Since we still don't know the causes of ALS and several other sporadic neurodegenerative diseases, it's interesting to find potential biomarkers associated with these diseases. Furthermore, since most clinical trials are negative, the pharmaceutical industry is seeking to change the definition of a successful trial by substituting biomarkers for clinical signs, as they are said to be more reliable. While there is some truth in this approach, it still seems highly questionable from an ethical perspective: The sole purpose is to validate clinical trials, even if there is no improvement in patient symptoms.
Researchers are therefore currently working to find biomarkers for neurodegenerative diseases, as the market for these tools appears immense and lucrative. The ideal biomarker would be an inexpensive blood test.
Researchers from Thomas Jefferson University examined two American GEO databases that collect blood samples (plasma and serum). https://www.ncbi.nlm.nih.gov/geo/info/overview.html
The Gene Expression Omnibus (GEO) is a public repository that archives and freely distributes comprehensive microarray, next-generation sequencing, and other forms of high-throughput functional genomics data submitted by the scientific community. These are digital data provided by microbiology tools and are therefore presumed to be reliable, but the associated metadata (provided by humans) may not be of high quality. Furthermore, some datasets submitted to GEO may have been contaminated.
Scientists focused on the small RNA fragments present in these samples. They focused on small non-coding RNAs because these molecules are stable and abundant in blood fluids such as plasma and serum.
These molecules were classified as follows:
isomiR: slightly altered versions of microRNAs
tRF: fragments from transfer RNA
rRF: fragments of ribosomal RNA
yRF: fragments of another type of RNA, Y RNAs
And a residual group, called "not-itrs," for sequences they initially could not categorize.
The scientists found that these small types of RNA do not appear in sufficient quantities in ALS patients as in healthy people.
Some of these differences were related to the patients' survival time, even after taking into account factors such as age, sex, and whether or not they were taking riluzole (a common treatment for ALS).
Interestingly, some "non-itrs" sequences did not match human DNA, but rather the ribosomal DNA of bacteria (Burkholderiales) or fungi. Some of these foreign sequences were also linked to patient survival. This is a worrying claim.
What should we make of these claims of non-human RNA discovery in patients? Initial contamination is a plausible explanation for the detection of small non-human RNAs (sncRNAs), and it is a known concern in studies of low-input samples such as plasma and serum. But this contamination would also be apparent in samples from people without ALS.
The tools used in microbiology use short reads that are algorithmically reassembled. These short reads are likely to match multiple genomes by chance, increasing the risk of false positives during alignment, particularly if the databases are large and noisy. Here too, contamination would be apparent in samples from people without ALS.
Another explanation is that the presence of non-human genomes in humans is completely normal: our skin, mucous membranes, and internal organs harbor an extensive variety of microorganisms. We don't live in a vacuum. What we do know is that these populations of microorganisms respond to the host's health, sometimes with significant variations. For example, it has been shown that in Alzheimer's disease and, more generally, in aging, the dental microbial population is very different from that of healthy people.
So what can we conclude? There's probably no reason to worry; ALS is probably not caused by specific microbes or microscopic fungi. But that doesn't change the fact that we know that certain cyanobacteria cause a disease similar to ALS.