This is Part 2 of a three-part series. See also Part 1 and Part 3. Read the entire series.
11 July 2011. When neuroanatomists Heiko Braak and Kelly Del Tredici proposed a new theory on the origin and progression of Parkinson’s disease (PD)—suggesting that it starts outside the central nervous system, induced by a virus or other pathogen, and then spreads to different areas of the brain in stages—they started a debate in the field.
The husband-and-wife team based their hypothesis on the locations throughout the nervous system of Lewy bodies—large protein aggregates inside nerve cells that consist primarily of the protein α-synuclein and are the pathologic mark of PD. Every aspect of their proposal, from the notion that disease pathology shows a predictable pattern of distribution, to the suggestion that the pathology starts in the gut and not the substantia nigra, to the idea that the pathology might spread across synapses from one neuron to the next, has found supporters and critics (see Part 1 of this series). “I am a major believer in Braak’s hypothesis,” said Richard Smeyne of St. Jude Children’s Hospital in Memphis, Tennessee. “But the field is definitely divided—half the people are saying yes and the other half is yelling no.”
The discovery of Lewy bodies in the intestine of patients led neuroanatomists Heiko Braak and Kelly Del Tredici to propose that the disease might start there. Image credit: H. Braak and K. Del Tredici
Not everything about this work rubs critics the wrong way, however. “There is good consensus in the field that if you take a patient who during life was diagnosed to have Parkinson’s and look at the brain postmortem, that patient will show not just the conventional loss of substantia nigra neurons, but also the presence of Lewy bodies in areas like the brainstem, dorsal motor nucleus of the vagus nerve, the forebrain, and so on,” said Robert Burke at Columbia University, New York City. “That simple scenario has a good consensus, but Braak’s hypothesis went far beyond that.”
Picking the Issues Apart
For one thing, Braak and colleagues proposed that the pathology associated with PD advances systematically through the nervous system in six stages, sequentially moving from the vagus nerve up the brainstem to the substantia nigra in the midbrain and eventually reaching the forebrain and cerebral cortical areas. They based this staging scheme on the assumption that disease pathology would not occur in an area of lower vulnerability without also being present in areas of higher vulnerability.
But since the hypothesis was first proposed in 2003, several groups have looked at the distribution of Lewy bodies in the nervous system of autopsied patients and found some that do not follow this scheme. “We looked not only at the presence and distribution [of α-synuclein], but also the intensity of deposition in cases coming to a Parkinson's Disease Brain Bank,” wrote Ronald Pearce of Imperial College London, U.K., in an e-mail to ARF. “Several cases did not have vagal nucleus pathology, and some had spinal cord pathology and no vagal nucleus pathology, thus 'skipping' the vagus nucleus” (Kalaitzakis et al., 2008). A review of the literature by the respected Austrian neuropathologist Kurt Jellinger revealed “sparing of medullary nuclei in 7-8.3 percent of clinically manifested PD cases with [α-synuclein] inclusions in midbrain and cortex corresponding to Braak Stages 4 and 5, whereas mild parkinsonian symptoms were already observed in Stages 2 and 3” (Jellinger, 2008).
But the fact that there are exceptions does not necessarily diminish the value of Braak’s staging system. “It is a useful construct that could apply to many but not all cases of disease,” said John Trojanowski at the University of Pennsylvania in Philadelphia. “I appreciate the staging scheme. It is not 100 percent absolute, but it is a good generalization.”
According to Braak, the staging system is reproducible in about 80-90 percent of Parkinson’s patients. Some of the discrepancies may also reflect variable protocols for assessing and diagnosing tissues. In his studies, Braak examined tissue samples that were 100 microns thick. That is much thicker than the samples most pathologists use, which typically span 10 or fewer microns. “If you see pathology in an area affected in Stage 5 but do not see it in regions corresponding to Stages 3 or 4, it might just mean that it’s not in the 10-micron section you examined,” said John Duda, at the Philadelphia VA Medical Center. “You would need to examine 10 of those 10 micron-thick sections to replicate what Braak did.” Even so, “very few Parkinson’s patients completely violate [Braak’s] staging,” Duda added. “It’s a very good scheme, if you accept the notion that it is biology and that there are always exceptions.”
Another issue complicating the staging system is the connection between Lewy body pathology and the clinical symptoms of PD. Braak had postulated that the early Stages 1 to 3 of disease develop in patients who have not yet been diagnosed with classical motor symptoms of disease. Those motor symptoms, according to Braak’s scheme, start to develop in Stage 3, when pathology has spread to the substantia nigra. Cognitive problems and dementia would occur when the neocortex becomes affected in Stages 5 and 6 (see Part 1).
However, a number of studies show widespread pathology in people who were not diagnosed with any clinical symptoms of disease. Irina Alafuzoff of Uppsala University Hospital in Sweden and colleagues examined brain tissues from 226 autopsied patients who had signs of Lewy body pathology in the brain. They found that only 50 percent of patients with a widespread distribution of Lewy bodies, corresponding to Braak Stages 5 or 6, had been diagnosed with PD or were demented (Parkkinen et al., 2008). The results raise two different questions, according to Alafuzoff: “1) Is this pathology significant regarding the symptoms, or 2) do those with symptoms have some other alteration that makes their brains sensitive to the pathology?” she wrote in an e-mail to ARF.
