Adoptively Transferred in vitro-Generated Myeloid-Derived Suppressor Cells Improve T-Cell Function and Antigen-Specific Immunity after Traumatic Lung Injury

Immune reactions after trauma are characterized by immediate activation of innate immunity and simultaneously downregulation of adaptive immunity leading to a misbalanced immunohomeostasis and immunosuppression of the injured host. Therefore, the susceptibility to secondary infections is strongly increased after trauma. Immune responses are regulated by a network of immune cells influencing each other and at the same time modifying their functions dependent on the inflammatory environment. Although myeloid-derived suppressor cells (MDSCs) are initially described as T-cell suppressors, their immunomodulatory capacity after trauma is mostly undefined. Therefore, in vitro-generated MDSCs were adoptively transferred into mice after blunt chest trauma (TxT). A single MDSC treatment-induced splenic T-cell expansion decreased apoptosis sensitivity and improved proliferation in the absence of T-cell exhaustion until 2 weeks after trauma. MDSC treatment had a long-lasting effect on the genomic landscape of CD4+ T cells by upregulating primarily Th2-associated genes. Remarkably, immune-activating functions of MDSCs supported the ability of TxT mice to respond to post-traumatic secondary antigen challenge. Secondary insults were mimicked by immunizing MDSC-treated TxT mice with ovalbumin (OVA), followed by OVA restimulation in vitro. MDSC treatment significantly increased the frequency of OVA-specific T cells, enhanced their Th1/Th2 cytokine expression, and induced upregulation of cytolytic molecules finally improving OVA-specific cytotoxicity. Overall, we could show that therapeutic MDSC treatment after TxT improves post-traumatic T-cell functions, which might enable the traumatic host to counterbalance trauma-induced immunoparalysis.

© 2022 The Author(s). Published by S. Karger AG, Basel

Introduction

Traumatic insults cause the release of multiple danger molecules, which immediately activate the innate immune system resulting in a local and systemic proinflammatory response. This acute unspecific immune response is counterbalanced by a systemic post-traumatic immune suppression. Victims of severe trauma frequently survive the initial result but are vulnerable to infections and multiorgan failure both by an over-activated innate immune response and by impaired adaptive immunity favoring secondary infections and sepsis [1, 2]. Ideally, therapeutic interventions should first aim to counterbalance the traumatic insult and second, restore and improve immunohomeostasis by reducing inflammation and strengthening adaptive immune responses.

Thoracic trauma occurs in 60% of patients with polytrauma, whereby blunt injuries cause 70% of all chest trauma. Blunt chest trauma (TxT) strongly impacts the morbidity and mortality of patients regardless of whether it is combined with other injuries or occurs in an isolated manner [3, 4]. Early innate immune responses in TxT patients and experimental TxT models are characterized by excessive release of proinflammatory mediators in bronchoalveolar lavage fluids and lung tissue [5, 6]. When innate immunity dissolves, splenic lymphocytes and macrophages of TxT mice, however, still exhibit reduced protective functions depicted by increased mortality after septic challenge [7, 8]. Likewise, increased mortality of patients was observed until 3 years after major trauma and associated with post-traumatic immunosuppression reflected by diminished cytokine secretion of T cells and monocytes as well as changes in surface marker expression of immune cells [9-12]. Most detailed analysis of changes in adaptive immunity is available from studies in septic patients and animals. T cells are massively depleted by apoptosis induction and the remaining T cells poorly respond to activation signals and exhibit an exhausted phenotype. At the same time, dendritic cells are decreased in numbers and skewed towards anti-inflammatory functions, and immunosuppressive regulatory T cell (Treg) numbers are increased [11, 13-15].

Myeloid-derived suppressor cells (MDSCs) represent another suppressive immune cell population, which strongly impacts cellular immunity. MDSCs are a heterogenous population of immature myeloid cells, which are induced and activated under inflammation and have been first described as suppressor cells in the tumor environment. Murine MDSCs co-express CD11b and Gr-1 and are divided into two major subsets: granulocytic/polymorphonuclear MDSCs (PMN-MDSCs) are defined by the CD11b+Ly-6G+Ly-6Clow, whereas monocytic MDSCs (M-MDSCs) express CD11b+Ly-6G−Ly-6Chigh [16]. T cells represent the primary target cell population of MDSCs, although B cell, Treg, NK cell, and mature myeloid cell functions can also be altered in their presence [17]. MDSCs exert their inhibitory function on T cells by various mechanisms. MDSCs secrete immunosuppressive cytokines, reactive oxygen, and nitrogen species; they catabolize amino acids essential for T-cell functions by enzymes such as arginase-1, iNOS, or IDO finally leading to inhibition of T-cell expansion and apoptosis induction. MDSC-mediated induction of Tregs as well as blocking T-cell invasion into lymphatic organs further underlines their inhibitor functions [18]. Recently, several reports, however, also show T cell-activating activities of MDSCs in inflammatory bowel disease and systemic lupus erythematosus [19, 20]. Even in the tumor environment, MDSCs do not exclusively block T-cell functions. CD11b+Gr-1+ cells found in the ascites of epithelial ovarian cancer-bearing mice at advance stages of disease augment the proliferation of functional CTLs in vitro [21] or T cells conditioned with MDSC exhibit an increased anti-tumor activity after adoptive T cell-based immunotherapy [22]. Besides acting as T-cell modulators, MDSCs promote liver regeneration and bone fracture healing or ameliorate renal fibrosis and heart failure [23-26], supposing that the multifarious functions of MDSCs are dictated by the local environment.

The physiological functions of MDSCs in traumatic injuries are still not well defined. Common to various trauma models such as traumatic brain and spinal cord injury, peripheral tissue trauma, sepsis, or TxT is the accumulation of MDSCs as a response to the post-traumatic inflammation. Trauma-induced MDSCs always show increased expression of immunosuppressive molecules and inhibit T-cell proliferation in vitro [27-33]. However, the in vivo functions of MDSCs are not well defined. Currently, mice deficient of MDSCs or antibodies specifically depleting MDSC are not available. Therefore, in vivo functions of MDSCs are analyzed either by adoptive transfer of ex vivo-isolated or in vitro-generated MDSCs, which are derived from bone marrow (BM) cells cultured in the presence of GM-CSF alone or in combination with, e.g., G-CSF, M-CSF, or IL-6. These in vitro-generated MDSCs are successfully used as cellular therapy to establish transplantation tolerance [34]. Administration of in vitro-generated MDSCs at the site of injury in a model of spinal cord trauma reduces local inflammation and promotes tissue regeneration [30]. Adoptive transfer of MDSCs, which are evolved during sepsis can either enhance or attenuate sepsis development dependent on the time of isolation [35, 36]. However, the effect of adoptive MDSCs therapy on post-traumatic T-cell functions and susceptibility to secondary infections is currently not known.

