ORIGINAL RESEARCH published: 17 November 2020 doi: 10.3389/fimmu.2020.582330 Impact of Long-Term Cryopreservation on Blood Immune Cell Markers in Myalgic Encephalomyelitis/Chronic Fatigue Syndrome: Implications for Biomarker Discovery Edited by: Elisabet Gómez-Mora 1, Jorge Carrillo 1, Vı́ctor Urrea 1, Josepa Rigau 2, José Alegre 3, Jorge Matias-Guiu, Cecilia Cabrera 1, Elisa Oltra 4, Jesús Castro-Marrero 5*† and Julià Blanco 1,6*† Complutense University of Madrid, Spain 1IrsiCaixa AIDS Research Institute, Institut d’Investigació en Ciències de la Salut Germans Trias i Pujol, Badalona, Spain, 2Rigau Private Clinic, Tarragona, Spain, 3 Division of Rheumatology, ME/CFS Clinical Unit, Vall d’Hebron University Hospital, Reviewed by: Aletta D. Kraneveld, Universitat Autònoma de Barcelona, Barcelona, Spain, 4 School of Medicine, Universidad Católica de Valencia San Vicente Utrecht University, Netherlands Mártir, Valencia, Spain, 5 Division of Rheumatology, ME/CFS Research Unit, Vall d’Hebron Hospital Research Institute, Paul Robert Fisher, Universitat Autònoma de Barcelona, Barcelona, Spain, 6 Chair in Infectious Diseases and Immunity, Centre for Health and La Trobe University, Australia Social Care Research (CESS), Faculty of Medicine, University of Vic, Central University of Catalonia (UVic–UCC), Vic, Spain *Correspondence: Jesús Castro-Marrero Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) is a complex jesus.castro@vhir.org neuroimmune disorder characterized by numerous symptoms of unknown etiology. The Julià Blanco jblanco@irsicaixa.es ME/CFS immune markers reported so far have failed to generate a clinical consensus, † These authors have contributed perhaps partly due to the limitations of biospecimen biobanking. To address this issue, we equally to this work and share senior performed a comparative analysis of the impact of long-term biobanking on previously authorship identified immune markers and also explored additional potential immune markers linked Specialty section: to infection in ME/CFS. A correlation analysis of marker cryostability across immune cell This article was submitted to subsets based on flow cytometry immunophenotyping of fresh blood and frozen PBMC Multiple Sclerosis and Neuroimmunology, samples collected from individuals with ME/CFS (n = 18) and matched healthy controls a section of the journal (n = 18) was performed. The functionality of biobanked samples was assessed on the Frontiers in Immunology basis of cytokine production assay after stimulation of frozen PBMCs. T cell markers Received: 13 July 2020 defining Treg subsets and the expression of surface glycoprotein CD56 in T cells and the Accepted: 20 October 2020 Published: 17 November 2020 frequency of the effector CD8 T cells, together with CD57 expression in NK cells, Citation: appeared unaltered by biobanking. By contrast, NK cell markers CD25 and CD69 were Gómez-Mora E, Carrillo J, Urrea V, notably increased, and NKp46 expression markedly reduced, by long-term Rigau J, Alegre J, Cabrera C, Oltra E, Castro-Marrero J and Blanco J (2020) cryopreservation and thawing. Further exploration of Treg and NK cell subsets failed to Impact of Long-Term identify significant differences between ME/CFS patients and healthy controls in terms of Cryopreservation on Blood Immune biobanked PBMCs. Our findings show that some of the previously identified immune Cell Markers in Myalgic Encephalomyelitis/Chronic Fatigue markers in T and NK cell subsets become unstable after cell biobanking, thus limiting their Syndrome: Implications for use in further immunophenotyping studies for ME/CFS. These data are potentially relevant Biomarker Discovery. Front. Immunol. 11:582330. for future multisite intervention studies and cooperative projects for biomarker discovery doi: 10.3389/fimmu.2020.582330 using ME/CFS biobanked samples. Further studies are needed to develop novel tools for Frontiers in Immunology | www.frontiersin.org 1 November 2020 | Volume 11 | Article 582330 Gómez-Mora et al. Biomarkers in Cryopreserved ME/CFS Immune Cells the assessment of biomarker stability in cryopreserved immune cells from people with ME/CFS. Keywords: cryopreservation, immunophenotyping, immune biomarkers, myalgic encephalomyelitis, chronic fatigue syndrome, freeze-thaw process INTRODUCTION This key service platform facilitates worldwide network collaboration by harmonizing clinical data and by providing Myalgic Encephalomyelitis (ME), also known as Chronic Fatigue high-quality standardized biosamples, which are obtained, Syndrome (CFS), is a multisystem, complex and extremely processed, and p rese rved through opt imized good debilitating chronic illness. The 2015 Institute of Medicine manufacturing practice methods. Specifically, in the field of (IOM) committee report estimated that ME/CFS could ME/CFS biomedical research, biobanking facilities such as potentially affect up to 2.5 million people in the US, and as SolveCFS (https://solvecfs.org) in the US and the ME/CFS many as 17-24 million worldwide, of whom roughly 90% are Biobank (https://cureme.lshtm.ac.uk) in the UK (11, 12), undiagnosed (1). There is no proven specific cause, no accurate already have large biospecimen collections and associated and objective diagnostic test, and no universally effective anonymized clinical information available for future treatment for ME/CFS. Beyond the puzzling post-exertional biomedical research, including biomarker discovery. However, malaise as the cardinal symptom that cannot be alleviated by biobanking involves storage, often long-term (>6 months), and rest, ME/CFS patients also present a plethora of unspecific downstream freeze/thaw processes, which may potentially symptoms including debilitating fatigue, unrefreshing sleep, explain the alterations in sensitive markers. Since our initial gastrointestinal problems, orthostatic intolerance, cognitive study was performed in fresh samples, it seemed necessary to impairments, and pain/myalgia that change in frequency and confirm the cryostability of the discriminant immune markers severity over time and differ from patient to patient (2). identified before proceeding with population validation. No Despite exhaustive research in the past decades, the previous studies have explored the effects of storage time of pathogenesis of ME/CFS is not yet completely understood; blood samples on the cryostability of immune markers in prognostic indicators, efficient targeted treatments or coping ME/CFS, and to the best of our knowledge, no studies have clinical approaches remain elusive. It has been hypothesized that attempted to establish standard operating procedures on ME/CFS is a multifactorial condition in which a combination of ME/CFS-focused sample quality control. immunogenetic and shared environmental factors coexist (3). The The aim of this study was to assess the impact of long-term immunological basis of the disease is highlighted by evidence biobanking on our previously identified immune markers in linking it to autoimmunity, and by the potential role of pathogens peripheral blood mononuclear cell (PBMC) subsets of as triggers (4–6), leading to the general belief that some disrupted individuals with ME/CFS and healthy controls and to explore immunometabolic and neuroinflammatory pathways may be additional immune markers in frozen Treg and NK cells from the associated with ME/CFS (7–9). same previously studied cohort. Ideally, study biomarkers Currently, there is no single accurate ‘gold standard’ unaffected by sample cryopreservation would be identified. biomarker for diagnosis of ME/CFS. Several immune mediators, such as circulating cytokines/chemokines or immune cell subset markers, have been investigated as candidate biomarkers for the disease. These biomarkers