Chapter Ten: From Laboratory to Patient Care By Kenny De Meirleir, Daniel L. Peterson, Pascale De Becker, and Patrick Englebienne
The last half of this chapter deals with the proper CFS patient work up – something so extensive that one wonders how anyone could afford it. The first half of this chapter, however, not only provides the most complete summary of this texts findings to date but adds new information and speculation on RNase L dysfunction in other diseases, and on the RNA fragments believed to jumpstart the process.
According to the case definition promulgated by the CDC, CFS is an illness defined by a set of symptoms. As such a definition of CFS can only be made on the basis of exclusion. (A patient has CFS only if all the other possible causes for the CFS-like symptoms can be excluded.)
It is not surprising given the wide variety of symptoms associated with CFS, and their occurrence in other diseases, including depression, that the organic basis of CFS has been questioned. A psychological basis for the disease has, however, been repeatedly challenged by reports that detail the physiological anomalies found in the disease. Despite this evidence, however, the etiology and pathogensis of CFS has remained obscure.
The discovery of abnormalities in the IFN/2-5A pathways in the mid 1990’s prompted investigations into the molecular basis of the disease. As detailed in this book, these investigations have not only produced the first biomarker for CFS, but have also ‘unraveled the etiology, pathogenesis and evolution of the syndrome’. These findings have major therapeutic implications.
Etiology, pathogenesis and evolution of CFS: a laboratory perspective
Upon infection of the cell ss or dsRNA induces the production of type I IFN’s and 2-5OAS. IFN then induces 2-5OAS, PKR and RNase L activity. (Despite a report by Vodjani and ?, that chemical toxins may activate the disregulated IFN/RNase L pathway, the authors indicate that intracellular infection is probably the initial insult.)
After polymers produced by 2-5OAS activate RNase L, RNase L begins to fragment mRNA and upregulates an apoptotic pathway involving caspases. A second apoptotic pathway induced via PKR activation acts by blocking mRNA translation (via EIF-2a), and by activating transcription (NF-kB) and regulatory factors (IRF-1).
RNase L, 2-5OAS and PKR are all upregulated in CFS. Levels of 2-5A, the adenylate oligomer produced by 2-5OAS that binds with RNase L, are high as well. The evidence suggests, however, that an abnormal form of 2-5A that is produced in CFS patients, is long enough to bind with RNase L, but is too short to induce it to dimerize (and become activated.
RNase L remains in a monomeric form that is susceptible to cleavage by proteolytic enzymes. Once apoptotic and/or inflammatory activity occurs, the native RNase L is broken up and the 37-kDa fragment is generated.
(The central question then is why does 2-5OAS produce 2-5A oligomers (dimers) that are not only too short to induce RNase L to dimerize but actually inhibit RNase L activity and leave it open to fragmentation?)
The authors report that 2-5OAS induction by type I IFN’s is a ‘very sensitive, tightly regulated process’. It requires protein kinase C (a kinase we have not come across before) and multiple regulatory elements that are either activated or repressed by interferon regulatory factors 1 or 2. 2-5OAS only produces oligomers long enough to induce RNase L dimerization if the pieces of ss or ds RNA are ‘aptamers’ ‘with little secondary structure’ or if the dsRNA is over 25 bp’s long.
That the 2-5OAS system works properly in CFS patients undergoing Ampligen therapy suggests that the activators not the system are impaired in CFS. (Ampligen is an engineered dsRNA able to reregulate the 2-5OAS system in some CFS patients).
The different viruses found in the CFS patient population could provide the triggers for the abnormal 2-5OAS response. (This very interesting statement broadens the search for the cause of the poor 2-5OAS inducers. Any of the viruses found in CFS could do it!
It begs the question, however, why viral attack would produce small or oligo deficient RNA in a person with CFS and not in one without it. This appears to push us to look ever more closely at the viral response. Just how does dsRNA appear in the cytosol?
Just how are the viral metabolites that turn on the IFN system created? Is there a problem with phagocytic processing or antigen processing in monocytes/macrophages?) The authors suggest that endogenous retroviral sequences that are part of our genome or short interspersed elements that contain ‘ALU’ sequences, could also be poor 2-5OAS inducers.
A ‘large body of evidence’ indicates that states of cellular stress, caused by toxins or ionizing radiation, can result in the transcription or expression of heretofore untranscribed or expressed elements in our genome called HERV’s or SINE’s.
The authors suggest that HERV or SINE elements may also improperly induce the 2-5OAS system to produce dimers instead of trimers. (HERV’s or endogenous retroviral sequences are sequences of retroviral genes that have found their way, probably in the very distant past, into our genome.
They are surprisingly enough, quite abundant, and make up about one percent of our genome. In most people, most of the time, they lie dormant and unexpressed. One HERV family (HERV k), however, retains the genetic sequences that express the proteins that enable viruses to escape from and enter into cells.
