The Upstream 1350~1250 Nucleotide Sequences of the Human ENDOU-1 Gene Contain Critical Cis-Elements Responsible for Upregulating Its Transcription during ER Stress

ENDOU-1 encodes an endoribonuclease that overcomes the inhibitory upstream open reading frame (uORF)-trap at 5′-untranslated region (UTR) of the CHOP transcript, allowing the downstream coding sequence of CHOP be translated during endoplasmic reticulum (ER) stress. However, transcriptional control of ENDOU-1 remains enigmatic. To address this, we cloned an upstream 2.1 kb (−2055~+77 bp) of human ENDOU-1 (pE2.1p) fused with reporter luciferase (luc) cDNA. The promoter strength driven by pE2.1p was significantly upregulated in both pE2.1p-transfected cells and pE2.1p-injected zebrafish embryos treated with stress inducers. Comparing the luc activities driven by pE2.1p and −1125~+77 (pE1.2p) segments, we revealed that cis-elements located at the −2055~−1125 segment might play a critical role in ENDOU-1 upregulation during ER stress. Since bioinformatics analysis predicted many cis-elements clustered at the −1850~−1250, we further deconstructed this segment to generate pE2.1p-based derivatives lacking −1850~−1750, −1749~−1650, −1649~−1486, −1485~−1350 or −1350~−1250 segments. Quantification of promoter activities driven by these five internal deletion plasmids suggested a repressor binding element within the −1649~−1486 and an activator binding element within the −1350~−1250. Since luc activities driven by the −1649~−1486 were not significantly different between normal and stress conditions, we herein propose that the stress-inducible activator bound at the −1350~−1250 segment makes a major contribution to the increased expression of human ENDOU-1 upon ER stresses.


Introduction
The endoplasmic reticulum (ER) is a critical organelle involved in the synthesis and folding of proteins in order to maintain normal cell function and homeostasis [1][2][3].The Integrated Stress Response (ISR) is a cellular stress response that aims to restore cellular homeostasis upon different types of extrinsic or intrinsic stresses.The ISR can be stimulated by different environmental and pathological conditions, including nutrient deprivation, oxidative stress, protein homeostasis defects, and viral infection [4][5][6][7][8][9][10][11].The ISR can also be stimulated by intrinsic ER stress, which is caused by the accumulation of unfolded or misfolded proteins within the ER [12].The ISR is primarily a pro-survival homeostatic program, aiming to optimize adaptive cellular response to stress.The ISR restores cellular balance by reprogramming gene expressions in response to different environmental and pathological conditions, including oxidative stress, proteostasis, viral infection, and nutrient deprivation [13,14].However, once the stress cannot be removed, the ISR drives signaling toward cell death, fro example, through apoptosis-inducing factors which will act as a prodeath effector to trigger DNA cleavage and parthanatos [15].For example, dysregulated ISR signaling can be involved in cognitive disturbances, neurodegenerative diseases, tumor malignancies, diabetes, and metabolic imbalances [16].
The Eukaryotic ENODU ribonuclease family consists of several enzymes that share amino acid homology with XendoU [17,18].This ENDOU family is broadly conserved from viruses to humans.For example, Nsp15, a viral orthologue of XendoU, plays an essential role in coronavirus replication [19] and host immune suppression [20].Xenopus XendoU is a Uridylate-specific, divalent cation-dependent enzyme that produces molecules with 2 ,3cyclic phosphate ends, a unique characteristic of this particular class of RNases involved in generating U16 and U86 small nucleolar RNAs (snoRNAs) through the cleavage of pre-mRNAs encoded within the third intron of the L1 ribosomal protein gene [17,18,[21][22][23].Human PP11 has been recently categorized as a member of the XendoU family, i.e., ENDOU.Despite its annotated function as a putative serine protease, PP11/ENDOU is a placentalspecific endoribonuclease.Moreover, the dysregulated expression of PP11/ENDOU in tumor tissues may be associated with carcinogenesis [24].Nevertheless, Xenopus XendoU and human ENDOU-2 play similar roles in cellular processes, including the regulation of ER structure, RNA degradation, and cell survival [25,26].Moreover, C. elegans ENDOU-2 was reported to be involved in regulating cold tolerance and synaptic remodeling through caspase signaling [27].Mammalian ENDOU-1 mRNAs are rapidly increased in the presence of stresses, such as heat shock and hypoxia, subsequently inducing eIF2α phosphorylation, suggesting that ENDOU-1 protein may play a role in stress response and serve as a novel stress factor [28].However, control over the transcription of ENDOU-1 gene in response to ER stresses remains to be elucidated.
In general, the eukaryotic promoter can roughly be divided into two parts: proximal and distal.The proximal region of eukaryotic promoter is believed to be responsible for correctly assembling the RNA polymerase II complex at the right position, and for directing a basal level of transcription.The distal region is believed to contain cis-elements that regulate spatiotemporal expression.It is well known that 98% of DNA sequences are non-protein-coding regions in the human genome [29].These non-coding regions contain a variety of cis-regulatory elements which interact with transcription factors to orchestrate the expression of specific genes spatiotemporally.Therefore, identification of cis-acting and trans-acting elements is important to gain insight into the regulatory mechanisms controlling ENDOU-1 expression In this study, we first cloned an upstream 2.1 kb (−2055~+77 bp) of human ENDOU-1 fused with luciferase (luc) reporter.Next, we compared the quantification of luc expression activities driven by different deleted upstream segments during normal and stress condition.Evidence from both in vivo and in vitro systems suggests two indispensable cis-acting elements within the −1850~−1250 segment.These cis-acting elements are involved in controlling the transcription of −2.1-kb human ENDOU-1 during normal and stress conditions.One is a repressor binding motif located within the −1649~−1486 region, but the other is an activator binding motif located within the −1250~−1350 region of ENDOU-1.Since the −1649~−1486 segment was independent of upregulated ENDOU-1 at stress, while gene expression driven by the −1250~−1350 segment could be significantly enhanced after ER stress, we herein propose that the stress-inducible cis-element bound by an activator within the −1350~−1250 segment makes a major contribution to the increased expression of ENDOU-1 upon ER stress treatment.

