Salivary DNA Loads for Human Herpesviruses 6 and 7 Are Correlated With Disease Phenotype in Myalgic Encephalomyelitis/Chronic Fatigue Syndrome

Abstract

Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) is a complex chronic condition affecting multiple body systems, with unknown cause, unclear pathogenesis mechanisms, and fluctuating symptoms which may lead to severe debilitation. It is frequently reported to have been triggered by an infection, but there are no clear differences in exposure to, or seroprevalence of, any particular viruses between people with ME/CFS and healthy individuals. However, herpes viruses have been repeatedly hypothesized to underlie the chronic relapsing/remitting form of MS/CFS due to their persistence in a latent form with periodic reactivation. It is possible that ME/CFS is associated with herpes virus reactivation, which has not been detectable previously due to insufficiently sensitive testing methods. Saliva samples were collected from 30 people living with ME/CFS at monthly intervals for 6 months and at times when they experienced symptom exacerbation, as well as from 14 healthy control individuals. The viral DNA load of the nine humanherpes viruses was determined by digital droplet PCR. Symptoms were assessed by questionnaire at each time point. Human herpesvirus (HHV) 6B, HHV-7, herpes simplex virus 1 and Epstein-Barr virus were detectable within the saliva samples, with higher HHV-6B and HHV-7 viral loads detected in people with ME/CFS than in healthy controls. Participants with ME/CFS could be broadly separated into two groups: one group displayed fluctuating patterns of herpesviruses detectable across the 6 months while the second group displayed more stable viral presentation. In the first group, there was positive correlation between HHV-6B and HHV-7 viral load and severity of symptom scores, including pain, neurocognition, and autonomic dysfunction. The results indicate that fluctuating viral DNA load correlates with ME/CFS symptoms: this is in accordance with the hypothesis that pathogenesis is related to herpesvirus reactivation state, and this should be formally tested. Herpesvirus reactivation might be a cause or consequence of dysregulated immune function seen in ME/CFS. The sampling strategy and molecular tools developed here permit such large-scale epidemiological investigations.

Keywords: Clinical specimens; DNA viral load; ME/CFS; digital droplet PCR; human herpesvirus.

Figure 1

DNA concentration measured by ddPCR assays for the nine human herpes viruses. (A) Example 1D dot plot data file for HHV-6A and HHV-6B ddPCR assays. The X-axes show quantities of commercial viral DNA in different assay plate wells (E02, F02, G02, etc.,). Assays were run either singly or in duplex as shown. The Y-axes indicate the fluorescence intensity within each PCR droplet, with the pink line indicating the threshold for positivity. (B) Example representation of 2D dot plot displaying duplex ddPCR of HHV-6B (FAM labelled, channel 1) and HHV-6A (HEX labelled, channel 2). (C) Standard curves obtained for each of the nine human herpesviruses, indicating the concentration of serial dilutions of commercial DNA on the X-axes, and the calculated concentration of viral DNA from the ddPCR assay on the Y-axes.

Figure 2

Flow diagram of participant recruitment and follow up in the longitudinal saliva clinical study. Serological test (n = 235) and consecutive ddPCR assay in the plasma (n = 189) and PBMC (n = 76) were performed for HHV screening purposes. In total, 46 participants were enrolled into a 6-month longitudinal study to determine salivary HHV. Saliva samples (monthly and day 1, 3, and 5 of an acute illness or worsening of disease episode in people with ME/CFS, n = 276) were then analysed for HHV DNA concentration by ddPCR assay, with results compared with symptoms scores. *30 participants were tested for both plasma and PBMC.

Figure 3

Herpes virus DNA prevalence and concentration in saliva from people living with ME/CFS, measured by ddPCR. (A) Percentage HHV prevalence in HC (n = 16) and people with mild/moderate ME/CFS (ME/CFS_MM: n = 14) or severe symptoms (ME/CFS_SA: n = 16) at month 1. (B) Persistence of viral DNA positivity throughout 6 months. The persistency (%) was calculated for each participant as the number of times HHV was detected divided by the total number of samples collected over 6 months, multiplied by 100. HHV viral DNA concentration comparison between HC and ME/CFS patients. Each monthly viral DNA concentration in saliva from participants with mild/moderate (ME/CFS_MM) or severe (ME/CFS_SA) disease was compared to HC across the 6 months for (C) HSV-1, (D) EBV, (E) HHV-6B, and (F) HHV-7. Scatter dot plots show the median and the interquartile range. P-values < 0.1 (Mann-Whitney) are shown.

Figure 4

Change in herpesvirus DNA concentration in saliva in individual participants over time. (A) The monthly viral load of HSV-1, EBV, HHV-6B, and HHV-7 detected in saliva from month 1 (Jan 2018) to month 6 (Jun 2018), measured by ddPCR assay, is shown for individual HC, ME/CFS_MM and ME/CFS_SA study participants. (B) The HHV viral load of HSV-1, EBV, HHV-6B, and HHV-7 in ME/CFS patients is shown on day 1, 3, and 5 of episodes when participants experienced an acute illness or worsening of disease symptoms.

Figure 5

Disease symptom and correlation in people living with ME/CFS and Healthy controls. (A) Median score and IQR of seven ME/CFS symptom domains across 6 months in people with mild/moderate (n = 14) or severe ME/CFS (n = 16) and HC (n = 16), based on questionnaires completed at the time of saliva sample collection by participants. (B) Correlation matrix between seven ME/CFS symptom domain scores in people with ME/CFS (n = 30).

Figure 6

Correlations of salivary HHV DNA concentration with disease symptoms scores. (A) Correlation between monthly symptoms score and monthly HHV viral DNA concentration in people with ME/CFS_SA (pattern 1, left panel; pattern 2, right panel). All participant's monthly HHV viral DNA concentration and symptoms score data were included in the correlation analysis. Each symbol represents an individual's monthly sample. NB: Zero values are not shown on the graph due to the logarithmic scale, but are included in the analysis. Spearman's coefficient is denoted as r. The best-fit regression line, 95% confidence interval (grey area) and prediction band (dotted line) are shown. (B) Correlation between symptom scores and EBV, HHV-6B, and HHV-7 DNA concentrations in saliva samples from participants experiencing an exacerbation of symptoms.