may MATERIALS AND METHODS become valuable tools for the diagnosis, prognosis and response to treatment in ME/CFS sufferers (10). To this end, our group Ethics Statement examined the phenotype and function of NK, B and T cell This study was carried out in accordance with the subsets in 22 Spanish ME/CFS patients, concluding that the recommendations of the local Research Ethics Committees. All cases showed an unaffected B-cell compartment and an altered T subjects provided written signed informed consent prior to and NK cell phenotype. Higher expression of NKp46 and CD69 participation in accordance with the Declaration of Helsinki. markers and lower expression of CD25 in NK cells were The study protocol was approved by the local Health Research evidenced, along with an apparently poor response of the T- Ethics Board (IrsiCaixa, IGTP Committee, Badalona, Spain; cell compartment, with increased numbers of regulatory T cells reference number: EO-10-007). (Tregs) and lower activation and frequency of effector memory CD8+ T cells. Furthermore, particular combinations of these Study Setting and Participants disease-biased parameters identified eight robust immune In this retrospective, cross-sectional, case-control cohort study, markers that were able to discriminate between ME/CFS 22 potentially eligible ME/CFS patients who met the 1994 patients and healthy controls with a sensitivity of 100% and CDC/Fukuda definition (13) and recruited from two cohorts specificity of 79% based on ROC analysis (7). of ME/CFS outpatient clinics (Clinics Omega Zeta, Tarragona, Biomarker validation studies involve large numbers of Spain and Vall d’Hebron University Hospital, Barcelona, biospecimens that could be effectively supplied by biobanks. Spain) were initially evaluated by two physicians specialized Frontiers in Immunology | www.frontiersin.org 2 November 2020 | Volume 11 | Article 582330 Gómez-Mora et al. Biomarkers in Cryopreserved ME/CFS Immune Cells in ME/CFS (JR and JA) as part of our initial study on immune washed and resuspended in a standard cryomedium containing biomarker discovery in 2013 (7). The 18 PBMC samples 90% fetal bovine serum (FBS, Gibco catalog no. 10091148) plus analyzed from each cohort were selected after three years of 10% dimethyl sulfoxide (DMSO, Sigma-Aldrich, catalog no. cryopreservation on the basis of cell availability for the D2650). One milliliters aliquots with 1x107 cells/ml were programmed assays. All routine laboratory testing yielded gradually frozen in Mr. Frosty device at −80°C for at least 48 h normal results (data not shown). As previously reported, and then stored into a cryo-tank containing liquid nitrogen at fatigue severity was assessed through the self-administered −196°C until use. 40-item Fatigue Impact Scale (FIS-40) validated in the Spanish ME/CFS population (14, 15). Unfortunately, no cells Cell Viability and Staining Assay were available from the same healthy controls of the original Cryopreserved PBMC-containing vials were rapidly thawed 2013 study. Therefore, eighteen age- and sex-matched healthy into pre-warmed bath at 37°C for 5 min and washed with controls from the local community recruited through word-of- PBS, and 1x106 cells were stained with eFluor™ 506 dye (200 mouth and advertisements and having cryopreserved PBMC mg/ml, ThermoFisher catalog no. 65-0866-18) to analyze cell samples (≥ 3 years) were selected. All participants were of viability using flow cytometry analysis. Cellular yield was Caucasian descent, from the same geographical area, and had a calculated as number of living cells/ml. Viability was defined sedentary lifestyle at the time of study. Table 1 shows the as % of living cells from total cell count. Cells negative for sociodemographic and clinical characteristics of the study eFluor™ 506 stains were regarded as viable. PBMCs were then participants. Exclusion criteria were acute infectious disorders washed with PBS and stained with different combinations of in the previous 4 weeks, past/present neurological, metabolic or fluorochrome-conjugated anti-human antibodies. Besides the infectious disorders, immune disorders such as IgA deficiency, antibodies previously used to compare expression between psychosis, major depression, cardiac, autoimmune and fresh blood and frozen PBMCs (7), four new combinations of hematological disorders, allergy-related diseases, dermal or antibodies were added in order to explore new immune chronic inflammation disorders; medical conditions that markers in the study (Table 2). required glucocorticoid, statin, or antiviral drug use; pregnancy/breastfeeding; or any symptom/signs that might be T Effector Cell Immunophenotyping confused with those of ME/CFS. PBMCs (0.5 million) were incubated for 15 min at room temperature with the anti-CD3-APC-Cy™7 (clone SP34-2), Blood Collection and Processing anti-CD4-PerCP (clone L78SK3), anti-CD8-V500 (clone SK1), Peripheral blood (25 ml) was collected by venipuncture directly anti-CD27-APC (clone M-T271) and anti-CD28-PE (clone into EDTA-containing tubes (K2E Vacutainer, BD) from each CD28.2) monoclonal antibodies (all from BD Biosciences) participant. An aliquot was used for immediate according to standard protocols. Cells were then washed with immunophenotyping on the day of blood collection; remaining PBS, fixed with 1% formaldehyde in PBS (Sigma-Aldrich, catalog whole blood was processed for PBMC isolation by density no. 1004960700) for 15 min at room temperature and analyzed gradient centrifugation using Ficoll ® Paque Plus (GE by flow cytometry. Healthcare, catalog no. 17-1440-02). Isolated PBMCs were TABLE 2 | Panel description of fluorochrome-conjugated monoclonal antibodies used in this study for staining by multicolor flow cytometry. TABLE 1 | Demographic and clinical characteristics of the study population at baseline. Panel 1 Panel 2 Panel 3 Panel 4 Variables ME/CFS HCs Fluorochromes Treg cells T cell function Effector T cells NK cells (n = 18) (n = 18) eFluor™ 506 Viability Viability Viability Viability Age (years) 50 (45–57) 48 (38–59) FITC Ki67 IFN-g – – Sex (female/male) 13/5 13/5 PE FOXP3 FOXP3 CD28 NKG2C Illness duration at diagnosis (years) 3.2 ± 7.2 – Fatigue severity 3 (2–3) – PE-Cy™7 – IL-4 – CD56 Triggers PerCP-Cy™5.5 CD3* CD3* CD4 CD69* Infectious 63 – APC – – CD27 CD57 Non-infectious 37 – APC-H7 CD4* CD4* CD3 CD16 Comorbid conditions Alexa Fluor® 700 CD8* CD8* – – Anxiety/depression 47 – PE-CF594 CD25* CD25* – CD25* Myalgia/tendinopathy 36 – Alexa Fluor® 647 CD127 CD127 – – Multiple Chemical Sensitivity 14 – PE-Cy™7 CD45RA – – – Medications V450 – – – NKp46 Analgesics/NSAIDs 33 – V500 – – CD8 – Antidepressants/anxiolytics 44 – BV421 CD73 TGF-b1 – – Homeopathy 44 – BV510 – – – CD3/CD19 Antioxidant supplements 44 – BV650 CD39 – – – BV786 – IL-17 – – Data are expressed as median ± standard deviation (SD) for continuous variables, and as number of cases (%) for categorical variables, except for age and fatigue severity (median, •Asterisks denote a change in the fluorochromes. interquartile range). NSAIDs, non-steroidal anti-inflammatory drugs. •Antibodies highlighted in bold denote new immune markers added in this study. Frontiers in Immunology | www.frontiersin.org 3 November 2020 | Volume 11 | Article 582330 Gómez-Mora et al. Biomarkers in Cryopreserved ME/CFS Immune Cells Regulatory T Cell Immunophenotyping (clone TW4-9E7) (all from BD Biosciences). At this point, cells To determine the Treg profile and function, 1x106 PBMCs were were washed twice with PBS and fixed with PBS containing 1% incubated for 15 min at room temperature with the anti-CD3- formaldehyde (Sigma-Aldrich, catalog no. 1004960700). As PerCP-Cy™5.5 (clone UCHT1), anti-CD4-APC-H7 (clone negative control, unstimulated cells were included in each