Stress-Responsive Element Was Inclusive
To understand whether the upregulation of human ENDOU-1 in response to ER stress is controlled at transcriptional level, we aimed to clone the upstream regulatory region of human ENDOU-1 and determine if this regulatory region could be activated during ER stress.First, we cloned an upstream 2.1 kb (−2.1 kb) fragment of human ENDOU-1 and subcloned it into the expression vector containing firefly-luc reporter (Flu) to generate pE2.1p plasmid.The dual-luc activity was performed in HEK-293T cells, in which pE2.1p (100 ng) was co-transfected with phRG-TK (20 ng, containing Renilla-luc reporter (Rlu), driven by thymidine kinase (TK) promoter which served as an internal control) for 24 h and treated with either DMSO (control group) or Thapsigargin (TH; stress group) for 2 or 6 h, followed by analysis of luc activity.Compared to DMSO-treated control cells, the luc activity driven by pE2.1p was increased 2.1-fold in cells treated with TH for 2 h (Figure 1A).Moreover, the stress-induced effect on the strength of ENDOU-1 promoter positively correlated with the duration of stress treatment.For example, luc activity was increased 3.1-fold after TH treatment for 6 h (Figure 1A).
stress is controlled at transcriptional level, we aimed to clone the upstream reg region of human ENDOU-1 and determine if this regulatory region could be ac during ER stress.First, we cloned an upstream 2.1 kb (−2.1 kb) fragment of hum DOU-1 and subcloned it into the expression vector containing firefly-luc reporter generate pE2.1p plasmid.The dual-luc activity was performed in HEK-293T c which pE2.1p (100 ng) was co-transfected with phRG-TK (20 ng, containing Ren reporter (Rlu), driven by thymidine kinase (TK) promoter which served as an i control) for 24 h and treated with either DMSO (control group) or Thapsigargi stress group) for 2 or 6 h, followed by analysis of luc activity.Compared to D treated control cells, the luc activity driven by pE2.1p was increased 2.1-fold treated with TH for 2 h (Figure 1A).Moreover, the stress-induced effect on the stre ENDOU-1 promoter positively correlated with the duration of stress treatment.For ple, luc activity was increased 3.1-fold after TH treatment for 6 h (Figure 1A).Values re ing DMSO-and TH-treated cells were calculated from four independent experiments and pr as mean ±SD (n = 4).Student's t-test was used to determine significant differences betwe group (**, p < 0.05; ***, p < 0.005).(B).Histograms presenting the relative luc activity obtain zebrafish embryonic cells microinjected at the one-cell stage with pE2.1p (6.9 pg/embryo) co with phRG-TK (2.3 pg/embryo), followed by heat shock at 72 hpf and luc quantification at Zebrafish embryos treated with microinjection, but not treated with heat shock, served as the group.The relative luc activity was represented by the fold increase in the Flu/Rlu ratio o obtained from the control group normalized as 1.Each value was averaged from three indep experiments and presented as mean ± SD (n = 3).Student's t-test was used to determine sig differences between each group (***, p < 0.005).Histograms presenting the relative luc activity obtained from HEK-293T cells co-transfected with pE2.1p (100 ng/well) and phRG-TK (20 ng/well) and then treated with either DMSO (control) or Thapsigargin (TH; stress inducer) for 2 or 6 h.HEK-293T cells treated with DMSO served as the control group.The relative luc activity was represented by the fold increase of Flu/Rlu ratio over that obtained from the control group normalized as 1.Values representing DMSO-and TH-treated cells were calculated from four independent experiments and presented as mean ±SD (n = 4).Student's t-test was used to determine significant differences between each group (**, p < 0.05; ***, p < 0.005).(B).Histograms presenting the relative luc activity obtained from zebrafish embryonic cells microinjected at the one-cell stage with pE2.1p (6.9 pg/embryo) combined with phRG-TK (2.3 pg/embryo), followed by heat shock at 72 hpf and luc quantification at 96 hpf.Zebrafish embryos treated with microinjection, but not treated with heat shock, served as the control group.The relative luc activity was represented by the fold increase in the Flu/Rlu ratio over that obtained from the control group normalized as 1.Each value was averaged from three independent experiments and presented as mean ± SD (n = 3).Student's t-test was used to determine significant differences between each group (***, p < 0.005).
Next, we employed zebrafish embryos to perform an in vivo study.Zebrafish embryos at the one-cell stage were simultaneously microinjected with a mixture of pE2.1p and phRG-TK and treated with heat shock at 72 h post-fertilization (hpf).We found that luc activity increased ~2.07-fold compared to that of embryos kept at a normal condition (28.5 • C) during 96 hpf (Figure 1B).Based on consistent results obtained from both in vitro and in vivo systems, we concluded that an upstream 2.1 kb segment of human ENDOU-1 is suitable for conducting a detailed investigation of cis-elements potentially involved in the transcription of human ENDOU-1 after the induction of stress.

