The wish for reliable biomarkers is particularly evident in diseases clinically silent in the early stages, as seems to be the case in multiple sclerosis (MS). In fact, in a recent study published in Neurology, Gasperi et al concluded that “patients with MS are frequently not diagnosed at their first demyelinating event but often years later. Symptoms and physician encounters before MS diagnosis seem to be related to already ongoing disease rather than a prodrome.”

In the early 2000s, ‘proteomics’ was a hot scientific buzzword. After the great success of the Human Genome Project, scientists and capitalists established ambitious goals for this new field. These included the discovery of biomarkers for multiple diseases. But progress in proteomics biomarkers seems to be slower than anticipated.

Unraveling disease-specific proteomes can undoubtedly help with that. In contrast to genomes, which are usually stable, proteomes are highly dynamic and prone to physiological and environmental changes. Yet, “despite many studies that have been aimed to identify biomarkers, no protein marker has yet been approved for MS,” wrote Jafari et al in a study published in Biomarker Insights.

Continue Reading

Read more about MS diagnosis

Initially, proteomic investigations mostly focused on the detection of canonical proteins. However, the identification of only canonical proteins misrepresents the complexity of proteomes. In fact, because of RNA splicing and post-translational modification events, more than 1 million proteoforms might occur in humans. All these variants should be considered when capturing the profile of a disease.

For instance, the amino acid arginine in the myelin basic protein is converted to citrulline through a process known as citrullination, or deamination. This alteration was found in MS lesions, but not in non-MS individuals. This and other observations suggest a role for proteoforms in MS pathogenesis.  

Sources of Biomarkers in MS

According to Sen et al, 29 studies investigated the proteome of MS patient samples. Most (19) explored the cerebrospinal fluid (CSF), followed by blood samples (5). A few studies also analyzed postmortem cerebral tissue samples from the central nervous system (CNS).

“The search for potential biomarkers via proteome analysis of CSF samples might be more informative than blood, as drainage of the proteins into the CSF more directly reflects the CNS pathological status,” explained the review authors. However, CSF sampling and handling entails substantial problems when compared to blood, including potential confounding blood contamination. 

In fact, when choosing the biomarkers source, one must consider several aspects, and accessibility is one of them. Sen et al found an overlap between proteins identified in the CSF and proteins identified in blood, primarily, but also in tears and urine. Therefore, they suggest that, besides blood, urine, tears, and saliva could be good sources of potential biomarkers for MS. However, studies remain scarce.

Several studies used rodent models to investigate MS pathogenesis. However, there is no ideal animal model for MS and the selection should be based on the purpose of investigation. 

According to the analysis by Sen et al, the overlap of differentially expressed proteins was greater between MS and the experimental autoimmune encephalomyelitis (EAE) animal model (total of 15 studies), than between MS and the cuprizone animal model (total of 5 studies). But, overall, neither of the models represented the proteome especially well. 

Besides the difficulty of transposing proteomic results from animal models to human research, additional sample-related issues compromise the consistency between proteomic studies and the search for MS biomarkers:

  • Proteomic analyses do not always accompany histological findings. For instance, they have failed to replicate the reduction in myelin proteins found in MS.
  • MS is a heterogeneous disease and proteomic changes might vary with different phenotypes. This is important since more than half of the proteomic studies in MS used relapsing-remitting MS patient samples.
  • MS shares many of the identified proteomic changes with Alzheimer’s and Parkinson’s diseases.

Despite the proteome source, most proteins identified in MS proteomic studies are blood proteins, proteins involved in metabolism, structural proteins, immune system-regulating proteins, proteases and protease inhibitors, chaperones, myelin-related proteins, and axonal injury markers.

Challenges in MS Proteomic Workflows

At the time of publication, Sen et al believed that “a lack of consistent procedures in proteomic analyses and the failure of journals to demand the necessary rigor in both methods and data reporting have yielded a literature of contradictory results.”

In fact, proteome analysis is technically challenging. Early proteomic studies in MS used a top-down approach. But the bottom-up/shotgun approach quickly conquered its place, since it required less sample.

Both methods have advantages and disadvantages, but Sen et al. recommend “the use of integrative top-down over bottom-up analyses, since this is sensitive, has the highest inherent capacity to resolve intact proteoforms (ie, quantitatively addresses the inherent complexity of proteomes), yields excellent sequence coverage, and provides a high degree of consistency across technical and biological replicates.”

Therefore, both sample- and method-related issues might contribute to discrepancies between studies.

In conclusion, the potential of proteomics in MS is still immense. But the consistency of analytical approaches and methodologies must be improved.


Sen MK, Almuslehi MSM, Shortland PJ, Mahns DA, Coorssen JR. Proteomics of multiple sclerosis: inherent issues in defining the pathoetiology and identifying (early) biomarkers. Int J Mol Sci. 2021;22(14):7377. doi:10.3390/ijms22147377

Jafari A, Babajani A, Rezaei-Tavirani M. Multiple sclerosis biomarker discoveries by proteomics and metabolomics approaches. Biomark Insights. Published online May 6, 2021. doi:10.1177/11772719211013352

Gasperi C, Hapfelmeier A, Daltrozzo T, Schneider A, Donnachie E, Hemmer B. Systematic assessment of medical diagnoses preceding the first diagnosis of multiple sclerosisNeurology. 2021;96(24):e2977-e2988. doi:10.1212/WNL.0000000000012074