Candidate proteins, metabolites and transcripts in the Biomarkers for Spinal Muscular Atrophy (BforSMA) clinical study

Abstract

Background: Spinal Muscular Atrophy (SMA) is a neurodegenerative motor neuron disorder resulting from a homozygous mutation of the survival of motor neuron 1 (SMN1) gene. The gene product, SMN protein, functions in RNA biosynthesis in all tissues. In humans, a nearly identical gene, SMN2, rescues an otherwise lethal phenotype by producing a small amount of full-length SMN protein. SMN2 copy number inversely correlates with disease severity. Identifying other novel biomarkers could inform clinical trial design and identify novel therapeutic targets.

Objective: To identify novel candidate biomarkers associated with disease severity in SMA using unbiased proteomic, metabolomic and transcriptomic approaches.

Materials and methods: A cross-sectional single evaluation was performed in 108 children with genetically confirmed SMA, aged 2-12 years, manifesting a broad range of disease severity and selected to distinguish factors associated with SMA type and present functional ability independent of age. Blood and urine specimens from these and 22 age-matched healthy controls were interrogated using proteomic, metabolomic and transcriptomic discovery platforms. Analyte associations were evaluated against a primary measure of disease severity, the Modified Hammersmith Functional Motor Scale (MHFMS) and to a number of secondary clinical measures.

Results: A total of 200 candidate biomarkers correlate with MHFMS scores: 97 plasma proteins, 59 plasma metabolites (9 amino acids, 10 free fatty acids, 12 lipids and 28 GC/MS metabolites) and 44 urine metabolites. No transcripts correlated with MHFMS.

Discussion: In this cross-sectional study, "BforSMA" (Biomarkers for SMA), candidate protein and metabolite markers were identified. No transcript biomarker candidates were identified. Additional mining of this rich dataset may yield important insights into relevant SMA-related pathophysiology and biological network associations. Additional prospective studies are needed to confirm these findings, demonstrate sensitivity to change with disease progression, and assess potential impact on clinical trial design.

Trial registry: Clinicaltrials.gov NCT00756821.

Figure 1. Illustration of the top 5 markers as candidate biomarkers by age.

The natural log intensity of the protein abundance of CILP2, TNXB, COMP, ADAMTSL4 and CLEC3B are shown by age (Panels A-E) across Types. Panels A-E generally show a trend for type but not age.

Figure 2. Illustration of the top 5 markers as candidate biomarkers by MFMHS.

The natural log intensity of the protein abundance of CILP2, TNXB, COMP, ADAMTSL4 and CLEC3B are shown by MHFMS (Panels A-E) across Types. Panels A-E again show a trend for type and MHFMS. Panels F-J shows the box plot distribution by type.

Figure 3. Illustration of the top 5 markers as candidate biomarkers by Type.

The natural log intensity of the protein abundance of CILP2, TNXB, COMP, ADAMTSL4 and CLEC3B are shown (Panels A-E) by Type. Error bars are expressed as standard errors.