PMC:4504148 / 21357-43690 JSONTXT

{"project":"2_test","denotations":[{"id":"26236191-22841758-32463226","span":{"begin":3551,"end":3555},"obj":"22841758"},{"id":"26236191-1324475-32463227","span":{"begin":3644,"end":3648},"obj":"1324475"},{"id":"26236191-1314207-32463228","span":{"begin":3667,"end":3671},"obj":"1314207"},{"id":"26236191-19306371-32463229","span":{"begin":3815,"end":3819},"obj":"19306371"},{"id":"26236191-22841758-32463230","span":{"begin":3846,"end":3850},"obj":"22841758"},{"id":"26236191-18154684-32463231","span":{"begin":7706,"end":7710},"obj":"18154684"},{"id":"26236191-18337823-32463232","span":{"begin":7730,"end":7734},"obj":"18337823"},{"id":"26236191-22052977-32463233","span":{"begin":7755,"end":7759},"obj":"22052977"},{"id":"26236191-18154684-32463234","span":{"begin":7876,"end":7880},"obj":"18154684"},{"id":"26236191-21907146-32463235","span":{"begin":7898,"end":7902},"obj":"21907146"},{"id":"26236191-21348809-32463236","span":{"begin":7919,"end":7923},"obj":"21348809"},{"id":"26236191-21522185-32463237","span":{"begin":10948,"end":10952},"obj":"21522185"},{"id":"26236191-22292025-32463238","span":{"begin":19906,"end":19910},"obj":"22292025"},{"id":"26236191-23124112-32463239","span":{"begin":19926,"end":19930},"obj":"23124112"},{"id":"26236191-19141680-32463240","span":{"begin":19978,"end":19982},"obj":"19141680"},{"id":"26236191-20158496-32463241","span":{"begin":19993,"end":19997},"obj":"20158496"},{"id":"26236191-8996080-32463242","span":{"begin":20136,"end":20140},"obj":"8996080"},{"id":"26236191-16123220-32463243","span":{"begin":20153,"end":20157},"obj":"16123220"},{"id":"26236191-9878250-32463244","span":{"begin":20483,"end":20487},"obj":"9878250"},{"id":"26236191-16784888-32463245","span":{"begin":20505,"end":20509},"obj":"16784888"},{"id":"26236191-20138239-32463246","span":{"begin":20529,"end":20533},"obj":"20138239"},{"id":"26236191-23935851-32463247","span":{"begin":20552,"end":20556},"obj":"23935851"}],"text":"Results\nTo study the biological pathways of cell death induced by DFNA5 in S. cerevisiae, we performed a transcriptomic study. RNA samples were collected in duplo in the mid-exponential phase and at the post-diauxic shift of yeast cells transformed with either wtDFNA5 of mutDFNA5. Figure 1 shows all the comparisons studied between the different RNA samples (comparisons are labeled 1a, 1b, 2a, and 2b). Analysis of the microarray data was performed using the LIMMA package provided in R and generated four lists of differentially expressed genes. For the GO-enriched term analysis, the cut-off for adjusted p-values of differentially expressed genes was set to 0.05 and the cut-off of the log2 (fold change) was respectively set to 1.5 and 0.5 for yeast and human cell lines.\nInvestigation of the differentially expressed genes in mid-exponential phase (comparison 1a) revealed no significantly up- or down-regulated genes. Therefore, this comparison was excluded and only comparisons 1b, 2a, and 2b (shown in Figure 1) will be taken into account and described.\n\nMitochondria-related processes are up-regulated in mutDFNA5 in post-diauxic shift\nComparison of the differentially expressed genes (adjusted p \u003c 0.05) in the post-diauxic shift (1b) resulted in 451 significantly up-regulated genes when expressing mutDFNA5 and using wtDFNA5 as a reference. The top 34 differentially up-regulated genes at the post-diauxic shift are shown in Table 1. Differentially expressed genes with a log2 (fold change) equal or higher than 1.5 were selected for GO-enriched annotations, generating 85 significantly up-regulated genes, which resulted in 65 significantly up-regulated enriched GO terms.\nTable 1 Top 34 of the differentially up-regulated genes upon transfection of mutDFNA5 in yeast at the post-diauxic shift. LogFC, logarithm of fold change, up-regulation in mutDFNA5 yeast samples using wtDFNA5 as a reference; adj.p.value, p-value adjusted for multiple hypothesis testing. Genes involved in processes related to the COX activity are indicated in bold. Analysis of the biological, cellular, and molecular GO annotations confirmed the role of the mitochondria in mutDFNA5-induced cell death (Supplemental Data Table 2, indicated in bold). Analysis of the GO annotations revealed that the molecular cytochrome-c oxidase activity (GO: 0004129) related process was the most significantly up-regulated mitochondrial process. Further down the list, several biological, molecular and cellular GO processes related to mitochondrial mechanisms, such as mitochondrial ATP synthesis-coupled electron transport (GO:0042775), aerobic respiration (GO:0009060), the mitochondrial respiratory chain (GO:0005746), mitochondrial respiratory chain complex IV (GO:0005751), and oxidative phosphorylation (GO:0006119) were significantly up-regulated suggesting mitochondrial dysfunction (Supplemental Data Table 2).\nNext we compared the identified GO terms with the list containing the highest differentially up-regulated genes generated by the R package LIMMA to evaluate the resemblances (Table 1). As shown in Table 1, several mitochondrial genes related to these GO processes were indeed present in the list, including COX/1/2/3 and AI5_ALPHA (Table 1, bold). COX1/2/3 are three main subunits of cytochrome c oxidase, the terminal enzyme of the mitochondrial electron transport chain, encoded by the mitochondrial genome. The electron transport chain is part of mitochondrial oxidative phosphorylation providing most of the cellular ATP (Srinivasan and Avadhani, 2012). AI5_ALPHA is an endonuclease encoding a mobile intron of the COXI gene (Moran et al., 1992; Seraphin et al., 1992). Up-regulation of the main COX genes suggests enhanced COX activity, which has been associated with increased oxidative stress (Singh et al., 2009; Srinivasan and Avadhani, 2012). Previous data indeed demonstrated a change in redox homeostasis due to mutDFNA5 expression in yeast (Van Rossom et al., 2012). The same study also showed increased oxidative stress measured by a dihydroethidium bromide staining (DHE). Moreover, preliminary experiments in human cell lines confirmed this and also revealed enhanced oxidative stress measured by a DHE staining (unpublished results).