Fig 1: UBE2D3 regulates the abundance of CRABP1 and TSPAN8.A, volcano plot showing changes in total protein levels upon UBE2D3 depletion with shRNA1 in the SILAC-based proteomics (n = 6). Blue dots represent proteins that are at least 1.5-fold upregulated or downregulated in abundance (log2 = 0.585 and log2 = -0.585, respectively) and have a p value of =0.05. Red dots indicate hits overlapping with LFQ proteomics with sh1 and/or sh2. Bold proteins overlap between SILAC Ube2d3 sh1, LFQ Ube2d3 sh1, and LFQ Ube2d3 sh2. B, volcano plot showing changes in total protein levels upon UBE2D3 depletion with shRNA1 in LFQ (n = 3). Blue dots represent proteins that are at least twofold upregulated or downregulated in abundance (log2 = 1.0 and log2 = -1.0, respectively) and have a p value of =0.05. Red dots indicate hits overlapping with SILAC proteomics with Ube2d3 sh1 and/or LFQ proteomics with Ube2d3 sh2. Bold proteins overlap between SILAC Ube2d3 sh1, LFQ Ube2d3 sh1, and LFQ Ube2d3 sh2. C, volcano plot showing changes in total protein levels upon UBE2D3 depletion with shRNA2 in LFQ (n = 3). Blue dots represent proteins that are at least twofold upregulated or downregulated in abundance (log2 = 1.0 and log2 = -1.0, respectively) and have a p value of =0.05. Red dots indicate hits overlapping with SILAC proteomics with Ube2d3 sh1 and/or LFQ proteomics with Ube2d3 sh1. Bold proteins overlap between SILAC Ube2d3 sh1, LFQ Ube2d3 sh1, and LFQ Ube2d3 sh2. D, Venn diagrams illustrating the overlap between proteins that go up in abundance in SILAC Ube2d3 sh1, LFQ Ube2d3 sh1, and LFQ Ube2d3 sh2. E, Venn diagrams illustrating the overlap between proteins that go down in abundance in SILAC Ube2d3 sh1, LFQ Ube2d3 sh1, and LFQ Ube2d3 sh2. F, table of overlapping proteins in (D and E), including their biological functions. Bold proteins overlap between SILAC Ube2d3 sh1, LFQ Ube2d3 sh1, and LFQ Ube2d3 sh2. Orange colored proteins are part of retinol metabolism and signaling pathways. G, immunoblot analysis of CRABP1, TSPAN8, and UBE2D3 protein levels in MEFs with or without depletion of UBE2D3 and untreated (DMSO) or treated with 10 µM of proteasome inhibitor MG132. For Ube2d3 shRNA1 (sh1), cells were treated with 10 µM MG132 for 16 h for assessment of CRABP1 protein levels and for 4 h for assessment of TSPAN8 protein levels. For Ube2d3 shRNA2 (sh2), cells were treated with 10 µM MG132 for 16 h for the assessment of CRABP1 and TSPAN8 protein levels. ß-Actin and GAPDH serve as loading controls. Representative blots of three independent experiments are shown (two additional biological replicates for CRABP1 and TSPAN8 levels in Ube2d3 sh1 cells and three additional biological replicates in Ube2d3 sh2 cells can be found in supplemental Fig. S2A). DMSO, dimethyl sulfoxide; LFQ, label-free quantitation; MEF, mouse embryonic fibroblast; SILAC, stable isotope labeling of amino acids in cell culture.
Fig 2: C32 and C4 dampen CaMKII activity in Day 1 and Day 3 P19-MN differentiation process. (A,B) Western blot of RA, C32, C4 effect and quantification on endogenous CaMKII activity marked by pThr 286/7 activity in Day 1 P19-differentiated MNs. Day 1 MNs were treated with 1uM atRA, C32, or C4 for 15 min and then harvested for western blot analyses. p = 0.02 (RA), p = 0.005 (C32), and p = 0.003 (C4) determined by paired Student’s t-test (n = 5–7). (C,D) Western blot and quantification of RA, C32, C4 effect on endogenous pCaMKII T286/7 activity in Day 3, P19-MNs. Day 3 MNs were treated with 1uM atRA, C32, or C4 for 30 min and then harvested for western blot analyses. Anti-CRABP1 was used to detect endogenous CRABP1. p= 0.04 (RA), p = 0.003 (C32), and p = 0.002 (C4) determined by paired Student’s t-test (n = 5–7). ß-actin was used as a protein loading control. * p = 0.05, ** p = 0.01. Error bars are presented as mean ± SD.