Reporter Gene Expression Driven by the Upstream 2.1-kb of Human ENDOU-1 Could Be Upregulated When the Stress Inducers Were Applied to Cells
We examined whether the upstream 2.1 kb fragment of ENDOU-1 was responsive to various stress-inducing drugs, as described below.Dual-luc activity was assessed in HEK-293T cells, in which pE2.1p (100 ng) was co-transfected with phRG-TK (20 ng) for 24 h and treated with DMSO (control group), TH, Anisomycin, CoCl 2 or Poly I:C for 4 h, followed by analysis of luc activity.
TH is a non-competitive inhibitor of the sarco/endoplasmic reticulum Ca 2+ ATPase (SERCA).Depletion of ER calcium stores leads to ER stress and activation of the unfolded protein response (UPR).Anisomycin is a potent protein synthesis inhibitor, which interferes with protein and DNA synthesis by inhibiting peptidyl transferase or the 80S ribosome system, resulting in ribotoxic stress.CoCl 2 treatment mimics hypoxic stress, since CoCl 2 inhibits prolyl hydroxylase domain (PHD) enzymes (oxygen sensors) through replacement of Fe 2+ with Co 2+ , making these enzymes unable to hydroxylate hypoxia-inducible factor-1 α (HIF-1α), thus activating HIF signaling, resulting in the induction of a hypoxia-like condition [30].Poly I:C is a synthetic double-stranded RNA that is used experimentally to model viral infections in vivo.Copper (Cu), a transition metal, is an essential trace element in human and animal nutrition at low concentration.However, Cu 2+ at high concentration can induce ER stress and promote apoptosis, since it activates the CHOP, JNK and Caspase-12 signaling pathways [31].
As shown in Figure 2, we found that HEK-293T cells treated with all stress-inducing drugs could increase promoter activity driven by pE2.1p (Figure 2A).Western blot analysis demonstrated that the expression levels of CHOP and p-eIF2α, two important markers for ER stress [28], were induced significantly (Figure 2B), indicating that the stress-inducing drugs used in this study caused the response of HEK-293T cells to stress.Interestingly, we noticed that Poly I:C could strongly induce ENDOU-1 promoter activity, whereas Rapamycin could only slightly induce it (Figure 2A), suggesting that promoter strength driven by the −2055~+77 DNA fragment of the human ENDOU-1 gene is dependent on the stress inducer applied to cells, in this case HEK-293T.