This ability – to successfully negotiate pathways through cellular membranes – gives HERV sequences the ability to leave one cell and infect another. Under conditions of cellular stress they can become activated. (Anti bodies to HERV k sequences have been found in stressful conditions such as pregnancy and cancer.)
(Repetitive short interspersed elements (SINES) are short genetic sequences derived from genes encoding for RNA. Because they possess the elements needed for retrotransposition they have been able to move about and insert themselves throughout our genome.
Usually expressed in low levels, SINE expression is increased greatly (up to 50x’s) durring cellular stress from viral infection, etc. Long thought to be the consequence of cellular stress, SINES are now believed to be integral components of the cellular stress response.)
These ‘hidden’ genetic elements may play a role in a host of autoimmune or malignant diseases. ALU sequences have been retrieved from Gulf War syndrome patients. Both HERV sequences and antibodies have been found in multiple sclerosis and are gaining validity as a possible etiological agent. Endogenous HLTV-II retroviral sequences were found in one study of CFS patients but not in another.
The authors postulate that the increased presence or activity of the 37-kDa fragment leads to the blockade of the apoptotic pathway found at the highest levels of RNase L fragmentation. Interestingly, the authors suggest that induction of the 2-5OAS (TRIP) proteins, which may interfere with thyroid hormone expression, may be responsible for the extreme fatigue found in CFS.
PKR is induced by the same polynucleotides that activate 2-5OAS. That PKR activity can either be suppressed or activated by these bits of RNA sets up a situation where a cell could be subject simultaneously to signals for both low and high apoptotic activity from PKR and RNase L respectively.
(Rather stunningly, the authors note that) 2-5OAS disregulation and RNase L fragmentation was found in MS patients who were having flareups. (RNase L fragmentation is not unique to CFS! What a shame! In fact we will see that CFS may be just one of several disorders which have RNase L dysfunction at their core.) Why then do MS patients not have CFS? The authors suggest that a HERV induced RNase L dysfunction lies at the core of several diseases. The direction a disease takes depends on how RNase L or RKR’s response to the HERV’s released in each patient. Interestingly, while HERV’s, SINE’S and ALU’S appear to induce the 2-5OAS pathway in the same way they can all have differing effects on PKR. The direction each disease takes then, differs depending on how the PKR pathway responds to the HERV’s or SINE’s that disregulated RNase L. The authors posit that improper 2-5OAS induction combined with an upregulated PKR leads to chronic fatigue syndrome; improper 2-5OAS induction and a down regulated PKR leads to multiple sclerosis (MS). How this might happen is described below. (This suggests the difference between CFS and MS may lie in the kind of HERV’s or SINE’s released in each case. In CFS the HERV’s cause a dysregulated RNase L and PKR; in MS an dysregulated CFS and downregulated PKR.)
Chronic fatigue syndrome:
In CFS an upregulated PKR activates the nuclear transcription factor NF-kB which can induce the production of a broad of inflammatory enzymes including nitric oxide synthetase (iNOS) and cyclooxygenase (COX2). NO regulates a wide variety of processes including neurotransmission, smooth muscle contraction, platelet activation and immune system ‘cytotoxicity’. (There is evidence that each of these could be dysregulated in CFS.)
NO also interacts with calcium channels to stop the contraction of skeletal and cardiac muscles and it enhances COX II activity.
If NO combines with the free radicals produced during phagocytosis, peroxynitrite, a very potent free radica that is particularly damaging to natural killer (NK) cells and lymphotoxic activated killer (LAK) cells can be formed. Oxidative damage, perhaps not surprisingly, is positively correlated with symptom levels in CFS. (Both NK cell function and 2-5OAS dysregulation, interestingly enough, ‘normalized’ during Ampligen treatment in breast cancer patients. HERV involvement is suspected in breast cancer as well.)
PKR upregulation contributes to the Th2 shift by prompting IgE switching (more allergic antibodies) and by inhibiting NK and LAK cell activity as well.
(Chronic immune activation can effect hypothalamic activity.) When activated, immune cells produce glutamate, a substance which ‘excites’ neurons via the NMDA (N-methyl-O-aspartate) receptors. NO is particularly effective at regulating glutamergic excitation in the hypothalamus. (NO also down regulates CRH (corticopin releasing hormone) in the hypothalamus.) (I believe the authors are saying that a scenario exists in CFS where excited neurons in the brain work at cross purposes with down regulated neurons in the HPA axis. No wonder we’re screwed up.) This would lead to the dysregulations in the central nervous system and HPA axis seen in CFS.
The simultaneous production of NO and COX 2 in CFS can cause dramatic increases in prostaglandin/prostacyclin (vasoconstrictor/relaxant) ratio which could result in two conditions commonly found in CFS; vasoconstriction and hypercoagulation.