Fig 3: A P19-derived, in vitro MN culture system. (A) Workflow of P19 MN differentiation. First, adherent P19 cells were suspended in +RA (0.5 μM) medium for embryoid body (EB) formation over a two-day period. EB’s were then exchanged into +RA, +Shh medium for neurosphere (NS) formation. Optimal EB and NS formation was achieved through the use of an up-right T75 flask. NS were then dissociated and seeded onto a Matrigel-coated 6-well plate for MN differentiation. Compound experiments were conducted by first depleting atRA and Shh on the relevant day (Day 1 or Day 3) for 18 h. Immediately after the 18-h depletion (indicated by an open circle), MNs were treated with compound at 1 μM for 15 min for Day 1 MNs and 1 μM for 30 min for Day 3 MNs. MNs were then harvested for western blot analyses. (B) Brightfield images of undifferentiated P19 cells (left) and P19-MN differentiated cells on Day 3 (right). Scale bar (white) indicates 100 μm length. (C) qPCR detecting the expression of MN-specific markers, HB9, ChaT, and Isl2. (D) qPCR detecting the expression of Crabp1. Error bars are presented as ± SD. qPCR was conducted in two independent experiments.
Fig 4: C32 and C4 dampen CaMKII activity in a CRABP1-dependent manner. (A,B) Western blot and quantification of pCaMKII activity, marked by pThr-286/7 (pCaMKII T286/7) in HEK293T cells treated with DMSO, atRA, C32, or C4 at 0.5–5 µM for 15 min. HEK293T cells were co-transfected with GFP-CaMKII and either empty vector control or CRABP1 expressing vector. Anti-CRABP1 was used to detect CRABP1 in empty vector and CRABP1 transfected samples. p = 0.03 (RA), 0.001 (C32), 0.001 (C4), determined by paired Student’s t-test (n = 4–7). (C,D) Western blot for detecting endogenous CaMKII activity, marked by pThr-286/7, in P19 cells treated with atRA, C32, or C4 at 0.5–5 µM for 15 min. Anti-CRABP1 was used to detect endogenous CRABP1. ß-actin was used as a protein loading control. p < 0.0001 (atRA, C32, and C4) determined by paired Student’s t-test (n = 12–19). * p = 0.05, *** p = 0.001, “n.s.” not significant. Error bars are presented as mean ± SD.
Fig 5: CRABP1 dampens CaMKII activity to protect against neurotoxic Ca2+ overload. (A) MTT cell viability assay of WT and CRABP1-MN1 exposed to ionomycin (5 µM, 18 h). Cell viability was measured via formazan generation with an absorbance at 570 nm. p = 0.01 determined by Student’s t-test. MTT assay was conducted in five independent experiments with 3–12 technical replicates per experiment. ** p < 0.01 determined by Student’s t-test. Values are mean ± SD. (B) Cell viability assay of WT MN1 to measure the protective effects of atRA, C32, and C4. WT-MN1 cells were pretreated with atRA, C32, or C4 (0.5–5 µM) for 1.5 h. Immediately after pre-treatment, ionomycin (4 µM) or DMSO (as vehicle control) was added to induce cell death and co-incubated with atRA, C32, or C4. p = 0.34 (RA), p = 0.006 (C32), p = 0.0004 determined with Student’s t-test. (C4) (**** p < 0.0001, DMSO vs. Ionmycin; ## p < 0.01 Ionomycin vs. C32; ### p < 0.001 Ionomycin vs. C4). Three independent experiments were performed. (C) Western blots of CaMKII activity marked by pThr 286/7 in wild-type (WT) and CRABP1 over-expressing MN1 cells treated with medium control, DMSO control, or ionomycin (10 µM, 5–10 min). Beta-actin was used as a loading control. Endogenous and over-expressed 3XFlag-HA-tagged CRABP1 expression was monitored with anti-CRABP1. In order to detect a much lower level of endogenous CRABP1 in WT-MN1, twice as much lysate was loaded for WT-MN1 as compared to the loaded CRABP1-MN1 (over-expression) cell lysate. (D) A model depicting the protective role of CRABP1 in MNs when they are exposed to (1) excitotoxicity which results in (2) pathological increases in [Ca2+]I (purple circles), and subsequent CaMKII over-activation and phosphorylation of Ca2+ permeable AMPA receptors. (3) CRABP1-RA could inhibit pCaMKII over-activation, ultimately protecting cells from AMPA-mediated Ca2+ overload and death (4). The model was illustrated using Biorender.com.
Supplier Page from MilliporeSigma for Anti-CRABP1 antibody produced in rabbit