Recently, we found that prophylactical adoptive transfer of in vitro-generated MDSCs into mice before TxT treatment induced immunostimulatory functions. Splenic T cells increased in numbers and exhibited an activated phenotype with an increased proliferative capacity [37]. In the present study, we, therefore, used a therapeutic approach by treating TxT mice one day post TxT with in vitro-generated MDSCs to define the impact of MDSCs on post-traumatic T-cell immune responses.

Materials and MethodsAnimals, TxT, i.v. Injection of MDSCs, and Immunization

Male C57BL/6 mice (B6, H-2b, CD45.2+) (Janvier, France) were used for TxT at an age between 11 and 15 weeks. Male B6.SJL-PtprcaPepcb/BoyJ (B6.SJL, H-2b, CD45.1+) mice (breeding pairs obtained from The Jackson Laboratory and bred at the University of Ulm) between 6 and 14 weeks were used for MDSC generation. TxT was induced by a single blast wave centered on the thorax under sevoflurane anesthesia as described previously [29]. In brief, compressed air was delivered in the upper chamber of the blast wave generator, which is divided by the lower chamber by a Mylar polyester film. As soon as the pressure in the upper part exceeded the defined resistance of the membrane, the film ruptured towards the nozzle and released a reproducible single blast wave and contusion of the lung, which is not associated with histological alterations in the liver or abdomen. The sternum cylinder distance was 1.5 cm. Mice received buprenorphine 0.03 mg/kg 30 min before TxT and additionally 8 and 24 h after TxT as analgetic treatment. Sham mice were treated exactly like TxT mice including anesthesia and analgesic treatment with the exception that no blast wave was given. 2 × 107 MDSCs were adoptively transferred into tail veins of B6 mice 24 h after TxT/sham treatment. Three and seven days after MDSC treatment, mice were immunized with a single i.m. injection of 100 μg plasmid ovalbumin (OVA)-DNA/100 µL PBS by injecting 50 μL into each tibialis anterior muscle. All animal experiments were performed according to the international regulations for the care and use of laboratory animals and were approved by the local Ethical Committee (Regierungspräsidium Tübingen, Germany).

Isolation of Cells from OrgansMagnetic Bead Isolation

CD3+ T cells were positively selected from spleens by CD3ε MicroBead Kit (Miltenyi) according to manufacturer’s protocol. Purity of all isolated cells ranged between 85 and 99%.

Leukocytes of different organs were isolated according to following protocols:

BM: BM was isolated from the femurs and tibias of mice, and a single-cell suspension was prepared by dissociating cell clumps with a syringe followed by the lysis of erythrocytes (0.15 M NH4Cl, 1 mM KHCO3, 0.1 mM Na2EDTA).

Spleen: The spleen was extracted from mice, and a single-cell suspension was prepared by gently pressing the spleen through a cell strainer (∅ 70 μm), and subsequently, the erythrocytes were lysed.

Lymph nodes: Axillary, mesenteric, and inguinal lymph nodes were gently pressed through a cell strainer to obtain a single-cell suspension.

Whole blood: Whole blood was collected into heparinized tubes, centrifuged, and subsequently, the erythrocytes were lysed.

Liver: The liver was perfused with liver perfusion medium (Invitrogen Life Technologies) followed by liver digest medium (Invitrogen Life Technologies). Afterwards, the liver was removed and digested for 30 min at 37°C. Liver cells were gently pressed through a cell strainer (∅ 70 μm) and lymphoid cells were separated by centrifugation 60 g for 5 min. Lymphoid cells in the supernatant were collected, washed, and resuspended in PBS supplemented with 1% fetal calf serum (FCS) and diluted in 70% Easycoll (Biochrome) in a 1:1 ratio (=35% Easycoll) and then overlaid onto 70% Easycoll and centrifuged for 20 min at 950 g. Lymphoid cells were collected from the interface, washed, and subsequently, erythrocytes were lysed.

Lung: The lung was extracted from mice and gently pressed through a cell strainer (∅ 70 μm). Lymphoid cells were separated by centrifugation 60 g for 5 min, and the lymphoid cell containing supernatant was collected, washed, and resuspended in PBS supplemented with 1% FCS and diluted in 70% Easycoll (Biochrom) in a 1:1 ratio (=35% Easycoll) and then overlaid onto 70% Easycoll and centrifuged for 20 min at 950 g. Lymphoid cells were collected from the interface, washed, and subsequently, erythrocytes were lysed.

Cell Culture and Restimulation of Splenocytes with E.G7-OVA

EL4 and E.G7-OVA cells were obtained from ATCC and grown in RPMI 1640, 10% FCS (Sigma), 2 mML-glutamine, 1 mM sodium pyruvate at 37°C, and 7.5% CO2. Primary cells were cultured in α-MEM, 10% FCS, 2 mML-glutamine, 1 mM sodium pyruvate, 100 U/mL penicillin-streptomycin, and 0.05 mM 2-mercaptoethanol. Splenocytes were restimulated with irradiated (100 Gy) E.G7-OVA cells at a splenocyte:E.G7-OVA ratio of 10:1.

MDSC Generation

3 × 105 BM cells/mL extracted from the femur and tibia were cultured with granulocyte-macrophage colony-stimulating factor (GM-CSF) (200 U/mL) (Peprotech) for 4 days. The cells were used for further experiments where more than 90% of the cells co-expressed CD11b and Gr-1 cells and were able to inhibit T-cell proliferation in mixed lymphocyte reactions as described previously [37].

CFSE-Labeling and T-Cell Activation

2 × 106 Spleen cells were labeled with 5 μM CFSE (Thermo Fisher Scientific) at 37°C for 10 min, immediately washed with ice-cold PBS-5%FCS, and subsequently used for proliferation assays. 2.5 × 106/mL CFSE-labeled spleen cells of TxT mice were activated with anti-CD3 (cl. 145-2C11) and anti-CD28 (cl. 37.51) (CD3/28) antibodies (BD Bioscience), phytohemagglutinin (PHA, Sigma-Aldrich), or incubated with medium, and proliferation of CD3+ T cells was determined on day 4 by flow cytometry. Percentage of specific proliferation = (% stimulus-induced proliferating T cells − % proliferating T cells in medium alone)/(100 − % proliferating T cells in medium alone) × 100.

Chromium Release Assay

51Cr release assays were performed in RPMI 1640, 10% FCS, 2 mML-glutamine, 1 mM sodium pyruvate, and 0.05 mM 2-mercaptoethanol according to Strauss et al. [38]. In brief, 2 × 106 target cells were labeled with 200 μCi 51Cr for 1 h. Subsequently, 5 × 103 51Cr labeled target cells were cocultured with effector cells at different effector:target ratios. After 6 h, the supernatant was assayed for 51Cr release in a Top CountNXT counter (Packard Bio Science, Meriden, CT, USA). Percentage of specific release = (experimental release − spontaneous release)/(maximum release − spontaneous release) × 100.