(Thus PKR upregulation in CFS patients could lead to (a) a TH1/TH2 imbalance that leaves the door open for intracellular invasion, (b) a dysfunctional HPA axis that causes CNS and endocrinological problems and (c) problems with blood pressure (orthostatic hypotension) and hypercoagulation problems.)
On the other hand, a dysregulated 2-5OAS/RNase L system and a de-activated PKR system could lead to Th1 dominance because of reduced NO levels. In this scenario a Th1 dominant immune system results in upregulated self-reactive T-cell activity that degrades the myelin found in the nervous system, resulting in MS.
Nitric oxide synthetase is deleted in monocytes only in acute cases of MS. Some attempts to control MS by reducing NO and other reactive nitrogen species have exacerbated the disease.
The authors suggest that CFS and chronic MS may, in effect, occupy the opposite ends of a gradient of PKR/2-5A disregulation that encompasses a variety of auto-immune diseases including lupus, Type I diabetes and acute MS.
PKR activation is not dependent on dsRNA. PACT, a PKR activating protein, appears to be activated by ceramide, a molecule that induces apoptosis in response to all stressful stimuli (including free radicals). Free radical production in activated phagocytes and macrophages in CFS could contribute, via PACT activation, to the PKR activation found in CFS.
(Free radical production, as explained earlier, accompanies phagocyte activation and is an essential component of phagocytosis. Chronic NO production as a result of upregulated PKR activity could deplete the cells ability to degrade free radicals.)
PKR activation induces apoptosis through eIF-2a inhibition and caspase 8 activation via the Fas associated death domain (FADD). The increased caspase 8 (and 2) activity seen as RNase L fragmentation progresses and the 37-kDa fragment increases in abundance, is paralleled by a rise in PKR activity.
PKR activation is further enhanced by calcium releases from the endoplasmic reticulum that occur during the apoptotic process. Rises in intracellular calcium activate calpain which in turn activates caspase 12, inhibits caspase 9, and fragments both the calpain inhibitor (calpastatin) and the monomeric RNaseenzyme. Thus a scenario can be formed that seats RNase L fragmentation in PKR activation.
Once calpastatin is inactivated the caspases are able to freely upregulate calpain activity. Caught in this soup of proteases, a host of other proteins (G-actin, STAT I, RLI and p53) are fragmented as well; they all contribute to the apoptotic blockage seen at the higher levels of the 37-kDa fragment.
RLI contributes indirectly by its inability to inhibit activity of the RNase L fragment. P53’s pro-apopototic activities cease once it is fragmented. P53 fragmentation is of particular interest because it may explain the increased cancer rates (particularly lymphomas) that may occur in some CFS patients. STAT 1 fragmentation inhibits apoptosis by blocking apoptotic signals produced by type I IFN’s (including those produced by RNase L and PKR).
(This very interesting point suggests that the 2-5OAS and PKR upregulation found in CFS is not, at least at some point in the disease, the result of IFN activity. IFN starts the process but because STAT I – which relays the IFN signal to the nucleus – appears to be degraded in CFS, the message apparently does not get hrough to 2-5OAS/PKR. (The primacy in the signaling pathways has recently been questioned, however,- see RNase L background).
(If I am reading this very difficult section correctly – and I very well may not be – it appears that upon infection the cell is given a signal to begin apoptosis by PKR and possibly RNase L. The rise of intracellular calcium levels caused through the apoptotic process activates calpain which fragments RNase L. Somehow, however, either accumulations of the RNase L fragment or calpains fragmentation of some pro-apoptotic proteins (p53, stat I) stops the apoptotic process. The pathogen ridden and severely damaged monocyte/macrophage cell survives (???). The pathogen spreads and a cycle of immune dysfunction is initiated.)
G-actin cleavage causes, through a disruption in antigen presentation and T-cell activation, a Th2 dominated immune system. Finally the ankyrin fragment released during cleavage of the native RNase L may cause channelopathies through its interaction with the ABC transporters.
The alternating phases of up and down apoptotic activity leave the system open in infections and/or viral reactivations. The chronic inflammation produced by this constant immune activity results in the production of elastase, a protease that further fragments RNase L.
Opportunistic infections, in particular mycoplasmas, may further contribute to the increased proteolytic activity found. ‘Cellular trash’ may build up as a result of RNase L’s fragmentation of a gene, ISG-15, which codes for a protein that targets proteins for degradation. STAT fragmentation not only blocks the induction of 2-5OAS and PKR but prevents IFN-y from inducing T-cell cytotoxicity. (Another ‘hole’ in the immune system defenses appears.)
These results indicate that a ‘vicious innate immune dysfunction induced by external stress stimuli and environmental factors’ occurs in CFS.