\nIn addition, two other groups of significantly enriched GO annotations could be distinguished, namely GO annotations related to catabolic and metabolic energy processes, such as oligosaccharide catabolic process (GO:0009313) or maltose catabolic processes (GO:0000025), and mechanisms related to transporter activity, such as cation (GO:0008324) and several sugar (GO:0005353 for example) transmembrane activities (Supplemental Data Table 2, respectively underlined and indicated in blue). Consistent with the previous results, comparison of these processes with the highest differentially up-regulated genes in Table 1 confirmed these identified GO terms (Table 1). Different maltose and sucrose genes like MAL12, MAL32, and SUC2, and transmembrane transporter genes like HXT4/7 were present in the list (Table 1).\nThese results revealed an important role for mitochondria-related processes in mutDFNA5 transformed yeast cells in the post-diauxic shift.\n\nProcesses associated with glycolysis are down-regulated in mutDFNA5 at post-diauxic shift\nTo investigate the significantly down-regulated processes and genes in the post-diauxic shift between wtDFNA5 and mutDFNA5, we used the same method as described in Section Mitochondria-related Processes are Up-regulated in mutDFNA5 in Post-diauxic Shift. This revealed 585 significantly down-regulated genes in cells expressing mutDFNA5 as compared to those expressing wtDFNA5 (adjusted p \u003c 0.05). The top 34 highest differentially down-regulated genes are shown in Table 2.\nTable 2 Top 34 of the differentially down-regulated genes upon transfection of mutDFNA5 in yeast at the post-diauxic shift. LogFC, logarithm of fold change, down-regulation in mutDFNA5 yeast samples using wtDFNA5 as a reference; adj.p.value, p-value adjusted for multiple hypothesis testing. Genes involved in processes related to the glycolysis or the pentose phosphate pathways are indicated in bold. The significantly down-regulated biological GO annotations can be divided in two main groups (Supplemental Data Table 3). One group is related to ribosomal processes and hence translation such as cytosolic ribosome (GO:0022626) and the positive regulation of translation fidelity (GO:0045903) (Supplemental Data Table 3, indicated in bold). This down-regulation is probably due to the fact that yeast is entering the post-diauxic shift and that mutDFNA5 has a growth defect compared to wtDFNA5. Hence, this is probably not due to mutDFNA5 expression. The second group was correlated with the biosynthesis and metabolism of glucose (GO:0006007), monosaccharide (GO:0046365) and glycolysis (GO:0006096).\nAgain, we compared the identified GO terms using the list containing the highest differentially down-regulated genes generated by the R package LIMMA to evaluate the resemblances (Table 2). As expected, this list contained several components associated with the glycolysis and several protein components of the small and large ribosomal subunit.\nInterestingly, the list also contained several genes such as TPI, TDH2/3, PGK1, and CDC19, which are all enzymes playing a role in either the glycolytic or the pentose phosphate pathway (PPP) (Table 2 indicated in bold). CDC19 is the yeast homolog of the human pyruvate kinase (PK) gene. Down-regulation of PK has been correlated with the activation of the PPP and the redirection of the metabolic flux from glycolysis to PPP both in human cell lines and in yeast (Ralser et al., 2007; Christofk et al., 2008; Anastasiou et al., 2011). This will enhance the anti-oxidant response and hence increase the tolerance for oxidative stress (Ralser et al., 2007; Gruning et al., 2011; Kruger et al., 2011). The down-regulation of genes involved in glycolysis and the PPP could suggest a link with oxidative stress providing a protection mechanism for mutDFNA5-transformed yeast cells.\n\nInduction of ER-related processes upon mutDFNA5 expression in yeast\nIn addition to the comparison of mutDFNA5 and wtDFNA5 in the post-diauxic shift, the modifications between mid-exponential phase and at the post-diauxic shift were investigated separately both in mutDFNA5-(comparison 2a Figure 1) and wtDFNA5-(comparison 2b Figure 1) transformed yeast cells. As both wtDFNA5 and mutDFNA5 cells demonstrated differentially expressed genes in exponential phase compared to post-diauxic shift, we expected the presence of many significantly up- or down regulated genes related to the post-diauxic shift but not solely due to mutDFNA5 expression. Therefore, genes which were differentially expressed at the post-diauxic shift upon mutDFNA5 transformation, but do not show any differences upon wtDFNA5 expression in post-diauxic phase, are potentially related to mutDFNA5-associated processes. These were assigned as mutDFNA5-related changes not associated with the post-diauxic shift in yeast.\nGenes with an adjusted p-value below 0.05 and a log2 (fold change) above 1.5 were selected for GO analysis. The GO-enriched processes significantly associated with up-regulated genes were very similar between comparison 2a and 2b and were associated with translation. GO-enriched terms significantly associated with down-regulated genes were related to ribosomes and RNA and were present both in wtDFNA5- and in mutDFNA5-transformed yeast cells. These processes were probably due to the shift to respiration and not in particular related to mutDFNA5 (data not shown). However, three main classes could be distinguished at the post-diauxic shift. Two of them were more prominent in mutDFNA5-transformed yeast cells. One class was related to the biosynthesis and the metabolism of lipids (GO:0008610), such as (ergo)sterols (GO:0016126), (phyto)steroids (GO:0006694), and fatty acids (GO:0006633) (Supplemental Data Table 4, indicated in bold). The other group was associated with the ER (GO:0005783), such as the ER membrane (GO:0005789) and protein targeting to ER (GO: 0045047) (Supplemental Data Table 4, processes are underlined).