|Home | About | Journals | Submit | Contact Us | Français|
Virus infections of the brain can lead to transient or permanent neurologic or psychiatric dysfunction. Some of the complexities in establishing the causal role of viruses in brain disease are explored here.
For most of the major neurological diseases afflicting us, including Alzheimer's, Parkinson's, multiple sclerosis, narcolepsy, schizophrenia, and others, we do not know the cause. We may know which neurons are involved, for instance that Parkinson's is caused by a loss of substantia nigra dopamine neurons, and narcolepsy is caused by loss of hypothalamic hypocretin neurons, but why these cells are lost remains a mystery. Viral infections of the brain may result in a clearly diagnosed fulminant encephalitis, sometimes resulting in lethal consequences. However, a growing understanding of the biology of the hundred or so viruses that infect the brain and the resultant host responses, together with improved diagnostic tools and the lessons learned from animal models, have raised the possibility of viral causation of a number of more subtle neurological conditions. Yet in contrast to the acute encephalitic presentation of viral infection of the brain, providing evidence for a causative link between viral infection and chronic neurological dysfunction remains challenging.
There are a number of underlying reasons. Much of experimental science today is based on testing hypotheses, with the expectation that if the hypothesis is correct, then the outcome is predictable. But even a single type of virus can act in an unpredictable manner in infected individuals, infecting different regions of the brain to evoke different symptoms, or causing CNS disease in only a small minority of infected individuals. Such diversity makes it difficult to provide consistent evidence in favor of viral origin hypotheses.
Robert Koch, who shared the Nobel prize in 1905, the year before the award was bestowed on Cajal and Golgi, was a brilliant German doctor who began his research in his bedroom and studied a number of potentially dangerous bacteria, including anthrax and tuberculosis. At a time when a debate still existed whether microorganisms could cause disease, Koch's seminal experiments led to a substantially clearer appreciation that microorganisms, particularly bacteria, can cause disease. Koch's postulates, the gold standard in establishing the cause of infectious disease, suggest in part that if a microorganism caused a disease, it should be found in every case of the disease, and conversely, that the microorganism should not be found in those not afflicted with that particular disease (Koch, 1890). But it is difficult to identify any virus that infects the CNS to cause brain dysfunction that behaves in a manner fully consistent with Koch's postulates. The virus that comes the closest is rabies, an enveloped RNA virus. But even with rabies, if the virus does not enter the CNS of an infected individual, then no serious symptoms will occur.
A case in point for a well-known neurotrophic virus for which a neurological symptom cannot be predicted is polio virus, a small positive-strand RNA virus. Fewer than one in a hundred non-immunized people that have a productive infection from polio virus show neurological symptoms. Despite an enormous focus on polio, we still do not understand how and why the polio virus selectively infects the motor system of these unlucky few that show neurological symptoms associated with poliomyelitis. Thanks to a world-wide immunization effort, polio has almost been eradicated, with some pockets remaining in less-affluent countries. Similarly, the unrelated West Nile virus, another RNA virus, also causes serious neurological dysfunction in fewer than one in a hundred people that become infected with the virus from a mosquito infusion. West Nile virus has become a concern in the last ten years after outbreaks were first noted in New York City; the suggested source of the current virus in the USA is from infected wild or captive birds, or an infected mosquito, arriving here from Africa or the middle east. Ironically, West Nile Virus was actually imported into New York City about 60 years ago when it was thought that it might serve in an oncolytic capacity, and a number of humans with cancer were purposefully infected with the virus in the hope of attenuating their cancer (Southam and Moore,1954). The virus was not particularly effective in cancer treatment, and its use was discontinued. That only a small number of infected individuals ever show neurological consequences with these two well-studied viruses underlines the difficult task of assigning cause for infections of the brain by the many other less-studied viruses that sporadically infect the CNS.
Infections of the brain are less common than that of other organs, and depend on rare events that allow the virus to penetrate the blood brain barrier. Most systemic viruses do not enter the brain. Those that do may take advantage of rare events that include break down of the blood brain barrier, or infection of Trojan horse-like immune cells that are competent to cross the blood brain barrier, but in doing so, subsequently release viruses within the brain. The route and site of entry may play a large role in the ultimate symptoms generated. Viruses can cause neurological problems due to a number of mechanisms including lytic effects on brain cells (cytomegalovirus), induced apoptosis (vesicular stomatitis virus, VSV), or secondary damage due to release of glutamate, DNA, and other inducers of further brain damage. Other viruses such as rabies do not kill neurons, but instead commandeer cellular transcriptional pathways to express viral rather than neuronal genes; this results in neurons that no longer function as neurons, but look normal upon routine pathological examination.
