Alcohol exposure alters pre-mRNA splicing of antiapoptotic Mcl-1L isoform and induces apoptosis in neural progenitors and immature neurons

Abstract

Alternative splicing and expression of splice variants of genes in the brain may lead to the modulation of protein functions, which may ultimately influence behaviors associated with alcohol dependence and neurotoxicity. We recently showed that ethanol exposure can lead to pre-mRNA missplicing of Mcl-1, a pro-survival member of the Bcl-2 family, by downregulating the expression levels of serine/arginine rich splicing factor 1 (SRSF1). Little is known about the physiological expression of these isoforms in neuronal cells and their role in toxicity induced by alcohol exposure during the developmental period. In order to investigate the impact of alcohol exposure on alternative splicing of Mcl-1 pre-mRNA and its role in neurotoxicity, we developed a unique primary human neuronal culture model where neurospheres (hNSPs), neural progenitors (hNPCs), immature neurons, and mature neurons were cultured from the matching donor fetal brain tissues. Our data suggest that neural progenitors and immature neurons are highly sensitive to the toxic effects of ethanol, while mature neuron cultures showed resistance to ethanol exposure. Further analysis of Mcl-1 pre-mRNA alternative splicing by semi-quantitative and quantitative analysis revealed that ethanol exposure causes a significant decrease in Mcl-1L/Mcl-1S ratio in a dose and time dependent manner in neural progenitors. Interestingly, ectopic expression of Mcl-1L isoform in neural progenitors was able to recover the viability loss and apoptosis induced by alcohol exposure. Altogether, these observations suggest that alternative splicing of Mcl-1 may play a crucial role in neurotoxicity associated with alcohol exposure in the developing fetal brain.

Fig. 1. Characterization of primary human neuronal culture model.

Primary human fetal neuronal cells (PHFNs) were cultured from matching human fetal donor brains (16–18 weeks of gestation). a Schematic presentation of human immature neurons (1 week in culture), mature neurons (4 weeks in culture), neurospheres (hNSPs), and hNSPs-derived neural progenitors (hNPCs). b Immature neurons, mature neurons, hNSPs and hNPCs were plated in chamber slides, fixed, and immunostained for Nestin (green; hNSPs and hNPCs), and Synaptophysin (green) and Map2 (red) co-staining (immature and mature neurons). Cells nuclei were also counterstained with DAPI (blue). c hNPCs were plated in neural differentiation media and local field potentials (LFP) were tested at 1, 3, and 4 weeks post differentiation by multielectrode arrays (MEA). hNPCs-derived neurons at 1 week were called immature neurons with no LFP recordings and 4 weeks cultures were classified as mature-functional neurons with high LFP recordings. Scale bar represent 50 μM

Fig. 2. Effect of EtOH exposure on viability of neuronal cells at different stages of differentiation.

hNSPs (a), hNPCs (b), immature neurons (c) and mature neurons (d) were plated in 12-well culture dishes. Cells were exposed to EtOH at 0, 10, 25, 50, and 75 mM concentrations for 6, 24, and 48 h. Cellular viability was assessed at each time point by MTT assay. e Relative cell viability (% control-untreated) were also compared for hNSPs, hNPCs, immature neurons (IN), and mature neurons (MN) exposed to 50 mM EtOH at 6, 24, and 48 h post exposures. Data are mean + SEM of three independent replicates. “P” values were calculated in comparison with control-untreated cells (a–d) or with mature neurons exposed to 50 mM EtOH (e). *p < 0.05, **p < 0.001, NS no significant difference (via t test)

Fig. 3. EtOH exposure induces cleaved caspase-3 activation in neurospheres, neural progenitors and immature neurons, but not in fully differentiated mature neurons in primary cultures.

hNSPs (a, e), hNPCs (b, f), immature neurons (c, g), and mature neurons (d, h) were isolated and cultured from matching human fetal brain in chamber slides. Cells were either treated or untreated with EtOH (50 mM) for 24 h, fixed, and processed for immunocytochemical determination of cleaved caspase-3 protein. Nuclei were also counterstained with DAPI. Cleaved caspase-3 activation was quantified as cl-caspase3 positive area (μm2) based on red fluorescein and presented as bar graph from three independent replicates. Data are mean + SEM of three independent replicates. *p < 0.05, **p < 0.001 (via t test). Scale bar represent 50 μM

Fig. 4. EtOH exposure leads to alternative splicing of Mcl-1S isoform in neurospheres, neural progenitors, and immature neurons but not in mature neurons.

a hNSPs, hNPCs, immature neurons, and mature neurons were isolated and cultured from matching human fetal brain and exposed to 50 mM EtOH. Total RNA was isolated from control cells (cont.) and from cells exposed to EtOH by using a commercial RNA extraction kit at 6 and 24 h post exposures. One microgram of total RNA was used in reverse transcription reactions and cDNA was synthesized. Splicing isoforms of Mcl-1 were analyzed by RT-PCR using specific primers at 6 and 24 h post exposures, and separated and monitored on agarose gel by ethidium bromide staining. The location of primer pairs and map of exon-intron structure of the Mcl-1 pre-mRNA are schematized on the top scheme. b hNSPs, hNPCs, immature neurons, and mature neurons were isolated and cultured from matching human fetal brain and exposed to 50 mM EtOH. Whole cell protein lysates were isolated from control cells (untreated-cont.) and from cells exposed to 50 mM EtOH for 24 h. Western blots of protein lysates were performed to access the expression levels of Mcl-1L, Mcl-1S, SRSF1, Bcl-2, Bax, Bad, and Puma. Tubulin was probed in the same blots as loading control. NS depicts “Nonspecific” band recognized by the antibodies

