Undiagnosable

UCSF Innovation Ventures Podcast Series
Over the last decade, genetic sequencing advances have exceeded Moore’s Law, and a biological revolution is underway. Dr. Joe DeRisi and Dr. Michael Wilson, are taking advantage of ultra-deep sequencing to uncover disease etiologies in previously undiagnosable neurological conditions. 6 min.
Or read the text...

 

“Over the last five to six years now, we've been incredibly successful at deploying metagenomic sequencing for infectious disease and neurological conditions…we can resolve cases that previously had no diagnosis, and in some cases provide life-saving therapy, where there was none before,” said Dr. Joe DeRisi, a Professor of Biochemistry and Biophysics at UCSF, and Co-President of the Chan Zuckerberg Biohub. Dr. DeRisi, in collaboration with DrMichael Wilson, an Associate Professor of Neurology at UCSF, have developed technologies that are now allowing scientists and clinicians to uncover disease etiologies in previously undiagnosable neurological conditions.

Diseases of the central nervous system are often harder to diagnose as they are the least accessible; brain biopsies are only conducted in extreme circumstances and a limited amount of biological material is obtained from a spinal tap. Driven to develop less invasive diagnostic tests, they have been applying genomic sequencing technologies to query cerebral spinal fluid (CSF) for evidence of infection or autoimmunity. As Dr. Wilson said, “there has been a big focus to mine the spinal fluid and get as much information from [the CSF] in a single shot.”

With their colleagues at the UCSF Center for Next-Gen Precision Diagnostics, they employed metagenomic next-generation sequencing (mNGS) to diagnose neurological infectious diseases, which resulted in the first and, to date, only CLIA-certified test for patients with idiopathic meningitis/encephalitis of unknown etiology1. Through this work they also noticed that a number of the meningitis/encephalitis patients did not have any indication of infectious disease and that some of these patients would get better with immune suppression, suggesting that for those patients, it wasn’t a primary infection, but rather an autoimmune process causing their disease. Thus, Drs. DeRisi and Wilson together with Dr. Sam Pleasure, also in Neurology, expanded their efforts to develop high-throughput ways for identifying autoantibodies to assist in diagnosing autoimmune causes of neurological illnesses.

“We can't really predict what people are going to be autoimmune to…[so] we need to cast the largest net possible to see what autoantibodies people are making,” said Dr. DeRisi. To detect all potential autoantibodies, they have iterated upon a technique pioneered in the laboratory of Stephen Elledge at Harvard Medical School – Phage Immunoprecipitation Sequencing (PhIP Seq). Dr. DeRisi and Dr. Wilson have expanded and customized a programmable phage display library to express basically every human protein and its isoforms and splice forms allowing for the simultaneous screen of more than 700,000 autoantibody targets. Utilizing their library, they recently discovered the underlying cause of testicular cancer-associated paraneoplastic encephalitis2.

When asked what’s next, Dr. DeRisi said, “proteins that aren't human are also on deck. We have a large project looking at all the viral proteins…and how the immune system develops antibodies to those and what they may tell us.” In an upcoming publication, they describe the discovery of antibodies specific to non-polio enteroviruses (EVs) in the CSF of patients with pediatric acute flaccid myelitis (AFM); EVs had been implicated in the disease etiology, but their role had been difficult to confirm as EV RNA is rarely detected in the (CSF)3. Drs. DeRisi and Wilson hypothesize that identifying evidence of viral infection, such as viral-specific antibodies, may also assist in revealing the etiology of CNS inflammation observed in other prominent neurological conditions like multiple sclerosis and psychiatric disorders. Furthermore, UCSF also has one of the largest collections of CSF from neuroinflammatory disease, which they believe will be a useful reference biobank. As Dr. Wilson said, “I think having this rich dataset of what inflammation really looks like will provide a great control dataset for investigating… inflammatory contributions to neurodegenerative disease and other neurologic diseases that have traditionally been thought to be non-inflammatory.”

By working across traditional silos, Drs. DeRisi and Dr. Wilson truly embody the cross-disciplinary work being conducted at UCSF. Dr. DeRisi concluded our interview by stating, “our general approach is to realize that patients with extraordinary phenotypes often times have amazing and exciting underlying biology… and we have the tools and technologies to discover what underlies their conditions. UCSF blends both clinical and basic sciences together in a perfect sort of virtuous cycle.”

...

1. Wilson MR, Sample HA, Zorn KC, et al. Clinical Metagenomic Sequencing for Diagnosis of Meningitis and Encephalitis. N Engl J Med. 2019;380(24):2327–40. PMID: 31189036

2. Mandel-Brehm C, Dubey D, Kryzer TJ, et al. Kelch-like protein 11 antibodies in seminoma-associated paraneoplastic encephalitis. N Engl J Med 2019;381:47-54. PMID: 31269365.

3. Schubert RD, et al. 2019. Serological and metagenomic interrogation of cerebrospinal fluid implicates enteroviruses in pediatric acute flaccid myelitis. bioRxiv doi: 10.1101/666230