Computational Analysis Proposes That the Distal Region of the −2.1 kb ENDOU-1 Fragment May Contain Critical Regulatory Elements
Next, we wanted to identify cis-regulatory responsive elements (CREs) located at the −2250~+77 DNA fragment of human ENDOU-1 that might play a role in upregulating ENDOU-1 transcripts during ER stress.To accomplish this, we employed Tfsitescan and PROMO to analyze potential transcription factor binding sites (TFBSs).Owing to the limited number of base pairs per scan, we divided the −2.1 kb fragments into two parts with an overlap of 90 bp.Thus, the −2055~−975 fragment was scanned first, followed by scanning the −1065~00 fragment.The results revealed that the distal region of the −2.1 kb fragment (the −2055~−975 fragment) contains more potential TFBSs of relative significance than those contained in the proximal region (the −1065~00 fragment) (Table S1).We further asked if important regulatory elements were harbored within the −2055~−975 region.Results showed that this region contained five putative transcription factors (Table S1), including nuclear factor Y (NF-Y), Yin Yang 1 (YY1), activator protein-1 (AP-1), c-Erythroblast transformation specific-1 (c-Ets-1), and nuclear factor-κB (NF-κB), all previously reported as critical CREs in the promoter regions of other ER stress-induced genes [32][33][34][35][36][37][38][39].Therefore, we hypothesized that regulatory elements contained in the −2055~−975 region may be critical for controlling the upregulation of ENDOU-1 in response to ER stress.

Computational Analysis Proposes That the Distal Region of the −2.1 kb ENDOU-1 Fragment May Contain Critical Regulatory Elements
Next, we wanted to identify cis-regulatory responsive elements (CREs) located at the −2250~+77 DNA fragment of human ENDOU-1 that might play a role in upregulating EN-DOU-1 transcripts during ER stress.To accomplish this, we employed Tfsitescan and PROMO to analyze potential transcription factor binding sites (TFBSs).Owing to the limited number of base pairs per scan, we divided the −2.1 kb fragments into two parts with an overlap of 90 bp.Thus, the −2055~−975 fragment was scanned first, followed by scanning the −1065~00 fragment.The results revealed that the distal region of the −2.1 kb fragment (the −2055~−975 fragment) contains more potential TFBSs of relative significance than those contained in the proximal region (the −1065~00 fragment) (Table S1).We further asked if important regulatory elements were harbored within the −2055~−975 region.Results showed that this region contained five putative transcription factors (Table S1), including nuclear factor Y (NF-Y), Yin Yang 1 (YY1), activator protein-1 (AP-1), c-Erythroblast transformation specific-1 (c-Ets-1), and nuclear factor-κB (NF-κB), all previously re-

The −2055~−1125 Segment of Human ENDOU-1 Contains Cis-Acting Elements Involved in Increasing Promoter Activity under Stress Conditions
Focusing on the distal region of the −2.1 kb fragment of ENDOU-1, we continued to seek to identify cis-regulatory elements capable of upregulating the expression of ENDOU-1 during stress.To accomplish this, we constructed serial deletion clones from the upstream fragment of ENDOU-1 fused with luc reporter cDNA.Based on the restriction enzyme map, we found a XhoI-cutting site at the −1125 nucleotide (nt) position of ENDOU-1.Taking advantage of this cutting site, XhoI was used to cut the −2250~+77 DNA fragment to obtain a resultant −1125~+77 fragment.This shortened DNA fragment was then inserted into the pGL3-basic vector to generate a plasmid pE1.2pcontaining this −1125~+77 fragment, allowing us to determine whether the increased ENDOU-1 expression during ER stress is independent of the −1125~+77 segment.In order to compare the luc activities driven by the −2055~+77 (pE2.1p) and −1125~+77 (pE1.2p)segments, we first co-transfected HEK-293T cells with pE1.2p (containing a −1125~+77 segment) and phRG-TK and then treated with TH.The results exhibited a ~0.77-fold decrease compared to that of cells treated with DMSO control (Figure 3A), suggesting that the luc activities did not display much difference among pE1.2p-injectedcells, TH-treated cells, and untreated cells.However, the luc activity of TH-treated HEK-293T co-transfected with pE2.1p (containing a −2055~+77 segment) and phRG-TK exhibited a ~2.36-fold increase (Figure 3A), suggesting that the cis-elements located at −2055~−1125 region might play a critical role in ENDOU-1 upregulation during ER stress.
then inserted into the pGL3-basic vector to generate a plasmid pE1.2pcontaining this −1125~+77 fragment, allowing us to determine whether the increased ENDOU-1 expres sion during ER stress is independent of the −1125~+77 segment.In order to compare the luc activities driven by the −2055~+77 (pE2.1p) and −1125~+77 (pE1.2p)segments, we firs co-transfected HEK-293T cells with pE1.2p (containing a −1125~+77 segment) and phRG TK and then treated with TH.The results exhibited a ~0.77-fold decrease compared to tha of cells treated with DMSO control (Figure 3A), suggesting that the luc activities did no display much difference among pE1.2p-injectedcells, TH-treated cells, and untreated cells.However, the luc activity of TH-treated HEK-293T co-transfected with pE2.1p (con taining a −2055~+77 segment) and phRG-TK exhibited a ~2.36-fold increase (Figure 3A) suggesting that the cis-elements located at −2055~−1125 region might play a critical role in ENDOU-1 upregulation during ER stress.Histograms presenting the relative luc activity obtained from zebrafish embryos microinjected at the one-cell stage with either pE2.1p (containing −2055~+77 fragment; 6.9 pg/embryo) combined with phRG-TK (2.3 pg/embryo) (above) or pE1.2p (containing −1125~+77 fragment; 6.9 pg/embryo) combined with phRG-TK (2.3 pg/embryo) (bottom), followed by no treatment (untreated; empty) or heat shock (solid) at 72 hpf and detection of luc activity at 96 hpf.Microinjected zebrafish embryos not exposed to heat shock served as the untreated control group.Relative luc activity was represented by the fold change of Flu/Rlu ratio over that obtained from the control group normalized as 1.Values representing untreated control and heat-shock-treated embryos were calculated from three independent experiments and presented as mean ± SD (n = 3).Student's t-test was used to determine significant differences between pE2.1pand pE1.2p-injected groups (***, p < 0.001; *, p < 0.05).
To further confirm the above in vitro finding, the dual-luc assay was also performed in zebrafish embryos to explore any difference between luc activities driven by pE2.1p and pE1.2p segments in vivo.The results demonstrated that luc activities did not display much difference (around 0.56-fold decrease) among pE1.2p-injectedembryos, THtreated embryos and untreated control embryos (Figure 3B).However, pE2.1p-injected zebrafish embryos treated with heat shock displayed ~2.29-fold increase in luc activity compared to that of embryos without TH treatment (Figure 3B).The consistency between in vitro and in vivo evidence strongly suggests that the upstream fragment of human ENDOU-1 that enables an increase in promoter activity under stress condition is independent of the −1125~+77 region.In other words, the cis-acting elements contained in the −2055~−1125 segment are responsible for the response of ENDOU-1 to ER stress.

