(We’re in the windup phase of Marco’s Neuroinflammatory Series as he explores the role the immune system and pathogens play in producing the neuroinflammatory state that may underlie chronic fatigue syndrome (ME/CFS)
C – IMMUNE FINDINGS, ETIOLOGY AND ONSET
Last time out I set out to show how a neuroinflammatory model could explain the major, minor and overlap (or even unreported) symptoms of ME/CFS. This time I’ll try to reconcile the model with the immune findings and type of onset.
Numerous immune abnormalities have been found in ME/CFS cohorts but findings are often inconsistent or even at times contradictory, are non-specific to ME/CFS and therefore not useful as a ‘biomarker’.
Nevertheless immune dysregulation of some sort does appear to be a consistent feature of ME/CFS although it has yet to be determined if the immune system dysregulation is a key driver or a downstream effect. Some immune findings suggest past or current viral infection involving a range of pathogens, i.e. no one pathogen has been implicated.
Amongst the few consistent findings are impaired NK cell cytotoxicity (Brenu et al, 2011) and increased pro-inflammatory cytokines such as TNF-alpha (Brenu et al 2011, Maes et al, 2012)
Except for the natural killer cell findings the immune findings haven’t been dramatic (but see Broderick’s work suggesting that studying immune networks is more effective than studying immune factors one by one.)
In a review of immune findings in ME/CFS to date, Bansal et all, 2011 concluded:
“Despite many years of intense investigation there is little consensus on the presence, nature and degree of immune dysfunction in this condition. However, slightly increased parameters of inflammation and pro-inflammatory cytokines such as interleukin (IL) 1, IL6 and tumour necrosis factor (TNF) α are likely present. Additionally, impaired natural killer cell function appears evident.”
Immune markers of inflammation
These findings are remarkably similar to Danzer et al’s description (discussed in Part I) of common immune markers found in “medically ill patients with a range of conditions in which inflammation played a key role”. Dantzer noted that “especially hs-CRP and IL-6 have been found to reproducibly identify the presence of an activated innate immune response”
High-sensitivity c-reactive protein (hs-CRP) is a common and non-specific marker for systemic inflammation which appears to be raised in ME/CFS (Raison et al, 2009).
“ Log-transformed mean plasma concentrations of hs-CRP were increased in subjects with (ME)CFS when compared to subjects who were well”
You may also recall that Rönnbäck and Hansson’s description of ‘mental fatigue’ discussed in Part II suggested the role that pro-inflammatory cytokines may play in mental fatigue :
“we present the hypothesis that the proinflammatory cytokines tumor necrosis factor-α, IL-1β and IL-6 could be involved in the pathophysiology of mental fatigue through their ability to attenuate the astroglial clearance of extracellular glutamate, their disintegration of the blood brain barrier, and effects on astroglial metabolism and metabolic supply for the neurons, thereby attenuating glutamate transmission.”
Primary biliary cirrhosis as a model for ME/CFS?
The work of Julia Newton’s team at Newcastle university examining fatigue and autonomic dysfunction in ME/CFS and the autoimmune disease primary biliary cirrhosis has been discussed here recently. Regarding fatigue in primary biliary cirrhosis Medscape states :
“The etiology of fatigue is unclear; although some evidence suggests that abnormalities of the hypothalamic-pituitary-adrenal axis, decreased release of serotonin, and increased production of proinflammatory cytokines (ie, interleukin-1 [IL-1], interleukin-6 [IL-6], tumor necrosis factor-α [TNF-α] ) may be responsible.”Medscape
Intriguingly, a just published paper (Raszeja-Wyszomirska et al, 2013) reports the association between polymorphisms in TRAF1-C5 – a gene regulating TNF-a with ‘sickness behaviour’ – fatigue, mental and physical function and mood in patients with primary biliary cirrhosis. This finding appears to parallel increased polymorphisms in a TNF-a promoter gene in ME/CFS patients (Carlo-Stella et al, 2006) which led the researchers to comment :
“We hypothesize that CFS patients can have a genetic predisposition to an immunomodulatory response of an inflammatory nature probably secondary to one or more environmental insults of unknown nature.”
Collectively then, the consistently noted immune disturbances found in ME/CFS all accord with the notion of a neuroinflammatory state.
NK cell cytotoxicity
Oxidative stress alone, in conditions such as type II diabetes and renal failure (Berrou et al, 2013) may also account for reduced NK cell cytotoxicity through dysfunction or apoptosis of NK cells (Betten et al, 2004; Wang et al, 2009).
However one NK cell finding requires further explanation. Brenu (Brenu et al, 2012) reported that “significantly lower numbers of CD56brightCD16- NK cells” may lead to a greater susceptibility to viral and other infections and may be a ‘biomarker’.