Similar to the real estate adage “location, location, location”, the same is important also for viral infections of the brain. Unlike other organs such as liver where the specific location within the liver infected by the virus does not substantively alter the symptoms, the precise region in the brain that is infected plays a key role in the type of resulting dysfunction. Limbic infections will manifest a completely different syndrome than infections of motor or sensory systems. Viruses such as cytomegalovirus, rubella, and lymphocytic choriomeningitis virus cause serious abnormalities if the developing brain is infected, and depending on the site and age of fetal infection, can generate overlapping but distinct symptoms such as deafness, blindness, epilepsy, hydrocephalus, and/or reduced IQ in a manner directly related to what part of the brain was infected. The age of the infected individual also plays a large role in some infections; different viruses cause neurologic dysfunction at different stages of life. Some viruses, for instance West Nile Virus, are more likely to cause neurological problems in the elderly. Conversely, the DNA virus, cytomegalovirus, is considered the most common infectious agent causing permanent neurologic dysfunction in the developing human brain, but presents little danger to the mature brain. Other viruses including wild type VSV show an age-dependent shift in the type of neurons infected. These factors all contribute to the difficulty in trying to demonstrate that the cause of an existing neurological syndrome may be viral in origin.
Must a single disease be associated with a single microorganism? This perspective may not hold for viral infections of the brain. Take, for example, the case of multiple sclerosis (MS). If MS is due to an autoimmune attack on oligodendrocytes, it might be initiated by a transient viral infection of those cells. While no single virus has consistently been associated with MS (Atkins et al,2000), morbilliviruses such as measles and/or other viruses that sometimes infect oligodendrocytes have been associated with the disease (Sips et al,2007). Viral diseases are particularly difficult to study and interpret in humans. But in animal models of MS, a surprisingly large number of unrelated viruses can induce an MS-like CNS disease. Viruses that can cause MS-like disease, associated with an autoimmune-mediated loss of oligodendrocytes, include measles, VSV, vaccinia, mouse hepatitis, Semliki Forest, Theiler's, Chandipura, and Venezuelan equine encephalitis virus (Johnson,1998). The animal studies give more credibility to the view that at least in some cases, MS in humans may be initiated by viral infection, and argue it is the host immune response to infection that matters for disease progression rather than the specific virus itself. In other cases, such as after lymphocytic choriomeningitis viral infection, damage to neurons may also be due to attack of the infected cells by killer T-cells of the systemic immune system.
The immune response to viruses is complex, but generally starts with the frontline innate neuronal and glial antiviral defenses including an upregulation of interferon and antiviral downstream genes including OAS/RNaseL, MxA, ISG15, and PKR within hours of an infection (van den Pol et al,2007); this is followed within days by an upregulation of acquired immunity with an increase in virus-specific B,T, natural killer cells, and macrophage/monocytes. In addition, many viruses, even those with only a handful of genes, have evolved an ability to block or subvert interferon-mediated cellular antiviral defenses. For instance, the M protein of VSV blocks nuclear pores, whereas Sindbis and Venezuelan equine encephalitis virus block interferon responses by attenuating phosphorylation of downstream STAT1 and STAT2 pathways (Yin et al,2009). Florey devoted part of his Nobel address for the discovery of penicillin to biological warfare, not the use of biological weapons by humans, but rather the idea that microorganisms are constantly battling one another for survival. In a sense, the struggle between viruses and the cells of the brain can also be viewed as a battlefield, with survival of the cells or virus dependent on complex strategies and counterstrategies that each employs.
A single mutation can dramatically alter a virus's ability to subvert cellular antiviral defenses, as in the case of the M51 mutation in VSV that not only reduces viral infection, but also increases the relative infectivity in cancer cells relative to normal cells (Stojdl et al,2003). Some viruses such as wild type VSV can mutate quickly, and exist in the wild not as a single genome, but as a population of closely related genomes, allowing the virus a survival advantage if a particular clone has an increased success rate at infecting particular types of cells, but also adding to the complexity in determining a preferred cellular target in the brain for the virus, or for determining if and which virus clone might be responsible for a neurological problem.
From a “host” perspective, our own antiviral defenses are also complex, and regulated by a wide variety of genes. Some gene alleles may enhance protection from specific viruses, others may protect from one virus, but increase susceptibility to another. Humans with CCR5delta 32, a mutated chemokine receptor, show a reduced potential for developing AIDS and consequent AIDS-related dementia from human immunodeficiency virus, whereas West Nile Virus encephalitis is more common in people with the same mutated receptor (Glass et al,2006). Several mutations in the interferon pathway affect the susceptibility for herpes simplex encephalitis; reported mutations altering herpes infections include those affecting UNC-93B, STAT1, and Toll-like receptor 3 (Arkwright and Abinun 2008; Sancho-Shimizu et al., 2007). We are only beginning to appreciate which genes might modulate viral susceptibility, but genetic differences in individuals undoubtedly contribute to the complexity of CNS responses to viral invasion.
