 in T regulatory cells from PD and control subjects²⁴. Hereby, we demonstrated that p11 expression level was elevated in the Th1, Th2, and Th17 subsets, while there was no difference within Th1/17 cell population when compared to control group. Upregulation of p11 in Th1 cells, Th1 bias together with increased IFNγ production but also specific increase level of p11 in Th2 and Th17 subsets suggests a potential of p11 to be used as a differential biomarker for PD. This work emphasizes that investigations in larger patient cohorts are needed to further elucidate the role of p11 in CD4⁺T cells and its alteration in PD.

One limitation of this study was the small cohort size in each of the experiments performed. Owing to limited sample availability, we were not able to perform all experiments on all subjects, hence we could not analyze the direct link between mitochondrial function, oxidative stress, p11, and migratory function in the same individuals. This also means the correlations between clinical parameters and functional experiments should be interpreted with caution. However, on the group level, the effect size of most experiments was large enough to make up for the limited sample size. In this study, we decided to focus on the role of CD4⁺ T cells in PD, however, the functional changes found could very well be extended to other immune cells and non-immune cells and should be explored in other follow-up studies.

Taken together, this work highlights functional changes in peripheral CD4⁺ T cells in PD. In particular, we show that compared to healthy subjects, CD4⁺ T cells derived from PD patients have altered migration potential along with differentially increased p11 levels across Th subsets, impaired mitochondrial positioning within the cell and reduced mitochondrial functionality. These results warrant replication in another larger PD cohort.