Diagnosis of prions in patients should utilize novel strategy, team says

By Jennifer O'Brien

A technique for detecting prions in tissue, developed in recent years by UCSF scientists, is significantly more sensitive than the diagnostic procedures currently used to detect the lethal particles in samples of brain tissue from patients, according to a study performed by a UCSF team.

The finding indicates that the diagnostic technique, known as the conformation-dependent immunoassay (CDI), should be established as the standard approach for brain biopsies of patients suspected of having the disease, they say. The team is exploring whether the CDI might be adapted to detect prions in blood and muscle.

The finding suggests that reliance on the current methods for detecting prions in human brain tissue—microscopic examination of tissue for the telltale vacuoles that form in brain cells and immunohistochemistry (IHC), which involves detecting prions in brain sections using prion protein-specific antibodies—may have led to an under diagnosis of the disease in patients in recent years, they say. (A definitive diagnosis of the disease in humans is made only on autopsy, when a neuropathologist can analyze multiple brain regions for vacuoles and evidence of prions by IHC, and it is estimated that only 50 percent of human cases are autopsied, in part because many pathologists do not want to risk infection during the autopsy.)

In the study, the team compared the ability of the CDI and the two traditional diagnostic techniques to detect prions in various brain samples from 28 patients diagnosed on autopsy as having one of several human forms of the disease—sporadic, familial or iatrogenic Creutzfeldt-Jakob disease (CJD). While the CDI detected the biochemical signal for prions in 100 percent of the samples studied, the traditional tests failed to detect the prion in a high proportion of cases. For example, in an experiment that focused on 18 brain regions from eight patients with sporadic CJD, the CDI detected prions in 100 percent of the samples, while IHC detected them in 22 percent and routine tissue examination in 17 percent.

“In about 80 percent of the different brain regions examined, prions were not consistently detected by either IHC or routine histology that measure vacuolation. In contrast, the CDI was always positive in all regions of the brain,” says the lead author of the study, Jiri Safar, MD, associate adjunct professor of neurology and a member of the UCSF Institute for Neurodegenerative Diseases, which is directed by senior author Stanley B. Prusiner, MD, UCSF professor of neurology and biochemistry.

“These findings indicate that histology and immunohistochemistry should no longer be used to rule out prion disease in single-site biopsy samples,” says Safar. “The superior performance of the CDI in diagnosing prion disease suggests that the CDI be used in future diagnostic evaluations of prion disease, particularly for single-site brain biopsies during life”

“If the traditional techniques are used at autopsy, they must be applied to many cortical and subcortical samples,” says co-author Stephen J. DeArmond, MD, PhD, UCSF professor of neuropathology.

Moreover, while the study examined the efficacy of the CDI in comparison to the two techniques routinely used by neuropathologists to detect prions in human brain tissue, previous studies at UCSF indicate that the CDI is also significantly more sensitive than Western blot analysis, the technology used with IHC to detect prions in brain tissue from cattle suspected of having bovine spongiform encephalopathy (BSE). That IHC and Western blot analysis are relatively insensitive methods, the researchers say, supports their ongoing assertion that the CDI should also be used to evaluate the brain tissue of cattle.

“The studies reported here are likely to change profoundly the approach to the diagnosis of prion disease in both humans and livestock,” says Safar.

More broadly, the scientists say, the high sensitivity of the CDI suggests that CDI-like tests could also prove useful for diagnosing other neurodegenerative diseases, such as Alzheimer’s disease, Parkinsons’s disease and fronto-temporal dementias, all of which, like prion diseases, involve various forms of protein misprocessing. These diseases currently are diagnosed by neuropathological analysis and immunohistochemistry.

“Whether immunohistochemistry underestimates the incidence of one or more of these common neurodegenerative diseases is unknown, but the CDI could shed light on these diseases,” says co-author Bruce Miller, MD, UCSF A.W. and Mary Margaret Clausen Distinguished Professor of Neurology and director of the UCSF Memory and Aging Center.

The finding will be printed on-line and in print on March 1, 2005 in Proceedings of the National Academy of Sciences.

The study brings into high relief the different detection strategies of immunohistochemistry and the CDI, both of which involve revealing the presence of prions, known as PrPsc, by applying antibodies to brain tissue.

Standard immunohistochemistry, developed in the DeArmond lab 20 years ago, involves using an enzyme known as a protease, or a combination of harsh acid and high temperature treatment, to destroy normal prion protein (PrPC), which is ubiquitous in brain tissue. Once this occurs, scientists apply fluorescently lit antibodies that react with residues of the relatively resistant abnormal prion protein (PrPSc), thereby highlighting it.