Cytokine Analysis

Cytokine secretion of restimulated splenocytes was defined on day 3 by analyzing 50 μL of culture supernatant using ProcartaPlex Multiplex immunoassays (Thermo Fisher Scientific) and analyzed on a BIO-RAD Bioplex.

Flow Cytometry and Cell Sorting

A total of 1–5 × 105 cells were stained in FACS-medium (PBS – 10% FCS – 0.2% NaN3). 7-Amino-actinomycin-D (7-AAD; Sigma-Aldrich) positivity was used for exclusion of dead cells. Apoptosis was defined by Annexin-V-FLUOS positivity (Roche). Antibodies used are specified in online supplementary Table S1 (see www.karger.com/doi/10.1159/000525088 for all online suppl. material). Flow cytometry samples were measured on LSR II flow cytometer (BD Bioscience). Gating of cell populations was done by hand. A total of 3–8 × 107 cells were stained with anti-CD3, CD4, and CD8 antibodies and sorted for CD3+CD4+ and CD3+CD8+ T cells with FACSAriaTMIII flow cytometer (BD Biosciences). Purity of sorted cells ranged between 95 and 99%.

Plasmid Preparation

DNA from the pCI/OVA expression vector encoding OVA antigen under the control of the CMV promotor was replicated in One Shot® TOP10 E. coli and isolated by QiAGEN® Plasmid Maxi Kit (QIAGEN, Hilden, Germany) and sequenced.

RNA Preparation and Quantitative Reverse-Transcription Polymerase Chain Reaction

RNA was isolated and complementary DNA was synthesized as previously described [29], and quantitative reverse-transcription polymerase chain reaction (qRT-PCR) was performed with a CFX ConnectTM Real-Time PCR Detection System (Bio-Rad) using a LightCyler FastStart DNA Master PLUS SYBR Green I Kit (Roche Diagnostics). The qRT-PCR results were normalized using mouse aryl hydrocarbon receptor-interacting protein (AIP) as housekeeping gene. Primer sets (Thermo Fischer Scientific) used are listed in online supplementary Table S2.

Microarray Analysis

After RNA isolation, the integrity of all samples was checked by Bioanalyzer measurements (Agilent), and all samples displayed RIN values ≥8.5. Microarray analyses were performed using 200-ng total RNA as starting material and 5.5-μg ssDNA per hybridization (GeneChip Fluidics Station 450; Affymetrix, Santa Clara, CA, USA). The total RNAs were amplified and labeled following the Whole Transcript (WT) Sense Target Labelling Assay (http://www.affymetrix.com). Labeled ssDNA was hybridized to Mouse Gene 1.0 ST Affymetrix GeneChip arrays (Affymetrix). The chips were scanned with an Affymetrix GeneChip Scanner 3000 and subsequent images analyzed using Affymetrix® Expression ConsoleTM Software (Affymetrix).

A transcriptome analysis was performed using BRB-ArrayTools developed by Dr. Richard Simon and BRB-ArrayTools Development Team (http://linus.nci.nih.gov/BRB-ArrayTools.html). Raw feature data were normalized and log2 intensity expression summary values for each probe set were calculated using robust multiarray average [39].

Filtering

Genes showing minimal variation across the set of arrays were excluded from the analysis. Genes whose expression differed by at least 1.5-fold from the median in at least 20% of the arrays were retained.

Class Comparison

We identified genes that were differentially expressed among the two classes using a 2-sample t test. Genes were considered statistically significant if their p value was less than 0.05 and displayed a fold change between the two groups of at least 1.5-fold.

We used the Benjamini and Hochberg correction to provide 90% confidence that the false discovery rate was less than 10%. Heat maps as well as hierarchical clustering analysis were done using Genesis (Alexander Sturn and Rene Snajder, TU Graz, version 1.8.1) [40].

Gene Ontology Analysis of Differentially Expressed Genes

To identify the most affected biological processes, as defined by Gene Ontology annotation, we used the GoMINER analysis tool [41]. This package allows the automatic analysis of multiple microarrays and then integrates the results, across all of them, to find the GO categories that were significantly over- or under-represented. Complete microarray data are available at Gene Expression Omnibus (GEO accession number: GSE 188171).

Statistical Analysis

Statistical significance between two groups was analyzed using the Mann-Whitney test. For multiple comparisons, statistical significance was determined by one-way ANOVA followed by a Sidak test or unpaired multiple t test one per row followed by a Holm-Sidak correction. Results were considered as significant if p ≤ 0.05. Statistical analysis used for each experiment is depicted in figure legends. Statistical tests were performed with GraphPad Prism version 8.

ResultsMDSCs Persist after Adoptive Transfer in TxT Mice and Increase Splenocyte Numbers

To define the immunomodulary effects of therapeutically applied MDSCs on post-traumatic immune responses in TxT mice, MDSCs were generated from BM cells of B6.SJL mice in the presence of GM-CSF as described previously [37, 42]. As shown in previous work [37], about 95% of in vitro-generated MDSCs co-express Gr-1 and CD11b, from which 70% are Ly-6GhighLy-6Clow (PMN-MDSCs) and 30% are Ly-6GnegLy-6Chigh (M-MDSCs), upregulate immunosuppressive molecules such as arginase-1, iNOS, and TGF-β, and most importantly suppress T-cell proliferation in mixed lymphocyte reactions. TxT was induced and MDSCs were adoptively transferred 24 h later. Since we know from previous studies that 2 × 107 MDSCs act immunostimulatory when transferred shortly before TxT, we used exactly the same number of MDSCs to determine their immunomodulating functions in the therapeutic approach. Using the congenic marker CD45.1 expressed by B6.SJL mice, transplanted CD45.1+ MDSCs were identified by flow cytometry in syngeneic CD45.2+ TxT-treated B6 mice in the spleen, liver, BM, lymph nodes, blood, and lung until 3 weeks after TxT. MDSCs preferentially homed into the spleen with the highest numbers of 6.6 ± 4.47 ×105 MDSCs on day 7 after TxT representing about 1% of total splenocytes (Fig. 1a; online suppl. Fig. S1A).

Fig. 1.

