Heliyon 7 (2021) e07665 Contents lists available at ScienceDirect Heliyon journal homepage: www.cell.com/heliyon Research article The SARS-CoV-2 receptor angiotensin-converting enzyme 2 (ACE2) in myalgic encephalomyelitis/chronic fatigue syndrome: A meta-analysis of public DNA methylation and gene expression data Jo~ ao Malato a, b, Franziska Sotzny c, **, Sandra Bauer c, Helma Freitag c, Andre Fonseca d, Anna D. Grabowska e, Luís Graça a, Clara Cordeiro b, d, Luís Nacul f, g, Eliana M. Lacerda f, Jesus Castro-Marrero h, Carmen Scheibenbogen c, Francisco Westermeier i, j, Nuno Sepúlveda b, c, k, * a Instituto de Medicina Molecular, Faculdade de Medicina, Universidade de Lisboa, Lisbon, Portugal b CEAUL – Centro de Estatística e Aplicaç~oes, Faculdade de Ci^encias, Universidade de Lisboa, Lisbon, Portugal c Charite - Universit€ atsmedizin Berlin, corporate member of Freie Universit€ at Berlin, Humboldt Universit€ at zu Berlin and Berlin Institute of Health, Institute of Medical Immunology, Berlin, Germany d Faculdade de Ci^encias e Tecnologia, Universidade do Algarve, Faro, Portugal e Department of Biophysics, Physiology, and Pathophysiology, Medical University of Warsaw, Warsaw, Poland f Department of Clinical Research, Faculty of Infectious and Tropical Diseases, London School of Hygiene & Tropical Medicine, London, United Kingdom g Complex Chronic Diseases Program, British Columbia Women's Hospital and Health Centre, Vancouver, British Columbia, Canada h Vall d’Hebron Hospital Research Institute, Division of Rheumatology, ME/CFS Unit, Universitat Autonoma de Barcelona, Barcelona, Spain i Institute of Biomedical Science, Department of Health Studies, FH Joanneum University of Applied Sciences, Graz, Austria j Centro Integrativo de Biología y Química Aplicada (CIBQA), Universidad Bernardo OHiggins, Santiago, Chile k Department of Infection Biology, Faculty of Infectious and Tropical Diseases, London School of Hygiene & Tropical Medicine, London, United Kingdom A R T I C L E I N F O A B S T R A C T Keywords: People with myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) often report a high frequency of viral Myalgic encephalomyelitis/chronic fatigue infections and flu-like symptoms during their disease course. Given that this reporting agrees with different syndrome immunological abnormalities and altered gene expression profiles observed in the disease, we aimed at answering SARS-CoV-2 whether the expression of the human angiotensin-converting enzyme 2 (ACE2), the major cell entry receptor for ACE2 Gene expression SARS-CoV-2, is also altered in these patients. In particular, a low expression of ACE2 could be indicative of a high DNA methylation risk of developing COVID-19. We then performed a meta-analysis of public data on CpG DNA methylation and gene expression of this enzyme and its homologous ACE protein in peripheral blood mononuclear cells and related subsets. We found that patients with ME/CFS have decreased methylation levels of four CpG probes in the ACE locus (cg09920557, cg19802564, cg21094739, and cg10468385) and of another probe in the promoter region of the ACE2 gene (cg08559914). We also found a decreased expression of ACE2 but not of ACE in patients when compared to healthy controls. Accordingly, in newly collected data, there was evidence for a significant higher proportion of samples with an ACE2 expression below the limit of detection in patients than healthy controls. Altogether, patients with ME/CFS can be at a higher COVID-19 risk and, if so, they should be considered a priority group for vaccination by public health authorities. To further support this conclusion, similar research is rec- ommended for other human cell entry receptors and cell types, namely, those cells targeted by the virus. * Corresponding author. ** Corresponding author. E-mail addresses: franziska.sotzny@charite.de (F. Sotzny), nunosep@gmail.com (N. Sepúlveda). https://doi.org/10.1016/j.heliyon.2021.e07665 Received 8 April 2021; Received in revised form 23 June 2021; Accepted 23 July 2021 2405-8440/© 2021 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). J. Malato et al. Heliyon 7 (2021) e07665 1. Introduction 2. Materials and methods Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) is 2.1. Eligible diagnostic criteria of ME/CFS a multifactorial and complex disease characterized by two key symp- toms: (1) persistent but unexplained fatigue that is not alleviated by In our meta-analysis, we selected public data from studies using either rest; and (2) post-exertional malaise upon minimal physical or even the 1994 Centre for Diseases Control criteria (1994 CDC/Fukuda) [1] or mental effort [1, 2]. Although its cause remains unknown, a growing the 2003 Canadian Consensus Criteria (2003 CCC) [2] for the disease body of evidence strongly associates ME/CFS with several microbial diagnosis. These criteria are defined by the presence of several key and viral infections, as potential triggering factors [3, 4]. In addition, it symptoms while excluding known medical conditions (e.g., multiple is currently hypothesized that reactivations of dormant viral infections sclerosis or lupus) that can also explain fatigue. The choice of using these also play a role [5, 6] due to several immunological abnormalities [7, 8, two criteria for study selection complies with the research standards set 9]. On the molecular basis of the disease, peripheral blood mononuclear by the European Network on ME/CFS [26]. cells (PBMCs) have altered gene expression profiles [10], including a decreased abundance of the human angiotensin-converting enzyme 2 (ACE2) [11], the main receptor of the severe acute respiratory syn- 2.2. Analysis of published DNA methylation association studies drome coronavirus-2 (SARS-CoV-2) for cell invasion [12, 13, 14]. Altogether, this evidence raises the question about the COVID-19 risk in Our meta-analysis was based on six genome-wide DNA methylation patients with ME/CFS. association studies (Table 1), four of which [27, 28, 29, 30] were pre- As basic information, ACE2 is encoded by the X-linked ACE2 gene viously reviewed [31], and other two published after this review [32, whose expression is predominant in the lungs, heart, skin, and kidneys 33]. Briefly, these studies aimed at identifying differentially methylated [15, 16, 17, 18]. Its expression can also be detected in monocytes [19] CpG dinucleotide sites between patients and healthy controls. Illumina and activated macrophages [20]. However, the percentage of methylation arrays were used to measure the respective DNA methyl- ACE2-expressing cells is below 5% in the main immune-cell populations ation levels with the exception of a single study (Table 1). In this study, [20]. Accordingly, current RNA-Seq studies suggest a residual ACE2 the measurements were made by the reduced representation bisulfite expression in PBMCs from healthy controls [18]. ACE2 has an sequencing [33]. amino-acid sequence identity of 41% with its homologous With respect to the exclusion criteria, one study excluded individuals angiotensin-converting enzyme (ACE) [21]. This sequence similarity who were taking beta-blockers or ACE inhibitors [30]. Three studies increases to 61% at the nucleotide level [21]. The enzymes ACE and excluded participants who were treated with immunomodulatory effects ACE2 are members of the renin-angiotensin-aldosterone system (RAAS), or affecting the underlying DNA methylation levels at the time of data which regulates blood pressure and vascular resistance [22]. In partic- collection [28, 29, 32]. ular, ACE and ACE2 have vasoconstriction and vasodilation effects, In four of the published DNA methylation studies, patients and respectively. Given this counteracting effect, high ACE:ACE2 ratios are healthy controls were matched for age, gender, and body mass index possible indicators of severe COVID-19 outcomes, linked to increased (Table 1) [28, 29, 30]. In two other studies, the matching was only based reactive oxygen species (ROS) production, vasoconstriction, and on age and gender [27, 33]. Ethnicity was also used for further matching inflammation [23]. [30, 32] or the same matching could be assumed in studies that only To answer our research question, we performed a meta-analysis of recruited white females [28, 29]. The DNA methylation levels were public DNA methylation and gene expression data of ACE2 and ACE in quantified in CD4þ T cells [27], PBMCs [28, 29, 30, 33], and T lym- PBMCs. Similar study was conducted on the DNA methylation pattern of phocytes [32]. ACE2 in the same cell type from patients with systemic lupus erythe- We conducted a joint analysis of the four array-based studies which matosus [24], an autoimmune disease whose symptoms overlap with the made