\nThe third group which could be distinguished was related to the cytoskeleton (GO:0005856) and was more pronounced in wtDFNA5-transformed yeast cells. Cellularly enriched GO terms such as the microtubule cytoskeleton (GO:0015630) and the microtubule organizing center (GO:0005815) were present in this list (Supplemental Data Table 5, indicated in bold).\n\nAssociation of DFNA5 with the MAPK-related mechanisms in HEK293T cells\nTo further elucidate the DFNA5-related pathways, a microarray experiment was performed in human HEK293T cells. As described previously, mutDFNA5 induced a growth defect in transfected HEK293T cells compared to wtDFNA5 and control (cells transfected with an empty vector) (Op de Beeck et al., 2011). These cell death events were evident from 9 h post-transfection and peaked at 12 h (data not shown). Therefore RNA samples of HEK293T cells were collected 12 h post-transfection. A transcriptomic analysis was performed on HEK293T cells transfected with either wtDFNA5 or mutDFNA5. Six biological replicates of every RNA sample were collected although one wtDFNA5-transfected sample did not survive quality control. Subsequent analyses, using wtDFNA5 as a reference, were therefore performed on five wtDFNA5- vs. six mutDFNA5-transfected samples. Analysis using “Beadarray” and “LIMMA” packages available in R identified 228 significantly up- and 222 significantly down-regulated genes after correction for multiple hypothesis testing (p \u003c 0.05). In addition to individual gene expression, GO analysis was performed to determine the biologically, cellularly, and molecularly enriched GO annotations linked to the differentially expressed genes.\nTable 3 shows the top 34 of the significantly up-regulated genes. It contains several genes related to the MAPK pathway such as EGR1/2, FOSB, andJUNB (indicated in bold). Interestingly, this list also contained the PMAIP1 gene. PMAIP1 encodes a BH3-only protein belonging to the BCL2 protein family, a family of important regulators of apoptotic cell death related to the mitochondria. The top 34 highest down-regulated genes are shown in Table 4 and contains several genes related to protein folding such as HSPA6, ATF3, and CTH (indicated in bold, Table 4).\nTable 3 Top 34 of the significantly up-regulated genes in mutDFNA5 transfected HEK293T cells. LogFC, logarithm of fold change, down-regulation in mutDFNA5 yeast samples using wtDFNA5 as a reference; adj.p.value, p-value adjusted for multiple hypothesis testing. The array address of the specific splice variant on the microarray is provided between parentheses. Genes involved in processes related to either the MAPK pathway, cAMP response or the mitochondria are indicated in bold.\nTable 4 Top 34 of the highest significantly down-regulated genes in mutDFNA5 transfected HEK293T cells. LogFC, logarithm of fold change, down-regulation in mutDFNA5 yeast samples using wtDFNA5 as a reference; adj.p.value, p-value adjusted for multiple hypothesis testing. The array address of the specific splice variant on the microarray is provided between parentheses. Genes involved in processes related to protein folding are indicated in bold. Subsequent GO analysis of the biological annotations revealed, in addition to the more general development processes, the up-regulation of the MAPK pathway (GO:0043407) and the cAMP response (GO:0051591) (Supplemental Data Table 6, indicated in bold). The response to protein folding (GO:0006986) and to topologically incorrect protein (GO:0035966) were the only two significantly down-regulated processes and both were related to protein folding (Supplemental Data Table 7). The most important genes that are involved in these processes were HSPA6, a heat shock protein and several chaperones proteins, such as DNAJB1 and DNAJB2.\nThese results demonstrate the association of mutDFNA5-induced cell death with the MAPK pathways. The identification of processes related to protein folding supports the results in yeast in which GO terms related to protein folding and the ER were significantly associated with mutDFNA5.\n\nValidation of the MAPK role in DFNA5-related cell death in HEK293T cells\nThe data generated by the transcriptomic analysis in HEK293T cells were validated by real-time rtPCR of newly collected RNA samples. EGR1 and FOSB gene expression was investigated on different time-points ranging from 3 to 72 h post-transfection. As shown in Figure 2, significantly up-regulated EGR1 and FOSB gene expression was observed in cells transfected with mutDFNA5 from 12 to 18 h post transfection (p \u003c 0.05) (Figure 2). Hence the data generated by the transcriptomic analyses were indeed confirmed by real time rtPCR as demonstrated by up-regulation of genes related to the MAPK pathway.\nFigure 2 Increased EGR1 and FOSB gene expression in mutDFNA5-transfected HEK293T cells. RNA samples were collected from HEK293T cells transfected with either wtDFNA5 or mutDFNA5 and gene expression was measured by real-time rtPCR. Significantly increased expression was seen in mutDFNA5 at 12 h [p(egr1) = 0.000; p(fosB) = 0.006], 15 h [p(egr1) = 0.017; p(fosB) = 0.000] and 18 h [p(egr1) = 0.004; p(fosB) = 0.026] post-transfection. *p \u003c 0.05; **p \u003c 0.01; ***p \u003c 0.001. CNRQ, calibrated normalized relative quantities. After confirmation by real time rtPCR, the significance of the activated MAPK pathway was further validated by two independent experiments. To investigate the significance of the MAPK pathway, we wondered whether inhibition of the MAPK pathway would attenuate this mutDFNA5-induced growth defect.\nTherefore, a specific JNK inhibitor, namely SP600125, was added, followed by a viability assay to determine the effect on cell survival. Different concentrations of the JNK inhibitor SP600125 were used to measure viability by flow cytometry (CyFlow ML, Partec, Germany) and these results were compared to untreated mutDFNA5-transfected HEK293T cells. Overnight treatment of the cells with different concentrations of SP600125 did not have any major effect on transfection efficiency, but significantly increased the viability of mutDFNA5-transfected cells. Although, addition of 12.5 and 25 μM SP600125 both significantly increased the viability with a p-value of 0.020 and 0.004 respectively SP600125 had the greatest effect with a concentration of 25 μM SP600125 since the viability was raised from 31.93 to 51.00% (Figure 3).