Two Indispensable Cis-Acting Elements within the −1850~−1250 Segment Controlled the Expression of −2.1-kb Human ENDOU-1 under Normal and Stress Conditions
The above results demonstrated that the −2055~−1125 fragment may contain ciselements responsible for controlling the transcription of ENDOU-1 during stress.Nevertheless, we employed Tfsitescan and PROMO to analyze potential TFBSs, and based on this analysis, many TFs predicted to target the ENDOU-1 promoter region were clustered at the −1850~−1250 region.Therefore, to further identify cis-regulatory elements involved in controlling the transcription of ENDOU-1, we employed a PCR-based internal deletion strategy to deconstruct the −1850~−1250 region of pE2.1p in more detail.To accomplish this, we constructed five internal deletion clones, including −1850~−1750, −1749~−1650, −1649~−1486, −1485~−1350 and −1350~−1250, and their promoter activities were analyzed, both in vitro and in vivo.According to our results, reporter activity driven by the −1649~−1486-deleted-pE2.1p exhibited a 4.11-fold increase over that of pE2.1p during normal condition (Figure 4A).However, unexpectedly, the reporter activity driven by this same deleted segment still exhibited about a 4-fold increase over that of pE2.1p under stress conditions (Figure 4A).Unlike pE2.1p and its deleted derivatives, including −1850~−1750, −1749~−1650 and −1485~−1350, which could significantly upregulate luc activity under stress, this −1649~−1486-deleted-pE2.1p fragment failed to induce reporter activity at higher levels under stress conditions.
Moreover, the results obtained from the above in vitro system were consistent with those obtained from our in vivo system.For example, when this −1649~−1486-deleted-pE2.1p fragment was injected into zebrafish embryos, the luc activity displayed a 5.9-fold increase over that of pE2.1p-injected embryos under normal conditions (Figure 4B).Again, the luc activity of the −1649~−1486-deleted-pE2.1p-injected embryos was maintained around a 5.4-fold increase, even after the induction of stress (Figure 4B).Consistent with the in vitro study, this −1649~−1486-deleted-pE2.1p was independent of further inducing luc expression to reach a higher level under stress.Therefore, to explain the failure of the −1649~−1486-deleted-pE2.1p segment to further induce luc expression and reach a higher level under stress conditions, as confirmed in both in vitro and in vivo studies, we looked to the repressor binding motif.Thus, we speculate that repressor bound at −1649~−1486 may block transcription under normal conditions, but exit the element exposed under stress conditions, resulting in the −1649~−1486-deleted-pE2.1p being unable to induce luc expression to reach a higher level.
To better understand a phenomenon whereby the repressor binds to an element under normal conditions, but exits during stress, we cite a convenient analogy.Inhibitors of DNAbinding (ID) proteins are a class of helix-loop-helix (HLH) transcription factors that serve as repressors to act as dominant negative antagonists of other basic HLH proteins through the formation of non-functional heterodimers [40].During ER stress, the activation of Inositol requires enzyme 1α (IRE1α) to reduce ID1 level through degradation of Id1 mRNA [41].Id-1 is thus downregulated in hypoxic cells via transcriptional repression mediated by ATF-3, and the effects would be expected to promote cell growth arrest [42].To more explicitly make the comparison, we speculate that the repressor bound at −1649~−1486 may block transcription under normal condition, but exit the element exposed during stress conditions, resulting in −1649~−1486-deleted-pE2.1p being able to induce luc expression to reach a higher level at under normal conditions (while unable to induce expression under stress conditions).The case of HLH transcription factors is analogous, because transcriptional repressor Id-1 is expressed under normal conditions, but reduced under stress conditions.Moreover, the results obtained from the above in vitro system were consistent with those obtained from our in vivo system.For example, when this −1649~−1486-deleted-pE2.1p fragment was injected into zebrafish embryos, the luc activity displayed a 5.9-fold increase over that of pE2.1p-injected embryos under normal conditions (Figure 4B).Again, Histograms presenting the relative luc activity obtained from zebrafish embryos microinjected at the one cell stage with either indicated plasmids (6.9 pg/embryo) and phRG-TK (2.3 pg/embryo) followed by heat-shock treatment at 72 hpf and luminescence detection at 96 hpf.Microinjected zebrafish embryos not exposed to heat shock served as the control group.The relative luc activity was represented by the fold change in the Flu/Rlu ratio over that obtained from the control group normalized as 1.Values representing control and heat-shock-treated embryos were averaged from three independent experiments.A one-way ANOVA, followed by Tukey's multiple comparison test, was used to perform statistical analysis (***, p < 0.001; error bars indicate mean ± SD).Additionally, it is theoretically possible that the proposed repressor might have dual functions.That is, it might repress transcription under normal conditions, but induce transcription under stress conditions by the recruitment of as-yet unknown cofactors.To clarify this hypothetical, we cite the activation of NF-κB to maintain myogenic cells in an undifferentiated state, whereas it acts as a repressor during myocyte differentiation [43,44].In another example, the role of Myc protein in the transcriptional activation and repression of target genes may depend on the unique factors it recruits, such as the constellation of other transcription factors proximal to the binding site, or even the chromatin context of the target gene [45][46][47].