A common treatment for a number of psychiatric diseases such as depression or anxiety disorders is the use of serotonin selective reuptake inhibitors (SSRIs) that increase extracellular serotonin. Interestingly, there are viruses that can exert the opposite action, and reduce serotonin levels, theoretically having the opposite effect of the SSRIs. Viruses such as wild type VSV can rapidly and selectively infect and destroy serotonin (and norepinephrine) neurons in young mice and rats. A subsequent immune system-mediated response eliminates all trace of the virus from the brain, leaving a permanent reduction in serotonin neurons with consequent behavioral alterations (Mohammed et al,1992) with no viral signature left behind. The neurotransmitter and behavioral changes resulting from this initial infection may last the lifetime of the organism, despite the lack of any trace of the virus in the brain, and with little detectable viral-mediated neuropathology evident by standard screening methods at the end of life.
Viral infections can also generate neurological problems years or decades after the initial infection. Chicken pox, actually not a pox virus but a double strand DNA enveloped herpes virus, can remain latent in neurons for decades, to return years later as shingles when the live virus is regenerated from its latent genome, causing protracted periods of itching or pain. The herpes simplex genome can also remain latent in peripheral neurons for long periods before the active, replication-competent virus is regenerated, and travels down the axon to once again infect the periphery. Some viruses leave a latent imprint on the brain that causes neurological dysfunction decades after the virus has been eliminated from the brain by the immune system. Postpolio syndrome is one example. Humans that have had symptoms of poliomyelitis and recovered can again show symptoms, such as weakness of affected limbs, 30–40 years later. In science and medicine, we are accustomed to a cause and effect within a relatively short time span. Consequences of viral infections that occur decades after the primary infection, often in the absence of any trace of a virus, is another serious problem is assigning cause postfacto to viral infections.
An interesting virus from a psychiatric perspective is Borna disease virus (BDV). BDV is a slowly replicating, enveloped, negative strand RNA virus that can enter the brain, infect neurons, often in the limbic system, and remain active for long periods in the CNS without generating neuronal lysis. BDV is unique in its mononegavirales order in that it replicates in the cell nucleus; other viruses of the order replicate in the cytoplasm. BDV causes behavioral disturbances in a remarkably wide variety of species, including mouse, rat, llama, ostrich, chicken, horse, cat, fox, dog, and monkey. These disturbances include falling, tremor, abnormal posture, hyperactivity or hypoactivity, increased aggression, and paralysis. In some rodent species, BDV may cause only mild symptoms or may be asymptomatic, whereas in other species such as horse, BDV can cause severe CNS symptoms often leading to death. It is difficult to imagine that a virus that can cause neurological dysfunction in so many species might not also do so in humans. Intriguingly, an increased incidence of BDV has been reported in patients in some psychiatric hospitals, where it has been associated with depression, bipolar disease, or schizophrenia. But a complication of correlating psychiatric disease symptoms with BDV (and other viruses) is that there are no standard tests for the presence of BDV infectivity. Tests widely ranging in sensitivity (including nested and non-nested RT-PCR, immunocytochemistry, and ELISA) have generated quite different sets of data on the relative occurrence and correlation of BDV with different psychiatric syndromes, making a clear determination of causation more difficult and adding fuel to the debate gauging the role of BDV in mental illness (http://www.cdc.gov/ncidod/eid/vol3no2/hatalski.htm). This confusion is further exacerbated by the fact that the diagnostic tests assay only for the presence of the virus in the blood or CSF, a step away from knowing where the virus might be within the brain. Additionally, the presence of BDV in some asymptomatic humans cannot necessarily be used to discount a BDV cause of brain disease, as illustrated by the example of the very small percent of individuals that show neurological symptoms after polio virus infections.
While the above arguments have highlighted that viruses can be potent adversaries, it is worth noting that some of these same features of viruses can also be harnessed to treat neurological disorders. Viruses are particularly good at bringing their genomes into a variety of cells. Beneficial genes can be engineered into viruses, particularly adeno-associated virus and the lentiviruses. These replication incompetent vectors can infect neurons with minimal complications, and lead to long-lasting transgene expression. Another potential role for viruses is in the treatment of brain cancer. Replication competent viruses such as VSV, myxoma, and herpes are being generated to target, infect, and kill glioblastoma (Chiocca,2008). Other viruses such as H1 have been designed to provoke an immune attack on cancer cells (Raykov et al, 2008).
In summary, the same mechanisms of infection described above that make it difficult to prove viral causation of brain disease also make it equally difficult to reject viral hypotheses for CNS dsyfunction. Viruses probably do constitute an unrecognized direct or indirect cause of a number of neurological syndromes. What percentage of these syndromes is due to viral infections, and what is due to many other potential causes remains to be determined.
I thank G.Wollmann for helpful suggestions.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.