the data available [28, 29, 30, 32]. We first retrieved the data from ones from ME/CFS [25]. To complement our findings, we also compared all the CpG probes located in the coding regions and the transcription the mRNA levels of these two genes in PBMCs from a new cohort of fe- starting sites (TSS) of ACE and ACE2, respectively. We then restricted our male patients with ME/CFS and healthy women. data analysis to the 27 probes shared between the Infinium Table 1. Summary of the six DNA methylation studies under analysis. Reference Sample ME/CFS patients Healthy Technology (manufacturer) NCBI GEO type controls, n Accession n Sample characteristics Case definition number [27] CD4þ T cells 25 Female/male adults 1994 CDC/Fukuda 18 Infinium HumanMethylation450K NA Mean age: 50 years old Array (Illumina) Mean BMI: not reported [28] PBMC 12 Female adults 1994 CDC/Fukuda & 12 Infinium HumanMethylation450K GSE59489 Mean age: 41 years old 2003 CCC Array (Illumina) Mean BMI: 23 kg/m2 [29] PBMC 49 Female adults 1994 CDC/Fukuda 25 Infinium HumanMethylation450 GSE93266 Mean age: 50 years old; Mean &2003 CCC Array (Illumina) BMI: 23 kg/m2 [30] PBMC 13 Female adults 1994 CDC/Fukuda & 12 Methylation EPIC Array (Illumina) GSE111183 Mean age: 50 years old Mean 2003 CCC BMI: 26 kg/m2 [32] T lymphocytes 61 Female/male adults 1994 CDC/Fukuda & 48 Infinium HumanMethylation450K GSE156792 Mean age: 32 years old 2003 CCC Array (Illumina) Mean BMI: 27 kg/m2 [33] PBMC 10 Female/male adults 2003 CCC 10 Reduced representation Bisulfite GSE153667 Mean age: Not reported sequencing Mean BMI: not reported 2 J. Malato et al. Heliyon 7 (2021) e07665 A B 0.05 1.00 0.75 Pearson's correlation coefficient 0.04 0.50 minor allele frequency 0.03 0.25 0.00 0.02 −0.25 −0.50 0.01 −0.75 0.00 −1.00 Europe North America Population C ACE − Overall ACE2 − Overall 1 1 adjusted p−value adjusted p−value 0.1 0.1 cg10468385 cg08559914 cg09920557 cg19802564 0.01 cg21094739 0.01 0.001 0.001 61.555 61.560 61.565 61.570 61.575 15.58 15.59 15.60 15.61 15.62 genomic position (chr 17) genomic position (chr X) D ACE − Females ACE2 − Females 1 1 adjusted p−value adjusted p−value 0.1 0.1 0.01 cg21094739 0.01 cg09920557 cg19802564 cg10468385 cg08559914 0.001 0.001 61.555 61.560 61.565 61.570 61.575 15.58 15.59 15.60 15.61 15.62 genomic position (chr 17) genomic position (chr X) E ACE − Males ACE2 − Males 1 1 adjusted p−value adjusted p−value 0.1 0.1 0.01 0.01 0.001 0.001 61.555 61.560 61.565 61.570 61.575 15.58 15.59 15.60 15.61 15.62 genomic position (chr 17) genomic position (chr X) Figure 1. DNA methylation analysis of 19 and 8 CpG probes located in the ACE and ACE2 genes, respectively. (A) Minor allele frequency in European and North American populations of SNPs located in the probes under analysis (see the respective data in Supplementary Table 2). (B) Boxplot of all possible Pearson's correlation coefficients (y axis) between the M-values of the probes under analysis. Horizontal dashed line represents the situation of lack of correlation. (C) Adjusted p-values for the overall association between each probe and ME/CFS. Adjusted p-values were calculated according to the Benjamini-Hochberg procedure with a false discovery rate of 5% (dashed line). Grey areas in the plots represent the TSS of the genes. (D) and (E) The same analyses as shown in C but for women and men separately. 3 J. Malato et al. Heliyon 7 (2021) e07665 HumanMethylation450K and the Infinium HumanMethylationEPIC ar- Pearson's correlation coefficient was used to check the validity of this rays (Supplementary Table 1). assumption (Figure 1B). Before conducting the statistical analysis itself, we checked whether We also repeated the same association analysis for women and men (1) the selected probes showed a high probability of detection, (2) they separately. Note that three studies only recruited women [28, 29, 30] were not cross-reactive with other genomic regions, and (3) they were while the remaining study recruited both men and women [32]. In the not affected by single nucleotide polymorphisms (SNPs) with high minor latter study, there was no information available about the gender of each allele frequencies [34]. In the latter criterion, the SNPs included in the participant. In this case, we estimated this missing information using the selected probes had a minor allele frequency less than 5% in Europeans function getSex of the R package minfi applied to the genome-wide DNA and North Americans (Figure 1A; Supplementary Table 2) referring to the methylation data [37]. The resulting frequencies of men and women sampled populations of the studies. All probes passed the remaining basic matched with those reported in the original study. quality control checks. In the women-specific analysis, we performed the same association We analyzed the M-values of a given probe instead of the respective analysis as described above. In the men-specific analysis, we compared a β-values to ensure a good approximation of the Normal distribution to linear regression model with the disease status as the single covariate the data [35]. Briefly, the β-values were calculated as the proportion of against another model without that covariate, when analyzing data from the methylation signal relative to the total signal for a given probe. The each probe. The comparison was done by the likelihood ratio test whose M-values were finally obtained by applying a logit transformation to the p-values were then adjusted for multiple testing in the same way as β-values. described above. To analyze the M-values of each probe, we initially estimated a linear Finally, for the study which did not share the respective data [27], we regression model where the respective covariates were the study indi- checked whether the reported differentially methylated CpG probes were cator and the disease status of the participants. In this model, we included located in either ACE or ACE2 (see Table 1 from this study). We did the the main effects of the covariates and the interaction. The model pa- same for the study based on the reduced representation bisulfite rameters were then estimated by the maximum likelihood method. Note sequencing technology [33] (see Additional File 1 from this study). that the main effect of the disease status is usually seen as the pooled effect of this covariate across all studies, as done in meta-analysis. 2.3. Analysis of gene expression studies We then simplified the model using a backward stepwise procedure based on Akaike's information criteria. Since the effect of the study in- Our meta-analysis of gene expression studies was focused on eight dicator was significant for the data of each probe, we tested the associ- reports using microarray technology (Table 2) [11,38–44]. These studies ation between ME/CFS and a given probe using a likelihood ratio test. In complied with the Minimum Information about a Microarray Experiment this test, we compared the model including the study indicator only with (MIAME) standard [45] and, therefore, they were considered to have the best model including that covariate and the one associated with sufficient quality for their inclusion in the meta-analysis. In particular, disease status (i.e., either the model only including the main effects or the these studies normalized the data which ensured comparability between model including both main effects and the interaction term). different samples and between different measurements of the same genes. To control for multiple testing, we adjusted the raw p-values using the Gene expression of these studies was performed in PBMCs (5 studies), Benjamini-Hochberg procedure [36]. This adjustment ensured a false whole blood (2 studies) and muscle biopsies (one study). One study discovery rate of 5% under the assumption of independent tests. excluded participants who were taking any regular medication [43]. Table 2. Summary of the 8 microarray-based gene expression studies under analysis, ordered by the year of publication. Reference Sample ME/CFS patients Healthy Technology (Manufacturer) ACE/ACE2 Data availability type controls, available (NCBI GEO Assession number) n Sample characteristics Case definition n [38] PBMCs 5 Female adults 1994 CDC/Fukuda 5 Atlas Glass Human 3.8 I No/No No (NA) Mean age: 42 years old Microarray (BD Biosciences Mean BMI: not reported Clontech) [39] PBMCs 25 Female/male adults 1994 CDC/Fukuda 25 Custom microarray Unclear No (NA) Mean age: 41 years old (Nimblegen) Mean BMI: not reported [40] Whole blood 25 Female/male adults 1994 CDC/Fukuda 50 GeneChip Human Genome Yes/Yes No (NA) Mean age: 43 years old U133 Plus 2.0 (Affymetrix) Mean BMI: not reported [41] Whole blood 11 Female/male adults 1994 CDC/Fukuda 11 Custom microarray (NA) Yes/No Yes (NA)a Mean age: 34 years old Mean BMI: 20.3 kg/m2 [42] Muscle biopsies 4 Female/male adults 1994 CDC/Fukuda 5 Operon V2.0 (CRIBI Yes/Yes No (NA) Mean age: 45/37 years old University of Padova) Mean BMI: not reported [43] PBMCs 8 Male adults 1994 CDC/Fukuda 7 GeneChip Human Genome Yes/Yes Yes (GSE14577) Median age: 36 years old U133 (Affymetrix) Mean BMI: not reported [11] PBMCs 37 Female/male adults 1994 CDC/Fukuda 25 MWG 20K human Array Yes/Yes No (NA) Mean age: 51 years old (Biotech MWG) Mean BMI:29.4 kg/m2 [44] PBMCs 33 Female/male adults 1994 CDC/Fukuda 21 GeneChip Human Gene ST Yes/No No (NA) Mean age: not reported (Affymetrix) Mean BMI: not reported a Data shared as a supplementary file in the online version of the study. 