Meanwhile, we noticed that the −1350~−1250 segment was also important for controlling ENDOU-1 promoter transcription, since we found that luc activity driven by −1350~−1250-deleted-pE2.1p was substantially reduced both under normal and stress conditions (Figure 4A).Similar results were also obtained from in vivo experiment.For example, Zebrafish embryos injected with −1350~−1250-deleted-pE2.1p showed a significant decrease in luc activity, exhibiting a 0.52 and 0.53-fold decrease compared to pE2.1p-injected embryos under normal and stressed conditions, respectively (Figure 4B).This evidence suggested that the −1350~−1250 segment might contain an enhancer or activator-binding motif which facilitates the upregulated transcription of human ENDOU-1 promoter.
Some putative activator-binding sites within the −1350~−1250 segment were also predicted (Figure S1).For example, we found binding sites for (1) YY1, which has been reported to increase the promoter activity of the GRP78/BiP gene, a prosurvival ER chaperone, during ER stress [32]; and (2) transcription factor c-Est-1, which can be upregulated by the IRE1α/XBP1 signaling.The increased c-Ets-1 can activate the Mcl-1 promoter in human melanoma cells against apoptosis upon ER stress [34].We also found binding sites for (3) the activator protein-1 (AP-1), which is up-regulated under ER stress conditions.Many studies reported that AP-1 is required for CHOP induction through binding to the CHOP promoter in response to various stresses [36,52,53].Moreover, Klymenko et al. [37] identified AP-1 (−246 to −240) and c-Ets-1 (−205 to −199) binding sites within the CHOP promoter, which are necessary for ER-stress-induced CHOP gene activation.Yet, it is important to note that when using bioinformatic tools to analyze the cis-regulatory elements within a promoter, any potential TFBSs may be shown.However, due to the short and variable nature of TFBSs, it is highly plausible that false-positive elements are included [54].Thus, after the number of putative cis-acting elements for studying is narrowed down through bioinformatic analysis, we should be able to firmly confirm that these predicted cis-acting elements do indeed have the ability to cause upregulation of reporter gene expression; an additional promoter activity assay must therefore be performed.
In this study, based on bioinformatics tools and deletion analysis of promoter activity, we conclude that the −1350~−1250 DNA fragment contains an activator-binding motif which is critical in upregulating the transcription of the human ENDOU-1 promoter under stress condition.AP-1 and c-Ets-1, as noted above, are two known activators bound within the −1350~−1250 region to control CHOP induction.These TFs are, moreover, upregulated under ER stress.Therefore, it is reasonable to speculate the presence of a signaling cascade in which AP-1 and c-Ets-1 are engaged in upregulating ENDOU-1 expression, as an intermediate step, leading to an increase in CHOP translation mediated by the increased ENDOU-1 through overriding the inhibitory uORF-trap at 5 -UTR of CHOP mRNA, thereby allowing the translation of downstream CHOP mRNA during ER stress.Needless to say, such an AP-1/c-Ets-1/ENDOU-1/CHOP axis would require further study and confirmation.Apart from the presumed AP-1/c-Ets-1 activators, it is possible that other novel activators might similarly interact with this stress-inducible motif, thereby enabling the control of ENDOU-1 expression.Further studies to validate these theories could include an electrophoretic mobility shift assay on gel, deletion analysis, mutagenesis experiments, or even yeast one-hybrid screening, as described by [55,56].
Lastly, taking advantage of using bioinformatics could lead to accelerate the discovery of novel promoters and their cognate cis-elements of genes.However, how to evaluate variants based on in vitro and in vivo experimental platforms, rule out false positives, and confirm true regulatory motifs is still a major challenge for researchers who are working on promoter analysis.In this study, we used the human ENDOU-1 gene as an example to demonstrate how to identify cis-elements that importantly impact ENDOU-1 transcription among hundreds of TFBSs predicted by bioinformatics.Our results do contribute to computer algorithms with better prediction of cis-elements; they will help to circumvent false-positive elements and efficiently predict cognate transcriptional regulators.Moreover, our studies may contribute to future applications like the design of next-generation of ENDOU gene therapies against head and neck squamous cell carcinoma affected by environmental factors [57].