SK3), anti-CD8-Alexa Fluor® 700 (clone RPA-T8), anti-CD25- experiment. All stained samples were acquired on an PE-CF594 (clone M-A251), anti-CD127-Alexa Fluor® 647 LSRFortessa flow cytometer using a plate HTS loader (BD (clone HIL-7R-M21), anti-CD45RA-PE-Cy™7 (clone HI100), Biosciences), except for T effector cell immunophenotyping anti-CD39-BV650 (clone TU66), and anti-CD73-BV421 (clone (LSR-II flow cytometer, BD Biosciences). Data analysis was AD2) monoclonal antibodies (all from BD Biosciences). Cells performed using FlowJo LLC software v10.4.2 (Tree Star, were then washed and fixed/permeabilized (eBioscience, catalog Ashland, OR, USA). A minimum of 10,000 total events were no. 00-5521-00) using FOXP3 staining buffer (eBioscience, recorded for each panel and condition. Although most catalog no. 00-5523-00), and finally stained with the antibodies were maintained from our original study, the combination of conjugated antibodies anti-Ki67-FITC (BD addition of new markers (highlighted in bold on Table 2) and Biosciences, catalog no. 556026) and anti-FOXP3-PE (BD the changes in configuration of the flow cytometer resulted in Biosciences, clone 259D/C7) for nuclear epitope staining. Next, fluorochrome changes in several markers (marked by asterisks cells were washed with PBS, fixed with 1% formaldehyde in PBS on Table 2). We tried to minimize the impact of these changes by (Sigma-Aldrich, catalog no. 1004960700) for 15 min at room restricting them to highly expressed molecules (i.e., CD3, CD4 temperature and analyzed by flow cytometry. or CD8). NK Cell Immunophenotyping Statistical Analysis 6 For NK characterization, 2x10 PBMCs were incubated for Continuous variables were expressed as medians ± IQR 15 min at room temperature with the following conjugated (interquartile range). Qualitative variables were expressed as antibodies: anti-CD3-BV510 (clone UCHT1), anti-CD19- percentages. Descriptive statistics and data visualization BV510 (clone SJ25C1), anti-CD16-APC-Cy™7 (clone 3G8), (graphs) were generated using GraphPad Prism version 7.0 anti-CD56-PE-Cy™7 (clone B159), anti-NKp46-V450 (clone (GraphPad Software Inc., San Diego, USA). Group 9E2), anti-NKG2C-PE (clone 134591), anti-CD69-PerCP comparisons were performed by either Chi-square test for (clone L78), anti-CD57-APC (clone NK-1), and anti-CD25- continuous variables or the Fisher’s exact test for categorical PE-CF594 (clone M-A251) monoclonal antibodies (all from variables. Differences between quantitative variables were BD Biosciences). Cells were then washed with PBS, fixed with compared using the non-parametric Mann-Whitney U test or 1% formaldehyde in PBS (Sigma-Aldrich, catalog no. c2 test, as appropriate. Comparisons between fresh and thawed 1004960700) for 15 min at room temperature, and analyzed by samples were assessed in paired data using the Wilcoxon signed- flow cytometry. rank test (two-tailed). Correlation analysis between continuous variables was calculated using the non-parametric Spearman Intracellular Cytokine Staining Assay rank test to explore the nature of the relationship between two Briefly, 2 × 106 PBMCs per condition were thawed, washed and continuous variables and multiple testing and further adjusted by placed in culture at 4 million cells/ml in RPMI-1640 medium the false discovery rate. The comparison of slope values with a (Gibco by Life Technologies, catalog no. 11875-093) full identity (slope = 1) was performed after linear regression supplemented with penicillin/streptomycin (Gibco by Life analysis using the F-test. All statistical analyses were performed Technologies, catalog no. 15140122) and 10% FBS (Seradigm, using SPSS v23.0 (SPSS, Chicago, IL, USA). All experiments catalog no. 1500-500). After 2 h, cells were stimulated “in vitro” included a minimum of three independent replicates. Statistical with PMA (62.5 ng/ml, Sigma-Aldrich, catalog no. P1585) and significance was established at a p-value ≤ 0.05. All p-values are ionomycin (0.6 µM, Sigma-Aldrich, catalog no. I9657) to induce documented in the corresponding figures and legends. cytokine production in the presence of brefeldin A (10 mg/ml, BD Biosciences, catalog no. 555029) and monensin (2 mM, BD Biosciences, catalog no. 554724) and incubated for 5 h at 37°C RESULTS as described (16). Cells were then stained for 15 min with anti- CD3-PerCP-Cy™5.5 (clone UCHT1), anti-CD4-APC-H7 (clone Population Characteristics RPA-T4), anti-CD8-Alexa Fluor® 700 (clone RPA-T8), anti- ME/CFS participants were selected according to sample CD25-PE-CF594 (clone M-A251), and anti-CD127-Alexa availability from the previously defined cohort (7). Control Fluor® 647 (clone HIL-7R-M21) conjugated antibodies (all individuals were selected from biobanked samples and matched from BD Bioscience), washed and fixed/permeabilized by age and gender; the overlap with the former control group was (eBioscience, catalog no. 88-8824-00) using FOXP3 staining minimal, which precluded direct comparison of fresh and buffer (eBioscience, catalog no. 00-5523-00), and finally stained cryopreserved samples in this group. As shown in Table 1, no with the following intracellular monoclonal antibodies: anti- differences in age or gender were observed among participants. IFN-g FITC (clone B27), anti-IL-17A-BV786 (clone N49-653), Most participants reported infection as the disease trigger (63%). anti-IL-4-PE-Cy™7 (clone 8D4-8), and anti-TGF-b1-BV421 Additional factors that may have altered the immune function in Frontiers in Immunology | www.frontiersin.org 4 November 2020 | Volume 11 | Article 582330 Gómez-Mora et al. Biomarkers in Cryopreserved ME/CFS Immune Cells ME/CFS (comorbidity and polypharmacy) were also recorded. and 12 healthy controls were no longer available. Table 2 shows Among comorbidities, anxiety/depression was the most prevalent the panel of fluorochrome-conjugated monoclonal antibodies (47%), although myalgia/tendinopathy (36%) and self-reported used in the immunophenotypic characterization of frozen multiple chemical sensitivity (14%) also appeared frequently. The PBMCs. For the previously identified T cell biomarkers (7), we main potential interferences of self-reported poly-drugs in the measured the frequency of the Treg cells CD3+CD4+CD25bright ME/CFS cohort studied were homeopathy (44%), antioxidant FOXP3+CD127- (natural Tregs) using the antibodies listed as supplements (44%), NSAIDs (33%, mainly paracetamol), and panel 1 and quantified them as % of CD4+ T cells. The same antidepressants/anxiolytics (44%), which were evenly distributed panel also allowed for the quantification of the percentage of across participants. CD4+ T cells expressing the proliferation marker Ki67. The intracellular cytokine staining assay was performed using panel Validation of Previously Described 2 antibodies. In addition, effector CD8+ T cells were quantified as Markers CD3+CD8+CD28+CD27- using panel 3 antibody combinations To evaluate the stability after cryopreservation of the ME/CFS as listed, and the frequency of CD56+CD3+ T cells was assayed immune markers identified in our previous study and thus to with panel 4 antibodies. The latter panel was also used to analyze determine the possibility of applying our diagnostic criteria to NK cell subpopulations, which were defined as CD3-CD19- biobanked ME/CFS collections with a view to population CD56+CD16+, which the expression of the potential ME/CFS validation, we selected the most prominent subpopulation markers CD25, CD69, NKp46 and CD57 was analyzed. frequencies and immune markers found to be altered in fresh ME/CFS blood samples. We reassessed them on thawed replica Analysis of T Cell Markers in Fresh and aliquots using a similar methodology to that applied in our Frozen Samples original 2013 study (7), after a