\nFigure 3 MAPK inhibitor effect on mutDFNA5 transfected HEK293T cells. MutDFNA5-transfected HEK293T cells were pretreated with different amounts of SP600125 (JNK inhibitor). Cell viability was measured and compared to untreated mutDFNA5-transfected HEK293T cells. *p \u003c 0.05; **p \u003c 0.01. Next, to evaluate the effect of MAPK up-regulation on protein level, different MAPK proteins were studied by western blotting. There are three main MAPK pathways in human cell lines represented by the ERK, JNK, and p38 MAPK branch. Consistent with the results obtained by real time rtPCR and the viability assay, activation of the MAPK pathway proteins was also demonstrated by western blotting. Total protein lysates were collected from HEK293T cells 12 h post-transfection. Three phosphorylated (activated) and non-phosphorylated (not activated) proteins of the MAPK pathway were studied using six different antibodies. No differences were seen in the expression level of non-phosphorylated ERK and JNK (Figure 4A). Activation of JNK and to a minor extent of ERK (p42/p44) was seen upon mutDFNA5 transfection compared to control and wtDFNA5 (Figure 4B). The expression of p38 was also evaluated but no difference in protein expression was observed between mutDFNA5 compared to wtDFNA5 and control (data not shown). β-Actin was used as a loading control.\nFigure 4 Activation of the MAPK pathway by mutDFNA5. (A) Western blot analysis of non- phosphorylated ERK (p42/p42) and JNK. No differences were seen in protein expression level of ERK and JNK between control, wtDFNA5 and mutDFNA5 transfected HEK293T cells. (B) Western blot analysis of phosphorylated, activated ERK (p42/p42) and JNK. Increased expression of JNK and to a lesser extent of ERK was seen in mutDFNA5 transfected HEK293T cells as compared to wtDFNA5 and control. These results suggest that DFNA5 induces PCD mediated through activation of the MAPK pathways. Addition of a MAPK inhibitor partially attenuated the mutDFNA5-induced growth defect identifying the MAPK pathway as an early event in mutDFNA5-associated cell death.\n\nComparison of the yeast microarray results with the gene expression in human cell lines\nIn order to study the significance of the yeast results, a comparison was made between the two microarray experiments. Human homologs of the significantly up- and down-regulated yeast genes at the post-diauxic shift (comparison 1b) were identified using Ensemble Biomart. Of the 451 significantly up- and 585 significantly down-regulated yeast genes, respectively 296 and 647 human homologs were identified. These specific human homologs were analyzed using the R package LIMMA, which generated a new list of human genes. The FC cut-off of the resulting gene list was set to FC 1.2, resulting in 16 up- and 14 down-regulated human genes (Supplemental Tables 8A,B). TM7SF2, UCP2, and VPS33B, three down-regulated human genes, were selected to verify the yeast results in human cell lines using real-time rtPCR. UCP2 and VPS33B were evaluated because they were the two most down-regulated genes present in the list. UCP2, an uncoupling protein, is a mitochondrial carrier located at the mitochondrial inner membrane. Suppression of UCP2 has been linked to increased ROS production (Deng et al., 2012; Dando et al., 2013) and lifespan regulation (Andrews and Horvath, 2009; Andrews, 2010). The Vacuolar protein sorting 33 homolog (VPS33B) gene is involved in intracellular vesicle Golgi-to-lysosome transport (Pevsner et al., 1996; Lo et al., 2005). TM7SF2 was selected based on its function in relation to the yeast results. TM7SF2 is a transmembrane protein present in the ER and associated with biosynthesis of cholesterol. In addition to its role in cholesterol synthesis, TM7SF2 appears to be involved in the inflammatory response upon cellular stress (Holmer et al., 1998; Bennati et al., 2006; Schiavoni et al., 2010; Bellezza et al., 2013).\nTo verify the yeast results in human cell lines, RNA was collected from HEK293T cells transfected with either wtDFNA5 or mutDFNA5 at different time-points starting at 12 h after transfection as this was the time-point of the human microarray experiment. The TM7SF2, UCP2, and VPS33B genes had a fold change of respectively 1.23, 1.31, and 1.27 on the microarray. Real-time rtPCR on RNA samples 12 h post-transfection confirmed these microarray results as all three genes were down-regulated in mutDFNA5 compared to wtDFNA5 (Figure 5A) with fold changes comparable to the microarray (respectively 1.35, 1.48, and 1.59). Although not significantly, these three genes were down-regulated in HEK293T cells at 12 h after transfection (Figure 5A). The down-regulation was still present at 20 h, peaked at 24 h after transfection and was even significant for TM7SF2 (p:0.01) and UCP2 (p:0.07) at respectively 20 h and 24 h after transfection (Figure 5B). Due to this down-regulation, we can conclude that there are some similarities between the yeast and the HEK293T microarray. Differences are seen when looking at the individual genes, but upon study of the different pathways a role for processes related to protein folding were seen in both model systems.\nFigure 5 Validation of the yeast microarray by real-time rtPCR in HEK293T cells. (A) Gene expression 12 h post-transfection in human HEK293T cells of three selected genes. (B) Gene expression 12, 20, and 24 h post-transfection with either wtDFNA5 or mutant DFNA5. UCP2, uncoupling protein 2 (mitochondrial, proton carrier); TM7SF2, transmembrane 7 superfamily member 2; VPS33B, vacuolar protein sorting 33 homolog. Light gray, wtDFNA5; Dark gray, mutDFNA5. *p \u003c 0.05. CNRQ, calibrated normalized relative quantities.\n\nDi"}

