Identification of KHDC1L, a DUX4-regulated protein, as a novel plasma biomarker in facioscapulohumeral muscular dystrophy

Figures

• 

  Figure 1. DUX4 induces expression of KHDC1L and has a ChIP-seq peak in the promoter region. Presentation of prior RNA sequencing (GSE85461) and DUX4 ChIP-seq (GSE33838) data showing the KHDC1L and KHDC1 gene regions. MB135 cells do not have detectable expression of DUX4 (RNA-seq no DUX4) and the expression of DUX4 robustly induces KHDC1L RNA (RNA-seq + DUX4). KHDC1 is not expressed in MB135 cells either with or without DUX4. A DUX4 ChIP-seq peak is present at the promoter region of KHDC1L, indicating DUX4 binding at that region.

• 

  Figure 2. (A) KHDC1L is released from MB135 myoblasts. MB135 cells transduced with a doxycycline-inducible 3xFLAG-tagged KHDC1L showed release of the tagged KHDC1L protein in the concentrated supernatant following doxycycline induction. Actin was present with or without DUX4 induction and is associated with exosomes but we did not determine whether the actin signal represented exosomes. (B) DUX4 induction of KHDC1L mRNA correlates with detection of KHDC1L protein by the KHDC1L 11–22 and the KHDC1L 114–128 monoclonal antibodies. Panels show immunoblot detection of KHDC1L protein in cells treated with doxycycline (+ DOX) to induce DUX4 and KHDC1L mRNA expression. Left panel: Antibody to KHDC1L peptide 11–22; right panel, antibody to KHDC1L peptide 114–128. (C, D) KHDC1L protein is present in the supernatant of cells expressing DUX4 but not in the same cells without DUX4 expression. (C) Doxycycline treatment (+) has been shown to induce KHDC1L mRNA and the KHDC1L 11–22 monoclonal antibody detects the KHDC1L protein in both the whole cell lysate proteins (WC lysate) and in the concentrated supernatant from the tissue culture media. GAPDH is a secreted protein during micronutrient starvation…

• 

  Figure 3. (A, B) specificity of anti-KHDC1L clones 11–22 and 114–128 assessed using immunoblots against KHDC1, a related KHDC1L family member with high sequence homology. Similar to Fig. 2, both mAbs detect KHDC1L upon DUX4 induction in MB135iDUX4 cell lysate (lanes 1 and 2, both blots; 20ug total cell lysate). (A) When tested against the two recombinant family members, clone 11–22 detected both recombinant His-tagged KHDC1L and His-tagged KHDC1. (B) Clone 114–128 is selective for recombinant His-tagged KHDC1L. Recombinant proteins were loaded on the gel at 62.5 ng for His-KHDC1L and 70.2 ng for His-KHDC1. (C) Specificity of anti-KHDC1L clones 11–22 and 114–128 evaluated by ECLIA. Recombinant His-tagged KHDC1 and KHDC1L were directly coated to plates at concentrations listed, followed by incubation with either anti-KHDC1L clone 11–22 or 114–128 (1ug/mL). Detection was with anti-mouse antibodies conjugated to SULFO tag (1ug/mL). Dotted line denotes background level.

• 

  Figure 4. Mass spectrometry detection of KHDC1L-specific peptides in FSHD patient plasma samples. Enrichment from plasma was performed using a biotinylated aptamer targeting KHDC1L coupled to Dynamag beads. Eluted proteins were digested with trypsin and analyzed using a PRM-PASEF targeted acquisition on a timsTOF HT mass spectrometer (Bruker).

• 

  Figure 5. ECLIA detection of circulating KHDC1L levels in FSHD patient plasma compared to healthy volunteer plasma. (left) standard curve for the ECLIA assay generated using recombinant KHDC1L spiked in 50% human plasma to concentrations of 500, 125, 31.25, 7.81, 1.95, 0.49, 0.12 and 0 ng/ml. The standard curve was interpolated to allow quantitation of endogenous KHDC1L levels in plasma. (right) analysis of the levels of plasma KHDC1L in plasma samples from FSHD patients (n = 20; samples from Resolve natural history study) versus unaffected controls (n = 20; BioIVT).

• 

  Figure 6. Aptamer-mediated detection of KHDC1L levels in plasma. The SomaScan 7K platform (SomaLogic, Boulder CO) was used to quantitate levels of KHDC1L in plasma samples from subjects with FSHD (n = 98; samples from ReSolve natural history study) versus age- and sex-matched healthy volunteers (n = 100; BioIVT). Levels of KHDC1L in plasma were interpolated from a standard curve of recombinant KHDC1L. Circulating levels in FSHD plasma were significantly higher than healthy volunteers (****P < 0.0001).