three-year interval. The cohorts Comparison of ME/CFS T cell markers in fresh blood and consisted of 18 patients with ME/CFS and 18 matched healthy cryopreserved PBMCs from ME/CFS participants is shown in controls, since aliquots from four of the original ME/CFS cases Figure 1. The percentage of natural Treg cells (CD3+CD4+ A B C D FIGURE 1 | Analysis of T cell subset markers in fresh and frozen samples from ME/CFS patients (A–D). Frequencies of the analyzed T cell subsets in fresh (2013) and frozen (2016) PBMC samples from the same participants are shown. Complete identity between fresh and frozen cells is illustrated by blue dotted lines. Linear regression of data is illustrated by red lines. Spearman correlation coefficients and P-values are shown for each panel. Frontiers in Immunology | www.frontiersin.org 5 November 2020 | Volume 11 | Article 582330 Gómez-Mora et al. Biomarkers in Cryopreserved ME/CFS Immune Cells FOXP3brightCD25+CD127-) showed significant correlations (r = To further confirm this possibility, frozen PBMC samples 0.50; p = 0.043) between frozen and fresh cells, with a trend from ME/CFS cases and healthy controls were compared. towards decreased frequencies in frozen samples (Figure 1A). Unexpectedly, none of the markers that had shown significant Similarly, significant correlations were obtained for classic Treg differences in our previous study performed in 2013 presented cells (CD3+CD4+FOXP3brightCD25+; r = 0.54, p = 0.032) (data statistically significant differences across groups (Figure 2). not shown). A non-significant correlation (r = 0.407; p = 0.106) Although ME/CFS patients showed a trend towards lower was observed for the frequency of CD4+Ki67+ cells which tended levels of effector CD8+ T cells and CD56+ T cells and higher to show higher values in frozen samples compared to fresh blood levels of natural Treg cells, consistent with the differences found data (Figure 1B). By contrast, CD56 expression in T cells (NKT in our 2013 study, on this occasion, no significant differences cells) and the frequency of effector CD8+ T cells showed better could be established between groups. The fact that two of the correlation between fresh and frozen samples (r = 0.715; p = immune cell subsets with lower p-values (Tregs and effector T 0.004 and r = 0.635; p = 0.007, respectively), with linear cells, p = 0.043 and p = 0.007; respectively) presented significant regressions close to identity values (slope closer to 1) (Figures correlations with fresh blood values (Figures 1A, D) argues in 1C, D). These results indicate the robust reproducibility, and favor of pursuing the evaluation of these cell subsets of larger sets thus cryostability, of some markers assayed during long-term of ME/CFS biobanked samples, in preference to the other two T storage periods, with the exception of Ki67. However, despite the cell subpopulations assayed (CD3+CD56+ and CD4+Ki67+) significant correlations, the slopes of linear regression analysis (Figure 2). indicated significant differences from identity (slope = 1) for natural Tregs and Ki67+ CD4 T cells (p = 0.0008 and p = 0.0035, Analysis of NK Cell Surface Markers in respectively, F-test). Slope value > 1 was found for natural Tregs Fresh and Frozen Samples indicating weaker expression in frozen compared to fresh cells, A similar cryostability analysis was performed of four additional while the slope value < 1 for Ki67+ cells, indicated the opposite. NK cell immune markers after long-term storage and freeze/ These changes were confirmed by comparing data obtained from thawing process (Figure 3). This time, particularly poor fresh and frozen samples using the Wilcoxon paired test, which correlations were observed for both CD69 and CD25 between showed significantly lower values for the frequency of natural fresh and frozen samples (Figures 3A, B) with slope values Tregs (p = 0.0002) and significantly higher values for Ki67+ CD4 significantly lower than one (p< 0.0001 for both, F-test), T cells (p = 0.0461) in frozen samples (Supplementary Figure indicating increased staining for both markers after the 1A, upper panel). However, no statistically significant differences thawing process (p = 0.0002 and p = 0.0003 respectively, were shown for the frequency of CD56+ cells (p = 0.5995) and Wilcoxon paired test, Supplementary Figure 1B, upper panel). effector CD8 T cells (p = 0.3289) (Supplementary Figure 1A, A similarly poor correlation was also observed for the expression lower panel). Despite the potential impact of cryopreservation, of the NKp46 marker (Figure 3C), although in this case the lack our data suggest that the use of biobanked PBMCs to validate of concordance in NKp46 expression between fresh and frozen previously identified immune markers may be feasible. NK cells was the consequence of a severe reduction of staining FIGURE 2 | Analysis of T cell subset markers in frozen PBMC samples from individuals with ME/CFS and healthy controls. The indicated T cell subsets were assessed in frozen PBMC samples from 18 ME/CFS individuals (brown) and 18 healthy controls (green). Data is shown as median values with interquartile range (boxes), plus minimal and maximal observations (bars). P-values are indicated for each set of groups compared. Frontiers in Immunology | www.frontiersin.org 6 November 2020 | Volume 11 | Article 582330 Gómez-Mora et al. Biomarkers in Cryopreserved ME/CFS Immune Cells A B C D FIGURE 3 | Analysis of NK cell subset markers in fresh and frozen samples from ME/CFS patients (A–D). Frequencies of the analyzed NK cell subsets in fresh (2013) and frozen (2016) blood samples from the same participants are shown. As in Figure 1, complete identity between fresh and frozen cells is illustrated by blue dotted lines. Linear regression of data is illustrated by red lines. Spearman correlation coefficients and P-values are shown for each panel. upon biobanking with a slope value significantly higher than one (Figure 3C) further argues against future attempts to validate (p = 0.0324, F-test), indicating decreased staining after the markers in samples other than fresh blood. thawing process (p = 0.0015, Wilcoxon paired test, Supplementary Figure 1B, lower panel). Only the percentage Assessment of New T Cell Immune of CD57+ NK cells showed a highly significant correlation (r = Markers 0.909, p< 0.001, Spearman test) (Figure 3D), with a slope value In addition to the immune markers described in our previous in linear regression analysis close to one (p = 0.3587, F-test). study (7), some authors have reported altered expression of However, significantly lower values were found in frozen PBMC other immune markers in ME/CFS patients (9, 17–19), some of samples (p = 0.0129, Wilcoxon paired test, Supplementary them related to Treg cells. Since the Treg compartment is one Figure 1B, lower panel). This suggests that three of the four of the most altered subsets in ME/CFS, we decided to include NK cell markers assayed were affected by sample biobanking, additional markers which may broaden ME/CFS biomarker and so their expression cannot be extrapolated to in vivo values. validation options with biobanked samples. In particular, we Thus, the validation of the differences of ME/CFS NK analyzed CD45RA expression to define subpopulations of subpopulations in our previous study with biobanked PBMCs naïve (CD45RA+) and induced/memory (CD45RA-) among is discouraged. Although the lack of differences for the highly natural Treg (CD25+ CD127- FOXP3 + ) cells (20). We also correlated CD57 marker between frozen PBMCs from identified FOXP3 bright CD45RA - cell subsets (Figure 4A) individuals with ME/CFS and healthy controls agreed with the which were previously reported to correspond to highly findings of our 2013 study, as