{"project":"MyTest","denotations":[{"id":"26236191-22841758-32463226","span":{"begin":3551,"end":3555},"obj":"22841758"},{"id":"26236191-1324475-32463227","span":{"begin":3644,"end":3648},"obj":"1324475"},{"id":"26236191-1314207-32463228","span":{"begin":3667,"end":3671},"obj":"1314207"},{"id":"26236191-19306371-32463229","span":{"begin":3815,"end":3819},"obj":"19306371"},{"id":"26236191-22841758-32463230","span":{"begin":3846,"end":3850},"obj":"22841758"},{"id":"26236191-18154684-32463231","span":{"begin":7706,"end":7710},"obj":"18154684"},{"id":"26236191-18337823-32463232","span":{"begin":7730,"end":7734},"obj":"18337823"},{"id":"26236191-22052977-32463233","span":{"begin":7755,"end":7759},"obj":"22052977"},{"id":"26236191-18154684-32463234","span":{"begin":7876,"end":7880},"obj":"18154684"},{"id":"26236191-21907146-32463235","span":{"begin":7898,"end":7902},"obj":"21907146"},{"id":"26236191-21348809-32463236","span":{"begin":7919,"end":7923},"obj":"21348809"},{"id":"26236191-21522185-32463237","span":{"begin":10948,"end":10952},"obj":"21522185"},{"id":"26236191-22292025-32463238","span":{"begin":19906,"end":19910},"obj":"22292025"},{"id":"26236191-23124112-32463239","span":{"begin":19926,"end":19930},"obj":"23124112"},{"id":"26236191-19141680-32463240","span":{"begin":19978,"end":19982},"obj":"19141680"},{"id":"26236191-20158496-32463241","span":{"begin":19993,"end":19997},"obj":"20158496"},{"id":"26236191-8996080-32463242","span":{"begin":20136,"end":20140},"obj":"8996080"},{"id":"26236191-16123220-32463243","span":{"begin":20153,"end":20157},"obj":"16123220"},{"id":"26236191-9878250-32463244","span":{"begin":20483,"end":20487},"obj":"9878250"},{"id":"26236191-16784888-32463245","span":{"begin":20505,"end":20509},"obj":"16784888"},{"id":"26236191-20138239-32463246","span":{"begin":20529,"end":20533},"obj":"20138239"},{"id":"26236191-23935851-32463247","span":{"begin":20552,"end":20556},"obj":"23935851"}],"namespaces":[{"prefix":"_base","uri":"https://www.uniprot.org/uniprot/testbase"},{"prefix":"UniProtKB","uri":"https://www.uniprot.org/uniprot/"},{"prefix":"uniprot","uri":"https://www.uniprot.org/uniprotkb/"}],"text":"Results\nTo study the biological pathways of cell death induced by DFNA5 in S. cerevisiae, we performed a transcriptomic study. RNA samples were collected in duplo in the mid-exponential phase and at the post-diauxic shift of yeast cells transformed with either wtDFNA5 of mutDFNA5. Figure 1 shows all the comparisons studied between the different RNA samples (comparisons are labeled 1a, 1b, 2a, and 2b). Analysis of the microarray data was performed using the LIMMA package provided in R and generated four lists of differentially expressed genes. For the GO-enriched term analysis, the cut-off for adjusted p-values of differentially expressed genes was set to 0.05 and the cut-off of the log2 (fold change) was respectively set to 1.5 and 0.5 for yeast and human cell lines.\nInvestigation of the differentially expressed genes in mid-exponential phase (comparison 1a) revealed no significantly up- or down-regulated genes. Therefore, this comparison was excluded and only comparisons 1b, 2a, and 2b (shown in Figure 1) will be taken into account and described.\n\nMitochondria-related processes are up-regulated in mutDFNA5 in post-diauxic shift\nComparison of the differentially expressed genes (adjusted p \u003c 0.05) in the post-diauxic shift (1b) resulted in 451 significantly up-regulated genes when expressing mutDFNA5 and using wtDFNA5 as a reference. The top 34 differentially up-regulated genes at the post-diauxic shift are shown in Table 1. Differentially expressed genes with a log2 (fold change) equal or higher than 1.5 were selected for GO-enriched annotations, generating 85 significantly up-regulated genes, which resulted in 65 significantly up-regulated enriched GO terms.\nTable 1 Top 34 of the differentially up-regulated genes upon transfection of mutDFNA5 in yeast at the post-diauxic shift. LogFC, logarithm of fold change, up-regulation in mutDFNA5 yeast samples using wtDFNA5 as a reference; adj.p.value, p-value adjusted for multiple hypothesis testing. Genes involved in processes related to the COX activity are indicated in bold. Analysis of the biological, cellular, and molecular GO annotations confirmed the role of the mitochondria in mutDFNA5-induced cell death (Supplemental Data Table 2, indicated in bold). Analysis of the GO annotations revealed that the molecular cytochrome-c oxidase activity (GO: 0004129) related process was the most significantly up-regulated mitochondrial process. Further down the list, several biological, molecular and cellular GO processes related to mitochondrial mechanisms, such as mitochondrial ATP synthesis-coupled electron transport (GO:0042775), aerobic respiration (GO:0009060), the mitochondrial respiratory chain (GO:0005746), mitochondrial respiratory chain complex IV (GO:0005751), and oxidative phosphorylation (GO:0006119) were significantly up-regulated suggesting mitochondrial dysfunction (Supplemental Data Table 2).\nNext we compared the identified GO terms with the list containing the highest differentially up-regulated genes generated by the R package LIMMA to evaluate the resemblances (Table 1). As shown in Table 1, several mitochondrial genes related to these GO processes were indeed present in the list, including COX/1/2/3 and AI5_ALPHA (Table 1, bold). COX1/2/3 are three main subunits of cytochrome c oxidase, the terminal enzyme of the mitochondrial electron transport chain, encoded by the mitochondrial genome. The electron transport chain is part of mitochondrial oxidative phosphorylation providing most of the cellular ATP (Srinivasan and Avadhani, 2012). AI5_ALPHA is an endonuclease encoding a mobile intron of the COXI gene (Moran et al., 1992; Seraphin et al., 1992). Up-regulation of the main COX genes suggests enhanced COX activity, which has been associated with increased oxidative stress (Singh et al., 2009; Srinivasan and Avadhani, 2012). Previous data indeed demonstrated a change in redox homeostasis due to mutDFNA5 expression in yeast (Van Rossom et al., 2012). The same study also showed increased oxidative stress measured by a dihydroethidium bromide staining (DHE). Moreover, preliminary experiments in human cell lines confirmed this and also revealed enhanced oxidative stress measured by a DHE staining (unpublished results).