CD56(bright) NK cells make up a minority of NK cell types but represent the majority in lymphatic tissues and are associated with sites of inflammation (Chan et al, 2007). While the exact proportion of these cells varies, the percentages are expanded or reduced in a certain number of diseases including coronary heart disease, allergic rhinitis and juvenile rheumatoid arthritis (Brenu et al, 2012). While, the significance of these variations is not yet clear (Poli et al, 2009), their presence does suggest that ME/CFS fits in a broad realm of inflammatory disorders.
Most viral studies in ME/CFS were done years ago and are dated and according to Bansal et al. the results are not particularly promising. However they do note some evidence that does suggest an intriguing herpesvirus/natural killer cell interaction :
“While the prevalence of positive serology for the common herpes viruses appears no different from healthy controls, there is some evidence of viral persistence and inadequate containment of viral replication. The ability of certain herpes viruses to impair the development of T cell memory NK cells may explain this viral persistence and the continuation of symptoms.”
Again viral persistence is consistent with impaired NK cell function and is discussed further below.
Similarly to the notion of ‘heterogeneity’ of symptoms, the pattern of onset reported by patients is also heterogeneous.
The various ‘epidemic ME’ clusters cast a long shadow over ME/CFS research (and how we conceive ME/CFS) and, as in the recent case of XMRV, a single pathogen model may initially appear attractive. Numerically, though, ‘epidemic’ cases must represent only a tiny proportion of those diagnosed with ME/CFS and research has shown that only a small proportion of people infected with common viruses will develop ME/CFS regardless of the pathogen, suggesting it is the host response to infection which matters. (Cameron et al, 2006; Katz, Jason, 2012)
Nonetheless, many (but far from all) people with ME/CFS report an acute infection immediately prior to the onset of ME/CFS indicating that viral (or bacterial) infections are a common trigger. As previously seen in Part III, various infections have also been seen to apparently trigger diabetes, stiff person syndrome and anti-NMDA receptor encephalitis.
The Japanese encephalitis virus (JEV) could provide something of a model for ME/CFS. This virus has been shown to cause encephalitis via a mechanism whereby TNF-alpha inhibits glutamate re-uptake leading to raised extracellular glutamate and excitotoxicity (Chen et al, 2012) :
“Our findings support a potential link between neuroinflammation and the development of excitotoxic neuronal injury in Japanese encephalitis. The link between neuroinflammation and excitotoxic death may involve a mechanism in which TNF-α released by microglia plays a facilitory role in glutamate excitoneurotoxicity via up-regulation of glutamate synthesis and down-regulation of glutamate uptake.”
Note that this process is remarkably similar to that proposed by Rönnbäck and Hansson to cause mental fatigue.
We also saw in Part I that ‘psychological’ stress (isolation rearing) can induce a sensory gating deficit in rats and that (as discussed in Part II) this deficit may be a measurable artefact of a glutamate/GABA imbalance. Many people with ME/CFS report a period of elevated stress prior to or around onset, a ‘correlation’ that led some to conclude that ME/CFS is largely a ‘mood’ or ‘stress’ related disorder. Others, however, report that there were no unusual feelings of stress or stressful events associated with onset.
As discussed above, both acute traumatic stress and chronic stress may cause lasting glutamate induced dysregulation of the HPA axis. We also saw in Part I that the dopamine and norepinephrine (and indirectly glutamate) modulating COMT gene has been associated with ME/CFS in three studies. The low activity met allele of COMT is associated with increased pain sensitivity, executive function that deteriorates under stress and may mediate the emotional response to negative stimuli and stress (Smolka et al, 2004).
Prevalence estimates for the met/met allele equate to approximately 25-30% of the European population (SNPedia) with the low activity met polymorphism more common in women (Tunbridge, 2010 – Press release). It may be that psychological stress represents a more significant stressor for people who have one or more met alleles of COMT.
Exercise may be another stressor that ‘kindles’ the vicious neuroinflammatory cycle. Many patients report having been very physically active prior to onset and a large scale prospective study (Harvey et al, 2008) suggests that those who exercised frequently are at greater risk of developing ME/CFS. Strenuous exercise of course also induces oxidative stress and high demands on mitochondria and this could be an initiating factor in the ‘vicious cycle’. As discussed above, ME/CFS patients appear to have an attenuated response from the normally protective heat shock proteins.