4 J. Malato et al. Heliyon 7 (2021) e07665 Another study reviewed the medications taken by the participants [11]. not explain how their mapping algorithms dealt with reads that could be However, it was unclear which medications were considered as a part of ambiguously mapped onto different locations in the transcriptome. the exclusion criteria. A third study reported that healthy controls were The selected studies were conducted in small cohorts of patients with free from any medication at the time of sampling [41]. ME/CFS (mean sample size ¼ 18.5; range ¼ 4–37) and healthy controls Three additional studies using microarray technology [46, 47, 48] (mean sample size ¼ 18.6; range ¼ 5–50 individuals) (Table 2). In these were excluded from our meta-analysis due to unclear or ineligible case studies, the patients and healthy controls were matched for age and definitions of ME/CFS. We also excluded four RNA-seq studies [49, 50, gender. Different commercial and custom microarray technologies were 51, 52], because of insufficient reporting on the basic quality control used for the respective gene expression quantification. There was only checks. In particular, these studies did not report the percentage of reads one study in which the microarray did not include any probe in the genes that could be mapped onto the reference transcriptome, the percentage of of interest [38]. Another study used a custom array based on 9,522 genes the transcriptome covered, the average number of mapped reads per from the RefSeq database, as available in August 2002 [39]. However, transcript, the relationship between the GC content and the mapped read this study did not provide the list of genes included in the respective distribution, as recommended elsewhere [53]. More importantly, given microarray. In terms of data sharing, one study made the data available in the high sequence homology between ACE and ACE2, these studies did the GEO database [43] and another one within the respective publication A cg09920557 − ACE cg19802564 − ACE −2 Females Males 6 Females Males −2.5 5.5 5 −3 4.5 −3.5 4 M values −4 M values 3.5 −4.5 3 −5 2.5 2 −5.5 1.5 −6 1 −6.5 0.5 −7 0 de Vega et al [39] de Vega et al [40] Trivedi et al [41] Herrera et al [42] Herrera et al [42] de Vega et al [39] de Vega et al [40] Trivedi et al [41] Herrera et al [42] Herrera et al [42] Study Study cg21094739 − ACE cg10468385 − ACE 6 Females Males 6 Females Males 5.5 5.5 5 5 4.5 4.5 4 4 M values M values 3.5 3.5 3 3 2.5 2.5 2 2 1.5 1.5 1 1 0.5 0.5 0 0 de Vega et al [39] de Vega et al [40] Trivedi et al [41] Herrera et al [42] Herrera et al [42] de Vega et al [39] de Vega et al [40] Trivedi et al [41] Herrera et al [42] Herrera et al [42] Study Study B cg08559914 − ACE2 6 Females Males 5.5 5 4.5 4 Groups M values 3.5 3 Healthy Controls 2.5 ME/CFS Patients 2 1.5 1 0.5 0 de Vega et al [39] de Vega et al [40] Trivedi et al [41] Herrera et al [42] Herrera et al [42] Study Figure 2. Boxplots per study, group and gender of the M-values referring to probes identified in Figures 1C and 1D. (A) Significant probes located in ACE. (B) Significant probe located in ACE2. 5 J. Malato et al. Heliyon 7 (2021) e07665 [41]. The latter study used a custom microarray that measured the Note that we were unable to quantify the ACE2 expression in 11 expression of stress-related genes including ACE but excluding ACE2. patients due to cDNA material below the limit of detection. These Before conducting a meta-analysis of the available data, we first re- problematic samples could be due to a lower expression of ACE2 in ME/ analyzed two studies where the normalized data were available [41, CFS patients than in healthy controls. To test this hypothesis, we 43]. In the first study [41], we calculated the mean of the log2(fold-- compared the respective proportion of samples below the limit of change) for ACE and the respective standard error. Note that the detection using the Pearson's χ2 test for two-way frequency tables. microarray used in this study did not include any probe in ACE2. In the The significance level of the statistical analysis was set at 5%. second study [43], we initially calculated the mean and the respective standard error of the log2(fold-change) for each probe located in ACE and 2.4.4. Ethical approval ACE2. We then pooled each pair of means for the same gene using the The protocol of this study was approved by the Ethics Committee of inverse-variance weighting method [54]. A third study reported the Charite-Universit€atsmedizin Berlin in accordance with the 1964 Decla- mean of the log2(fold-change) for ACE2 and the respective p-value using ration of Helsinki and its later amendments (reference number EA2/067/ a two-tailed Student's test [11]. In this case, we determine the quantile of 20). All patients and healthy controls gave written informed consent to the t-distribution associated with half of the reported p-value, equated it participate in the study. to the test statistic, and solved the resulting equation as a function of the standard error. No information was available from this study concerning 2.5. Statistical software the expression levels of ACE. Finally, we pooled the different estimates for the same gene from We performed our statistical analysis in the R software version 4.0.3. different studies using the inverse-variance weighting method [54]. In this analysis, we used the following Bioconductor packages: hgu133a.db, hgu133plus2.db, IlluminaHumanMethylation450kanno.ilmn12. hg19, and IlluminaHumanMethylationEPICanno.ilm10b2.hg19 to retrieve 2.4. Analysis of new RNA data on the ACE/ACE2 gene expression in ME/ the annotation of the GeneChip HG-U133A, GeneChip U133 þ 2, Infin- CFS ium HumanMethylation450K Array and HumanMethylationEPIC arrays, respectively; minfi to estimate the sex of each individual from DNA 2.4.1. Study participants methylation data [37]. The R scripts are freely available from the first Thirty-seven women with ME/CFS were recruited in 2020 from the and last authors upon request. outpatient clinic for immunodeficiencies at the Institute for Medical Immunology at the Charite-Universit€atsmedizin Berlin, Germany. These 3. Results patients were diagnosed according to the 2003 CCC while excluding other medical or neurological diseases which could explain fatigue [2]. 3.1. Meta-analysis of ACE/ACE2 DNA methylation in ME/CFS patients Thirty-four women with self-reported healthy status were recruited from staff. The oldest DNA methylation study [27] did not make the data available and hence, we screened the list of 120 differentially methylated 2.4.2. Experimental procedure for RNA isolation and expression probes (see table 1 from this study). Although located in 70 genes, these Consistently with previous studies of ME/CFS, the gene expression probes were neither located in ACE nor ACE2. We also screened the list of quantification was performed in PBMCs. These cells were isolated from differentially methylated probes reported by the study based on the heparinized whole blood by density gradient centrifugation using Biocoll reduced representation bisulfite sequencing technology (see Additional Separating Solution (Merck Millipore). Total RNA was isolated and File 1 from ref. [33]). Again, none of these probes was in the ACE or ACE2 extracted from 2106 PBMCs according to the manufacturer's in- loci. structions (NucleoSpin RNA Kit, Macherey-Nagel, cat. nr. 740955.50). For the four array-based studies [28, 29, 30, 32], we conducted a joint Afterwards cDNA was prepared by reverse transcription (High-Capacity analysis of the respective data in accordance with a meta-analysis. We cDNA Reverse Transcription Kit, Applied Biosystems, cat. nr. 4368814) first observed that the M-values of the 27 probes under investigation and real-time PCR was performed using TaqMan® Universal PCR Master tended to be uncorrelated with each other (Figure 1B). This observation Mix (cat. nr. 4305719) and TaqMan® Gene Expression Assays (cat. nr. supported the use of the Benjamini-Hochberg