\nIn addition, two other groups of significantly enriched GO annotations could be distinguished, namely GO annotations related to catabolic and metabolic energy processes, such as oligosaccharide catabolic process (GO:0009313) or maltose catabolic processes (GO:0000025), and mechanisms related to transporter activity, such as cation (GO:0008324) and several sugar (GO:0005353 for example) transmembrane activities (Supplemental Data Table 2, respectively underlined and indicated in blue). Consistent with the previous results, comparison of these processes with the highest differentially up-regulated genes in Table 1 confirmed these identified GO terms (Table 1). Different maltose and sucrose genes like MAL12, MAL32, and SUC2, and transmembrane transporter genes like HXT4/7 were present in the list (Table 1).\nThese results revealed an important role for mitochondria-related processes in mutDFNA5 transformed yeast cells in the post-diauxic shift.\n\nProcesses associated with glycolysis are down-regulated in mutDFNA5 at post-diauxic shift\nTo investigate the significantly down-regulated processes and genes in the post-diauxic shift between wtDFNA5 and mutDFNA5, we used the same method as described in Section Mitochondria-related Processes are Up-regulated in mutDFNA5 in Post-diauxic Shift. This revealed 585 significantly down-regulated genes in cells expressing mutDFNA5 as compared to those expressing wtDFNA5 (adjusted p \u003c 0.05). The top 34 highest differentially down-regulated genes are shown in Table 2.\nTable 2 Top 34 of the differentially down-regulated genes upon transfection of mutDFNA5 in yeast at the post-diauxic shift. LogFC, logarithm of fold change, down-regulation in mutDFNA5 yeast samples using wtDFNA5 as a reference; adj.p.value, p-value adjusted for multiple hypothesis testing. Genes involved in processes related to the glycolysis or the pentose phosphate pathways are indicated in bold. The significantly down-regulated biological GO annotations can be divided in two main groups (Supplemental Data Table 3). One group is related to ribosomal processes and hence translation such as cytosolic ribosome (GO:0022626) and the positive regulation of translation fidelity (GO:0045903) (Supplemental Data Table 3, indicated in bold). This down-regulation is probably due to the fact that yeast is entering the post-diauxic shift and that mutDFNA5 has a growth defect compared to wtDFNA5. Hence, this is probably not due to mutDFNA5 expression. The second group was correlated with the biosynthesis and metabolism of glucose (GO:0006007), monosaccharide (GO:0046365) and glycolysis (GO:0006096).\nAgain, we compared the identified GO terms using the list containing the highest differentially down-regulated genes generated by the R package LIMMA to evaluate the resemblances (Table 2). As expected, this list contained several components associated with the glycolysis and several protein components of the small and large ribosomal subunit.\nInterestingly, the list also contained several genes such as TPI, TDH2/3, PGK1, and CDC19, which are all enzymes playing a role in either the glycolytic or the pentose phosphate pathway (PPP) (Table 2 indicated in bold). CDC19 is the yeast homolog of the human pyruvate kinase (PK) gene. Down-regulation of PK has been correlated with the activation of the PPP and the redirection of the metabolic flux from glycolysis to PPP both in human cell lines and in yeast (Ralser et al., 2007; Christofk et al., 2008; Anastasiou et al., 2011). This will enhance the anti-oxidant response and hence increase the tolerance for oxidative stress (Ralser et al., 2007; Gruning et al., 2011; Kruger et al., 2011). The down-regulation of genes involved in glycolysis and the PPP could suggest a link with oxidative stress providing a protection mechanism for mutDFNA5-transformed yeast cells.\n\nInduction of ER-related processes upon mutDFNA5 expression in yeast\nIn addition to the comparison of mutDFNA5 and wtDFNA5 in the post-diauxic shift, the modifications between mid-exponential phase and at the post-diauxic shift were investigated separately both in mutDFNA5-(comparison 2a Figure 1) and wtDFNA5-(comparison 2b Figure 1) transformed yeast cells. As both wtDFNA5 and mutDFNA5 cells demonstrated differentially expressed genes in exponential phase compared to post-diauxic shift, we expected the presence of many significantly up- or down regulated genes related to the post-diauxic shift but not solely due to mutDFNA5 expression. Therefore, genes which were differentially expressed at the post-diauxic shift upon mutDFNA5 transformation, but do not show any differences upon wtDFNA5 expression in post-diauxic phase, are potentially related to mutDFNA5-associated processes. These were assigned as mutDFNA5-related changes not associated with the post-diauxic shift in yeast.\nGenes with an adjusted p-value below 0.05 and a log2 (fold change) above 1.5 were selected for GO analysis. The GO-enriched processes significantly associated with up-regulated genes were very similar between comparison 2a and 2b and were associated with translation. GO-enriched terms significantly associated with down-regulated genes were related to ribosomes and RNA and were present both in wtDFNA5- and in mutDFNA5-transformed yeast cells. These processes were probably due to the shift to respiration and not in particular related to mutDFNA5 (data not shown). However, three main classes could be distinguished at the post-diauxic shift. Two of them were more prominent in mutDFNA5-transformed yeast cells. One class was related to the biosynthesis and the metabolism of lipids (GO:0008610), such as (ergo)sterols (GO:0016126), (phyto)steroids (GO:0006694), and fatty acids (GO:0006633) (Supplemental Data Table 4, indicated in bold). The other group was associated with the ER (GO:0005783), such as the ER membrane (GO:0005789) and protein targeting to ER (GO: 0045047) (Supplemental Data Table 4, processes are underlined).