In the context of a possible autoimmune etiology of ME/CFS discussed in Part II, it’s notable that CD56 bright NK cell counts discussed above may be used to determine how effective the monoclonal antibody daclizumab is in the treatment of relapsing/remitting multiple sclerosis (RRMS). This supports the immunomodulatory role natural killer cells play in reducing brain inflammation and slowing disease progression in RRMS. (Bielekova et al, 2006) :
“…daclizumab therapy was associated with a gradual decline in circulating CD4+ and CD8+ T cells and significant expansion of CD56 bright natural killer (NK) cells in vivo, and this effect correlated highly with the treatment response”
This suggests that the low levels of CD56 bright natural killer (NK) cells seen in ME/CFS could reflect similar ‘brain inflammation’ and a possible autoimmune etiology.
As seen in Parts II and III, a number of conditions which involve a glutamate/GABA imbalance also now appear to have an autoimmune component including stiff person syndrome, anti-NMDA receptor encephalitis, late onset autism, autoimmune encephalitis and potentially type II diabetes, fibromyalgia and schizophrenia/schizoaffective disorders. As discussed above primary biliary cirrhosis is an autoimmune disease sharing many symptoms with ME/CFS where TNF-a may contribute to neuroinflammation.
An endogenous retrovirus?
Having apparently got one retrovirus out of our systems (pardon the pun), the results of Brigette Huber’s research on the involvement of the endogenous retrovirus HERV-K18 in ME/CFS have been long-awaited.
Human endogenous retroviruses (HERV’s) are the legacy of previous, often ancient, encounters with retroviruses which have become part of the human genome. While previously considered to be ‘inactive’ they can be reactivated by various environmental influences (including common pathogens) and are now implicated in the pathology of a range of diseases including cancers, nervous system diseases and autoimmune rheumatic and connective tissue diseases (Balada et al , 2009). The (not exhaustive) list includes rheumatoid arthritis , Systemic Lupus Erythematosus (SLE), Sjögren’s syndrome (Balada et al , 2009), multiple sclerosis (Laska et al, 2012), schizophrenia and bipolar disorder (Diem et al, 2012).
HERVs and ME/CFS
Similarities in the symptoms between ME/CFS and certain autoimmune diseases such as multiple sclerosis and Systemic Lupus Erythematosus (where HERVs may play a role) have led some investigators to look for and find a putative association between HERVs and CFIDS/ME/CFS (Urnowitz and Murphy, 1991; Huber – Conference press release, 2008).
Huber has previously presented data that suggests that HERV K18 may be reactivated in some ME/CFS and multiple sclerosis patients and that this reactivation may result from an initial infection by Human Herpes Virus -6 or Epstein Barr virus. These preliminary findings appear not to have been replicated in a more stringent study that failed to find evidence of higher levels of either HHV-6 or HERVs in the blood or saliva of ME/CFS patients (Oakes et al, 2013) :
“We fail to demonstrate a difference in HERV-K18 env transcripts, HHV-6 viral copy number, and HHV-7 viral copy number between CFS patients and healthy controls. Our data do not support the hypothesis of reactivation of HHV-6 or HHV-7 in CFS.”
Notwithstanding this negative finding, HERV’s have been implicated in a number of the neuroinflammatory conditions discussed previously and recently antibodies to a range of HERVs (including HERV K) have been detected in the tissues of the gut in a small number of ‘ME’ patients but not in controls (Lombardi et al, 2013).
HERV involvement in human diseases and the specific mechanisms involved is too vast for this review however, in the context of the current discussion, I’ll make just a few observations.
HERVs in ‘Neuroinflammatory’ diseases
Variants of HERV-W and HERV-K have been associated with schizophrenia and bipolar disorder (Diem et al, 2012) and may be reactivated by common pathogens including influenza virus, herpesvirus and Toxoplasma gondii (Perron et al, 2012) :
“HERV-W envelope gene (env) is activated by environmental factors and encodes a protein displaying inflammation and neurotoxicity”
“The seroprevalence for Toxoplasma gondii yielded low but significant associations with HERV-W transcriptional level in a subgroup of BD and SZ (bipolar disorder and schizophrenia), suggesting a potential role in particular patients.”
In the context of the association between ‘neuroinflammatory’ diseases and diabetes, polymorphisms in HERV K may be associated with a higher incidence of type II diabetes in schizophrenics (Dickerson et al, 2008; Nyegaard et al, 2012).
HERV impact on glutamate and GABA?
HERVs are known to ‘make use of’ many elements of the human biological machinery in order to replicate – including binding to the receptors of several neurotransmitters including the glutamate transporter ASCT2 receptor (Marin et al, 2003) lending some ‘construct validity’ to the role of HERVs in a ‘glutamate model’ of schizophrenia. Needless to say parking an endogenous retrovirus on top of neurotransmitter receptors might very well interfere with the normal functioning of those receptors. This ability has been hypothesized as being the result of the (ongoing) ‘arms race’ between viruses and the host immune system and enables HERVs to gain access to a wider range of human tissues.