Adoptively transferred in vitro-generated MDSCs preferentially home to the spleen and increase the number of splenic leukocytes. a B6.SJL-derived MDSCs (H-2b, CD45.1+) were injected 24 h post TxT in B6 mice (H-2b, CD45.2+), and homing was analyzed in spleen (S), liver (L), BM, and lung (LU) by calculating CD45.1+-expressing cells. b, c Mice received TxT and were adoptively transferred with MDSCs 24 h post TxT (TxT + MDSC) or left untreated (TxT w/o MDSC). b Number of splenocytes were determined at different time points. c The distribution of the splenic T cells was analyzed by flow cytometry, and subsequently, the total cell numbers were calculated. a–c Show the mean value ± SD of 3–48 mice/group analyzed. *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001. Significance was calculated by either using one-way ANOVA followed by a Sidak test as a post hoc test for multiple comparisons (a) or Mann-Whitney test comparing TxT w/o MDSC with TxT + MDSC for each time point (b, c). Number of mice used: a 3 d: S/L/BM n = 6, LU n = 3; 7 d S/L/BM n = 9, LU n = 6; 14 d: S/L/BM n = 6, LU n = 3; 21 d: S/L/BM n = 3. b 3 d: T/TM n = 6 mice; 7 d: T n = 48, TM n = 47; d14: T n = 14, TM n = 16; d21: T n = 6, TM n = 7. c 3 d: T/TM n = 6 mice; 7 d: T n = 20, TM n = 21; d14: T n = 8, TM n = 10; d21: T n = 6, TM n = 7.

/WebMaterial/ShowPic/1437713

Surprisingly, although MDSC numbers in the spleen account for less than 2 × 106 cells, spleens were visibly enlarged in MDSC-treated TxT mice. On day 3 after TxT, splenocyte numbers increased from 5.1 ± 0.9 ×107 cells in untreated mice to 7.7 ± 1.9 ×107 in MDSC-treated mice and peaked 14 days after TxT injection with 8.4 ± 1.7 ×107 cells. On day 21 after TxT, the difference in MDSC-treated and -untreated mice disappeared (Fig. 1b). Next, we determined which cell types are expanded in spleens of MDSC-treated TxT mice. Elevated cell numbers were detected for CD3+, CD3+ CD4+, CD3+CD8+ T cells 7 and 14 days after TxT (Fig. 1c). Especially B cells (CD19+) also expanded until day 14 after TxT, while Bregs (CD19+C1dhighCD5+), NKT cells (NK1.1+ CD3+), Tregs (CD4+CD25+FoxP3+), NK cells (CD3−NK1.1+), and CD11b+ myeloid cell numbers increased only at single time points (online suppl. Fig. S1B). The percental distribution of different splenocyte subsets, however, is only slightly affected by MDSC treatment (online suppl. Fig. S1C, D). Taken together, MDSCs preferentially home into the spleen, survive in vivo at least 3 weeks after transfer, and induce splenic lymphocyte expansion.

Adoptive Transfer of in vitro-Generated MDSCs in TxT Mice Supports T-Cell Survival and Expansion

Since T cells are the most prominent targets of MDSC-mediated immunomodulation and are strongly expanded in MDSC-treated TxT mice, we analyzed the effect of adoptively transferred MDSCs on T-cell survival and proliferation 3, 7, 14, or 21 days after TxT in vitro. Splenocytes from untreated and MDSC-treated mice were cultivated in vitro and stained for CD3 and apoptosis induction by Annexin-V positivity after 24 h. Annexin-V positivity was significantly decreased in T cells derived from TxT + MDSC mice at all time points indicating a survival advantage of T cells in the presence of MDSCs (Fig. 2a). Decrease in apoptosis might be influenced by increased proliferation rates. Therefore, CFSE-labeled splenocytes from MDSC-treated and -untreated TxT mice were cultured in medium in the absence of further activation signals, and proliferation was determined 4 days later. Seven days after TxT, CD3+ T cells from MDSC-treated mice exhibited strongly increased proliferation compared to T cells from untreated mice. T-cell expansion in the absence of exogenous stimulation was detectable until 21 days after TxT (Fig. 2b). Although constitutive activation is often associated with T-cell exhaustion, exhaustion markers were hardly upregulated on CD4+ and CD8+ T cells independent of whether mice received MDSCs or not. Fourteen days after TxT, BTLA (CD272) was expressed on CD4+ and CD8+ T cells, while PD-1 (CD279) expression was only slightly elevated on CD4+ T cells in both groups. Other exhaustion markers such as CTLA-4 (CD152), LAG-3 (CD223), or TIM3 were not induced (Fig. 2c). Exhaustion marker expression profiles of T cells analyzed 3, 7, or 21 days after TxT were comparable (data not shown). In summary, these data clearly show that adoptively transferred MDSCs support T-cell survival and expansion in the absence of T-cell exhaustion.

Fig. 2.

Adoptive transfer of in vitro-generated MDSCs activates TxT-induced T cells without inducing exhaustion. Mice received TxT and were adoptively transferred with MDSCs 24 h post TxT (TxT + MDSC) or left untreated (TxT w/o MDSC). a Three to twenty-one days later, splenocytes were cultured for 24 h, and the amount of apoptotic CD3+ T cells was analyzed by flow cytometry gating on CD3+Annexin-V+ cells. b Three to twenty-one days after TxT, CFSE-labeled splenocytes were cultured in medium, and 4 days later, proliferation of CD3+ T cells was analyzed. c On day 14 after TxT, the expression of exhaustion markers was defined on CD4+ and CD8+ T cells. Mean fluorescence intensity (MFI) was plotted for each exhaustion marker comparing TxT w/o MDSC and TxT + MDSCs. Data represent the mean value ± SD of 3–7 animals/group (a) and 3–4 mice/group (b). FACS stainings and MFI analysis show the expression of exhaustion markers for 3 mice/group. *p ≤ 0.05; ***p ≤ 0.001; ns, non-significant. Significance was calculated by one-way ANOVA followed by a Sidak test as a post hoc test for multiple comparisons (a, b) or Mann-Whitney test comparing TxT w/o MDSC with TxT + MDSC (c). Number of mice used: (a) 3 d: T/TM n = 3; 7 d T n = 7, TM n = 5; 14 d: T n = 3, TM n = 4; 21 d: T n = 3, TM n = 4. b 3 d: T/TM n = 3; 7 d T/TM n = 3; 14 d: T n = 3, TM n = 4; 21 d: T n = 3, TM n = 4. c 14 d: T/TM n = 3 mice.

/WebMaterial/ShowPic/1437711MDSC Treatment Increased T-Cell Responsiveness in vitro

To define whether T cells from MDSC-treated TxT mice maintain their capacity to respond to polyclonal stimulation, splenocytes from MDSC-treated and -untreated mice were CFSE-labeled 7 and 14 days after TxT and stimulated with anti-CD3/CD28 antibodies or PHA. At both time points, MDSC treatment increased the proliferative capacity after CD3/28 and PHA activation (Fig. 3a, b) about 2–5-fold. Similar results were obtained with T cells isolated 3 and 21 days after TxT (data not shown). Splenocytes used as effector cells, however, still contain low numbers of adoptively transferred MDSCs, which might interfere with T-cell proliferation. Therefore, we compared from the same mice the proliferative capacity of unseparated splenocytes containing low amounts of MDSCs with the proliferative capacity of isolated T cells toward CD3/CD28 activation 7 days after TxT. Increased proliferation was observed in T cells derived from MDSC-treated TxT mice compared to T cells from untreated TxT mice independent of whether unseparated splenocytes or isolated T cells were used as effector cells (Fig. 3c) pointing to sustained intrinsic changes in T cells by MDSC treatment. To clarify whether the traumatic environment drives the immunoactivating effects of MDSCs, we transplanted MDSCs into sham-treated mice receiving anesthesia and analgetic treatments but no TxT. Seven days after sham treatment, MDSC-treated sham mice showed no elevated splenocyte or T cell numbers (online suppl. Fig. S2A, B). Most importantly, splenocytes from MDSC-treated sham mice exhibited no increased proliferative capacity after CD3/CD28 or PHA activation (online suppl. Fig. S2C).