procedure to adjust the raw 4331182) for ACE (Hs00174179_m1), ACE2 (Hs01085333_m1) and the p-values under a multiple testing scenario. housekeeping gene HPRT1 (Hs02800695_m1) (Applied Biosystems). The The subsequent analysis suggested four CpG probes in ACE to be amplification of ACE and HPRT1 was based on 20 ng template cDNA. For associated with ME/CFS (Figure 1C). The probe cg09920557 belongs to the amplification of ACE2, this quantity was increased to 100 ng. All the TSS region of the gene while the remaining probes (cg19802564, measurements were performed with the ABI7200 and software Step One cg21094739, and cg10468385) are located in the gene body. The best Plus as absolute quantification according to manufacturer's instruction. linear regression models for each probe included both the main effects of Relative gene expression was analysed using the ΔCT method. the study indicator and of the disease status and the respective interaction term (Supplementary Table 3). The statistical interaction between these 2.4.3. Statistical analysis two covariates could be seen when plotting the whole data set (Figure 2A). We first tested whether patients and healthy controls were matched Although not significant, the estimated main effect of the disease status for age using the Kolgomorov-Smirnov test for two independent samples. was negative for each of the significantly associated probes. For statistical convenience, gene expression values were independently Concerning the probes in ACE2, the only significant association with transformed for ACE and ACE2 using a Box-Cox transformation [55]. The ME/CFS was obtained for cg08559914 located in the TSS region of the parameter estimates of this transformation were 0.303 and 0.225 for ACE gene (Figure 1C). According to the best linear regression model for this and ACE2, respectively. The transformed values for each gene were then probe, there was a negative association between the respective M-values analyzed as the outcome variable of a linear regression model specifying and ME/CFS (coefficient estimate ¼ -0.141 with a standard error of age and disease status of the participants as the respective covariates. The 0.048; Figure 2B and Supplementary Table 3). Given that a hypo- linear regression model was estimated using the maximum likelihood methylated promoter region is typically indicative of an increased method. After estimating the models, we tested the normal distribution in expression of the respective gene, this finding suggested an increased the resulting residuals using the Shapiro-Wilk test. We also visually ACE2 expression in patients with ME/CFS. inspected the assumption of constant variance of the same residuals as a We then repeated the same analysis for women and men separately. function of the covariates. For women, we obtained the same disease associations, as described 6 J. Malato et al. Heliyon 7 (2021) e07665 A Saiki et al (2008) Gow et al (2009) B Studies Saiki et al (2008) Estimate (95% CI) 0.265 (0.089;0.441) 14 ACE ACE ACE2 Probe 1 Probe 2 Probe 1 Probe 2 Gow et al (2009) − probe 1 0.012 (−0.012;0.036) ACE 7.0 6.5 − probe 2 0.004 (−0.014;0.022) 12 6.0 Summary effect 0.008 (−0.005;0.021) log2( values) log2( values) 5.5 5.0 Gow et al (2009) − probe 1 −0.038 (−0.085;0.009) 10 4.5 − probe 2 −0.037 (−0.083;0.008) ACE2 4.0 Smith et al (2011) −2.396 (−4.518;−0.273) 8 3.5 ME ME ME ME Summary effect −0.038 (−0.068;−0.009) HC ME/CFS HC CFS HC CFS HC CFS HC CFS −4.5 −4 −3.5 −3 −2.5 −2 −1.5 −1 −0.5 0 0.5 1 log2( fold change) Figure 3. Analysis of ACE/ACE2-related data from eligible microarray-based gene expression studies. (A) Boxplots of the data from these studies (Saiki et al (2008), ref. [41]; Gow et al (2009); ref. [43]). (B) Forest plot for the study-specific and pooled estimate of the mean of the log2(fold-change) between patients with ME/CFS and healthy controls using data shown in A. above (Figure 1D and Supplementary Table 3). For men, we did not find complete case scenario (Table 4). However, there were 11 (29.7%) of the any significant associations, probably due to data from a single study 37 samples from patients in which the expression level of ACE2 was [32] (Figure 1E). below the limit of detection. This proportion of samples was significantly higher than that for healthy controls given that the expression of ACE2 3.2. Meta-analysis of ACE/ACE2 gene expression in ME/CFS patients could be quantified in all the samples (29.7% versus 0%; Pearson's χ2 test, p ¼ 0.002). Consequently, we could not rule out that the patients with We first conducted a re-analysis of the two studies in which the ME/CFS from this cohort have a decreased expression of ACE2 when expression levels of ACE or ACE2 were available for each participant compared to healthy controls. Finally, in accordance with our meta- (Figure 3A) [41,43]. In the first study [41], there was evidence for an analysis, there was no evidence of differential expression of ACE be- increased expression of ACE in patients with ME/CFS (mean of the tween patients and healthy controls from this cohort. log2(fold-change) ¼ 0.265; 95% CI¼(0.089; 0.441)). In the second study [43], the means of the log2(fold-change) were estimated at 0.012 (95% 4. Discussion CI¼(-0.012; 0.036)) and 0.004 (95% CI¼(-0.014; 0.022)) for the two probes in ACE. The corresponding estimates for the two probes in ACE2 In this work, we investigated potential differences in ACE/ACE2 DNA were -0.038 (95% CI¼(-0.085; 0.009)) and -0.037 (95% CI¼(-0.083; methylation and expression levels between patients with ME/CFS and 0.008)) (Figure 3A). The pooled estimates for this study were 0.007 (95% CI¼(-0.006; 0.020)) and -0.038 (95% CI¼(-0.067; -0.008)) for ACE and ACE2, respectively. Table 3. Summary statistics for the gene expression of ACE and ACE2 from the Although not sharing the data, there was a study [11] that reported a German female study participants where data of ACE2 were only available for 26 significant negative association between ME/CFS and ACE2 expression affected patients. (see online Supplementary Table 2 of this study). In this case, we ob- tained the following mean of the log2(fold-change) ¼ -2.396 and 95% Summary statistic Healthy controls ME/CFS patients CI¼(-4.518; -0.273). N 34 37 We then pooled the estimates from different studies for the same Mean age (range), years 37.4 (23; 65) 41.1 (19; 60) gene: 0.008 (95% CI¼(-0.005; 0.021)) and -0.038 (95% CI¼(-0.068; Mean disease duration NA 5.4 (0; 24) -0.009)) for ACE and ACE2, respectively (Figure 3B). Therefore, our since diagnostic (range), months meta-analysis suggested a reduced expression of ACE2 but not of ACE in ACE patients with ME/CFS when comparing to healthy controls. Geometric mean 0.153 0.144 Finally, the remaining gene expression studies neither shared the Interquartile range 0.087 0.073 respective data nor reported any differential ACE/ACE2 expression be- ACE2 tween patients and healthy controls. Geometric mean 0.002 0.001 Interquartile range 0.005 0.004 3.3. Analysis of ACE/ACE2 gene expression from a new female cohort To complement our findings from the above meta-analysis, we Table 4. Analysis of the linear regression models for the Box-Cox-transformed measured the ACE and ACE2 mRNA levels in PBMCs from 37 women ACE and ACE2 mRNA levels where data were only available for 26 ME/CFS with ME/CFS (mean age ¼ 41.1 years old) and 34 healthy women (mean patients. age ¼ 37.4 years old) (Table 3). Patients and healthy participants were Analysis Estimate (SE) P-value matched for age (Kolmogorov-Smirnov test, p ¼ 0.38). There was no Box-Cox transformed ACE information about the disease duration for 4 patients. The average dis- (Intercept) 0.541 (0.032) <0.001 ease duration for the remaining patients was 5.4 months in relation to the Age 0.001 (0.001) 0.328 time of diagnosis (range ¼ 0–24 months). Disease Status (ME/CFS) -0.013 (0.018) 0.481 We observed higher mRNA levels of ACE than of ACE2 (Table 4, Box-Cox transformed ACE2 Figure 4A). There was no evidence for a significant correlation between (Intercept) 0.307 (0.038) <0.001 ACE and ACE2 expression levels (Spearman's correlation coefficient ¼ Age -0.001 (0.001) 0.137 -0.120) (Figure 4B). In contrast to the above meta-analysis, we could not Disease Status (ME/CFS) -0.006 (0.021) 0.789 find a reduced expression of ACE2 in patients with ME/CFS using the 7 J. Malato et al. Heliyon 7 (2021) e07665 A ACE ACE2 B Raw Raw 0.5 0.04 0.5 λ 1 = 0.303 0.4 0.03 λ 2 = 0.225 0.3 0.02 0.4 0.2 0.01 0.1 0.3 mRNA level 0.0 0.00 ACE2λ 2 ACE ACE2 Box−Cox Box−Cox 0.8 0.5 0.2 0.4 0.6 0.3 0.1 0.2 0.4 0.1 Control ME/CFS 0.2 0.0 0.0 0.3 0.4 0.5 0.6 0.7 0.8 Control ME/CFS Control ME/CFS ACEλ 1 Figure 4. Analysis of ACE and ACE2 expression levels from the German