Model of Post Infective Fatigue Forming: the Dubbo Studies

The Dubbo studies are a series of studies funded by the CDC and the Australian government that examine what happens to the subset of CFS patient whose disease is triggered by an infection (Epstein-Barr Virus (EBV), Ross-River virus, Coxiella burneti). Because these studies focus on producing an infectious model of CFS they may not reflect processes occurring in CFS patients with a different type of onset.

The three studies reviewed here focus on Infectious mononucleosis (IM), an EBV caused disease found in adolescents and adults. IM is a particularly apt disease to study in reference to CFS. Like CFS it is characterized by extreme fatigue, and like CFS, viral loads cannot explain the lingering problems found.

Two studies have shown that about 10% of healthy young adults that come down with infectious mononucleosis (IM) still experience disabling symptoms, in particular fatigue, six months after the disease’s onset. Since the CDC has stated that almost all the chronically ill patients in the Dubbo studies met the criteria for CFS they will be referred to as CFS patients in these overviews.

There’s a lot at stake in these studies. The proponents of the ‘biopsychosocial’ model of CFS propose that it mostly occurs in people who for one reason or another just cope poorly with disease. If these studies find no differences between the people who get ill and stay ill and those who recover from the infection, then the case of the CBT proponents will be enhanced.

If, on the other hand, these studies find that the still ill patients undergo unique physiological changes in response to infection then not only will the ‘biopsychosocial’ model of CFS take a hit but the beginnings of unique disease process may be uncovered.

The first study

Cameron, B., Bharadwaj, M., Burrows, J., Fazou, C., Wakefield, D., Hickie, I., Ffrench, R., Khanna, R. and A. Lloyd. 2006. Prolonged Illness after infectious mononucleosis is associated with altered immunity but not with increased viral load. Journal of Infectious Diseases 193, 664-671.

At least two general possibilities could explain the ongoing debility of the still ill or ‘CFS’ patients in these studies; they could have trouble eliminating the virus, or they could display an response to the virus that persists after the virus has been eliminated.

This study examined both possibilities; it examined the viral loads to see if the CFS patients were unable to eliminate the virus, and cytokine levels to see if the CFS patients had a different response to the pathogen. Pro-inflammatory cytokines such as TNF-a, IL-1b and IL-6 are known to cause the symptoms of ‘sickness behavior’ (malaise, fatigue, aching muscles, etc.) so similar to those found in CFS that commonly occur during the early stages (acute phase) of infectious illness

Immune response:  The successful resolution of IM appears to derive from an individual’s ability to mount a ‘broad-based’ cytotoxic T-cell response to EBV. Studies have found that increased levels of EBV specific cytotoxic T-cells are paralleled by declines in the levels of circulating EBV.

Viruses usually enter cells, use the cell’s RNA machinery to produce more viruses and then leave the cell and swarm through the bloodstream looking for more cells to infect. Large numbers of circulating viruses (virions) in the blood stream indicate virus replication is in full swing.

An examination of the cytotoxic T-cell response found that neither people who recovered or who didn’t recover from IM, had a significantly different T-cell response; both groups were able to fairly quickly produce large numbers of EBV specific T-cells that were able to eventually knock down the infection.

The researchers also examined how well the immune system responded to the production of two EBV proteins associated with EBV replication. The proteins viruses produce help it to assemble new virions. The T-cells should take note of these new proteins and react by producing a pro-inflammatory cytokine, IFN-7, that helps to regulate the immune response. Both the still ill and the healthy IM patients reacted similarly to the presence of these viral proteins; these patients did not appear to be ill because of their inability to mount an immune response.

Because cytokines can be responsible for many of the symptoms associated with infection, the researchers also examined the two groups to see if the ‘CFS’ patients demonstrated increased or prolonged levels of pro-inflammatory cytokines. The cytokine levels of the ‘CFS’ patients did not, however, differ from those of the recovered patients.

The humoral branch (B-cells) of the immune system has long been thought, because of its delayed appearance, to play a relatively minor role in EBV suppression. Indeed, the antibodies produced by B-cells (IgG anti-EBNA – Epstein-Barr nuclear antigen) appear to signal the beginning of the latency phase when EBV takes up its lifelong residence in B-cells. The presence of anti-EBNA antibodies is believed to be associated with the period of convalescence that follows the acute immune response.

