The flu that never ends

This is the translation to English of an article I wrote in Italian several months ago.

Introduction

In three previous articles (here, here and here) – commenting on a recent work published by Fluge and colleagues (Fluge O et al. 2017) – I described a scenario in which a reduced activity of the enzyme pyruvate dehydrogenase (PDH) in ME/CFS patients, leads to an inefficient energy synthesis within the TCA cycle. The reduced function of PDH was inferred from the presence of a phenomenon of amino acids catabolism and from the over-expression of the enzymes pyruvate dehydrogenase kinase (PDK), in particular of the isoforms 1, 2, and 4. Now the question is: what causes this metabolic shift? The authors, on the basis of their success with Rituximab in about 2/3 of their patients, speculated that an autoantibody might – in some patients – turn off some crucial pathway related to energy metabolism. In what follows, I propose an alternative scenario based on a study on mice with influenza.

First act: influenza A and pyruvate dehydrogenase

In 2014, a Japanese group (Yamane K et al. 2014) inoculated the virus of influenza A (IAV) in mice, and for 7 days conducted a study on the unfortunate animals, similar to the one performed by Fluge and Mella on ME/CFS patients, except for the fact that the mice were sacrificed, in order to carry out the measurements directly in their tissues. As you can see in Figure 1.A, after a week from the start of the infection the activity of pyruvate dehydrogenase is reduced in all tissues examined, with the sole exception of the brain; at the same time (Figure 1.B) ATP level drops everywhere, except for the brain. This first part of this experiment can be considered equivalent to the first part of the study by Fluge and Mella, that I discussed here. There is a difference in the type of measurements carried out between the two studies though, but the outcome is the same: energy metabolism is depressed and you have a loss of activity of pyruvate dehydrogenase.

Figure 1. Pyruvate dehydrogenase activity in various tissues (A) and concentration of ATP in the same tissues (B).

Second act: pyruvate dehydrogenase kinase, the usual suspect

Just as Fluge and Mella did, Japanese researchers wondered whether unusually high gene expression of the enzymes pyruvate dehydrogenase kinase (there are four, indicated PDK1, PDK2…) could have been responsible for the decreased activity of pyruvate dehydrogenase they observed. In fact, these four enzymes have precisely the function of inhibiting pyruvate dehydrogenase. And as you can see in Figure 2, PDK4 increases rapidly in the heart, lungs, liver and skeletal muscles, as days pass by.

Figure 2. Expression of PDK4 in various tissues, depending on the days counted from the moment the influenza virus was inoculated.

This second experiment is similar to the one performed by Fluge and Mella on gene expression in peripheral blood mononuclear cells from ME/CFS patients (here). In both cases we have an over-expression of PDK4. Nevertheless, while PDK1 and 2 are over-expressed in humans with ME/CFS, they are normal in mice.

Of mice and men

On the basis of what has been seen during the first 7 days after the influenza A virus inoculation, the mice begin to develop a metabolic dysfunction similar to the one described by Fluge and Mella in ME/CFS patients: an increase in pyruvate dehydrogenase kinase is associated with a loss of function of pyruvate dehydrogenase and an overall shut off of the energy metabolism. What does that mean? It is difficult to draw conclusions, but we might perhaps venture the hypothesis that:

• the metabolic alteration described in ME/CFS patients by Fluge and Mella is nothing other than the one that occurs during an infection.

Since IgMs are produced only one or two weeks after the beginning of an infection, we can exclude that the alterations observed in mice by Japanese researchers are due to antibodies. The authors attribute them to various cytokines (see Figure 3).

Figure 3. Influenza A virus induces the synthesis of cytokines which, in turn, trigger the overexpression of PDK4, that inhibits pyruvate dehydrogenase.

This means that a possible scenario for the defect in pyruvate dehydrogenase activity in ME/CFS patients can simply be a depletion due to the presence of an ongoing infection (in agreement with what proposed by Antony Komaroff, among others, see here), or to a process mistaken for an infection. Of course, this is only one of many possible hypotheses.

So, what about rituximab?

If antibodies are not involved, then why has the drug rituximab – which depletes CD20 expressing B cells – a therapeutic effect in more than half of ME/CFS patients? This is an excellent question, and if we knew the answer, we would be closer to the solution. However, we have to keep in mind that B cells are not just factories for antibodies, but they are also antigen-presenting cells, they produce cytokines (Frances E. Lund 2009) and release mitochondrial DNA (unpublished data, presented by Anders Rosén in this video, minute 14:00) just like mast cells do (Zhang et al. 2012). It is believed that mitochondrial DNA is strongly inflammatory (it is very close to bacterial DNA) and therefore it could be the source for several disorders (Zhang et al. 2012), including even perhaps the dysregulation of pyruvate dehydrogenase. So the effect of rituximab in ME/CFS might not be necessarily linked to autoantibodies depletion.

Conclusion

We have seen that the metabolic disorder recently hypothesized in ME/CFS patients (Fluge O et al. 2016) is also present during the first 7 days of a viral infection in mice. So the common saying according to which “CFS is like flu that never ends” seems correct also from a metabolic point of view.