Zebrafish Husbandry and Microscopy
The zebrafish AB strain was cultured as previously described [58].hpf is defined as normalized hours after fertilization at 28.5 • C, an optimal temperature of culturing.To administer heat-shock treatment to embryos, we followed the protocol described by Lee et al. [28].Briefly, a 2 mL centrifuge tube filled with 30 dechorionated zebrafish embryos at 72 hpf was subjected to 40 • C for 1 h.Then, the treated embryos were collected into a 3 cm Petri dish and incubated at 28.5 • C until they developed at 96 hpf.
The experiments and treatments of the zebrafish model have been reviewed and approved by the Fu Jen Catholic University and MacKay Medical College Institutional Animal Care and Use Committee with ethics approval numbers A11065 and A1040008, respectively.No specific ethics approval was required for this project as all zebrafish used in this study were between 0 to 5 days post-fertilization.Given the age of the embryos, pain perception has not yet developed at these earlier stages; therefore, this is not considered a painful procedure.

Dual-Luc Assay
The dual-luc assay performed on cells [59] and zebrafish embryos [60] has previously been described, except a 5 ng/µL of phRG-TK (internal control) and puORF chop -luc reporter construct were used.For in vitro studies, we did not assay the luc activity longer than 6 h; this was based on previous studies published by Lee et al. [28], which demonstrated that CHOP protein was first detected after stress treatment for 2 h, while the CHOP protein reached a relatively high level after stress treatment for 6 h.For each experiment in zebrafish, embryos at 96 hpf were collected and divided into three groups to carry out in the vivo Dual-luc assay (n = 60).

Statistics and Reproducibility
All experiments were performed in triplicate, and data were averaged from three independent experiments and presented as mean ± SD.A one-way ANOVA, followed by Tukey's multiple comparisons test or Student's t-test for comparisons was used to perform statistical analysis.Three levels of significance at * (p < 0.05), ** (p < 0.005), and *** (p < 0.001) were determined.