Here, however, a difference was found; the humoral response occurred earlier and was stronger in the CFS-like patients. The researchers posited this meant EBV was able to mount a quicker attack in these patients. They noted that although the both sets of patients mounted an equally powerful cytotoxic T-cell attack that it was ‘somewhat slower’ in the CFS-like ones.

They suggested that the earlier humoral response in CFS patients may have been due to enhanced Th2 immune responses. CFS patients have long been thought to display an enhanced Th2 immune response. Cytokines produced during the Th2 immune response activate B-cells, among other things, and increased B-cell activity and antibody production is one of the hallmarks of the Th2 immune response. Interestingly EBV is able to produce a cytokine, IL-10, that induces a Th2 response.

This was not enough to satisfy the authors, however, that a cause had been found for the protracted illness seen in these patients. None of the critical factors involving EBV suppression; the cytotoxic T-cell response, IFN-y production, or the cytokine response were found to be altered.

Neither the viral load in the bloodstream or in the cells themselves was higher in the chronically ill patients; they appeared to have successfully resolved the infection! Based on this data authors were unable to determine why these patients were still so sick.

Almost coyly, however, they suggested in the last sentence of the paper, that they did have a idea where the problem lay but aren’t going to share the details of it with us yet. They stated “we propose that alternative neurobiological mechanisms triggered during the severe, acute illness and sustained in the absence of peripheral inflammation underpin prolonged illnesses after EBV infection”.

This suggests the problem lies not in the immune system but in the nervous system (probably the brain) and that it is triggered during the initial phase of the illness and then sustained even after the infection has been resolved. It suggests this team has found markers of nervous system dysfunction both early in the illness and later on in the CFS patients.

Indeed Lloyd later in an interview said “We believe that the parts of the brain that control the perception of fatigue and pain get damaged during the acute infection phase of glandular fever. If you’re still sick several weeks after infection, it seems that the symptoms aren’t being driven by the activity of the virus in body, it’s happening in the brain.

“It’s not too big a leap of faith to say after that, it’s in the brain, because of the nature of the symptoms – it’s fatigue, it’s pain, sleep disturbance, concentration and memory difficulties and mood disturbance. They’re very much brain symptoms.”

While the Lloyd team did look at many of the parameters of EBV infection and the immune response their research was not necessarily conclusive. They did not, for instance, search for the antibodies to the early EBV enzymes the Glaser’s believe are creating havoc in CFS patients. It is also possible that the virus could have been vanquished in the periphery but became established in the brain.

Several studies have suggested that immune activation may be occurring in the brains of CFS patients. These include the increased white blood cell levels in Natelson’s cerebrospinal fluid study, a distinctive proteome that could reflect glial cell activation in the Baraniuk cerebrospinal fluid study, and the increased choline levels seen in the basal ganglia in several magnetic resonance spectroscopy studies.

Albashi believes HHV-6A infection of the CNS contributes to the symptoms seen in CFS (see The Pathogens in CFS, Part I, HHV-6). The Kauschik/Kerr and Vernon gene expression studies have found evidence of both central nervous system and immune dysregulation in CFS. Hopefully the Lloyd team’s next paper explaining their rather cryptic comment will be out soon.

The second study

The next Dubbo study examined the gene expression patterns in chronically ill and recovered IM patients. Oddly enough, at first glance at least, it appeared to come to a conclusion opposite than that of the Cameron study.

Vernon, S., Whistler, T., Cameron, B., Hickie, I., Reeves, W. and A. Lloyd. 2006. Preliminary evidence of mitochondrial dysfunction associated with post-infective fatigue after acute infection with Epstein Barr Virus. BMC Infectious Diseases 6: 15.

This study found that the gene response in the chronically ill IM (or CFS) patients differed early in the illness, and this difference appeared to reflect the ability of EBV to better evade or exploit these patients immune systems and replicate in the body.

Amazingly all eight of the genes more highly expressed in the CFS-like patients in the early stages of the IM were involved the activation of a phase of the cell  cycle.  The cell cycle is invoked when cells begin to  proliferate. I believe EBV induces B-cells to proliferate in order to create the B-memory cells it hides out in.

Four of the genes differentially expressed in the chronically ill group were involved in parts of the immune response EBV is known to evoke or exploit. One, an interferon stimulated gene (ISG20) is evoked by cells infected with viruses. Another, involved in apoptosis, is evoked when the immune system kills off infected cells by invoking their suicide program.