\nThe third group which could be distinguished was related to the cytoskeleton (GO:0005856) and was more pronounced in wtDFNA5-transformed yeast cells. Cellularly enriched GO terms such as the microtubule cytoskeleton (GO:0015630) and the microtubule organizing center (GO:0005815) were present in this list (Supplemental Data Table 5, indicated in bold).\n\nAssociation of DFNA5 with the MAPK-related mechanisms in HEK293T cells\nTo further elucidate the DFNA5-related pathways, a microarray experiment was performed in human HEK293T cells. As described previously, mutDFNA5 induced a growth defect in transfected HEK293T cells compared to wtDFNA5 and control (cells transfected with an empty vector) (Op de Beeck et al., 2011). These cell death events were evident from 9 h post-transfection and peaked at 12 h (data not shown). Therefore RNA samples of HEK293T cells were collected 12 h post-transfection. A transcriptomic analysis was performed on HEK293T cells transfected with either wtDFNA5 or mutDFNA5. Six biological replicates of every RNA sample were collected although one wtDFNA5-transfected sample did not survive quality control. Subsequent analyses, using wtDFNA5 as a reference, were therefore performed on five wtDFNA5- vs. six mutDFNA5-transfected samples. Analysis using “Beadarray” and “LIMMA” packages available in R identified 228 significantly up- and 222 significantly down-regulated genes after correction for multiple hypothesis testing (p \u003c 0.05). In addition to individual gene expression, GO analysis was performed to determine the biologically, cellularly, and molecularly enriched GO annotations linked to the differentially expressed genes.\nTable 3 shows the top 34 of the significantly up-regulated genes. It contains several genes related to the MAPK pathway such as EGR1/2, FOSB, andJUNB (indicated in bold). Interestingly, this list also contained the PMAIP1 gene. PMAIP1 encodes a BH3-only protein belonging to the BCL2 protein family, a family of important regulators of apoptotic cell death related to the mitochondria. The top 34 highest down-regulated genes are shown in Table 4 and contains several genes related to protein folding such as HSPA6, ATF3, and CTH (indicated in bold, Table 4).\nTable 3 Top 34 of the significantly up-regulated genes in mutDFNA5 transfected HEK293T cells. LogFC, logarithm of fold change, down-regulation in mutDFNA5 yeast samples using wtDFNA5 as a reference; adj.p.value, p-value adjusted for multiple hypothesis testing. The array address of the specific splice variant on the microarray is provided between parentheses. Genes involved in processes related to either the MAPK pathway, cAMP response or the mitochondria are indicated in bold.\nTable 4 Top 34 of the highest significantly down-regulated genes in mutDFNA5 transfected HEK293T cells. LogFC, logarithm of fold change, down-regulation in mutDFNA5 yeast samples using wtDFNA5 as a reference; adj.p.value, p-value adjusted for multiple hypothesis testing. The array address of the specific splice variant on the microarray is provided between parentheses. Genes involved in processes related to protein folding are indicated in bold. Subsequent GO analysis of the biological annotations revealed, in addition to the more general development processes, the up-regulation of the MAPK pathway (GO:0043407) and the cAMP response (GO:0051591) (Supplemental Data Table 6, indicated in bold). The response to protein folding (GO:0006986) and to topologically incorrect protein (GO:0035966) were the only two significantly down-regulated processes and both were related to protein folding (Supplemental Data Table 7). The most important genes that are involved in these processes were HSPA6, a heat shock protein and several chaperones proteins, such as DNAJB1 and DNAJB2.\nThese results demonstrate the association of mutDFNA5-induced cell death with the MAPK pathways. The identification of processes related to protein folding supports the results in yeast in which GO terms related to protein folding and the ER were significantly associated with mutDFNA5.\n\nValidation of the MAPK role in DFNA5-related cell death in HEK293T cells\nThe data generated by the transcriptomic analysis in HEK293T cells were validated by real-time rtPCR of newly collected RNA samples. EGR1 and FOSB gene expression was investigated on different time-points ranging from 3 to 72 h post-transfection. As shown in Figure 2, significantly up-regulated EGR1 and FOSB gene expression was observed in cells transfected with mutDFNA5 from 12 to 18 h post transfection (p \u003c 0.05) (Figure 2). Hence the data generated by the transcriptomic analyses were indeed confirmed by real time rtPCR as demonstrated by up-regulation of genes related to the MAPK pathway.\nFigure 2 Increased EGR1 and FOSB gene expression in mutDFNA5-transfected HEK293T cells. RNA samples were collected from HEK293T cells transfected with either wtDFNA5 or mutDFNA5 and gene expression was measured by real-time rtPCR. Significantly increased expression was seen in mutDFNA5 at 12 h [p(egr1) = 0.000; p(fosB) = 0.006], 15 h [p(egr1) = 0.017; p(fosB) = 0.000] and 18 h [p(egr1) = 0.004; p(fosB) = 0.026] post-transfection. *p \u003c 0.05; **p \u003c 0.01; ***p \u003c 0.001. CNRQ, calibrated normalized relative quantities. After confirmation by real time rtPCR, the significance of the activated MAPK pathway was further validated by two independent experiments. To investigate the significance of the MAPK pathway, we wondered whether inhibition of the MAPK pathway would attenuate this mutDFNA5-induced growth defect.\nTherefore, a specific JNK inhibitor, namely SP600125, was added, followed by a viability assay to determine the effect on cell survival. Different concentrations of the JNK inhibitor SP600125 were used to measure viability by flow cytometry (CyFlow ML, Partec, Germany) and these results were compared to untreated mutDFNA5-transfected HEK293T cells. Overnight treatment of the cells with different concentrations of SP600125 did not have any major effect on transfection efficiency, but significantly increased the viability of mutDFNA5-transfected cells. Although, addition of 12.5 and 25 μM SP600125 both significantly increased the viability with a p-value of 0.020 and 0.004 respectively SP600125 had the greatest effect with a concentration of 25 μM SP600125 since the viability was raised from 31.93 to 51.00% (Figure 3).