in the case of T cell markers, no active Tregs (21) and analyzed the expression of CD39 (ecto- statistical differences for the previously identified ME/CFS nucleoside triphosphate diphosphohydrolase-1) and CD73 NKp46 upregulated markers could be confirmed in frozen (ecto-5′-nucleotidase), since these markers play a major role PBMCs (Supplementary Figure 2). The lack of correlation in ATP catabolism and adenosine production (22). Both between NKp46 values in fresh and frozen ME/CFS samples enzymes are important mediators of Treg activity (23) and Frontiers in Immunology | www.frontiersin.org 7 November 2020 | Volume 11 | Article 582330 Gómez-Mora et al. Biomarkers in Cryopreserved ME/CFS Immune Cells finding that identifies this new subset of Treg cells, which are also relevant in other chronic pathologies (25) as possible A candidate markers that should be explored further in biobanked specimens. Intracellular Cytokine Analysis in Frozen Samples As changes were found in immune markers associated with the process of cryopreservation of ME/CFS PBMCs, we were interested in testing whether biobanking also affected their cell function. Besides Treg frequency or function, altered levels of circulating inflammatory cytokines and a skewed Th1/Th2 balance have been associated with ME/CFS in recent years (26– 29). Therefore, we evaluated the ability of thawed T cells to B produce several relevant cytokines. With this aim in mind, cultured T cells were stimulated “in vitro” to produce classical pro-inflammatory (IFN-g and IL-17) and anti-inflammatory cytokines (IL-4 and TGF-b1). TGF-b1 was of particular interest due to its relevance in ME/CFS and in the generation and function of Treg cells (30). Our data show that stimulating cytokine expression from frozen ME/CFS and healthy controls PBMCs is feasible (Figure 5A); however, the comparison of cytokine production by stimulated T cells did not show statistically significant differences between cases and healthy controls for any of the cytokines assayed (Figure 5B). The observation that T cells could respond to the stimulus C applied supports the potential use of biobanked samples for further functional studies of cytokine profiling in ME/CFS. Unfortunately, no data from the corresponding fresh samples were available and therefore the question of whether biobanking affects ME/CFS T cell response to stimulation and cytokine production in different ways remains unanswered. Analysis of New Immune Markers in NK Cells in Frozen Samples Since over 60% of the ME/CFS cohort reported an infectious trigger FIGURE 4 | Analysis of new Treg subset markers. (A) Gating strategy of (Table 1), we expanded the previous characterization of NK cells in Treg cells, first identified by high CD25+ and FOXP3+ expression (classic Treg) and defined as natural Tregs by the lack of CD127 expression. ME/CFS cases and healthy controls by analyzing the potential Coexpression of CD45RA, FOXP3 or co-expression of CD39 and CD73 as biomarker value of the CD57 expression with NKp46 or NKG2C. new potential Treg markers were assessed as indicated. (B, C) Show the The latter is considered an NK-cell surface marker that is strongly frequencies of the indicated cell subsets obtained for healthy controls (n = 18; affected during infection with herpesvirus and/or cytomegalovirus green) and ME/CFS patients (n = 18; brown). Data are presented as median (31). Although our previous study did not find significant with interquartile range (boxes) plus minimal and maximal values (bars). P- values are indicated for each set of groups compared. differences in the number of CD57+ NK cells between ME/CFS cases and healthy controls, we were still interested in evaluating the CD57 marker further, based on the observation that CD57 MFIs have been associated with ME/CFS (24). Our data show that (mean fluorescence intensity) yielded significant differences these markers can be readily quantified in frozen cells (Figure between groups (7), perhaps indicating the presence of other as 4A). However, the comparison of the different Treg cell subsets yet unidentified NK cell subsets in ME/CFS (i.e., lack of CD94/ defined either by CD45RA and FOXP3 or by CD39 and CD73 NKG2C receptor expression). Interestingly, NK cells expressing surface markers did not show statistically significant CD57 and NKG2C or CD57 and NKp46 were detected on thawed differences between ME/CFS individuals and healthy controls cells (Figure 6A). Additionally, a clear negative correlation of the (Figures 4B, 4C). Interestingly, despite the reduced number of latter set of markers (CD57 and NKp46) was evidenced (r = -0.58; samples analyzed, a trend towards increased levels of p< 0.0001) (Figure 6B), suggesting that they are expressed by CD39+CD73+ Treg cells in ME/CFS cases compared with different NK cell subsets. Although this observation may link healthy controls was observed (p = 0.06, Figure 4C), a increased NKp46 expression to lower CD57 levels in ME/CFS, Frontiers in Immunology | www.frontiersin.org 8 November 2020 | Volume 11 | Article 582330 Gómez-Mora et al. Biomarkers in Cryopreserved ME/CFS Immune Cells A B FIGURE 5 | Analysis of cytokine expression in frozen/thawed stimulated samples. (A) Representative example of cytokine production from ME/CFS PBMCs by gated CD4+ T cells. IFN-g and IL-17 pro-inflammatory cytokines, and IL-4 and TFG-b1 anti-inflammatory cytokines were evaluated as indicated. (B) Percentages of CD4+ cytokine producing T cells are shown for healthy controls (n = 18; green) and ME/CFS patients (n = 18; brown). Data are shown as median with interquartile range (boxes) and minimal and maximal values (bars). Unstimulated conditions were included as negative controls in each experiment. P-values are indicated for each set compared. the comparison of different subsets of NK cell markers defined by CD57 in NK cells, CD8+ effector T cells, CD56 expression in T NKG2C and NKp46 with CD57 did not show significant cells, and Treg frequency. In contrast, other markers showed differences between cases and controls (Figures 6C, D). The inconsistent expression in biobanked PBMCs when compared to observed effect of PMBC freeze/thawing on NKp46 expression their fresh counterparts: namely NKp46, CD69 and CD25 (Figure 3C) might have limited the discrimination potential of this expression on NK cells. The reasons for the lack of stability previously identified ME/CFS marker, either when analyzed alone may differ in each case. Indeed, cell surface markers show or in combination with the CD57. varying sensitivity to cryopreservation, with some markers, such as CD62L, clearly decreasing due to protein shedding (31). Main lineage markers may also be susceptible to DISCUSSION cryopreservation, mainly for NK cells (32, 33) however, the analysis of cryopreserved PBMCs has been standardized for Among the different immunological markers previously shown several diseases, in particular HIV infection (34, 35). In to be useful for discriminating between ME/CFS and healthy contrast, less information is available on other markers, in individuals (7), at least four showed notable stability in frozen particular those with a low expression, which may increase cells compared to fresh cell samples. These stable markers were: their signal due to nonspecific binding to damaged cells. Our Frontiers in Immunology | www.frontiersin.org 9 November 2020 | Volume 11 | Article 582330 Gómez-Mora et al. Biomarkers in Cryopreserved ME/CFS Immune Cells biobanked samples. In addition, the validation of CD25 and CD69 expression in NK cells as potential markers in frozen cells A B was also clearly compromised by the inconsistent and highly variable levels of expression observed in the biobanked cells, which effectively ruled out their use in