Fig. 3.

T cells from MDSC-treated TxT mice exhibit an increased proliferative capacity in vitro. Mice received TxT and were adoptively transferred with MDSCs 24 h post TxT (TxT + MDSC) or left untreated (TxT w/o MDSC). On day 7 (a) and 14 (b) after TxT, spleen cells of MDSC-treated or -untreated TxT mice were CFSE-labeled and stimulated with anti-CD3/28 antibodies or PHA, and after 4 days, proliferation of CD3+ T cells was analyzed and percentage of specific proliferation was calculated. c Seven days after TxT, splenocytes or isolated CD3+ T cells were CFSE-labeled and stimulated with CD3/28 antibodies, and after 4 days, proliferation of CD3+ T cells was analyzed, and percentage of specific proliferation was calculated. *p ≤ 0.05; ***p ≤ 0.001. Data represent the mean value ± SD of 7 mice/group (a) and 3–4 mice/group (b, c). Significance was calculated by one-way ANOVA followed by a Sidak test as a post hoc test for multiple comparisons. % specific proliferation = (% proliferation of activated T cells − % proliferation of T cells in medium)/(100 − % proliferation of T cells in medium) × 100. Number of mice used: (a) 7 d: T/TM n = 7 mice. b 14 d: T n = 3, TM n = 4. c 7 d: T/TM n = 3.

/WebMaterial/ShowPic/1437709Adoptively Transferred MDSCs Preferentially Promote Th2 Development in TxT Mice

Since T-cell functions are changed over an extended period of time in TxT + MDSC mice, splenic CD4+ and CD8+ T cells were FACS-sorted from MDSC-treated and -untreated mice 7 days after TxT and used for microarray analysis to define transcriptomic changes. Gene expression profiles and functional gene set enrichment analysis defined major changes only in CD4+ T cells. Differentially regulated genes expressed in CD4+ T cells derived from MDSC-treated TxT mice compared to untreated mice are shown in Figure 4a. 168 genes were differentially regulated (p < 0.05. FC > 1.5×) 7 days post TxT (154, FDR < 0.1, 168 < 0.17) in which 160 genes were upregulated (155, FDR < 0.1), and 8 genes (7, FDR < 0.1) were downregulated (Fig. 4b). Sixty-two genes were still differentially expressed 14 days after TxT (online suppl. Fig. S3A, B), but only 9 of them displayed an FDR below 0.1 indicating a strong reduction in the number of genes accompanied by a higher variability. Ingenuity pathway analysis (IPA) revealed that changes occurred preferentially in the Th2 activation pathways (Fig. 4c; online suppl. Fig. S3C). GO enrichment analysis of target genes identified the top 10 pathways of GO terms and linked target genes (online suppl. Table S3). Genes associated with positive regulation of interleukin-13 production (GATA3, IL-4) and C-C chemokine binding (CCR2, CCR5) were strongly upregulated in CD4+ T cells 7 days after TxT treatment, suggesting that MDSCs promote Th2 polarization. However, expression of genes associated with the C-C chemokine receptor activity (CCR3, CCR4, CCR2) was unchanged in CD4+ T cells 14 days after TxT treatment (online suppl. Table S4). Array-defined gene expression changes of CCR3, CCR4, CCR5, CxCR6, GATA3, IL-4, and Tbx21 were confirmed by qRT-PCRs (Fig. 4d). While GATA3 and Tbx21 represent the master transcription factor responsible for Th2 and Th1 differentiation, respectively [43], CCR4 is preferentially expressed on Th2 cells in contrast to CCR5 expressed on Th1 cells [44]. MDSC treatment differentially induced the expression of 186 genes in CD8+ T cells 7 days after TxT (130 FDR <0.1, 186 FDR <0.15) in which 94 (81, FDR < 0.1) genes were upregulated and 92 genes (49, FDR < 0.1) were downregulated (online suppl. Fig. S4A, B). However, only 6 genes (all with a FDR of 0.74) were differently regulated in CD8+ T cells 14 days after TxT (online suppl. Fig. S4C, D), also demonstrating a strong reduction in the number of perturbated genes or even just background noise. Although 186 genes were differentially regulated, IPA analysis revealed that no pathway was enriched in CD8+ T cells 7 days after TxT (data not shown).

Fig. 4.

MDSC treatment induces differentially expressed genes (DEGs) in CD4+ T cells 7 days after TxT. Mice received TxT and were adoptively transferred with MDSCs 24 h post TxT (TxT + MDSC) or left untreated (TxT w/o MDSC). Seven days after TxT, CD4+ T cells were sorted from splenocytes. a Heat map shows DEGs in CD4+ T cells. b Volcano plot comparing the transcriptional profile of CD4+ T cells isolated from TxT w/o MDSCs and TxT + MDSCs mice. Blue dots indicate genes differentially expressed between both groups. c Top 15 significantly enriched canonical pathways analyzed by IPA software in DEGs (adjusted p value <0.05 & |fold change| ≥ 0.5) found in CD4+ T cells of MDSC-treated TxT mice. d Fold change of expression of genes enriched by IPA analysis in CD4+ T cells from TxT + MDSC mice compared to TxT w/o MDSC mice in microarray (indicated by arrow heads in (a)). The expression of genes was further corroborated with qRT-PCR. Number of mice used for (a–c): n = 3 mice in TxT w/o MDSC and n = 3 mice TxT + MDSC group. d n = 5 mice.

/WebMaterial/ShowPic/1437707

MDSC-mediated changes in T-cell polarization were confirmed by intracellular cytokine measurements. PMA and ionomycin stimulation of splenocytes from MDSC-treated animals 7 days after TxT showed an increased intracellular expression of Th2-associated cytokines (IL-4, -5, -10, -13) in CD4+ T cells. While expression of TNFα and IL-2 were hardly altered, the Th1-associated cytokine IFN-γ was upregulated in CD4+ and CD8+ T cells (online suppl. Fig. S5), further underlining the effect of MDSC treatment on T-cell polarization.