\nFigure 3 MAPK inhibitor effect on mutDFNA5 transfected HEK293T cells. MutDFNA5-transfected HEK293T cells were pretreated with different amounts of SP600125 (JNK inhibitor). Cell viability was measured and compared to untreated mutDFNA5-transfected HEK293T cells. *p \u003c 0.05; **p \u003c 0.01. Next, to evaluate the effect of MAPK up-regulation on protein level, different MAPK proteins were studied by western blotting. There are three main MAPK pathways in human cell lines represented by the ERK, JNK, and p38 MAPK branch. Consistent with the results obtained by real time rtPCR and the viability assay, activation of the MAPK pathway proteins was also demonstrated by western blotting. Total protein lysates were collected from HEK293T cells 12 h post-transfection. Three phosphorylated (activated) and non-phosphorylated (not activated) proteins of the MAPK pathway were studied using six different antibodies. No differences were seen in the expression level of non-phosphorylated ERK and JNK (Figure 4A). Activation of JNK and to a minor extent of ERK (p42/p44) was seen upon mutDFNA5 transfection compared to control and wtDFNA5 (Figure 4B). The expression of p38 was also evaluated but no difference in protein expression was observed between mutDFNA5 compared to wtDFNA5 and control (data not shown). β-Actin was used as a loading control.\nFigure 4 Activation of the MAPK pathway by mutDFNA5. (A) Western blot analysis of non- phosphorylated ERK (p42/p42) and JNK. No differences were seen in protein expression level of ERK and JNK between control, wtDFNA5 and mutDFNA5 transfected HEK293T cells. (B) Western blot analysis of phosphorylated, activated ERK (p42/p42) and JNK. Increased expression of JNK and to a lesser extent of ERK was seen in mutDFNA5 transfected HEK293T cells as compared to wtDFNA5 and control. These results suggest that DFNA5 induces PCD mediated through activation of the MAPK pathways. Addition of a MAPK inhibitor partially attenuated the mutDFNA5-induced growth defect identifying the MAPK pathway as an early event in mutDFNA5-associated cell death.\n\nComparison of the yeast microarray results with the gene expression in human cell lines\nIn order to study the significance of the yeast results, a comparison was made between the two microarray experiments. Human homologs of the significantly up- and down-regulated yeast genes at the post-diauxic shift (comparison 1b) were identified using Ensemble Biomart. Of the 451 significantly up- and 585 significantly down-regulated yeast genes, respectively 296 and 647 human homologs were identified. These specific human homologs were analyzed using the R package LIMMA, which generated a new list of human genes. The FC cut-off of the resulting gene list was set to FC 1.2, resulting in 16 up- and 14 down-regulated human genes (Supplemental Tables 8A,B). TM7SF2, UCP2, and VPS33B, three down-regulated human genes, were selected to verify the yeast results in human cell lines using real-time rtPCR. UCP2 and VPS33B were evaluated because they were the two most down-regulated genes present in the list. UCP2, an uncoupling protein, is a mitochondrial carrier located at the mitochondrial inner membrane. Suppression of UCP2 has been linked to increased ROS production (Deng et al., 2012; Dando et al., 2013) and lifespan regulation (Andrews and Horvath, 2009; Andrews, 2010). The Vacuolar protein sorting 33 homolog (VPS33B) gene is involved in intracellular vesicle Golgi-to-lysosome transport (Pevsner et al., 1996; Lo et al., 2005). TM7SF2 was selected based on its function in relation to the yeast results. TM7SF2 is a transmembrane protein present in the ER and associated with biosynthesis of cholesterol. In addition to its role in cholesterol synthesis, TM7SF2 appears to be involved in the inflammatory response upon cellular stress (Holmer et al., 1998; Bennati et al., 2006; Schiavoni et al., 2010; Bellezza et al., 2013).\nTo verify the yeast results in human cell lines, RNA was collected from HEK293T cells transfected with either wtDFNA5 or mutDFNA5 at different time-points starting at 12 h after transfection as this was the time-point of the human microarray experiment. The TM7SF2, UCP2, and VPS33B genes had a fold change of respectively 1.23, 1.31, and 1.27 on the microarray. Real-time rtPCR on RNA samples 12 h post-transfection confirmed these microarray results as all three genes were down-regulated in mutDFNA5 compared to wtDFNA5 (Figure 5A) with fold changes comparable to the microarray (respectively 1.35, 1.48, and 1.59). Although not significantly, these three genes were down-regulated in HEK293T cells at 12 h after transfection (Figure 5A). The down-regulation was still present at 20 h, peaked at 24 h after transfection and was even significant for TM7SF2 (p:0.01) and UCP2 (p:0.07) at respectively 20 h and 24 h after transfection (Figure 5B). Due to this down-regulation, we can conclude that there are some similarities between the yeast and the HEK293T microarray. Differences are seen when looking at the individual genes, but upon study of the different pathways a role for processes related to protein folding were seen in both model systems.\nFigure 5 Validation of the yeast microarray by real-time rtPCR in HEK293T cells. (A) Gene expression 12 h post-transfection in human HEK293T cells of three selected genes. (B) Gene expression 12, 20, and 24 h post-transfection with either wtDFNA5 or mutant DFNA5. UCP2, uncoupling protein 2 (mitochondrial, proton carrier); TM7SF2, transmembrane 7 superfamily member 2; VPS33B, vacuolar protein sorting 33 homolog. Light gray, wtDFNA5; Dark gray, mutDFNA5. *p \u003c 0.05. CNRQ, calibrated normalized relative quantities.\n\nDi"}