future ME/CFS studies with frozen samples. Although some authors report robust stability among markers, including CD3, CD4, CD8, CD45RO, CD16, CD19, and CD56 after PBMC freeze/thawing (32), others stress the particular sensitivity of T cells (33, 34). Differences in freezing protocols, such as snap-freezing on liquid nitrogen versus the use of isopropyl-alcohol containers and slow pre-freeze at -80°C, or differences in the amounts of DMSO in the freezing medium are known to impact marker cryostability and cell function (32–34). C Some authors report that even the time-lapse between cell thawing and surface marker assays by cytometry can determine marker level outcomes and hamper biomarker discovery (33). Even anticoagulant use may impact downstream features of cryopreserved blood cells (32–35). Although the sensitivity to cryopreservation and freeze/thaw under identical conditions might be expected to impact PBMCs from any individual to the same extent, this is in fact not the case: the individual disease state itself may in principle also influence lymphocyte cryosensitivity, particularly in diseases related to immunometabolic imbalance such as ME/CFS. This is D an area that deserves future evaluation. We also analyzed several other parameters such as Treg cell phenotype (naïve, memory and effector, defined by CD45RA and FOXP3 expression levels), Treg functionality assessed by the expression of CD39 and CD73, CD4+ T cell cytokine expression profile and the combined expression of CD57 and NKp46 or NKG2C receptors in NK cells. Some of these parameters have previously been explored in ME/CFS patients, with inconsistent findings (6, 18, 24, 36). Our data indicated that the values obtained in frozen cells are similar to those described in the FIGURE 6 | Analysis of new NK cell markers. (A) Representative dot-plots literature for fresh cells (31, 37). However, none of the for the co-expression of CD57 and either NKG2C or NKp46 in gated parameters analyzed showed significant differences between CD56+CD16+NK cells. (B) Correlation between CD57 and NKp46 expression in CD56+CD16+ NK cells from frozen PBMC samples (18 healthy controls in ME/CFS cases and healthy controls. This is of particular green and 18 ME/CFS patients in brown). Linear regression of data is relevance in the case of Th17 cells, which represent a key pro- illustrated by a red line. Spearman correlation coefficients and P-values are inflammatory subset (38) with functional links to the gut shown. (C, D) Marker coexpression for healthy controls (n = 18; green) and microbiome (39, 40), a field of particular interest in ME/CFS individuals with ME/CFS (n = 18; brown) are shown as indicated. Data are shown as median with interquartile ranges (boxes) plus minimal and maximal that has been the subject of intense research over the past few values (bars). P-values are indicated for each set compared. years (41, 42). Similarly, no differences in pro- or anti- inflammatory cytokines between ME/CFS and control individuals were noticed in this study. Nevertheless, the low data show a clear decrease in NKp46 expression on frozen cells, number of PBMC samples included in the present study and the suggesting a certain level of instability of this molecule on the experimental approach (analyzing cytokine-producing cells surface of NK cells that can be explained by its already reported instead of functional cytokine release) may have hampered the susceptibility to cleavage by metalloproteases during NK cell detection of differences between ME/CFS and healthy controls. activation (31). This instability shows that distinguishing Exploring other specific (microbial) stimuli, particularly acting between ME/CFS and healthy individuals using biobanked on Pattern Recognition Receptors (such as PAMS, DAMPS and PBMCs is a particularly challenging endeavor; since NKp46 TLRs), could be a new strategy for detecting immune was among the best potential candidate biomarkers identified dysfunction in ME/CFS. in fresh blood samples (7), it was one of the main focuses in the In conclusion, this study was unable to validate immune present study. Thus, it seems that this major NK-cell activating biomarkers in ME/CFS, either alone or in combination with receptor cannot be validated as a biomarker of ME/CFS using other previously defined markers (such as the tandem NKp46 Frontiers in Immunology | www.frontiersin.org 10 November 2020 | Volume 11 | Article 582330 Gómez-Mora et al. Biomarkers in Cryopreserved ME/CFS Immune Cells and CD57), which unfortunately showed a high degree of immune markers which may contribute to the diagnosis and instability in frozen PBMC samples. These data are in prognosis of ME/CFS. The findings indicate the loss of specific agreement with a recent report of a large cohort of frozen PBMC subpopulations or their markers, particularly on T and PBMC samples from individuals with ME/CFS, and multiple NK cells. These observations draw attention to the challenges sclerosis and healthy individuals. In that study, similar levels of facing immune biomarker discovery in ME/CFS, as they limit the CD57, NKG2C, and NKp46 expression in NK cells were use of frozen specimens from ME/CFS biobanks worldwide. This described in each group; no differences in cytokine profile potentially critical limitation of sample storage should be expression by T cells (IL-2 and IFN-g) or NK cells (not considered in the design of future studies aimed at exploiting assessed in our study) were found (18). In contrast, recent the vast potential of ME/CFS biobanks for exploring large reports highlight the relevance of different metabolic populations of ME/CFS and for further validating new determinants or immune markers in ME/CFS (43, 44), candidate immune biomarkers of the disease. suggesting that functional analysis may provide further relevant diagnostic tools, although those assays may require le ss stand a rd ize d tech nique s th an fl ow cytometry. DATA AVAILABILITY STATEMENT Alternatively, immunometabolic dysfunction may affect cell The raw data supporting the conclusions of this article will be recovery from ME/CFS frozen PBMC samples (44); however, made available by the authors, without undue reservation. no differences in cell viability after thawing were observed in our cohort. The immunophenotypic analysis of frozen samples is a useful ETHICS STATEMENT tool to characterize immune responses against pathogens and has been used in numerous large intervention multisite studies All subjects provided written signed informed consent prior to for a range of medical conditions. However, this kind of participation in accordance with the Declaration of Helsinki. The approach has limitations, mainly due to the fact that some study protocol was approved by the Health Research Ethics Board markers are not stable after the freeze/thawing process of of IrsiCaixa, IGTP Committee (reference number: EO-10-007). biosamples. For this reason, we believed that it was essential to evaluate the behavior of the immune biomarkers identified for ME/CFS (7) before proceeding with validation studies that would AUTHOR CONTRIBUTIONS require large numbers of samples and would therefore benefit greatly from the use of biobanked samples. Our data suggest that EG-M, JC, CC, and JB initiated the concept of the project, it is impossible (or at best extremely difficult) to accurately supervised sampling, and freezing/thawing processing. EO, JC- measure certain markers in frozen immune cells, especially M, and JB collected and interpreted the data and wrote the CD25 and CD69 in NK cells, but also NKp46, whose manuscript. EG-M, VU, JR, and JA contributed to the study expression was strongly reduced after biobanking. The use of design, supervised the laboratory experiments, analyzed