One recent hypothesis (Hegyi, 2013) suggests that the long terminal region (which is known to silence, enhance or regulate gene activity) of HERV-W may downregulate the GABA receptor B1 creating a GABAergic deficit in schizophrenia. The author further suggests that the immune response to suppress HERV transcription may also suppress the activity of GABA.
While schizophrenia is obviously not ME/CFS it may share some sensory gating issues with it and other disorders including ME/CFS which might suggest common biological pathologies. Both of these findings offer a plausible mechanism for an association between HERVs and ‘neuroinflammatory’ conditions involving ‘mood’, sensory processing problems (and potentially movement disorders).
Antiretrovirals as a treatment for multiple sclerosis?
A clinical trial is currently planned to investigate the efficacy of an antiretroviral in multiple sclerosis (Giovannoni, Queen Mary University of London) where, as discussed above, CD56bright NK cell counts may act as a biomarker for the efficacy of daclizumab treatment :
“There is accumulating research evidence that Human Endogenous Retrovirus (HERV) and herpes viruses (in particular Epstein-Barr Virus) are involved in the pathogenesis of multiple sclerosis”.
People with active MS have higher levels of HERVs than people either without MS or who have other neurological conditions. It has been shown that HERVs may produce neurotoxic proteins/antigens associated with MS activity and disease progression.
This is the first clinical trial investigating the hypothesis that the antiretroviral drug raltegravir may suppress HERV activity and ameliorate progression of relapsing remitting MS.”
Significant gender imbalances are common in autoimmune disease (more females), autism spectrum disorder (ASD) (males) and ME/CFS (females). The protective effects of the estrogen mediated RORA gene may help explain the gender imbalance in ASD. Assuming that gender imbalances in ME/CFS are not merely an artifact (e.g. perhaps resulting from under-reporting or alternative diagnoses in men), as previously discussed, COMT gene polymorphisms that affect dopamine levels (and indirectly glutamate levels) could play a role in the gender imbalance in ME/CFS. As discussed in Part II recent research strongly suggests that gender (or more specifically the effects of estrogen) impacts on COMT functioning could lead to sexually dimorphic effects on brain function.
Estrogen also appears to have significant effects on some of the genes associated with ME/CFS – i.e. if you’re female, estrogen’s effects on the RORA gene may protect you against autism but if you have a certain form of the COMT gene, your brain may more vulnerable to ME/CFS.
Sudden Vs Gradual Onset – Were We Ever ‘Well’?
Once again, ME/CFS patients report different types of onsets with some reporting a sudden dramatic onset of symptoms (often following a viral illness) or a more gradual onset over a period of time. Some also report a ‘morphing’ of symptoms over time – often described in the literature as symptoms that are ‘protean’. In both cases most patients however report few prior physical complaints.
As discussed in Part III, infection appears to be able to trigger a number of conditions involving a glutamate/GABA imbalance and from this respect, if the neuroinflammatory cycle also underpins these cases, may represent a ‘de novo’ pathology.
However the ‘Dubbo’ and post-infective mononucleosis studies discussed above do also suggest that developing ME/CFS requires a combination of a pathogen and predisposing factors.
My introductory biography did state that I was fully physically healthy prior to the onset of ME/CFS. This is true, but it is also an error of omission in that I didn’t state that I had always experienced a high level of anxiety and an enhanced stress response plus some ‘characteristics’ (poor social skills, poor face recognition, sensory issues, delayed speech) that might suggest some ASD like tendencies.
I would propose that at least some ME/CFS patients might recognise similar characteristics prior to the onset of ‘physical illness’ but and that some of these symptoms may have increased post onset (others may have developed some of these characteristics afterwards and others may not have them at all). I suggest that these ‘atypical’ symptoms may be early indications of a predisposition to developing this neuroinflammatory cycle.
A one, two or multiple ‘hits’ etiology
More than likely all of the above factors act as stressors individually or in concert which may overwhelm a system that is predisposed to illness of various forms via the development of a neuroinflammatory vicious cycle. This predisposition may be genetic and could involve antioxidant, immune, mitochondrial integrity, neurotransmitter balance or an autoimmune or perhaps even a HERV component.
In this respect the model proposed is similar to that set out by Beatrice Golomb for Gulf War Illness and overlapping conditions such as CFS and ASD where a wide range of putative stressors in GWI may all result in the same pathological process involving oxidative stress/mitochondrial dysfunction with many downstream effects including a vulnerability to autoimmunity (Golomb, 2012).