study. (A) Violin plots of ACE (left side) and ACE2 (right side) mRNA raw data (upper row) and transformed data using a Box-Cox transformation (lower row). (B) Scatterplot between the transformed ACE and ACE2 expression levels (Spearman's correlation coefficient ¼ -0.120). healthy controls. With the identification of these differences, we ex- rule out that SARS-CoV-2 might use alternative receptors when infecting pected to determine the health risk of patients with ME/CFS if infected by PBMCs. SARS-CoV-2. However, we stumbled upon hurdles related to (i) data Among the alternative receptors for SARS-CoV-2, the human trans- unavailability for a possible re-analysis, (ii) availability of data derived membrane protease serine 2 (TMPRSS2) was suggested as a strong from PBCMs and related subsets in which ACE2 is not particularly candidate [63] due to its role on SARS-CoV-1 infection [64, 65]. This expressed, (iii) studies with unclear data quality, and (iv) studies using protease seems to induce SARS-CoV-2 cell entry through endocytosis via disease case definitions that are not recommended for research. As a a mechanism of ACE2 cleavage [14]. Another candidate receptor is the A consequence, we could not provide a more definite answer to our main disintegrin and metallopeptidase domain 17 protein (ADAM17) recog- research question. nized by the immune system as a stress-response signal [66]. Like Notwithstanding these difficulties, we could identify four CpG probes TMPRSS2, ADAM17 can also cleave ACE2 but with a reduced viral in- on ACE and another one on ACE2 with decreased DNA methylation levels vasion efficiency [67]. in patients with ME/CFS. This finding suggested an increased expression With respect to the role of these proteases in ME/CFS, a targeted gene of the respective genes. However, our meta-analysis of public data sug- expression study analyzed ADAM17 and other stress-response proteins gested the opposite. Such decrease in ACE2 expression was partially [41]. This study did not report any differential expression of this protease confirmed by new data in which there was a significant higher proportion between patients with ME/CFS and healthy controls. However, this study of samples below the limit of detection in patients with ME/CFS than in is likely to be affected by a low statistical power due to small sample sizes healthy controls. Nonetheless, it was clear that ACE2 is not particularly for both groups. In addition, one of the selected DNA methylation studies expressed in PBMCs from both patients with ME/CFS and healthy con- suggested a decrease in the DNA methylation levels of one trols, as mentioned in the introduction. ADAM17-related CpG probe in patients with ME/CFS [30]. In general, ACE2 downregulation is known to occur after host-cell Dipeptidyl peptidase-4 (DPP4), also known as the lymphocyte cell entry by SARS-CoV-2 [56]. This downregulation is particularly prob- surface protein CD26, was found to be the main receptor for the Middle lematic in individuals affected by cardiovascular diseases, diabetes, and East respiratory syndrome–related coronavirus [68, 69]. In contrast to other medical conditions, due to their low ACE2 levels before the ACE2, this surface protein is highly abundant in PBMCs including CD4þ infection [57]. SARS-CoV-2 infection is then expected to further increase and CD8þ T cells [18]. Bioinformatic analysis also suggested a strong the ACE:ACE2 ratio, thus, promoting vasoconstriction, increased pro- interaction potential between this protein and SARS-CoV-2 [70,71]. duction of ROS and inflammation in patients with these co-morbidities Finally, DPP4 inhibitors were found to be protective against severe [23]. In this scenario, a putative reduction of the ACE2 expression COVID-19 in patients with diabetes mellitus when compared to RAAS makes patients with ME/CFS similar to these patients with a high risk for blockers [72]. After initial concerns, this finding combined with others COVID-19. As a consequence, patients with ME/CFS could be considered suggested an interesting therapeutic avenue against COVID-19 using a priority group for vaccination by public health authorities. The DPP4 blockers [73]. fundamental question is then to know whether our findings based on Interestingly, there is evidence for an increased proportion of natural PBMCs could recreate what occurs in pulmonary epithelial and endo- killer cells and T cells expressing DPP4/CD26þ in patients with ME/CFS thelial cells, the main targets of SARS-CoV-2. Future research should be [7, 74]. However, the number of DPP4/CD26 molecules was significantly conducted to answer this question, as similarly done in past studies reduced in T lymphocytes and natural killer cells of these patients [74]. If aiming at understanding how the gene expression profiles from PBMCs DPP4 is indeed a relevant receptor for immune-cell invasion by could mimick those present in other tissues affecting by a given disease SARS-CoV-2, research about this receptor should be prioritized when [58, 59, 60]. analyzing PBMCs from patients with ME/CFS. Given the residual ACE2 expression in PBMCs under normal condi- Sialic acids were also hypothesized as binding receptors used by tions, one is tempted to say that SARS-CoV-2 does not infect these cells. SARS-CoV-2, as reported for other human coronaviruses [75]. These However, earlier studies on SARS-CoV-1 found this virus within T lym- acids are highly expressed in the epithelium cells of the lungs and oral phocytes, macrophages, and dendritic cells [61]. More recently, an in cavity [76]. In vitro and in silico studies demonstrated the same binding vitro study was able to infect PBMCs with SARS-CoV-2 [62]. Monocytes potential for SARS-CoV-2 [77, 78, 79]. However, the ACE2 glycosylation are particularly susceptible to such infections. In this context, one cannot inhibition studies suggested that sialic acids on ACE2 receptor prevent 8 J. Malato et al. Heliyon 7 (2021) e07665 ACE2–virus interaction [80, 81]. Again, detailed research on these pu- Carmen Scheibenbogen: Analyzed and interpreted the data; tative receptors could help to determine the health risk of patients with Contributed reagents, materials, analysis tools or data; Wrote the paper. ME/CFS when infected by SARS-CoV-2. Nuno Sepúlveda: Conceived and designed the experiments; Analyzed It was suggested that the arousal state experienced by patients with and interpreted the data; Wrote the paper. ME/CFS protects them against microbial infections [82]. This sugges- tion came from a clinical trial where patients were treated with cloni- Funding statement dine to decrease such a state. Treated patients got their symptoms worsened and had their inflammation markers increased during the Jo~ao Malato and Andre Fonseca were fully funded by FCT – Fundaç~ao trial. In contrast, basic epidemiological studies reported many patients para a Ci^encia e Tecnologia, Portugal (ref.grant: SFRH/BD/149758/ with frequent viral infections and flu-like symptoms [3, 4, 83]. The 2019 and SFRH/BD/147629/2019, respectively). Nuno Sepúlveda and question is how an infection by SARS-CoV-2 lies in this contrasting Clara Cordeiro were partially funded by FCT – Fundaç~ao para a Ci^encia e evidence. A possible answer can be given with the assistance of the a Tecnologia, Portugal (ref. grant: UIDB/00006/2020). Luís Nacul and so-called sustained arousal model of ME/CFS [84]. According to this Eliana M Lacerda acknowledge the funding from the National Institute of model, a sustained arousal state promotes in the long-run deleterious Allergy and Infectious Diseases (NIAID) of the National Institutes of alterations of different body systems, including the immune system. Health (NIH -Award Number: R01AI103629), and from the ME Associ- Similar prediction was made by a recent study discussing the natural ation (Award number: PF8947) for their studies on ME/CFS. history of ME/CFS [85]. If so, patients with longer disease durations are more likely to show these immunological alterations than patients at Data availability statement the early stages of the disease. However, we could not analyze the effect of disease duration on our results, because this variable was not Data are available from the GEO database under the accession available in the public data sets included in our meta-analyses. number GSE59489, GSE93266, GSE111183, GSE156792, GSE153667, Finally, our original idea was also to include a meta-analysis of ACE/ GSE14577. ACE2 data from published genome-wide association studies on ME/CFS [11, 32, 86, 87, 88]. However, we could not materialize this idea, because such studies did not make their data publicly available. Never- Declaration of interests statement theless, evidence is scarce for a putative role of ACE/ACE2 poly- morphisms on ME/CFS. Two studies reported many candidate SNPs for The authors declare no conflict of interest. such association, but none was located in ACE or ACE2 [11,86]. Two other studies did not find any significant SNPs associated with ME/CFS Additional information [32, 88]. The most optimistic study reported thousands of SNPs related to the disease [87]. However, this study did not perform all the basic quality Supplementary content related to this article has been published control checks [89]. online at https://doi.org/10.1016/j.heliyon.2021.e07665. 