Three of the genes were involved in a cellular process called the cell cycle that EBV hijacks in order to replicate in the bodies cells.

An analysis of genes that were upregulated both early and late in the disease process found that over half of them were involved in fatty acid metabolism and mitochondria functioning and several of these genes are known to be associated with increased EBV activity.

One of the early proteins produced by EBV (BRLFI), for instance, induces the infected cell to produce an enzyme, called fatty acid synthase (FAS), that produces a fatty acid (palmitate) found in cellular membranes. EBV uses the FAS enzyme and palmitate to awake from latency and begin replication in the cell. Li et. al. noted that the active ingredient in green tea (epigallocatechin gallate) has been found to inhibit FAS activity and EBV replication*.

The authors posited that the mitochondrial abnormalities found may have been responsible for the immune alterations seen in  the first study.  Indeed, EBV infected monocytes display reduced production of TNF-a, an important pro-inflammatory cytokine.

The tight fit between the genes found and EBV replication was impressive and suggested increased EBV replication had occurred in the CFS patients. Indeed the authors stated that the gene expression changes seen:

“potentially implicates a failure of the host to adequately control viral replication.”

Unfortunately the very small sample size (5 CFS patients) impeded the researchers from doing a more thorough gene expression analysis. Because they had to pool both the late and the early gene results we don’t know, for instance, if the mitochondrial activity was induced early or late in the disease process or both.

The number of genes assessed (3,800) was also quite small for a gene expression study. These factors, however, make the impressive results found all the more startling. Hopefully larger follow up studies are in process.

Why was the patient sample so small? The researchers followed the cohort of IM patients that presented themselves in the township of Dubbo, Australia over a period of time. Either they didn’t realize that so few people would get IM or they simply were satisfied with such a small study.

Since only about 10% of IM patients come down with ‘CFS’ the researchers must have known that their odds of getting an adequate sample size were small. We can be thankful to the CDC for pursuing such innovative studies yet one can only rue their decision not to commit the resources needed for this investigations results to have been other than ‘preliminary’.

These studies take a long time; the Dubbo studies have be ongoing since 1999 – will it take another five years to get other than a ‘preliminary’ study?

How to explain the differing results from the two studies? The first study suggested the CFS patients successfully contained the virus; the second one suggested they did not. One explanation might involve the range of the different studies.

The Cameron/Lloyd study looked strictly at peripheral indices of infection and immune activation while the Vernon study examined the gene expression of immune cells that are able to reflect both peripheral and central processes. Could the Vernon study be picking up evidence of a CNS infection or disruption? Is the fact that EBV is able to infect the central nervous system important?

EBV may be a factor in another fatiguing disease of the central nervous system, multiple sclerosis (MS). A recent study found that EBV infection was a risk factor for MS. Most intriguingly, those at the highest risk of MS encountered EBV later in life; i.e. they came down with infectious mononucleosis.

(Sooner or later most people get exposed to EBV. If you do so as a child you usually have a mild infection, if you’re an adolescent or adult you have a good chance of getting infectious mononucleosis and if you get infectious mononucleosis you apparently stand a pretty good chance of getting CFS (1/10) and may have an increased (but still low) chance of getting MS.)

The third study

The next paper from the Dubbo group, which looked at gene expression during the early (acute) phase of infectious mononucleosis may give us some clues as to what could go wrong in the chronically ill patients.

Vernon, S, Nicholson, A., Rajeevan, M., Dimulescu, I., Cameron, B., Whistler, T. and A. Lloyd. 2006. Correlation of psycho-neuroendocrine-immune (PNI) gene expression with symptoms of acute infectious mononucleosis. Brain Research 1068, 1-6.

Most gene expression studies have focused on genes found in immune cells in the blood called peripheral blood mononuclear cells (monocytes and lymphocytes). This seems fine for examining immune activity in CFS but what about uncovering information on metabolic or nervous system processes? How in the world can one get information on these processes from immune cells?

Researchers at the CDC recently demonstrated – to their surprise – that almost 2/3rds of 1600 genes known to mediate psychological – neuroendocrine – and immune (PNI) processes were expressed in PBMC’s. This indicates gene expression in the PBMC’s does give a credible snapshot of the biological activity in the body. PBMC gene expression analysis has, in fact, been described as being a kind of molecular biopsy of the body.