Antigen-Specific Immunity Develops after OVA Immunization in MDSC-Treated TxT Mice

As trauma induces a general immunosuppression, leading to an increased sensitivity for post-traumatic secondary infections, we defined whether the adoptive transfer of MDSCs interferes with the development of antigen-specific immunity in vivo after trauma induction. Immunization of untreated or MDSC-treated TxT mice with OVA-DNA was used to mimic a post-traumatic secondary insult. OVA immunization was performed in both groups either 3 or 7 days after MDSC transplantation. Twelve days after immunization, liver lymphocytes and splenocytes were analyzed for T-cell composition and induction of OVA-specific T cells (Fig. 5a). MDSC-treated or -untreated TxT mice showed a comparable ratio of splenic and liver CD4+, CD8+, and Tregs (Fig. 5b, c). Most interestingly, MDSC treatment does not affect the generation of OVA-specific CD8+ T cells identified by H-2Kb OVA-specific tetramers. 17.5% of CD8+ T cells in the liver of TxT + MDSC mice were OVA specific compared to 13.6% in untreated TxT w/o MDSC mice (Fig. 5b). The percentage of OVA-specific CD8+ T cells in the spleen was low with about 1% in both treatment groups (Fig. 5c). Lymphocyte distribution and percentage of OVA-specific CD8+ T cells were comparable if mice were immunized 7 days after MDSC treatment (online suppl. Fig. S6), indicating that MDSC treatment does not impair the induction of antigen-specific T cells in vivo.

Fig. 5.

OVA immunization of TxT mice after MDSC treatment does not alter the frequency of lymphocytes and OVA-specific CD8+ T cells in the liver and spleen ex vivo. a Schematic representation of experimental execution and analysis. After 24 h of TxT, mice were treated with MDSCs. Following 3 days of MDSC transplantation, mice with/without MDSC treatment were immunized with plasmid DNA coding for chicken OVA. After 12 days of immunization, the frequency of T lymphocyte subsets in the liver (b) and spleen (c) were investigated by flow cytometry. No statistical significance was detected between TxT w/o MDSC and TxT + MDSC groups. Statistical significance between two groups was determined by Mann-Whitney test. Data represent the mean ± SD. Number of mice used: n = 4 mice in TxT w/o MDSC and n = 5 mice TxT + MDSC group. ns, non-significant.

/WebMaterial/ShowPic/1437705MDSC Treatment Increases Frequency and Cytotoxicity of OVA-Specific T Cells after Prime-Boost Immunization

Mice underwent TxT and were treated with MDSCs 24 h later and subsequently received DNA immunizations with OVA encoding plasmids 3 days after MDSC treatment. To increase numbers and define the functionality of OVA-induced T cells, spleen cells of immunized mice were restimulated in vitro with E.G7-OVA cells 12 days after immunization. Induction and function of OVA-specific T cells in TxT mice treated with or without MDSCs were analyzed either 3 or 6 days after in vitro restimulation (TxT + MDSC + OVA; TxT w/o MDSC + OVA) by tetramer staining and proliferation and cytotoxicity assays (Fig. 6a). Of note, MDSC-treated TxT mice exhibited a 2-fold increase of OVA-specific CD8+ T cells on day 6 after restimulation compared to untreated mice (60 % vs. 30%), while difference was less pronounced on day 3 (Fig. 6b). Elevation of OVA-specific CD8+ T cells after MDSC treatment was reflected by an increased proliferative rate of CFSE-labeled CD3+, CD4+, and CD8+ T cells in restimulation cultures, which was most pronounced on day 6 after restimulation (Fig. 6c). Most importantly, cytotoxicity of T cells was strongly increased in T cells derived from TxT + MDSC + OVA mice. In chromium release assays performed 6 days after restimulation, T cells derived from TxT + MDSC + OVA mice lysed about 45% of OVA-expressing E.G7-OVA cells at an effector:target ratio of 12:1, while only 10% of cells are lysed by T cells derived from TxT w/o MDSC + OVA mice. Lysis subsequently decreased with reduced effector:target ratios. OVA cytotoxicity was already detectable in T cells derived from TxT + MDSC + OVA mice 3 days after restimulation. EL-4 cells devoid of OVA expression served as a negative control. T cells derived from TxT + MDSC + OVA mice showed slight background lysis of EL4 cells at higher effector:target ratios, while T cells from TxT + OVA did not lyse EL4 cells independent of the amount of effector cells (Fig. 6d). Similar results were obtained if OVA immunization was performed 7 days after MDSC transfer (online suppl. Fig. S7). Subsequently, CD8+ T cells were sorted 3 or 6 days after in vitro restimulation to compare the expression of cytotoxic molecules and death-inducing ligands. Granzyme B and IFN-γ levels were strongly increased in CD8+ T cells derived from TxT + MDSC + OVA. Granzyme A and perforin expression were elevated on day 6 after restimulation (Fig. 6e). Changes in the expression of death-inducing ligands CD95L were not detected. TNFα and TRAIL expression revealed a different response dependent on the time point post-trauma: elevated TNFα levels are expressed on day 3 and decreased on day 6 after restimulation, while TRAIL expression was reduced on day 3 after restimulation (Fig. 6f). CD4+ T cells sorted from the restimulation cultures derived from TxT + MDSC + OVA on day 3 or day 6 showed strongly increased expression of Th1- (IFNγ, TNFα) but also Th2-(IL-4, -5, -13)associated cytokines. IL-2 as an early marker for T-cell activation was elevated on day 3 (Fig. 7a). Induction of Th1 and Th2 cytokines in restimulated T-cell cultures derived from TxT + MDSC + OVA mice was confirmed by increased cytokine levels in the supernatant of the cultures (Fig. 7b). Despite increased T-cell activation, proliferation, and effector functions, exhaustion marker expression on CD4+ and CD8+ T cells derived from restimulated splenocytes of TxT + MDSC + OVA mice were not elevated (online suppl. Fig. S8A, B). In summary, these results clearly show that the therapeutic transfer of MDSCs after TxT enhances antigen-specific immunity and might counterbalance post-traumatic immunosuppression.

Fig. 6.