Fig. 5. Real-time qRT-PCR of Mcl-1L and Mcl-1S isoforms in response to EtOH exposure in neuronal cells.

a Schematic representation of Mcl-1L probe (spanning exons 2/3 junction) with a FAM fluorophore and MCL-1S probe (spanning exons 1/3 junction) with a HEX fluorophore. B. Specificity of Mcl-1L-FAM (b) and Mcl-1S-HEX (c) probes were analyzed by Q-PCR utilizing expression vectors encoding Mcl-1L and Mcl-1S isoforms as standards. Sensitivity of both probes was determined in dual probe reactions in the presence of both Mcl-1L (d) or Mcl-1S standards (e) by Q-PCR. Quantitative amplification and differential detection of both isoforms were also analyzed in a single reaction in the presence of both standards (f). E1–E5 depicts the log scale of plasmid copy numbers. Cq represents the quantitative PCR cycle in which fluorescence can be detected. g hNSPs, hNPCs, immature neurons, and mature neurons were exposed to EtOH (50 mM) for 24 h. Total RNA was extracted and processed by real-time qRT-PCR by utilizing Mcl-1L probe for exon 2/3 junction with FAM fluorophore and Mcl-1S probe for exon 1/3 junction with HEX fluorophore. Percent of Mcl-1L/Mcl-1S ratio was presented as bar graph from three independent assays. h hNPCs were exposed to increasing concentrations of EtOH for 24 h. Total RNA was extracted and processed by real-time qRT-PCR by utilizing Mcl-1L-FAM and Mcl-1S-HEX probes. Percent of Mcl-1L/Mcl-1S ratio was presented as bar graph. Data are mean + SEM of three independent replicates

Fig. 6. Droplet digital-PCR analysis of Mcl-1L and Mcl-1S isoforms in hNPCs exposed to EtOH.

a, b Representative ddPCR reads of cDNA samples from untreated (a) and EtOH treated (b) hNPCs for Mcl-1L (FAM) and Mcl-1S (HEX) transcript copies. c ddPCR reads of Mcl-1L copies from untreated and EtOH treated cells were represented as bar graph. d ddPCR reads of Mcl-1S copies from untreated and EtOH treated cells were represented as bar graph. e Percent of Mcl-1L/Mcl-1S copies was presented as bar graph

Fig. 7. Ectopic expression of Mcl-1L isoform is protective against toxic effects of EtOH in neural progenitors.

a hNPCs were transiently transfected with increasing concentration of expression vectors encoding Mcl-1L or Mcl-1S isoforms. At 24 h post transfections, cells were either exposed to EtOH (50 mM) or left untreated. At 24 h post EtOH exposure, cell viability was assessed by MTT assay, normalized to control (lane 1), and shown as bar graph. b hNPCs were transiently transfected with increasing concentration of an expression vector encoding human Bcl-2. At 24 h post transfections, cells were either exposed to EtOH (50 mM) or left untreated. At 24 h post EtOH exposures, cell viability was assessed by MTT assay, normalized to control (lane 1), and shown as bar graph. Whole cell protein lysates from hNPCs either untransfected (cont.) or transfected with an expression vector encoding human Bcl-2 were processed by western blotting using anti-Bcl-2 antibody and shown as integrated into the bar graph. NS depicts “Nonspecific” band recognized by the antibody. c hNPCs were transiently transfected with expression vectors encoding Mcl-1L or Mcl-1S isoforms. At 24 h post transfections, cells were either exposed to EtOH (50 mM) or left untreated. At 24 h post EtOH exposure cells were fixed and processed by immunocytochemistry for cleaved caspase-3 immunostaining. Nuclei were also counterstained with DAPI. d Cleaved caspase-3 activation was quantified as cl-caspase3 positive area (μm2) based on red fluorescein and presented as bar graph from three independent replicates. Data are mean + SEM of three independent replicates. **p < 0.001 (via t test). Scale bar represent 50 μM. e Whole cell protein lysates from hNPCs either untransfected (cont.) or transfected with expression vectors encoding Mcl-1L or Mcl-1S isoforms were processed by western blotting using anti-Mcl-1 and anti-Tubulin antibodies. Data are mean + SEM of three independent replicates. P values were calculated in comparison with untreated cells (0 mM). *p < 0.05, **p < 0.001 (via t test). Scale bar represents 50 μM

Fig. 8. Schematic representation of proposed model of Mcl-1 pre-mRNA missplicing under the influence of alcohol exposure leading to sensitization of neural progenitors and immature neurons for the toxic effects of alcohol.

EtOH exposure in neural progenitors leads to transcriptional suppression of SRSF1^, that leads to a shift in Mcl-1S splicing over Mcl-1L isoform with exon 2 exclusion resulting reduced copies of Mcl-1L transcript. Compared to Mcl-1L, Mcl-1S lacks the BH1 and BH2 domains which are critical for interacting and sequestering proapoptotic genes, such as NOXO, BIM, PUMA, and BAX, for the inhibition of apoptotic stimulus induced by EtOH