Large-scale serum protein biomarkers discovery associated with function and clinical milestones in Duchenne muscular dystrophy

Abstract

Duchenne muscular dystrophy (DMD) is characterized by progressive muscle wasting and weakness. Serum proteins may offer insight into disease processes and clinical decline. This observational study uses the 7 K SomaScan® assay to discover serum proteins associated with muscle function and disease milestones. In total 702 serum samples from 153 male patients, collected across two centers (2009-2022), are analyzed. Using linear mixed effects modelling, we evaluate age and corticosteroid use as covariates affecting protein levels and assess protein correlations with longitudinal clinical function. Here we show 318 aptamers (294 proteins) significantly associated with motor performance across the two sites, with most associations found with lower limb functional tests (NSAA, 10MRW, and 6MWT). Thirty-six proteins are associated with milestones including RGMA, ART3, ANTXR2, and DLK1. These proteins show promise as prognostic biomarkers, and could potentially be used for patient stratification in clinical trial design and for monitoring interventions.

Fig. 1. Associations of serum proteins with age in individuals with DMD.

a Volcano plots showing the strength and significance of the association with age in the LUMC cohort, and b in the UF cohort. For each probe (point), the -log₁₀ FDR (y-axes) and coefficient (x-axes) of age as determined by an LMEM probe expression from age are shown. The horizontal line represents the threshold for significance, FDR < 0.05. c Venn diagram showing the overlap of protein numbers associated with age in the LUMC (orange) and UF cohorts (blue). d Scatterplot of the coefficients for age shown in (a, b) in the LUMC (x-axis) versus UF (y-axis) cohorts. Dark green dots show proteins significantly associated with age in both cohorts. e–h Trajectory plots of selected probes decreasing with age such as CK-MM, MYOM3, TTN, and TNNI2, colored by site. i–l Trajectory plots of selected probes increasing with age such as CNTN3, CNDP1, LEP, and TIMP4, colored by site. (RFU relative fluorescence units, DMD Duchenne muscular dystrophy, LUMC Leiden University Medical Center, UF University of Florida, CK-MM muscle creatine kinase, MYOM3 myomesin 3, TTN, titin TNNI2, troponin I2 fast skeletal type, CNTN3 contactin 3, CNDP1 carnosine dipeptidase 1, LEP leptin, TIMP4 tissue inhibitor of metalloproteinase 4). Source data are provided as a Source Data file.

Fig. 2. Associations of serum proteins with corticosteroid use in individuals with DMD.

Volcano plots showing strength and significance of the associations between proteins and corticosteroid use in a the LUMC cohort, and b the UF cohort. For each probe (point), the −log₁₀ FDR (y-axes) and coefficient (x-axes) of CS use as determined by an LMEM probe expression from CS use and age are shown. A few probes are highlighted by arrows. The horizontal line represents the threshold for significance, FDR < 0.05. c Venn diagram showing the overlap of proteins associated with corticosteroid use in common between LUMC (orange) and UF (blue) cohorts. d Scatterplot of the coefficients for CS use shown in (a, b) in LUMC (x-axis) and UF (y-axis) cohorts. Box plots of probe expression in CS treated versus untreated patients (LUMC 405 samples, UF 79 samples) for two proteins previously reported in association with CS use (e, f) and two proteins identified in this study (g, h). For each box, the center line represents the median and the lower and upper edges represent the 25th (Q1) and 75th (Q3) percentiles respectively. Whiskers extend to the most extreme probe expression values within 1.5 × interquartile range from Q1 and Q3. i Scatterplot showing the relationship between age coefficients (x-axis) and CS coefficients (y-axis) for probes significantly associated with CS treatment. Proteins showing discordant coefficients for age and CS treatments were considered as efficacy biomarkers, while those with concordant coefficients were considered as safety biomarkers (shaded gray). Orange dots represent estimates for the LUMC cohort, while blue dots represent estimates for the UF cohort. (CS corticosteroids, DMD Duchenne muscular dystrophy, LUMC Leiden University Medical Center, UF University of Florida, RFU relative fluorescence units, MMP3 matrix metalloproteinase-3, IGLL1 immunoglobulin lambda like polypeptide 1, RGMA repulsive guidance molecule A). Source data are provided as a Source Data file.

Fig. 3. Associations of serum proteins with performance on tests of motor function in individuals with DMD.

a–d Longitudinal trajectory plots of NSAA scores, 10MRW velocity, 6MWT distance, and PUL 2.0 scores from both cohorts. Error bands represent 95% confidence intervals. e–h Scatterplots of coefficients, indicating the direction of protein association with motor function test performance (subpanels) in the LUMC (x-axes) versus UF (y-axes) cohorts. Bolded data points represent significant associations with the number of significant associations listed at the top-left in each subpanel. i UpSetR plot of significant probes for test of motor function. Bar plots at the bottom-left show the number of probes associated with each individual motor function test, colored by site. j Trajectory plots showing the relationship between ART3 levels (y-axes) and NSAA score (left panels) or 6MWT distance (right panels) by age for the LUMC cohort (top panels) and the UF cohort (bottom panels). The color code (depicted on the top-right in each panel) represents NSAA scores and 6MWT distances. (CS corticosteroids, DMD Duchenne muscular dystrophy, LUMC Leiden University Medical Center, NSAA North Star Ambulatory Assessment, PUL Performance of the Upper Limb 2.0, UF University of Florida, RFU relative fluorescence units). Source data are provided as a Source Data file.

Fig. 4. Associations of serum proteins with clinical milestones.

LoA: LUMC n = 74, UF n = 78; OHR LUMC n = 74, UF n = 76; HTM LUMC n = 74, UF n = 77. a Survival curves showing the probability (y-axis) of never achieving LoA, loss of OHR, and loss of HTM (solid/dashed/dotted lines) by age in each cohort. b UpSetR plot of probes significantly associated with each clinical milestone by the Cox model. Bar plots at the bottom-left show the number of probes associated with each individual milestones, color coded by site. c Forest plot showing lnHR (x-axes) of protein probes (y-axes) significantly associated (FDR < 0.05) with at least one clinical milestone. Points represent the estimated lnHR, error bars represent the 95% confidence intervals. Each panel shows the proteins associated with an individual milestone. d Survival curves for the 3 clinical milestones (columns) stratified by RGMA expression quartiles at first visit (colors), by years since first visit (x-axes). The first row represents curves for LUMC, the second row represents curves for UF. (HTM hand-to-mouth, LoA loss of ambulation, OHR overhead reach, UF University of Florida, LUMC Leiden University Medical Center). Source data are provided as a Source Data file.