MDSC treatment increases frequency and cytotoxicity of OVA-specific T cells after OVA-specific restimulation in vitro. a Schematic representation of experimental execution and analysis. Mice received TxT and were left untreated or transplanted with MDSCs 24 h post TxT. After 3 days of MDSC treatment, TxT w/o MDSC and TxT + MDSC mice were immunized with OVA-encoding plasmid DNA. Following 12 days of immunization, splenocytes were restimulated with E.G7-OVA cells (TxT w/o MDSC + OVA; TxT + MDSC + OVA). Three or six days after restimulation, the frequency of OVA-specific CD8+ T cells (b) and proliferating T-cell subsets (c) were determined by flow cytometry. d OVA-specific cytotoxicity of restimulated splenocytes from TxT w/o MDSC + OVA and TxT + MDSC + OVA mice was determined on day 6 or 3 after restimulation in 51Cr release assays using E.G7-OVA or EL4 cells as target cells at various effector: target ratios. e, f Three or six days after restimulation of splenocytes with E.G7-OVA cells, CD8+ T cells were FACS sorted. The expression of cytotoxic molecules and death-inducing ligands was determined by qRT-PCR. Relative expression of cytotoxic molecules to aryl hydrocarbon receptor-interacting gene (AIP) was calculated. *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; ****p ≤ 0.0001; ns, non-significant. Statistical significance was determined by unpaired multiple t test one per row followed by a Holm- Sidak correction for multiple comparisons. Data represent the mean ± SD. Number of mice used (b): n = 5 mice in TxT w/o MDSC and n = 5 mice TxT + MDSC group. c, d: n = 4 mice in TxT w/o MDSC and n = 5 mice TxT + MDSC group. e, f Day 3: n = 5 mice in TxT w/o MDSC and n = 4 mice TxT + MDSC group; day 6: each data point represents the mean of two technical replicates from CD8+ T cells pooled from two different mice and n = 8 mice in TxT w/o MDSC and n = 10 mice TxT + MDSC group.

/WebMaterial/ShowPic/1437703Fig. 7.

MDSC treatment induces expression of Th1 and Th2 cytokines after OVA-specific restimulation in vitro. Mice received TxT and were left untreated or transplanted with MDSCs 24 h post TxT. After 3 days of MDSC treatment TxT w/o MDSC and TxT + MDSC mice were immunized with OVA-encoding plasmid DNA. Following 12 days of immunization, splenocytes of treated mice were each restimulated with E.G7-OVA cells (TxT w/o MDSC + OVA; TxT + MDSC + OVA). Three or six days after restimulation, CD4+ T cells were FACS sorted. a The expression levels of cytokines were determined by qRT-PCR. Expression of cytokines relative to expression of aryl hydrocarbon receptor-interacting gene (AIP) was calculated. b Three days after restimulation, the culture supernatants were analyzed for cytokine expression. *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; ****p ≤ 0.0001. a Statistical significance was determined by unpaired multiple t test one per row followed by a Holm- Sidak correction for multiple comparisons. b Statistical significance between two groups was determined by Mann-Whitney test. Number of mice used for (a) day 3: n = 5 mice in TxT w/o MDSC and n = 4 mice TxT + MDSC group. a Day 6: each data point represents the mean of two technical replicates from CD4+ T cells pooled from two different mice, and n = 8 mice in TxT w/o MDSC and n = 10 mice TxT + MDSC group. b: n = 5 mice in TxT w/o MDSC and n = 5 mice TxT + MDSC.

/WebMaterial/ShowPic/1437701Discussion

After trauma, the immune response is characterized by a balance between the proinflammatory immune responses to prevent infections and eliminate tissue debris and the compensatory anti-inflammatory responses to limit further tissue damage, preserve organ functions, and quiet the proinflammatory state. A prolonged compensatory anti-inflammatory response leads to immunosuppression through maladaptive immunity, placing injured patients at risk of secondary infections. Immunomodulatory therapies aim to reverse immune dysregulation and improve patient outcomes. To our knowledge, we show here for the first time, that treating mice after TxT with in vitro-generated MDSCs supports expansion, activation, and effector functions of T cells. Most notably, adoptive post-traumatic MDSC transfer substantially improves the formation of antigen-specific T-cell responses after TxT indicating that MDSC treatment helps to promote immunohomeostasis and immunity towards post-traumatic secondary infections.

To define the modulation of the post-traumatic cellular immune response by MDSCs, mice underwent TxT and were therapeutically treated with in vitro-generated MDSCs 24 h later. This considerable time gap between trauma and the treatment start is clinically meaningful since trauma patients are admitted to the hospital with a significant time delay. In vitro-generated MDSC consist of about 70% of PMN-MDSCs and 30% of M-MDSCs as shown previously [37]. In vitro-generated MDSCs exhibited a long-term survival until 21 days after adoptive transfer, which is in contrast to endogenously induced MDSCs. In former studies [29], we were able to show that TxT-induced MDSCs are also preferentially PMN-MDSCs but their increase is only of short duration and is no longer detectable 3 days after TxT. Extended half-lives of in vitro-generated MDSCs might be explainable by a combination of a highly activated phenotype associated with apoptosis resistance and in vivo proliferation [42, 45, 46].

Most importantly, MDSC-treated mice showed a strong expansion of splenocytes including T cells, which are the major targets of MDSCs. T-cell increase by a single MDSC treatment was long-lasting until 14 days after TxT. In MDSC-treated TxT mice, T-cell apoptosis was decreased, while proliferative capacity in the absence of further stimulation was increased. Most importantly, T cells from TxT-treated animals showed improved reactivity towards polyclonal stimulation pointing to immune-activating MDSC functions. Elevated proliferative capacity was not associated with an increased exhaustion status. The strong impact of MDSCs on T-cell immunity was underlined by a significant change in the transcriptomic landscape of T cells with a more pronounced effect on CD4+ T cells affecting preferentially Th2 polarization.

The immune-activating functions of MDSCs are counterintuitive considering their primarily described function of immunosuppression. In cancers, MDSC numbers positively correlate with tumor burden, cancer stage, metastasis, response to anti-tumor therapy, and survival [47-49]. MDSC appearance after chronic viral, bacterial, or fungal infections is also associated with impaired T-cell immunity and adverse outcome [50-52]. However, MDSC functions in autoimmune diseases are dependent on the type of autoimmune disease, since MDSCs can either accelerate or extenuate disease progression [53].

Likewise, the role of MDSCs in the context of trauma is also versatile. Injuries commonly cause MDSC increase in experimental trauma models such as focal brain and spinal cord injury, peripheral tissue trauma and TxT [27-29, 31]. Although ex vivo-isolated trauma-induced MDSCs suppress the proliferation of polyclonally activated T cells in vitro, it remains elusive whether they mediate immunosuppression also in vivo and whether they are beneficial or detrimental for the injured host. Discrepancies between in vitro and in vivo functions of MDSCs have been already reported in tumor and autoimmunity models showing that MDSCs, which suppress T-cell functions in vitro do not necessarily confer immunosuppression in vivo [19, 54]. Due to the fact that MDSCs currently cannot be selectively and specifically depleted or inhibited in vivo, approaches using adoptive transfer of ex vivo-isolated or in vitro-generated MDSCs try to define the role of MDSCs after traumatic injuries. Intraspinal transplantation of in vitro-generated MDSCs in spinal cord-injured hosts promotes tissue generation and improves neurological outcome by reducing inflammation at the injury site [30]. Effects on adaptive immunity, however, are not analyzed. More detailed studies are available from preclinical sepsis models. In the early stage of sepsis, MDSCs dramatically increase in number, and blocking their expansion aggravates mortality. Likewise, survival in septic mice could be improved by the adoptive transfer of MDSCs [55]. Brudecki et al. [35] also reported massive MDSC increase and protective effects of adoptively transferred MDSCs in septic mice. Interestingly, their protective role is dependent on the time point of isolation. MDSCs isolated from early septic mice (day 3) and subsequently transferred to septic mice support disease progression. If MDSCs, however, are isolated at late time points (day 12) from septic mice for adoptive transfer, they dramatically improve survival rates [35] indicating that timing of MDSC expansion in correlation to disease development defines their beneficial or detrimental potential. Elevated levels of MDSCs are only detectable until 24 h after TxT due to a limited inflammatory response compared to sepsis. Therefore, possible alterations in MDSC functions at different time points after TxT cannot be defined.