and metalloprotease inhibitors (45) or specific freezing protocols to performed the statistical analysis of data, and also contributed protect NK cells (46) may reverse this loss of signal and increase to writing the manuscript. All authors contributed to the article the potential of NKp46 as a biomarker for ME/CFS. and approved the submitted version. To the best of our knowledge, this is the first study to allow a direct comparison of PBMC immune markers in frozen and fresh FUNDING ME/CFS samples. It has served to identify important limitations of PBMC storage for use as potential ME/CFS immune markers. This study was partially supported by the Association of However, the small number of frozen samples used may have Healthcare Workers for ME/CFS Research (ASSSEM) in limited the statistical power to detect differences between groups. Catalonia, Spain. Funders had no role in the study design, data In addition, only limited information was available on serological collection and analysis, interpretation, decision of publish, or testing, which might have served to identify potential confounding writing of the manuscript. factors. Further investigations are required to clarify more precisely the immune pathomechanisms in people with ME/CFS. Research activity in this area is clearly increasing; with particularly promising ACKNOWLEDGMENTS data on immunometabolism of “exhausted” phenotype and function, and integrated multi-omics signatures in people with The authors thank all ME/CFS patients and healthy volunteers ME/CFS (10, 18, 47). for participating and supporting this study. We are also grateful to Rafaela Ayé n, Tania Gonzá lez and Eulalia Grau for processing and storage of blood specimens. The authors are particularly CONCLUSIONS AND FUTURE grateful to Marco A. Fernandez (IGTP Flow Cytometry Core DIRECTIONS Facility) for his technical support and advice designing the cytometry panels. Vall d’Hebron Hospital Research Institute In summary, our study suggests for the first time that PBMC (VHIR), IrsiCaixa and IGTP are members of the CERCA sample biobanking impacts some of the previously defined programme (Generalitat de Catalunya, Spain). We also thank Frontiers in Immunology | www.frontiersin.org 11 November 2020 | Volume 11 | Article 582330 Gómez-Mora et al. Biomarkers in Cryopreserved ME/CFS Immune Cells Michael Maudsley for linguistic advice regarding the final SUPPLEMENTARY FIGURE 1 | Frequency of different T cell subsets (panel A) and NK cell subsets (panel B) was assesed in fresh and frozen PBMC samples from version of the manuscript. individuals with ME/CFS (n = 18). Data are shown as individuals values with means (bars). P-values are shown for Wilcoxon paired test when comparing fresh and frozen samples for each immune marker. SUPPLEMENTARY MATERIAL SUPPLEMENTARY FIGURE 2 | Frequency of NK cell subsets was assessed in frozen PBMC samples from 18 healthy controls (HC, green) and 18 ME/CFS The Supplementary Material for this article can be found online patients (CFS, brown) as indicated. Data are displayed as median percentage with at: https://www.frontiersin.org/articles/10.3389/fimmu.2020. interquartile range (boxes), plus minimal and maximal observations (bars). P-values 582330/full#supplementary-material using the Mann-Whitney test are shown for each set compared. Prostaglandins Leukot Essent Fat Acids (2018) 139:20–4. doi: 10.1016/ REFERENCES j.plefa.2018.11.006 1. Institute of Medicine (IOM). Beyond myalgic encephalomyelitis/chronic 16. Massanella M, Negredo E, Puig J, Puertas MC, Buzó n MJ, Pé rez-Á lvarez N, fatigue syndrome: Redefining an illness. Washington, DC: The National et al. Raltegravir intensification shows differing effects on CD8 and CD4 T Academies Press (2015). cells in HIV-infected HAART-suppressed individuals with poor CD4 T-cell 2. Castro-Marrero J, Sá ez-Francàs N, Santillo D, Alegre J. Treatment and recovery. AIDS (2012) 26:2285–93. doi: 10.1097/QAD.0b013e328359f20f management of chronic fatigue syndrome/myalgic encephalomyelitis: All 17. Sotzny F, Blanco J, Capelli E, Castro-Marrero J, Steiner S, Murovska M, et al. roads lead to Rome. Br J Pharmacol (2017) 174:345–69. doi: 10.1111/ Myalgic Encephalomyelitis/Chronic Fatigue Syndrome: evidence for an bph.13702 autoimmune disease. Autoimmun Rev (2018) 17:601–9. doi: 10.1016/ 3. Carruthers BM, van de Sande MI, De Meirleir KL, Klimas NG, Broderick G, j.autrev.2018.01.009 Mitchell T, et al. Myalgic encephalomyelitis: International Consensus Criteria. 18. Cliff JM, King EC, Lee J-S, Sepú lveda N, Wolf A-S, Kingdon C, et al. Cellular J Intern Med (2011) 270:327–38. doi: 10.1111/j.1365-2796.2011.02428.x immune function in Myalgic Encephalomyelitis/Chronic Fatigue Syndrome 4. Bansal AS, Bradley AS, Bishop KN, Kiani-Alikhan S, Ford B. Chronic fatigue (ME/CFS). Front Immunol (2019) 10:796. doi: 10.3389/fimmu.2019.00796 syndrome, the immune system and viral infection. Brain Behav Immun 19. Rivas JL, Palencia T, Ferná ndez G, Garcı́a M. Association of T and NK cell (2012) 26:24–31. doi: 10.1016/j.bbi.2011.06.016 phenotype with the diagnosis of Myalgic Encephalomyelitis/Chronic Fatigue 5. Klimas NG, Salvato FR, Morgan R, Fletcher MA. Immunologic abnormalities Syndrome (ME/CFS). Front Immunol (2018) 9:1028. doi: 10.3389/ in chronic fatigue syndrome. J Clin Microbiol (1990) 28:1403–10. doi: fimmu.2018.01028 10.1128/JCM.28.6.1403-1410.1990 20. Cuadrado E, van den Biggelaar M, de Kivit S, Chen Y-Y, Slot M, Doubal I, 6. Brenu EW, van Driel ML, Staines DR, Ashton KJ, Hardcastle SL, Keane J, et al. et al. Proteomic analyses of human regulatory T cells reveal adaptations in Longitudinal investigation of natural killer cells and cytokines in chronic signaling pathways that protect cellular identity. Immunity (2018) 48:1046– fatigue syndrome/myalgic encephalomyelitis. J Transl Med (2012) 10:88. doi: 59. doi: 10.1016/j.immuni.2018.04.008 10.1186/1479-5876-10-88 21. Sugiyama D, Nishikawa H, Maeda Y, Nishioka M, Tanemura A, Katayama I, 7. Curriu M, Carrillo J, Massanella M, Rigau J, Alegre J, Puig J, et al. B and T cell et al. Anti-CCR4 mAb selectively depletes effector-type FoxP3+CD4+ phenotype and function in patients suffering from Chronic Fatigue regulatory T cells, evoking antitumor immune responses in humans. Proc Syndrome. J Transl Med (2013) 11:68. doi: 10.1186/1479-5876-11-68 Natl Acad Sci (2013) 110:17945–50. doi: 10.1073/pnas.1316796110 8. Brenu EW, Staines DR, Baskurt OK, Ashton KJ, Ramos SB, Christy RM, et al. 22. Blanco J, Canela EI, Sayó s J, Mallol J, Lluis C, Franco R. Adenine nucleotides Immune and hemorheological changes in chronic fatigue syndrome. J Transl and adenosine metabolism in pig kidney proximal tubule membranes. J Cell Med (2010) 8:1. doi: 10.1186/1479-5876-8-1 Physiol (1993) 157:77–83. doi: 10.1002/jcp.1041570110 9. Brenu EW, van Driel ML, Staines DR, Ashton KJ, Ramos SB, Keane J, et al. 23. Ohta A, Sitkovsky M. Extracellular adenosine-mediated modulation of Immunological abnormalities as potential biomarkers in chronic fatigue regulatory T cells. Front Immunol (2014) 5:304. doi: 10.3389/ syndrome/myalgic encephalomyelitis. J Transl Med (2011) 9:81. doi: fimmu.2014.00304 10.1186/1479-5876-9-81 24. Brenu EW, Huth TK, Hardcastle SL, Fuller K, Kaur M, Johnston S, et al. Role 10. Strawbridge R, Sartor M-L, Scott F, Cleare AJ. Inflammatory proteins are altered in of adaptive and innate immune cells in Chronic Fatigue Syndrome/Myalgic chronic fatigue syndrome: A systematic review and meta-analysis. Neurosci Encephalomyelitis. Int Immunol (2014) 26:233–42. doi: 10.1093/intimm/ Biobehav Rev (2019) 107:69–83. doi: 10.1016/j.neubiorev.2019.08.011 dxt068 11. Irlbeck DM, Vernon SD, McCleary KK, Bateman L, Klimas NG, Lapp CW, 25. Umansky V, Shevchenko I, Bazhin AV, Utikal