The limitation of this technique is that scientists have since learned that there is a large part of the abnormal prion protein that is protease sensitive, and that portion escapes detection by the standard technique. Thus, this traditional method underestimates the level of PrPSc in tissue.

The CDI addresses this limitation by revealing the region of PrPSc that is exposed in the normal PrPC but is buried in infectious PrPSc, using high affinity, newly generated antibodies that identify PrPSc through the distinct shape of the molecule, independent of proteolytic treatments. This makes it possible to detect potentially large concentrations of protease sensitive PrPSc molecules.

Detractors would say that it is not necessary to detect the minute level of infectious agent that the CDI is capable of revealing, as it would be unlikely to be lethal, says Safar. But Prusiner and his colleagues maintain that any risk is too great when it comes to having prions in the food supply. In addition, because even low levels of prions are extremely resistant to inactivation, they may contaminate the environment for many years.

Prusiner won the 1997 Nobel Prize in Physiology or Medicine for discovering that a class of neurodegenerative diseases known as spongiform encephalopathies was caused by prions. Prion diseases develop in humans, cattle, sheep, deer, elk and mink.

The CDI was developed by members of the Prusiner lab. The CDI methodology has been licensed to InPro Biotechnology, Inc.

Prusiner, Safar, DeArmond and other members of the Institute for Neurodegenerative Diseases are scientific advisors to, or own stock in, InPro.

Other co-authors of the study were Michael D. Geschwind, Camille Deering, Svetlana Didorenko, Mamta Sattavat, Henry Sanchesz, Ana Serban, Kurt Giles, of UCSF, and Martin Vey, of Behring, Marburg, Germany, and Henry Baron, of Behring, Paris.

The study was funded by the National Institutes of Health, the John Douglas French Foundation for Alzheimer’s research, the McBean Foundation, the State of California, Alzheimer’s Disease Research Center of California and the RR00079 General Clinical Research Center.

The UCSF Institute for Neurodegenerative Diseases: http://ind.medschool.ucsf.edu/.


Explanation as to why the CDI is more sensitive than Western blot analysis: Studies at UCSF during development of the CDI showed that CDI could detect prions in brain homogenates at levels that fail to produce disease in animals (bioassay for prions).  Therefore, the CDI is more sensitive than the bioassay method, which was considered to be the most sensitive technique for detecting prions. In contrast, Western blot analysis for prions is significantly less sensitive than the bioassay and is, therefore, significantly less sensitive than the CDI. Currently, the USDA uses a combination of Western blot analysis of brainstem homogenates and immunohistochemistry of the medulla to test cattle suspected of having bovine spongiform encephalopathy (“mad cow disease”). The relative insensitivity of IHC and Western blot analysis, says DeArmond, supports the UCSF scientists’ ongoing assertion that the CDI should also be used to evaluate the brain tissue of cattle.

DeArmond cites additional evidence about Western blot analysis from a World Health Organization (WHO) study group, which compared the CDI method with Western blots for detection of prions in sporadic and variant CJD brains. Based on the smallest amount of prions that could detected by the two techniques, they found that the CDI was from 1000- to 100,000-fold more sensitive than Western blot analysis performed in six different research laboratories (Minor et al. Standards for the assay of Creutzfeldt-Jakob disease specimens. J. Gen. Virol. 85: 1777-1784, 2004).

Explanation as to why IHC for prions is less sensitive than the CDI:  IHC is routinely performed on formalin-fixed, paraffin-embedded samples of brain. Formalin fixation markedly decreases the ability of antibodies to bind to proteins in general, which greatly weakens the IHC signal for prions (PrPSc). In contrast, homogenates for the CDI are not treated with reagents that decrease prion antigenicity.  Moreover, to concentrate the PrPSc for measurement by the CDI, the homogenates are exposed to phosphotungstic acid, which selectively precipitates both protease-sensitive and protease-resistant PrPSc that comprise prions, but not the normal prion protein conformer found in uninfected animals, PrPC.  This step results in a higher concentration of PrPSc for detection by the CDI.  Because the PrPSc was not exposed to proteases, the CDI measures all forms of abnormally folded PrPSc molecules. Protease-sensitive PrPSc can account for 50 percent of the total PrPSc. For Western analysis, homogenates of brain are treated with protease to eliminate PrPC; however, this step also eliminates protease-sensitive PrPSc leaving only protease-resistant PrPSc for Western blot detection and decreasing the PrPSc signal at least in half.