This study examined the expression of 1058 genes known to be engaged in PNI processes. First the researchers assessed the symptom presentation of people with infectious mononucleosis. Then they correlated the symptom expression data with gene expression data in order to determine which genes may have contributed to which symptoms.

It is important to note that they did not look at gene expression in CFS patients; this study simply examined what happens to the body to create the symptoms of infectious mononucleosis. Since CFS and infectious mononucleosis have such similar symptoms, however, it is possible that this study uncovered genes involved in generating the symptoms found in CFS.

Because this study did not differentiate between chronically fatigued and recovered patients some genes that could be uniquely upregulated in CFS patients may not be presented here. We don’t know if they examined these genes in the second Dubbo study or not. This study, then, does give us a preliminary idea of what might have gone wrong in CFS.

The researchers must have beamed at their results. No genes were significantly associated with fever, malaise or irritability/depression but they found a nice tight fit between gene expression and fatigue, sleep disturbance and neurocognitive problems. This means there is little doubt that these genes play a role in causing these symptoms.

Surprisingly, none of these genes were directly involved in cytokine production. Many researchers believe the symptoms (i.e. sickness behavior) seen in the early stages of an infection are cytokine related. This is an important finding.

Skeptics of a physiological interpretation of a post-infective fatigue state have been able to point to the conflicting findings in CFS cytokine studies to assert there is no evidence of an aberrant physiological process. But what if we’re looking in the wrong place? This study strongly suggests that cytokines are not the only agents involved in producing ‘sickness behavior’.

Sleep problems (Hypocretin/orexin)

How nicely this gene fits this symptom! Low hypocretin levels occur in narcoplepsy, a disease characterized by extreme sleepiness during the daytime and insomnia at night. Narcoleptics get about as much sleep as everyone else but they can’t get it in one shot – they are unable to stay awake or asleep for significant periods of time.

Hypocretin is produced by neurons that extend throughout the brain with especially heavy concentrations in the hypothalamus and brainstem – two regions of interest in CFS. The hypocretin system regulates the roles the monaminergic (dopamine, serotonin, histamine) and cholinergic systems play in producing ‘vigilance’ or alertness.

Injecting hyporexin into the brains of rats results in increased wakefulness for hours, probably through the induction of histamine. Researchers have recently demonstrated that glutamate drives neuronal hypocretin activity. Some believe increased glutamate levels could play a role in cognitive as well as other problems found in CFS.

This seems like a strange finding though; increased hypocretin mRNA in EBV patients would lead, one would think, toincreased alertness. High hypocretin levels at night, however, could lead to insomnia and that could theoretically lead to increased daytime sleepiness. A disturbance in the hypocretin system by EBV could therefore play a role in the extreme daytime sleepiness found in CFS patients.

Fatigue (MEF2C)

Remarkably, one gene (MEF2C) accounted for almost 2/3rds of the variance associated with fatigue in these patients! A further analysis found that this gene was highly expressed in infectious mononucleosis patients with high fatigue and hardly expressed in patients with low fatigue (p<.0015). It was also highly correlated with muscle/joint and cognitive problems. This is an important gene!

So what is MEF2C? MEF2C is a transcription factor, a protein that regulates gene activity; transcription factors enter the nucleus and then either promote or inhibit the transcription of DNA into mRNA. Many CFS patients know of the STAT transcription factor that regulates the transcription of immune genes. One of the genes MEF2C activates promotes the expression of a key viral protein (BZLFI) important in the EBV replication.

This doesn’t appear to tell us much about the fatigue found in infectious mononucleosis patients, though. About half of the IM patients – those without fatigue – did not have high MEF2C levels. Since there is no evidence of increased viral replication in the more fatigued IM patients MEF2C’s role in producing fatigue during IM is something of a mystery.

MEF2C also plays a role in muscle generation. A recent paper reported that MEF2C upregulation induced the cardiac muscle cells in mice to elongate or hypertrophy, a finding which appears to implicate MEF2C overexpression in cardiomyopathy (heart disease). MEF2C also plays a critical role in the survival of the endothelial cells that line the blood vessels. There is growing evidence that vascular problems may contribute to CFS. This study groups was examining immune not endothelial or cardiac cells but as we have seen all sorts of genes are present in immune cells. Could high MEF2C expression over time contribute to the vascular or cardiac problems in CFS patients?