5. Conclusions References [1] K. Fukuda, S.E. Straus, I. Hickie, M.C. Sharpe, J.G. Dobbins, A. Komaroff, The Notwithstanding the low expression of ACE2 in PBMCs in general, chronic fatigue syndrome: a comprehensive approach to its definition and study, there is evidence for a decreased expression of the gene in these cells Ann. Intern. Med. 121 (1994) 953–959. from patients with ME/CFS. If PBMCs can qualitatively recreate what is [2] B.M. Carruthers, A.K. Jain, K.L. De Meirleir, D.L. Peterson, N.G. Klimas, A.M. Lemer, occurring in the main cellular targets of SARS-CoV-2, then patients with A.C. Bested, P. Flor-Henry, P. Joshi, A.C.P. Powles, J.A. Sherkey, M.I. Van de Sande, Myalgic encephalomyelitis/chronic fatigue syndrome: clinical working case this disease could be at a higher COVID-19 risk. In this regard, a recent definition, diagnostic and treatment protocols, J. Chronic Fatigue Syndrome 11 preliminary report suggested that patients with ME/CFS got their (2003) 7–115. symptoms worsened upon SARS-CoV-2 infection [91]. Altogether, these [3] S.C. Johnston, D.R. Staines, S.M. Marshall-Gradisnik, Epidemiological characteristics of chronic fatigue syndrome/myalgic encephalomyelitis in patients could be considered a priority group for vaccination against Australian patients, Clin. Epidemiol. 8 (2016) 97–107. COVID-19, even though vaccines could trigger ME/CFS [92, 93] or even [4] L. Chu, I.J. Valencia, D.W. Garvert, J.G. Montoya, Onset patterns and course of exacerbate ME/CFS symptoms as the case of the natural immunization by myalgic encephalomyelitis/chronic fatigue syndrome, Front. Pediatr. 7 (2019) 12. [5] S. Rasa, Z. Nora-Krukle, N. Henning, E. Eliassen, E. Shikova, T. Harrer, SARS-CoV-2. To further consolidate the existing evidence, future C. Scheibenbogen, M. Murovska, B.K. Prusty, Chronic viral infections in myalgic research should prioritize the collection of data from the main cellular encephalomyelitis/chronic fatigue syndrome (ME/CFS), J. Transl. Med. 16 (2018), targets in patients with ME/CFS. Further investigation should be also 268. [6] M.E. Ariza, Myalgic encephalomyelitis/chronic fatigue syndrome: the human conducted on alternative SARS-CoV-2 receptors (i.e., DPP4 and sialic herpesviruses are back!, Biomolecules 11 (2021) 1–17. acids). At last, future research should also consider investigating putative [7] N.G. Klimas, F.R. Salvato, R. Morgan, M.A. Fletcher, Immunologic abnormalities in sex differences in patients with ME/CFS given that, in general, men are chronic fatigue syndrome, J. Clin. Microbiol. 28 (1990) 1403–1410. [8] L. Lorusso, S.V. Mikhaylova, E. Capelli, D. Ferrari, G.K. Ngonga, G. Ricevuti, more affected by COVID-19 than women [90]. Immunological aspects of chronic fatigue syndrome, Autoimmun. Rev. 8 (2009) 287–291. Declarations [9] E.W. Brenu, M.L. van Driel, D.R. Staines, K.J. Ashton, S.B. Ramos, J. Keane, N.G. Klimas, S.M. Marshall-Gradisnik, Immunological abnormalities as potential biomarkers in chronic fatigue syndrome/myalgic encephalomyelitis, J. Transl. Med. 9 (2011), 81. Author contribution statement [10] J.R. Kerr, Gene profiling of patients with chronic fatigue syndrome/myalgic encephalomyelitis, Curr. Rheumatol. Rep. 10 (2008) 482–491. Jo~ao Malato, Andr e Fonseca, Anna D Grabowska, Luís Graça, Clara [11] A.K. Smith, H. Fang, T. Whistler, E.R. Unger, M.S. Rajeevan, Convergent genomic studies identify association of GRIK2 and NPAS2 with chronic fatigue syndrome, Cordeiro, Luís Nacul, Eliana M Lacerda, Jesus Castro-Marrero, Francisco Neuropsychobiology 64 (2011) 183–194. Westermeier: Analyzed and interpreted the data; Wrote the paper. [12] W. Li, M.J. Moore, N. Vasllieva, J. Sui, S.K. Wong, M.A. Berne, M. Somasundaran, Franziska Sotzny: Conceived and designed the experiments; Per- J.L. Sullivan, K. Luzuriaga, T.C. Greeneugh, H. Choe, M. Farzan, Angiotensin- converting enzyme 2 is a functional receptor for the SARS coronavirus, Nature 426 formed the experiments; Analyzed and interpreted the data; Contributed (2003) 450–454. reagents, materials, analysis tools or data; Wrote the paper. [13] X.Y. Ge, J.L. Li, X. Lou Yang, A.A. Chmura, G. Zhu, J.H. Epstein, J.K. Mazet, B. Hu, Sandra Bauer, Helma Freitag: Performed the experiments; Contrib- W. Zhang, C. Peng, Y.J. Zhang, C.M. Luo, B. Tan, N. Wang, Y. Zhu, G. Crameri, S.Y. Zhang, L.F. Wang, P. Daszak, Z.L. Shi, Isolation and characterization of a bat uted reagents, materials, analysis tools or data; Wrote the paper. SARS-like coronavirus that uses the ACE2 receptor, Nature 503 (2013) 535–538. 9 J. Malato et al. Heliyon 7 (2021) e07665 [14] M. Hoffmann, H. Kleine-Weber, S. Schroeder, N. Krüger, T. Herrler, S. Erichsen, [40] J.R. Kerr, R. Petty, B. Burke, J. Gough, D. Fear, L.I. Sinclair, D.L. Mattey, T.S. Schiergens, G. Herrler, N.H. Wu, A. Nitsche, M.A. Müller, C. Drosten, S.C.M. Richards, J. Montgomery, D.A. Baldwin, P. Kellam, T.J. Harrison, S. P€ ohlmann, SARS-CoV-2 cell entry depends on ACE2 and TMPRSS2 and is blocked G.E. Griffin, J. Main, D. Enlander, D.J. Nutt, S.T. Holgate, Gene expression subtypes by a clinically proven protease inhibitor, Cell 181 (2020) 271–280.e8. in patients with chronic fatigue syndrome/myalgic encephalomyelitis, J. Infect. Dis. [15] I. Hamming, W. Timens, M.L.C. Bulthuis, A.T. Lely, G.J. Navis, H. van Goor, Tissue 197 (2008) 1171–1184. distribution of ACE2 protein, the functional receptor for SARS coronavirus. A first [41] T. Saiki, T. Kawai, K. Morita, M. Ohta, T. Saito, K. Rokutan, N. Ban, Identification of step in understanding SARS pathogenesis, J. Pathol. 203 (2004) 631–637. marker genes for differential diagnosis of chronic fatigue syndrome, Mol. Med 14 [16] K.F. To, A.W.I. Lo, Exploring the pathogenesis of severe acute respiratory syndrome (2008) 599–607. (SARS): the tissue distribution of the coronavirus (SARS-CoV) and its putative [42] T. Pietrangelo, R. Mancinelli, L. Toniolo, G. Montanari, J. Vecchiet, G. Fan o, receptor, angiotensin-converting enzyme 2 (ACE2), J. Pathol. 203 (2004) 740–743. S. Fulle, Transcription profile analysis of vastus lateralis muscle from patients with [17] M.Y. Li, L. Li, Y. Zhang, X.S. Wang, Expression of the SARS-CoV-2 cell receptor gene chronic fatigue syndrome, Int. J. Immunopathol. Pharmacol. 22 (2009) 795–807. ACE2 in a wide variety of human tissues, Infect. Dis. Poverty 9 (2020), 45. [43] J.W. Gow, S. Hagan, P. Herzyk, C. Cannon, P.O. Behan, A. Chaudhuri, A gene [18] U. Radzikowska, M. Ding, G. Tan, D. Zhakparov, Y. Peng, P. Wawrzyniak, M. Wang, signature for post-infectious chronic fatigue syndrome, BMC Med. Genomics 2 S. Li, H. Morita, C. Altunbulakli, M. Reiger, A.U. Neumann, N. Lunjani, C. Traidl- (2009), 38. Hoffmann, K.C. Nadeau, L. O’Mahony, C. Akdis, M. Sokolowska, Distribution of [44] M.G. Jeffrey, L. Nathanson, K. Aenlle, Z.M. Barnes, M. Baig, G. Broderick, ACE2, CD147, CD26, and other SARS-CoV-2 associated molecules in tissues and N.G. Klimas, M.A. Fletcher, T.J.A. Craddock, Treatment avenues in myalgic immune cells in health and in asthma, COPD, obesity, hypertension, and COVID-19 encephalomyelitis/chronic fatigue syndrome: a split-gender pharmacogenomic risk factors, Allergy 75 (2020) 2829–2845. study of gene-expression modules, Clin. Ther 41 (2019) 815–835.e6. [19] M. Rutkowska-Zapała, M. Suski, R. Szatanek, M. Lenart, K. Weglarczyk, [45] A. Brazma, P. Hingamp, J. Quackenbush, G. Sherlock, P. Spellman, C. Stoeckert, R. Olszanecki, T. Grodzicki, M. Strach, J. Gasowski, M. Siedlar, Human monocyte J. Aach, W. Ansorge, C.A. Ball, H.C. Causton, T. Gaasterland, P. Glenisson, subsets exhibit divergent angiotensin I-converting activity, Clin. Exp. Immunol. 181 F.C.P. Holstege, I.F. Kim, V. Markowitz, J.C. Matese, H. Parkinson, A. Robinson, (2015) 126–132. U. Sarkans, S. Schulze-Kremer, J. Stewart, R. Taylor, J. Vilo, M. Vingron, Minimum [20] X. Song, W. Hu, H. Yu, L. Zhao, Y. Zhao, X. Zhao, H.H. Xue, Y. Zhao, Little to no information about a microarray experiment (MIAME) - toward standards for expression of angiotensin-converting enzyme-2 on most human peripheral blood microarray data, Nat. Genet. 