J. Extracellular adenosine et al. No association found between the detection of either xenotropic murine metabolism in immune cells in melanoma. Cancer Immunol Immunother leukemia virus-related virus or polytropic murine leukemia virus and chronic (2014) 63:1073–80. doi: 10.1007/s00262-014-1553-8 fatigue syndrome in a blinded, multi-site, prospective study by the 26. Hornig M, Montoya JG, Klimas NG, Levine S, Felsenstein D, Bateman L, et al. establishment and use of the SolveCFS Biobank. BMC Res Notes (2014) Distinct plasma immune signatures in Mylagic Encephalomyelitis/Chronic 7:461. doi: 10.1186/1756-0500-7-461 Fatigue Syndrome are present early in the course of illness. Sci Adv (2015) 12. Lacerda EM, Mudie K, Kingdon CC, Butterworth JD, O’Boyle S, Nacul L. The 1:1400121. doi: 10.1126/sciadv.1400121 UK ME/CFS biobank: a disease-specific biobank for advancing clinical 27. Broderick G, Fuite J, Kreitz A, Vernon SD, Klimas N, Fletcher MA. A formal research into myalgic encephalomyelitis/chronic fatigue syndrome. Front analysis of cytokine networks in chronic fatigue syndrome. Brain Behav Neurol (2018) 9:1026. doi: 10.3389/fneur.2018.01026 Immun (2010) 24:1209–17. doi: 10.1016/j.bbi.2010.04.012 13. Fukuda K, Straus SE, Hickie I, Sharpe MC, Dobbins JG, Komaroff A. The 28. Montoya JG, Holmes TH, Anderson JN, Maecker HT, Rosenberg-Hasson Y, chronic fatigue syndrome: a comprehensive approach to its definition and Valencia IJ, et al. Cytokine signature associated with disease severity in study. International Chronic Fatigue Syndrome Study Group. Ann Intern Med chronic fatigue syndrome patients. Proc Natl Acad Sci U S A (2017) 114: (1994) 121:953–9. doi: 10.7326/0003-4819-121-12-199412150-00009 e7150–8. doi: 10.1073/pnas.1710519114 14. Fisk JD, Ritvo PG, Ross L, Haase DA, Marrie TJ, Schlech WF. Measuring the 29. Moneghetti KJ, Skhiri M, Contrepois K, Kobayashi Y, Maecker H, Davis M, functional impact of fatigue: Initial validation of the fatigue impact scale. Clin Infect et al. Value of circulating cytokine profiling during submaximal exercise Dis (1994) 18(Suppl. 1):S79–83. doi: 10.1093/clinids/18.supplement_1.s79 testing in Myalgic Encephalomyelitis/Chronic Fatigue Syndrome. Sci Rep 15. Castro-Marrero J, Zaragozà MC, Domingo JC, Martinez-Martinez A, Alegre J, (2018) 8:2779. doi: 10.1038/s41598-018-20941-w von Schacky C. Low omega-3 index and polyunsaturated fatty acid status in 30. Zhang H-YY, Liu Z-DD, Hu C-JJ, Wang D-XX, Zhang Y-BB, Li Y-ZZ. Up- patients with chronic fatigue syndrome/myalgic encephalomyelitis. regulation of TGF-b1 mRNA expression in peripheral blood mononuclear Frontiers in Immunology | www.frontiersin.org 12 November 2020 | Volume 11 | Article 582330 Gómez-Mora et al. Biomarkers in Cryopreserved ME/CFS Immune Cells cells of patients with chronic fatigue syndrome. J Formos Med Assoc (2011) individuals with myalgic encephalomyelitis/chronic fatigue syndrome. 110:701–4. doi: 10.1016/j.jfma.2011.09.006 Microbiome (2016) 4:30. doi: 10.1186/s40168-016-0171-4 31. Björkström NK, Svensson A, Malmberg K-JJ, Eriksson K, Ljunggren H-GG. 42. Karhan E, Gunter CL, Ravanmehr V, Horne M, Kozhaya L, Renzullo S, et al. Characterization of natural killer cell phenotype and function during Perturbation of effect and regulatory T cell subsets in Myalgic recurrent human HSV-2 infection. PLoS One (2011) 6:e27664. doi: 10.1371/ Encephalomyelitis/Chronic Fatigue Syndrome. BioRxiv. doi: 10.1101/ journal.pone.0027664 2019.12.23.887505 32. Lauer FT, Denson JL, Burchiel SW. Isolation, cryopreservation, and 43. Esfandyarpour R, Kashi A, Nemat-Gorgani M, Wilhelmy J. Davis RW. A immunophenotyping of human peripheral blood mononuclear cells. Curr nanoelectronics-blood-based diagnostic biomarker for myalgic Protoc Toxicol (2017) 74:18201–16. doi: 10.1002/cptx.31 encephalomyelitis/chronic fatigue syndrome. Proc Natl Acad Sci U S A 33. Lemieux J, Jobin C, Simard C, Né ron S. A global look into human T cell (2019) 116:10250–7. doi: 10.1073/pnas.1901274116 subsets before and after cryopreservation using multi-parametric flow 44. Missailidis D, Sanislav O, Allan CY, Annesley SJ, Fisher PR. Cell-based blood cytometry and two-dimensional visualization analysis. J Immunol Methods biomarkers for Myalgic Encephalomyelitis/Chronic Fatigue Syndrome. Int J (2016) 434:73–82. doi: 10.1016/j.jim.2016.04.010 Mol Sci (2020) 21:1142. doi: 10.3390/ijms21031142 34. Sattui S, de la Flor C, Sá nchez C, Lewis D, Ló pez G, Rizo-Patró n E, et al. 45. Parsons MS, Tang C-CC, Jegaskanda S, Center RJ, Brooks AG, Stratov I, et al. Cryopreservation modulates the detection of regulatory T cell markers. Cytom Anti-HIV antibody-dependent activation of NK cells impairs NKp46 B Clin Cytom (2012) 82:54–8. doi: 10.1002/cyto.b.20621 expression. J Immunol (2014) 192:308–15. doi: 10.4049/jimmunol.1301247 35. Diks AM, Bonroy C, Teodosio C, Groenland RJ, de Mooij B, de Maertelaere E, 46. Veluchamy JP, Delso-Vallejo M, Kok N, Bohme F, Seggewiss-Bernhardt R, et al. Impact of blood storage and sample handling on quality of high van der Vliet HJ, et al. Standardized and flexible eight colour flow cytometry dimensional flow cytometric data in multicenter clinical research. panels harmonized between different laboratories to study human NK cell J Immunol Methods (2019) 475:112616. doi: 10.1016/j.jim.2019.06.007 phenotype and function. Sci Rep (2017) 7:43873. doi: 10.1038/srep43873 36. Metzger K, Fré mont M, Roelant C, De Meirleir K. Lower frequency of IL- 47. Naviaux RK, Naviaux JC, Li K, Bright AT, Alaynick WA, Wang L, et al. 17F sequence variant (His161Arg) in chronic fatigue syndrome patients. Metabolic features of chronic fatigue syndrome. Proc Natl Acad Sci U S A Biochem Biophys Res Commun (2008) 376:231–3. doi: 10.1016/ (2016) 113:e5472–80. doi: 10.1073/pnas.1607571113 j.bbrc.2008.08.135 37. Moreno-Fernandez ME, Rueda CM, Rusie LK, Chougnet CA. Regulatory T Conflict of Interest: JR is an employee at the Rigau Private Clinic. JB is the cells control HIV replication in activated T cells through a cAMP-dependent founder and CEO and JC is also co-founder and CSO (AlbaJuna Therapeutics, mechanism. Blood (2011) 117:5372–80. doi: 10.1182/blood-2010-12-323162 Barcelona, Spain). 38. Patel DD, Kuchroo VK. Th17 cell pathway in human immunity: Lessons from genetics and therapeutic interventions. Immunity (2015) 43:1040–51. doi: The remaining authors declare that the research was conducted in the absence of 10.1016/j.immuni.2015.12.003 any commercial or financial relationships that could be construed as a potential 39. Chege D, Sheth PM, Kain T, Kim CJ, Kovacs C, Loutfy M, et al. Sigmoid Th17 conflict of interest. populations, the HIV latent reservoir, and microbial translocation in men on long-term antiretroviral therapy. AIDS (2011) 25:741–9. doi: 10.1097/ Copyright © 2020 Gómez-Mora, Carrillo, Urrea, Rigau, Alegre, Cabrera, Oltra, QAD.0b013e328344cefb Castro-Marrero and Blanco. This is an open-access article distributed under the terms 40. Li Y, Wei C, Xu H, Jia J, Wei Z, Guo R, et al. The immunoregulation of Th17 of the Creative Commons Attribution License (CC BY). The use, distribution or in host against intracellular bacterial infection. Mediators Inflamm (2018) reproduction in other forums is permitted, provided the original author(s) and the 2018:6587296. doi: 10.1155/2018/6587296 copyright owner(s) are credited and that the original publication in this journal is 41. Giloteaux L, Goodrich JK, Walters WA, Levine SM, Ley RE, Hanson MR. cited, in accordance with accepted academic practice. No use, distribution or Reduced diversity and altered composition of the gut microbiome in reproduction is permitted which does not comply with these terms. Frontiers in Immunology | www.frontiersin.org 13 November 2020 | Volume 11 | Article 582330