Deep Sequencing of Primary Progressive Multiple Sclerosis and Encephaltitis Brain

Background:   If a triggering infection persists or leaves traces of a pathogen's RNA within demyelinated brain, then these sequences should be identifiable in the afflicted tissue.  

Objectives: Pathogen RNA signatures may now be detectable by using relatively new molecular techniques called “deep sequencing.”

Methods:   Deep sequencing on 14 primary progressive MS (PPMS), 14 normal and 7 encephalitis control frozen brain specimens was performed using the Illumina HiSeq 2000. The resulting 50 bp sequences (“reads”) were compared to a non-redundant, panviral database that includes 1.2 million sequences from 2,480 viral species. Sequences that aligned to this database were considered “hits.”  The number of hits to each viral taxon divided by the total number of reads in the specimen was used to define normalized viral taxon hit rates (HRs).  Viral taxon HRs were compared between the groups using the Mann-Whitney test.

Results:   Fifty-80 million reads per brain specimen were obtained. Over representation of specific viral taxa was present in 5 of 7 encephalitis brain samples, including one specimen misdiagnosed by neuropathology.  HSV1 was identified in 3 samples and measles in 2.  These findings were confirmed by pathogen-specific PCR and Sanger sequencing of amplicons.  Sequence comparisons between normal controls (N = 14) and PPMS (N=14) specimens revealed highly significant (p < 0.001) HR differences for 6 retroviral taxa, including several human endogenous retroviruses (HERV-K, HERV-W, and ERV-9).

Conclusions:   Deep sequencing is feasible for detecting the etiology of viral encephalitis in frozen brain.  Retrovirus-like sequences are significantly increased in PPMS brain compared to normal controls, consistent with the findings of some other groups.