29 (2001) 365–371. immune cells but highly expressed on tissue macrophages, Cytometry (2020) 1–10. [46] S. Galbraith, B. Cameron, H. Li, D. Lau, U. Vollmer-Conna, A.R. Lloyd, Peripheral [21] S.R. Tipnis, N.M. Hooper, R. Hyde, E. Karran, G. Christie, A.J. Turner, A human blood gene expression in postinfective fatigue syndrome following from three homolog of angiotensin-converting enzyme: cloning and functional expression as a different triggering infections, J. Infect. Dis. 204 (2011) 1632–1640. captopril-insensitive carboxypeptidase, J. Biol. Chem. 275 (2000) 33238–33243. [47] S.D. Vernon, E.R. Unger, I.M. Dimulescu, M. Rajeevan, W.C. Reeves, Utility of the [22] F. Westermeier, M. Bustamante, M. Pavez, L. García, M. Chiong, M.P. Ocaranza, blood for gene expression profiling and biomarker discovery in chronic fatigue S. Lavandero, Novel players in cardioprotection: insulin like growth factor-1, syndrome, Dis. Markers 18 (2002) 193–199. angiotensin-(1-7) and angiotensin-(1-9), Pharmacol. Res. 101 (2015) 41–55. [48] C.B. Nguyen, L. Alsøe, J.M. Lindvall, D. Sulheim, E. Fagermoen, A. Winger, [23] P. Pagliaro, C. Penna, ACE/ACE2 ratio: a key also in 2019 coronavirus disease M. Kaarbø, H. Nilsen, V.B. Wyller, Whole blood gene expression in adolescent (Covid-19)? Front. Med. 7 (2020) 335. chronic fatigue syndrome: an exploratory cross-sectional study suggesting altered B [24] A.H. Sawalha, M. Zhao, P. Coit, Q. Lu, Epigenetic dysregulation of ACE2 and cell differentiation and survival, J. Transl. Med. 15 (2017), 102. interferon-regulated genes might suggest increased COVID-19 susceptibility and [49] J. Bouquet, J.L. Gardy, S. Brown, J. Pfeil, R.R. Miller, M. Morshed, A. Avina- severity in lupus patients, Clin. Immunol. 215 (2020) 108410. Zubieta, K. Shojania, M. McCabe, S. Parker, M. Uyaguari, S. Federman, P. Tang, [25] J. Blomberg, C.G. Gottfries, A. Elfaitouri, M. Rizwan, A. Rosen, Infection elicited T. Steiner, M. Otterstater, R. Holt, R. Moore, C.Y. Chiu, D.M. Patrick, RNA-seq autoimmunity and Myalgic encephalomyelitis/chronic fatigue syndrome: an analysis of gene expression, viral pathogen, and B-cell/T-cell receptor signatures in explanatory model, Front. Immunol. 9 (2018), 229. complex chronic disease, Clin. Infect. Dis. 64 (2017) 476–481. [26] D.F.H. Pheby, D. Araja, U. Berkis, E. Brenna, J. Cullinan, J.-D. de Korwin, L. Gitto, [50] J. Bouquet, T. Li, J.L. Gardy, X. Kang, S. Stevens, J. Stevens, M. VanNess, C. Snell, D.A. Hughes, R.M. Hunter, D. Trepel, X. Wang-Steverding, The development of a J. Potts, R.R. Miller, M. Morshed, M. McCabe, S. Parker, M. Uyaguari, P. Tang, consistent europe-wide approach to investigating the economic impact of myalgic T. Steiner, W.S. Chan, A.M. De Souza, A. Mattman, D.M. Patrick, C.Y. Chiu, Whole encephalomyelitis (ME/CFS): a report from the European Network on ME/CFS blood human transcriptome and virome analysis of ME/CFS patients experiencing (EUROMENE), Healthcare 8 (2020) 88. post-exertional malaise following cardiopulmonary exercise testing, PloS One 14 [27] E.W. Brenu, D.R. Staines, S.M. Marshall-Gradisnik, Methylation profile of CD4þ T (2019), e0212193. cells in chronic fatigue syndrome/myalgic encephalomyelitis, J. Clin. Cell. [51] E. Sweetman, M. Ryan, C. Edgar, A. Mackay, R. Vallings, W. Tate, Changes in the Immunol. 5 (2014) 1–14. transcriptome of circulating immune cells of a New Zealand cohort with myalgic [28] W.C. De Vega, S.D. Vernon, P.O. McGowan, DNA methylation modifications encephalomyelitis/chronic fatigue syndrome, Int. J. Immunopathol. Pharmacol. 33 associated with Chronic Fatigue Syndrome, PloS One 9 (2014), e104757. (2019) 1–8. [29] W.C. De Vega, S. Herrera, S.D. Vernon, P.O. McGowan, Epigenetic modifications [52] R.P.H. Raijmakers, A.F.M. Jansen, S.P. Keijmel, R. Ter Horst, M.E. Roerink, and glucocorticoid sensitivity in myalgic encephalomyelitis/chronic fatigue B. Novakovic, L.A.B. Joosten, J.W.M. Van Der Meer, M.G. Netea, C.P. Bleeker- syndrome (ME/CFS), BMC Med. Genomics 10 (2017), 11. Rovers, A possible role for mitochondrial-derived peptides humanin and MOTS-c in [30] M.S. Trivedi, E. Oltra, L. Sarria, N. Rose, V. Beljanski, M.A. Fletcher, N.G. Klimas, patients with Q fever fatigue syndrome and chronic fatigue syndrome, J. Transl. L. Nathanson, Identification of myalgic encephalomyelitis/chronic fatigue Med. 17 (2019), 157. syndrome-associated DNA methylation patterns, PloS One 13 (2018), e0201066. [53] A. Conesa, P. Madrigal, S. Tarazona, D. Gomez-Cabrero, A. Cervera, A. McPherson, [31] E. Almenar-Perez, T. Ovejero, T. Sanchez-Fito, J.A. Espejo, L. Nathanson, E. Oltra, M.W. Szczesniak, D.J. Gaffney, L.L. Elo, X. Zhang, A. Mortazavi, A survey of best Epigenetic components of myalgic encephalomyelitis/chronic fatigue syndrome practices for RNA-seq data analysis, Genome Biol. 17 (2016), 13. uncover potential transposable element activation, Clin. Ther 41 (2019) 675–698. [54] J. Hartung, G. Knapp, B.K. Sinha, Statistical Meta-Analysis with Applications, first [32] S. Herrera, W.C. de Vega, D. Ashbrook, S.D. Vernon, P.O. McGowan, Genome- ed., John Wiley & Sons, New Jersey, 2008. epigenome interactions associated with myalgic encephalomyelitis/chronic fatigue € Asar, O. Ilk, O. Dag, Estimating Box-Cox power transformation parameter via [55] O. syndrome, Epigenetics 13 (2018) 1174–1190. goodness-of-fit tests, Commun. Stat. Simulat. Comput. 46 (2017) 91–105. [33] A.M. Helliwell, E.C. Sweetman, P.A. Stockwell, C.D. Edgar, A. Chatterjee, W.P. Tate, [56] P.K. Datta, F. Liu, T. Fischer, J. Rappaport, X. Qin, SARS-CoV-2 pandemic and Changes in DNA methylation profiles of myalgic encephalomyelitis/chronic fatigue research gaps: understanding SARS-CoV-2 interaction with the ACE2 receptor and syndrome patients reflect systemic dysfunctions, Clin. Epigenet. 12 (2020) 167. implications for therapy, Theranostics 10 (2020) 7448–7464. [34] S. Dedeurwaerder, M. Defrance, M. Bizet, E. Calonne, G. Bontempi, F.F. Fuks, [57] P. Verdecchia, C. Cavallini, A. Spanevello, F. Angeli, The pivotal link between ACE2 A comprehensive overview of Infinium HumanMethylation 450 data processing, deficiency and SARS-CoV-2 infection, Eur, J. Intern. Med. 76 (2020) 14–20. Brief. Bioinform. 15 (2013) 929–941. [58] I.C. Gerling, R.A. Ahokas, G. Kamalov, W. Zhao, S.K. Bhattacharya, Y. Sun, [35] P. Du, X. Zhang, C.C. Huang, N. Jafari, W.A. Kibbe, L. Hou, S.M. Lin, Comparison of K.T. Weber, Gene expression profiles of peripheral blood mononuclear cells reveal Beta-value and M-value methods for quantifying methylation levels by microarray transcriptional signatures as novel biomarkers of cardiac remodeling in rats with analysis, BMC Bioinf. 11 (2010), 587. aldosteronism and hypertensive heart disease, JACC Hear. Fail. 1 (2013) 469–476. [36] Y. Benjamini, Y. Hochberg, Controlling the false discovery rate: a practical and [59] T. Takamura, M. Honda, Y. Sakai, H. Ando, A. Shimizu, T. Ota, M. Sakurai, H. Misu, powerful approach to multiple testing, J. R. Stat. Soc. Ser. B 57 (1995) 289–300. S. Kurita, N. Matsuzawa-Nagata, M. Uchikata, S. Nakamura, R. Matoba, M. Tanino, [37] M.J. Aryee, A.E. Jaffe, H. Corrada-Bravo, C. Ladd-Acosta, A.P. Feinberg, Kichi Matsubara, S. Kaneko, Gene expression profiles in peripheral blood K.D. Hansen, R.A. Irizarry, Minfi: a flexible and comprehensive Bioconductor mononuclear cells reflect the pathophysiology of type 2 diabetes, Biochem. package for the analysis of Infinium DNA methylation microarrays, Bioinformatics Biophys. Res. Commun. 361 (2007) 379–384. 30 (2014) 1363–1369. [60] F.S. Manoel-Caetano, D.J. Xavier, A.F. Evangelista, P. Takahashi, C.V. Collares, [38] T. Whistler, J.F. Jones, E.R. Unger, S.D. Vernon, Exercise responsive genes D. Puthier, M.C. Foss-Freitas, M.C. Foss, E.A. Donadi, G.A. Passos, E.T. Sakamoto- measured in peripheral blood of women with Chronic Fatigue Syndrome and Hojo, Gene expression profiles displayed by peripheral blood mononuclear cells matched control subjects, BMC Physiol. 5 (2005), 5. from patients with type 2 diabetes mellitus focusing on biological processes [39] N. Kaushik, D. Fear, S.C.M. Richards, C.R. McDermott, E.F. Nuwaysir, P. Kellam, implicated on the pathogenesis of the disease, Gene 511 (2012) 151–160. T.J. Harrison, R.J. Wilkinson, D.A.J. Tyrrell, S.T. Holgate, J.R. Kerr, Gene [61] M.Z. Tay, C.M. Poh, L. Renia, P.A. MacAry, L.F.P. Ng, The trinity of COVID-19: expression in peripheral blood mononuclear cells from patients with chronic fatigue immunity, inflammation and intervention, Nat. Rev. Immunol. 20 (2020) 363–374. syndrome, J. Clin. Pathol. 