Links Between Immune Genes and Infections in Chronic Fatigue Syndrome and Similar Post-Infection Illnesses

ME/CFS is a chronic, debilitating condition that affects approximately 1% of the population¹, probably an underestimate². Although case definitions vary^3,4, symptoms of ME/CFS include impairing fatigue, post-exertional malaise, cognitive dysfunction, and joint pain, among others. Dysregulation of multiple physiological processes has been reported, affecting immune, inflammation, metabolic, and mitochondrial function, among others⁵. Moreover, the presence of widespread neuroinflammation in cortical areas and subcortical nuclei has been documented using positron emission tomography⁶. Overall, ME/CFS has major impacts on quality of life, including significant social and occupational impairment and disability^3,7. A comprehensive and detailed investigation and assessment of multiple clinical, behavioral, immunological and various biomarker aspects of post-infectious ME-CFS has been published recently⁸.

The etiology of ME/CFS is uncertain; viral infections, including several human herpes viruses, have been prominently implicated as a potential trigger^9,10, and there is evidence of genetic predisposition to ME/CFS^11,12. In 2022, we¹³ proposed that ME/CFS may be, at least in part, due to persistent viral antigens (i.e., fragments of viral proteins) resulting from low binding affinity of those antigens to the Human Leukocyte Antigen (HLA) molecules carried by the patient. The HLA region, which is located on chromosome 6, is the most highly polymorphic region of the human genome¹⁴ and is aimed at host protection against viruses and other foreign antigens. Each individual carries 12 HLA alleles—two of each of the classic HLA Class I (HLA-I) genes (HLA-A, HLA-B, and HLA-C) and two of each of the HLA Class II (HLA-II) genes (HLA-DPB1, HLA-DQB1, HLA-DRB1)—that determine the repertoire of foreign antigens that can bind with sufficient affinity to HLA molecules to mount an immune response, namely killing the infected cell and production of suitable antibodies. Here, we formally test the hypothesized role of HLA-antigen binding with putative viral antigens in ME/CFS risk and protection.

In a recent thorough and well-powered study¹² of HLA and ME/CFS, two HLA alleles were identified that occurred in significantly higher frequency in the ME/CFS group (N = 426) than the control group (N = 4511 healthy and ethnically matched participants), namely C*07:04 and DQB1*03:03. In addition, 2 alleles with independent, significantly higher frequencies in the control group, presumably protective against ME/CFS were identified: B*08:01 and DPB1*02:01. Based on our proposal that ME/CFS is influenced by persistent viral antigens due to HLA-antigen mismatch¹³, we hypothesized that the latter alleles were effective in facilitating elimination of pathogenic viral persistent antigens, thus rendering exposed individuals less susceptible to ME/CFS whereas, in contrast, the risk alleles were hypothesized to have low binding affinity to virus antigens, hindering effective immune response to facilitate their elimination, thus allowing their persistence. Here we tested that hypothesis by determining in silico the predicted best binding affinity (PBBA) of the 4 alleles above to antigens of the 9 known Human Herpes Viruses (HHV), all of which have been implicated in ME/CFS¹⁰, with the expectation that the HLA molecules of the risk alleles (C*07:04, DQB1*03:03) would bind weakly to the viral antigens, in contrast to those of the protective alleles (B*08:01, DPB1*02:01) which would bind strongly.

The resemblance in symptomatology between Long COVID and ME/CFS (fatigue, neurocognitive problems, postexertional tiredness, etc.) was noted early on^13–15. In 2022 we hypothesized that Long COVID might be due to persistent pathogenic SARS-CoV-2 fragments¹³, a hypothesis that has since been supported by the identification of SARS-CoV-2 RNA and viral fragments in long COVID^16,17. In that paper¹³ we hypothesized that the persistence of pathogenic antigens is due to the inability of the patient's HLA to effectively bind and present these antigen fragments, and proposed this as a common mechanism underlying Long COVID, ME/CFS, and Gulf War Illness, a disorder with symptomatology highly similar to ME/CFS, that is linked to anthrax vaccine administration during the 1990–91 Gulf War¹⁸. The HLA-antigen persistence hypothesis was proposed in 2018¹⁹. In this study we tested the hypothesis that the same HLA alleles that were identified as risk or protective for ME/CFS¹² might also share this property for Long COVID by determining exhaustively their binding affinities to peptides of the SARS-CoV-2 spike glycoprotein.

Finally, here we applied the same rationale for another syndrome with symptomatology overlapping those of ME/CFS and Long COVID, namely the PTLDS. Lyme disease is a tick-born disease caused by the spirochete Borrelia burgdorferi with very similar symptomatology to ME/CFS²⁰. In the large majority of cases, if treated early, recovery is complete. However, in a small fraction of patients, "The constellation of symptoms such as fatigue, cognitive dysfunction, and musculoskeletal pain that persist beyond 6 months and are associated with disability have been termed post-treatment Lyme disease syndrome (PTLDS)"²¹. A persistent antigen from B. burgdorferi associated with persistent post-Lyme arthritis is peptidoglycan, a major component of the B. burgdorferi envelope²². It is not known whether, or to what extent, additional antigens may exist and be involved in PTLDS, affecting the brain and other organs, besides the joints. In this study we tested the hypothesis that the ME/CFS-related HLA alleles identified by Lande et al.¹² might also share this property for PTLDS by determining exhaustively their binding affinities to peptides of 5 B. burgdorferi proteins.

Here we evaluated the PBBA of five B. burgdorferi antigens with the ME/CFS alleles by testing exhaustively 2164 peptides (Table 1). For HLA-I, the PBBA (across the 5 antigens) for the ME/CFS risk allele C*7:04 (1410.7 nM) was 20.7 × weaker than the protective B*08:01 allele (68.3 nM) (Table 5). For HLA-II, the PBBA for the ME/CFS risk allele DQB1*03:03 (1028.2 nM) was 14.4 × weaker than the protective allele DPB1*02:01 (71.4 nM).

We tested a total of 15 protein antigens of pathogens (Table 1), including 9 antigens of human herpes viruses (HHV), the spike glycoprotein of SARS-CoV-2 virus, and 5 antigens of Borrelia burgdorferi (B. burgdorferi). The amino acid (AA) sequences of those proteins were obtained from the Uniprot website (https://www.uniprot.org/uniprotkb/↗)⁴² and are shown in Table S1 (Supplementary Material).

We investigated 2 ME/CFS risk HLA alleles (C*07:04, DQB1*03:03) and 2 protective alleles (B*08:01, DPB1:02:01) reported by Lande et al.¹².

The IBM-SPSS statistical package (version 30.0.0.0 172) was used for implementing statistical analyses. Standard statistical methods were used; all P-values reported are 2-sided,. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$a=0.05$$\end{document}