Neurocognitive Problems (VACHT)

VACHT is a transporter takes up acetylcholine from the synapses of the nerves and returns it for use in the neurons. Many of the acetylcholine containing neurons occur in a part of the brain called the basal ganglia which some CFS researchers believe is disrupted in CFS (See Choline on the Brain? A Guide to Choline in Chronic Fatigue Syndrome).

Could VACHT be overexpressed in order to make up for low acetylcholine levels in the basal ganglia? Acetylcholine blockage in the brain has been shown to cause memory loss.

As was stated it is unclear if any of these genes are chronically upregulated in CFS patients; none of the gene expression studies have thus far stated they are. We don’t know, however, if they have been tested yet – the gene studies only specify which genes have an altered expression, not which ones are normal.

Conclusions: the Dubbo studies and EBV

At this point we are unable to build a coherent model of post-infective fatigue out of these studies. The Lloyd study found evidence of a somewhat delayed immune response and perhaps an initially stronger pathogenic attack in the still ill IM patients but it did not find evidence of increased viral activity or an prolonged or aberrant immune response.

Lloyd has indicated the problem lies in the brain and, importantly that it occurs early in the infectious process. Could the slightly delayed immune response and/or increased pathogenic attack in the chronically ill patients give EBV an opportunity to make it to the brain? The Vernon study clearly suggested the still ill IM patients had trouble shutting down EBV replication. Is that replication occurring in the periphery or in the brain?

And what about EBV? Does it play a special role in CFS or not? Is it just one of many pathogens that can trigger CFS or does it have certain features that make more dangerous than others? Is its ability to infect central nervous system cells important or necessary? Are CFS patients at special risk from neurotropic pathogens? Several of the pathogens of interest in CFS (HHV-6, EBV, some mycoplasma’s, coxsackie B) can infect the CNS.

The whole idea of ‘sickness behavior’ i.e. the CFS-like symptoms commonly found in the acute stages of an infection (fatigue, muscle aches, poor thinking ability, fever) is that infections in the periphery cause the brain to produce symptoms via cytokines and other agents that force people to rest.

It’s not the pathogen that’s make people so sick initially, it’s the immune response. You’d don’t need, therefore, to have a CNS infection to have CNS problems. This next series of papers summarized, in fact, suggest ways in which infections in the periphery may affect brain activity.

Information on the Dubbo Studies from the CDC website suggests it doesn’t matter which pathogen is present: “Two hundred fifty-three subjects have been enrolled and followed for at least 12 months¼ Protracted and disabling PIF (post-infectious fatigue)/CFS, characterized by fatigue, musculo-skeletal pain, neurocognitive difficulties, and mood disturbance, occurred in 12% of subjects at 6 months and in 9% at 12 months. We specifically identified CFS in 11% of participants at 6 months.”

(Look at these numbers; if 10% of the subjects got CFS then we have about 25 people with CFS spread among three diseases. Three studies/three papers – 8 patients a study! This is great stuff but it looks like we are in store for more preliminary findings!)

The risk of PIF/CFS was similar for all three agents, and although each of the acute infectious diseases had unique clinical features, the PIF/CFS phenotype was uniform and independent of the initial infection. (What an interesting statement – unique clinical features but a common endpoint – CFS. These studies looked two very different kinds of pathogens; viruses and bacteria. Since the immune response to each is likely to differ this appears to suggests that something rather fundamental in the immune response or in the reaction to the immune response goes awry in CFS patients.) PIF/CFS was predicted largely by the severity of the acute illness rather than by demographic, microbiological, immunological, or psychological factors.

This ‘severity of the acute illness’ appears to refer to how debilitating it was, e.g. people who got really sick had the best chance of getting CFS. Since the amount of pathogen present was not a factor the problem appears lies in the body’s response to it. Since the immune response does not appear to be the problem either, it seems we are probably left with the body’s response to the immune response.

Since the immune system interacts closely with the brain the most likely candidate here appears to be a problem in the brain triggered by the immune system). These infections clearly can have an etiological role in triggering CFS, and it appears that host response to infection (rather than the specific pathogen) determines the occurrence of PIF/CFS.


*Li,Y., Cyriaque, J., Tomlinson, C., Yohe, M. and S. Kenney. 2004. Fatty acid synthase expression is induced by the Epstein-barr virus immediate early protein BRLFI and is required for lytic viral gene expression. Journal of Virology 78; 4197-4205

Share this!