Serum from Myalgic encephalomyelitis/chronic fatigue syndrome patients causes loss of coherence in cellular circadian rhythms, 2023, Wei et al

Abstract
Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) is a disabling disorder characterized by disrupted daily patterns of activity, sleep, and physiology. Past studies in ME/CFS patients have examined circadian rhythms, suggested that desynchronization between central and peripheral rhythms may be an important pathological feature, and identified associated changes in post-inflammatory cytokines such as transforming growth factor beta (TGFB). However, no previous studies have examined circadian rhythms in ME/CFS using cellular models or studied the role of cytokines on circadian rhythms.

In this study, we used serum samples previously collected from ME/CFS patients (n = 20) selected for the presence of insomnia symptoms and matched controls (n = 20) to determine the effects of serum factors and TGFB on circadian rhythms in NIH3T3 mouse immortalized fibroblasts stably transfected with the Per2-luc bioluminescent circadian reporter. Compared to control serum, ME/CFS serum caused a significant loss of rhythm robustness (decreased goodness of fit) and nominally increased the rate of damping of cellular rhythms. Damping rate was associated with insomnia severity in ME/CFS patients using the Pittsburgh Sleep Quality Index (PSQI).

Recombinant TGFB1 peptide applied to cells reduced rhythm amplitude, caused phase delay and decreased robustness of rhythms. However, there was no difference in TGFB1 levels between ME/CFS and control serum indicating the effects of serum on cellular rhythms cannot be explained by levels of this cytokine. Future studies will be required to identify additional serum factors in ME/CFS patients that alter circadian rhythms in cells.

https://www.sciencedirect.com/science/article/abs/pii/S0165572823001285

 

If they are focusing on circadian rhythms I'd be interested in Cortisol readings as well.

 

It's odd they didn't select controls with insomnia.

 

I have a monoallelic frameshift mutation in the PER1 gene which is part of the the clock machinery that creates circadian rythms in the brain and peripheral tissues. My understanding is that it would not cause the serum to have the effect described here but might further worsen circadian rythms. A risk factor?

 

Quite selective inclusion criteria but no requirement for matching CCC or IOM criteria (unless that's built into the NINDS repository). Per2 is a transcription factor whose expression levels oscillate over a roughly 24 hour cycle and who regulates circadian rhythm related activity as I understand it. They're using mouse cells with a version of Per2 fused to luciferase, which gives a fluorescent signal which they can read off.

Considering these are mouse fibroblasts, they get large differences in amplitude and periodicity of per2 expression between using bovine and human control serum:

So between these 3 different organisms it's hard to tell how well this experimental system will reflect things in vivo. Panel E here is showing example traces of per2-luc expression over multiple days after adding the serum and medium and leaving them be. In this paper they are fitting dampened sine waves to the traces which reflect this periodicity and a gradually reducing amplitude.

The key part of this paper for ME/CFS seems to be figure 2:
Panels B, C and D show there is no difference between controls and ME/CFS in the amplitude or periodicity. You can also see there is a very strong dose dependent effect of the serum, in particular in reducing the amplitude and variance of the amplitude. So again hard to know what would be relevant in vivo.


They are arguing differences between controls and ME/CFS can be seen in panels E, F, G, H. Panel E shows when they model a dampened sine wave over the data, the ME/CFS data @ 20% serum do not fit as well as the controls. This, they argue, is because of ME/CFS 20% serum, which demonstrate a more rapid flattening out of the signal as shown in their examples traces panels G and H, and therefore a higher damping constant (Panel F, not significant)

Alltogether no difference in the circadian rhythm itself, and a possible subtly faster loss of signal in an experimental system which behaves very differently in different configurations - not a strong case for a positive result imo.

 

How disappointing that the effect is so weak.