An immunostimulatory effect of MDSCs was recently also found in TxT mice treated prophylactically with in vitro-generated MDSCs 1 h before TxT [37]. However, prophylactical treatment increases T-cell numbers but not T-cell reactivity. This discrepancy might be explained by the fact that MDSCs transplanted before TxT encounter an environment deficient of trauma-induced proinflammatory factors, while MDSC transplanted post TxT are confronted with a proinflammatory setting favoring phenotypical and functional changes of MDSCs. This is supported by our data showing that immunostimulatory functions of MDSCs are not detectable in MDSC-treated sham animals.

Improvement of T-cell immunity is a prerequisite to efficiently fight post-traumatic infections. In order to mimic a secondary insult and define antigen-specific immune responses after trauma, we performed immunization with OVA-coding plasmid DNA in TxT mice treated with MDSCs or left untreated. First, MDSC transfer does not prevent the development of antigen-specific immunity since OVA-specific T cells were present in the spleen and liver. OVA-specific T-cell numbers were comparably independent of MDSC treatment. Second, when spleen cells of immunized mice were restimulated with OVA-expressing cells in vitro, the antigen-specific immune response was strongly improved in MDSC-treated animals. OVA-specific T cells increased 2-fold and exhibited massively improved reactivity reflected by a strongly elevated cytotoxic capacity to lyse OVA-expressing cells and increased expression of cytotoxic mediators such as IFNγ, perforin, granzyme A, and granzyme B. An improved activation status was further confirmed by elevated Th1- and Th2-specific cytokine expression.

Of course, elevated OVA-specific immune responses do not necessarily reflect improved immunity towards frequently occurring post-traumatic secondary infections. Further studies will show, whether MDSC treatment will improve immunity towards pathogens such as murine cytomegalovirus, Escherichia coli, or Candida albicans. Additionally, it will be worthwhile to test whether adoptively transferred MDSCs exhibit similar functions in other trauma models. Especially sepsis models might be particularly well-suited due to the strong impairment and exhaustion of septic T cells.

However, at the moment, it is totally unclear how MDSCs perform their immunostimulatory effects on T-cell immunity. One possibility might be the maturation of in vitro-generated MDSCs into T cell-activating antigen-presenting cells in the post-traumatic proinflammatory environment. This would be in line with the observation that ex vivo-isolated inhibitory macrophages could be maintained as inhibitory cells or differentiated into fully mature, highly activated dendritic cells with T cell-activating functions dependent on the cytokine milieu that prevails during maturation [56]. This might also explain why MDSCs, which are induced in the exact same manner, inhibit the overwhelming T-cell response after allogeneic BM transplantation [42].

To identify the underlying mechanisms by which MDSCs change T-cell responses in the context of trauma, re-isolation of MDSCs, and subsequent RNAseq analysis and GO terms enrichment will identify the hierarchical clustering of up- and downregulated genes, signaling pathways, and soluble factors potentially involved in the immunoactivating functions of MDSCs. Identification of genes/molecules and signaling pathways is a prerequisite to target defined candidate genes or pathways in order to substitute time and cost-consuming individual cellular therapies by therapeutically more feasible and controllable strategies. In the course of these investigations, it will be interesting to define whether MDSC-induced immunoactivation in TxT mice depends on different MDSC subpopulations. Another possibility to elucidate the mechanism by which in vitro-generated MDSCs activate the cellular immune response in comparison to naturally induced MDSCs would be the adoptive transfer of MDSCs, which are isolated from TxT mice. However, this experiment is technically not feasible due to very low splenic MDSC numbers induced after TxT [29] and the high numbers of MDSCs required for adoptive transfer. Finally, the effect of T-cell activation in the presence of in vitro-generated MDSCs have to be confirmed in the human setting. First, induction of MDSCs from blood cells in the presence of serum derived from patients with sepsis or trauma might indicate, whether serum-derived proinflammatory factors turn immunosuppressive MDSCs into immunoactivating MDSCs. Second, ex vivo septic models, in which LPS stimulation of whole blood cells or isolated PBMCs induce suppression of T-cell proliferation and IL-2 production [57] will show whether the presence of human in vitro-generated MDSCs restores T-cell reactivity.

The immunostimulatory functions of MDSCs, however, have already been reported in other pathological settings. T cells conditioned with MDSCs in vitro exhibit increased anti-tumor activity after adoptive transfer into tumor-bearing mice [22]. CD11b+Gr-1+ derived from ascites of ovarian carcinoma-bearing mice augment CTL functions and inhibit tumor growth after adoptive transfer [21]. Immunostimulatory roles of MDSCs have currently also been reported in various human pathologies. The study by Li et al. [58] showed a strong increase of PMN-MDSC in patients with mild to severe trauma including fractures, traffic injuries, or brain injuries, which correlates with lower post-traumatic inflammation and positive prognosis. Ex vivo-isolated PMN-MDSCs from systemic lupus erythematosus patients do not inhibit T-cell proliferation but induce IFNγ and TNFα production in CD4+ T cells [20]. Likewise, isolated PMN-MDSCs from inflammatory bowel disease patients promote autologous T-cell proliferation [19]. On the other hand, MDSC accumulation in septic or COVID-19 patients positively correlates with adverse clinical outcome [59-62]. These obviously contrary findings further point to the enormous plasticity of MDSCs, which exhibit their expansion and functions dependent on the environment they encounter. Disease-specific compositions of pro- and anti-inflammatory factors might determine whether MDSCs retain in an immature state attributed to immunosuppression or further develop into more mature myeloid cells with different functions. Taken together, our results indicate that in vitro-generated MDSCs exhibit suppressive capacity in vitro but turn under traumatic conditions into immunostimulatory cells improving post-traumatic T-cell functions and possibly counteracting trauma-induced immunoparalysis.

Acknowledgments

The authors thank Linda Wolf (Department of Pediatrics and Adolescent, Ulm), Annette Palmer, and Sonja Braumüller (both from the Institute of Experimental Trauma-Immunology, Ulm) for excellent technical assistance.

Statement of Ethics

This study protocol was reviewed and approved by the local Ethical Committee (Regierungspräsidium Tübingen, Germany), approval number (1297).

Conflict of Interest Statement

The authors dec

留言 (0)

沒有登入
gif