58 (2005) 826–832. [62] A.C. Codo, G.G. Davanzo, L.B. de Monteiro, G.F. de Souza, S.P. Muraro, 10 J. Malato et al. Heliyon 7 (2021) e07665 J.V. Virgilio-da-Silva, J.S. Prodonoff, V.C. Carregari, C.A.O. de Biagi Junior, [77] E. Milanetti, M. Miotto, L. Di Rienzo, M. Monti, G. Gosti, G. Ruocco, In-Silico F. Crunfli, J.L. Jimenez Restrepo, P.H. Vendramini, G. Reis-de-Oliveira, K. Bispo dos evidence for two receptors based strategy of SARS-CoV-2, BioRxiv (2020) 6197, Santos, D.A. Toledo-Teixeira, P.L. Parise, M.C. Martini, R.E. Marques, H.R. Carmo, 2020.03.24. A. Borin, L.D. Coimbra, V.O. Boldrini, N.S. Brunetti, A.S. Vieira, E. Mansour, [78] M. Awasthi, S. Gulati, D.P. Sarkar, S. Tiwari, S. Kateriya, P. Ranjan, S.K. Verma, The R.G. Ulaf, A.F. Bernardes, T.A. Nunes, L.C. Ribeiro, A.C. Palma, M.V. Agrela, sialoside-binding pocket of SARS-CoV-2 spike glycoprotein structurally resembles M.L. Moretti, A.C. Sposito, F.B. Pereira, L.A. Velloso, M.A.R. Vinolo, A. Damasio, MERS-CoV, Viruses 12 (2020), 909. J.L. Proença-M odena, R.F. Carvalho, M.A. Mori, D. Martins-de-Souza, H.I. Nakaya, [79] A.N. Baker, S.J. Richards, C.S. Guy, T.R. Congdon, M. Hasan, A.J. Zwetsloot, A.S. Farias, P.M. Moraes-Vieira, Elevated glucose levels favor SARS-CoV-2 infection A. Gallo, J.R. Lewandowski, P.J. Stansfeld, A. Straube, M. Walker, S. Chessa, and monocyte response through a HIF-1α/Glycolysis-Dependent Axis, Cell Metabol. G. Pergolizzi, S. Dedola, R.A. Field, M.I. Gibson, The SARS-COV-2 spike protein 32 (2020) 437–446.e5. binds sialic acids and enables rapid detection in a lateral flow point of care [63] W. Sungnak, N. Huang, C. Becavin, M. Berg, R. Queen, M. Litvinukova, C. Talavera- diagnostic device, ACS Cent. Sci. 6 (2020) 2046–2052. Lopez, H. Maatz, D. Reichart, F. Sampaziotis, K.B. Worlock, M. Yoshida, J.L. Barnes, [80] H. Chu, B. Hu, X. Huang, Y. Chai, D. Zhou, Y. Wang, H. Shuai, D. Yang, Y. Hou, N.E. Banovich, P. Barbry, A. Brazma, J. Collin, T.J. Desai, T.E. Duong, O. Eickelberg, X. Zhang, T.T.T. Yuen, J.P. Cai, A.J. Zhang, J. Zhou, S. Yuan, K.K.W. To, C. Falk, M. Farzan, I. Glass, R.K. Gupta, M. Haniffa, P. Horvath, N. Hubner, D. Hung, I.H.Y. Chan, K.Y. Sit, D.C.C. Foo, I.Y.H. Wong, A.T.L. Ng, T.T. Cheung, S.Y.K. Law, N. Kaminski, M. Krasnow, J.A. Kropski, M. Kuhnemund, M. Lako, H. Lee, S. Leroy, W.K. Au, M.A. Brindley, Z. Chen, K.H. Kok, J.F.W. Chan, K.Y. Yuen, Host and viral S. Linnarson, J. Lundeberg, K.B. Meyer, Z. Miao, A.V. Misharin, M.C. Nawijn, determinants for efficient SARS-CoV-2 infection of the human lung, Nat. Commun. M.Z. Nikolic, M. Noseda, J. Ordovas-Montanes, G.Y. Oudit, D. Pe’er, J. Powell, 12 (2021), 134. S. Quake, J. Rajagopal, P.R. Tata, E.L. Rawlins, A. Regev, P.A. Reyfman, [81] Q. Yang, T.A. Hughes, A. Kelkar, X. Yu, K. Cheng, S.J. Park, W.C. Huang, J.F. Lovell, O. Rozenblatt-Rosen, K. Saeb-Parsy, C. Samakovlis, H.B. Schiller, J.L. Schultze, S. Neelamegham, Inhibition of SARS-CoV-2 viral entry upon blocking N-and O- M.A. Seibold, C.E. Seidman, J.G. Seidman, A.K. Shalek, D. Shepherd, J. Spence, glycan elaboration, Elife 9 (2020) 1–44. A. Spira, X. Sun, S.A. Teichmann, F.J. Theis, A.M. Tsankov, L. Vallier, M. van den [82] D. Sulheim, E. Fagermoen, A. Winger, A.M. Andersen, K. Godang, F. Muller, Berge, J. Whitsett, R. Xavier, Y. Xu, L.E. Zaragosi, D. Zerti, H. Zhang, K. Zhang, P.C. Rowe, J.P. Saul, E. Skovlund, M.G. Oie, V.W. Bruun, Disease mechanisms and M. Rojas, F. Figueiredo, SARS-CoV-2 entry factors are highly expressed in nasal clonidine treatment in adolescent chronic fatigue syndrome a combined cross- epithelial cells together with innate immune genes, Nat. Med. 26 (2020) 681–687. sectional and randomized clinical trial, JAMA Pediatr. 168 (2014) 351–360. [64] I. Glowacka, S. Bertram, M.A. Muller, P. Allen, E. Soilleux, S. Pfefferle, I. Steffen, [83] J. Slomko, J.L. Newton, S. Kujawski, M. Tafil-Klawe, J. Klawe, D. Staines, T.S. Tsegaye, Y. He, K. Gnirss, D. Niemeyer, H. Schneider, C. Drosten, S. Pohlmann, S. Marshall-Gradisnik, P. Zalewski, Prevalence and characteristics of chronic fatigue Evidence that TMPRSS2 activates the severe acute respiratory syndrome syndrome/myalgic encephalomyelitis (CFS/ME) in Poland: a cross-sectional study, coronavirus spike protein for membrane fusion and reduces viral control by the BMJ Open 9 (2019), e023955. humoral immune response, J. Virol. 85 (2011) 4122–4134. [84] V.B. Wyller, H.R. Eriksen, K. Malterud, Can sustained arousal explain the chronic [65] S. Matsuyama, N. Nagata, K. Shirato, M. Kawase, M. Takeda, F. Taguchi, Efficient fatigue syndrome? Behav. Brain Funct. 5 (2009) 1–10. activation of the severe acute respiratory syndrome coronavirus spike protein by the [85] L. Nacul, S. O’Boyle, L. Palla, F.E. Nacul, K. Mudie, C.C. Kingdon, J.M. Cliff, transmembrane protease TMPRSS2, J. Virol. 84 (2010) 12658–12664. T.G. Clark, H.M. Dockrell, E.M. Lacerda, How myalgic encephalomyelitis/chronic [66] S. Düsterh€oft, J. Lokau, C. Garbers, The metalloprotease ADAM17 in inflammation fatigue syndrome (ME/CFS) progresses: the natural history of ME/CFS, Front. and cancer, Pathol. Res. Pract. 215 (2019), 152410. Neurol. 11 (2020), 826. [67] A. Heurich, H. Hofmann-Winkler, S. Gierer, T. Liepold, O. Jahn, S. Pohlmann, [86] K.A. Schlauch, S.F. Khaiboullina, K.L. De Meirleir, S. Rawat, J. Petereit, TMPRSS2 and ADAM17 cleave ACE2 differentially and only proteolysis by A.A. Rizvanov, N. Blatt, T. Mijatovic, D. Kulick, A. Palotas, V.C. Lombardi, TMPRSS2 augments entry driven by the severe acute respiratory syndrome Genome-wide association analysis identifies genetic variations in subjects with coronavirus spike protein, J. Virol. 88 (2014) 1293–1307. myalgic encephalomyelitis/chronic fatigue syndrome, Transl. Psychiatry 6 [68] N. van Doremalen, K.L. Miazgowicz, S. Milne-Price, T. Bushmaker, S. Robertson, (2016), e730. D. Scott, J. Kinne, J.S. McLellan, J. Zhu, V.J. Munster, Host species restriction of [87] M. Perez, R. Jaundoo, K. Hilton, A. Del Alamo, K. Gemayel, N.G. Klimas, Middle East respiratory syndrome coronavirus through its receptor, dipeptidyl T.J.A. Craddock, L. Nathanson, Genetic predisposition for immune system, peptidase 4, J. Virol. 88 (2014) 9220–9232. hormone, and metabolic dysfunction in myalgic encephalomyelitis/chronic fatigue [69] W. Widagdo, N.M.A. Okba, W. Li, A. de Jong, R.L. de Swart, L. Begeman, J.M.A. van syndrome: a pilot study, Front. Pediatr. 7 (2019), 206. den Brand, B.-J. Bosch, B.L. Haagmans, Species-specific colocalization of Middle [88] J.J. Dibble, S.J. McGrath, C.P. Ponting, Genetic risk factors of ME/CFS: a critical East respiratory syndrome coronavirus attachment and entry receptors, J. Virol. 93 review, Hum. Mol. Genet. 29 (2020) R118–R125. (2019) e00107–19. [89] A.D. Grabowska, E.M. Lacerda, L. Nacul, N. Sepúlveda, Review of the quality [70] Y. Li, Z. Zhang, L. Yang, X. Lian, Y. Xie, S. Li, S. Xin, P. Cao, J. Lu, The MERS-CoV control checks performed by current genome-wide and targeted-genome association receptor DPP4 as a candidate binding target of the SARS-CoV-2 spike, IScience 23 studies on myalgic encephalomyelitis/chronic fatigue syndrome, Front. Pediatr. 8 (2020), 101160. (2020), 293. [71] N. Vankadari, J.A. Wilce, Emerging WuHan (COVID-19) coronavirus: glycan shield [90] N. Gadi, S.C. Wu, A.P. Spihlman, V.R. Moulton, What’s sex got to do with COVID- and structure prediction of spike glycoprotein and its interaction with human CD26, 19? Gender-based differences in the host immune response to coronaviruses, Front. Emerg. Microb. Infect. 9 (2020) 601–604. Immunol. 11 (2020), 2147. [72] S.Y. Rhee, J. Lee, H. Nam, D.S. Kyoung, D.W. Shin, D.J. Kim, Effects of a DPP-4 [91] http://www.meaction.net/wp-content/uploads/2021/04/Report-on-the-impact-of- inhibitor and RAS blockade on clinical outcomes of patients with diabetes and Covid-19-on-ME.pdf. (Accessed 1 August 2021). COVID-19, Diabetes Metab. J. 45 (2021) 251–259. [92] R.K. Gherardi, G. Crepeaux, F.J. Authier, Myalgia and chronic fatigue syndrome [73] A.J. Scheen, DPP-4 inhibition and COVID-19: from initial concerns to recent following immunization: macrophagic myofasciitis and animal studies support expectations, Diabetes Metab. 47 (2021), 101213. linkage to aluminum adjuvant persistency and diffusion in the immune system, [74] M.A. Fletcher, X.R. Zeng, K. Maher, S. Levis, B. Hurwitz, M. Antoni, G. Broderick, Autoimmun. Rev. 18 (2019) 691–705. N.G. Klimas, Biomarkers in chronic fatigue syndrome: evaluation of natural killer [93] J. Phelan, A.D. Grabowska, N. Sepúlveda, A potential antigenic mimicry between cell function and dipeptidyl peptidase IV/CD26, PloS One 5 (2010), e10817. viral and human proteins linking Myalgic Encephalomyelitis/Chronic Fatigue [75] X.-L. Sun, The role of cell surface sialic acids for SARS-CoV-2 infection, Syndrome (ME/CFS) with autoimmunity: the case of HPV immunization, Glycobiology (2021), cwab032. Autoimmun. Rev. 19 (2020), 102487. [76] B.W. Cross, S. Ruhl, Glycan recognition at the saliva – oral microbiome interface, Cell. Immunol. 333 (2018) 19–33. 11