Braak based his staging system on the detection of Lewy bodies. However, whereas smaller aggregates of α-synuclein have been shown to cause neuronal cells to die, a growing number of scientists now believe the formation of the larger Lewy bodies may actually serve to protect neurons from damage. “Lewy bodies are the pathological hallmark used in diagnosis, but they appear not to be associated with cell loss or to correlate with the severity of clinical symptoms,” says Walter Schultz-Schaeffer, a neuropathologist at the University Medical Center in Göttingen, Germany.
Thus, a number of researchers have suggested that other markers of disease, such as cell death or synaptic dysfunction, would be more appropriate for developing a staging scheme for PD. Schultz-Schaeffer is examining the relevance of smaller deposits of α-synuclein found at the synapse of neurons as markers of disease progression (Schulz-Schaeffer, 2010). “Braak’s staging is reproducible,” says Schulz-Schaeffer. “But now we should put more focus on the synapses.” In Alzheimer’s disease, by comparison, the development of fluid and imaging biomarkers have transformed the focus of disease staging from postmortem biology to preclinical measures in living people (Jack et al., 2010).
Time Course on Trial
One of the more provocative claims of Braak’s proposal is that, based on the pathology detected in postmortem tissues, the disease starts far away from the substantia nigra. Critics say the basic premise for this conclusion is flawed. “It is a basic error to take single snapshots of different patients and put them all together in a timeline,” argued Burke. “You can’t say ‘Based on what I see, I would propose that the disease begins here.’”
Others, however, counter that this limitation is inevitable when trying to describe a dynamic process based on such samples. “Unless we find a suitable animal model of disease or figure out how to bring people back to life, we will never be able to prove it,” said Smeyne. None of the available animal models of PD accurately mimic the disease in humans (see ARF AD/PD 2011 story), and no radioactive tracers exist that can detect protein aggregates in living patients by PET scans or other methods. Such tests are in clinical development for Alzheimer’s disease (see ARF related news story), and there is significant interest in doing the same for PD.
The two sites where PD pathology begins, Braak claimed, are the enteric nervous system (ENS) of the gut and the olfactory bulb (see image above). The ENS is particularly intriguing because it is relatively straightforward to take biopsies of this portion of the nervous system in living patients, unlike the brain or the olfactory bulb. Thus, pathology in the ENS could serve as a marker of early disease or disease progression.
Several investigators are starting to examine Lewy body pathology in the ENS of living PD patients and healthy volunteers. Pascal Derkinderen’s group at the Inserm U913 of the University of Nantes, France, examined colonic biopsies from 29 PD patients and 10 controls. They found Lewy pathology in one of the two layers of the gut (the submucosal plexus) in 21 of the patients and none of the controls (Lebouvier et al., 2010). “We started this work based on the Braak hypothesis,” said Michel Neunlist, a coauthor on the paper. “The first thing we wanted to do was to see whether we could detect the lesions, and we showed that the ENS bears the same lesions as the brain. We now have some interesting observations to suggest that we can anticipate over time the stage of the disease based on ENS examination.” Other groups, including those of Jeffrey Kordower at Rush University Medical Center in Chicago and of Patrik Brundin at Lund University in Sweden are conducting similar studies focusing on patients with early symptoms of disease, but those results have not yet been published.
Results from these studies should yield important information about the value of ENS pathology as a possible biomarker for PD, but whether or not the disease truly starts in the gut, as proposed by Braak, is still fodder for debate. “There are all kinds of studies now showing that the enteric nervous system is extremely sensitive to environmental insults,” said Smeyne. Epidemiological research suggests that in human patients, constipation might be one of the early signs of PD, preceding motor symptoms by decades (e.g., Abbott et al., 2001; Abbott et al., 2003). Animal studies provide some support for this conclusion as well. In collaboration with Smeyne, Robert Nussbaum at the University of California in San Francisco developed a transgenic mouse expressing a mutated version of the human α-synuclein gene. This strain develops α-synuclein aggregation in the gut at three months of age. The mice have signs of constipation and reduced defecation similar to what is seen in PD patients. Because this mutation causes some familial forms of PD, the results provide a link between PD and a disease process in the gut. “This mouse model mimics what we see in the early stages of PD,” says Smeyne. In the model, the pathology does not, however, progress to the central nervous system (Kuo et al., 2010).
On the other hand, Thomas Beach and colleagues at the Sun Health Research Institute in Arizona examined autopsy tissue from 92 people obtained through the Brain and Body Donation Program, a longitudinal study of elderly volunteers who are neurologically normal or have diseases like PD and Alzheimer's (AD). These researchers found no evidence that PD starts in the gut. Beach and colleagues looked for aggregates of phosphorylated α-synuclein, a form of α-synuclein typically found in Lewy bodies, throughout the volunteers’ bodies. They found that such aggregates are widely distributed, affecting virtually all organ systems and tissues. In no case, however, did phosphorylated α-synuclein aggregates appear in the spinal cord or any peripheral site without also being present in the brain (Beach et al., 2010). “I think we showed quite conclusively that the histopathology in the body does not generally occur prior to that in the brain,” wrote Beach in an e-mail to ARF.
A Viral Spread?
Perhaps the most speculative aspect of Braak’s hypothesis is that PD is caused by a pathogen, possibly a virus or another agent that spreads through the nervous system in a prion-like manner. Part 3 in this series discusses accumulating evidence showing that this idea is not entirely far fetched.—Laura Bonetta.
This is Part 2 of a three-part series. See also Part 1 and Part 3. Read the entire series.
Laura Bonetta is a freelance writer in Garrett Park, Maryland.