Figure 1 .
Figure 1.Effect of the −2055~+77 DNA fragment of human ENDOU-1 gene on reporter g pression in stress conditions.(A).Histograms presenting the relative luc activity obtain HEK-293T cells co-transfected with pE2.1p (100 ng/well) and phRG-TK (20 ng/well) and then with either DMSO (control) or Thapsigargin (TH; stress inducer) for 2 or 6 h.HEK-293T cells with DMSO served as the control group.The relative luc activity was represented by the crease of Flu/Rlu ratio over that obtained from the control group normalized as 1.Values re ing DMSO-and TH-treated cells were calculated from four independent experiments and pr as mean ±SD (n = 4).Student's t-test was used to determine significant differences betwe group (**, p < 0.05; ***, p < 0.005).(B).Histograms presenting the relative luc activity obtain zebrafish embryonic cells microinjected at the one-cell stage with pE2.1p (6.9 pg/embryo) co with phRG-TK (2.3 pg/embryo), followed by heat shock at 72 hpf and luc quantification at Zebrafish embryos treated with microinjection, but not treated with heat shock, served as the group.The relative luc activity was represented by the fold increase in the Flu/Rlu ratio o obtained from the control group normalized as 1.Each value was averaged from three indep experiments and presented as mean ± SD (n = 3).Student's t-test was used to determine sig differences between each group (***, p < 0.005).

Figure 1 .
Figure 1.Effect of the −2055~+77 DNA fragment of human ENDOU-1 gene on reporter gene expression in stress conditions.(A).Histograms presenting the relative luc activity obtained from HEK-293T cells co-transfected with pE2.1p (100 ng/well) and phRG-TK (20 ng/well) and then treated with either DMSO (control) or Thapsigargin (TH; stress inducer) for 2 or 6 h.HEK-293T cells treated with DMSO served as the control group.The relative luc activity was represented by the fold increase of Flu/Rlu ratio over that obtained from the control group normalized as 1.Values representing DMSO-and TH-treated cells were calculated from four independent experiments and presented as mean ±SD (n = 4).Student's t-test was used to determine significant differences between each group (**, p < 0.05; ***, p < 0.005).(B).Histograms presenting the relative luc activity obtained from zebrafish embryonic cells microinjected at the one-cell stage with pE2.1p (6.9 pg/embryo) combined with phRG-TK (2.3 pg/embryo), followed by heat shock at 72 hpf and luc quantification at 96 hpf.Zebrafish embryos treated with microinjection, but not treated with heat shock, served as the control group.The relative luc activity was represented by the fold increase in the Flu/Rlu ratio over that obtained from the control group normalized as 1.Each value was averaged from three independent experiments and presented as mean ± SD (n = 3).Student's t-test was used to determine significant differences between each group (***, p < 0.005).

Figure 2 .
Figure 2. Promoter activity driven by the −2055~+77 DNA fragment of human ENDOU-1 could be induced by various stress-inducing drugs.(A) HEK-293T cells co-transfected with pE2.1p and phRG-TK plasmid for 24 h and then treated with DMSO, Thapsigargin (TH), Anisomycin (Am), CoCl2 (Co), Poly I:C (IC), or CuSO4 (Cu) for 6 h, followed by analysis of luc activity via a dual-luc assay.Cells treated with DMSO served as a control group.The relative luc activity was represented by the fold increase in the Flu/Rlu ratio over that obtained from the pCS2-transfected control group, normalized as 1.The values were averaged from three independent experiments and presented as mean ± S.D. (n = 3).The Student's t-test was used to perform statistical analysis (***, p < 0.001; **, p < 0.005) (B) Western blot analysis of proteins, as indicated, in HEK-293T cells.The α-tubulin served as internal control.

Figure 2 .
Figure 2. Promoter activity driven by the −2055~+77 DNA fragment of human ENDOU-1 could be induced by various stress-inducing drugs.(A) HEK-293T cells co-transfected with pE2.1p and phRG-TK plasmid for 24 h and then treated with DMSO, Thapsigargin (TH), Anisomycin (Am), CoCl 2 (Co), Poly I:C (IC), or CuSO 4 (Cu) for 6 h, followed by analysis of luc activity via a dual-luc assay.Cells treated with DMSO served as a control group.The relative luc activity was represented by the fold increase in the Flu/Rlu ratio over that obtained from the pCS2-transfected control group, normalized as 1.The values were averaged from three independent experiments and presented as mean ± S.D. (n = 3).The Student's t-test was used to perform statistical analysis (***, p < 0.001; **, p < 0.005) (B) Western blot analysis of proteins, as indicated, in HEK-293T cells.The α-tubulin served as internal control.

Figure 4 .
Figure 4. Comparison of the promoter activity of ENDOU-1 driven by different internal deletion clones from −2.1 kb of the ENDOU-1 fragment.(A) In vitro system.Histograms presented the relative luc activity obtained from HEK-293T cells cotransfected either phRG-TK (20 ng/well) combined with pE2.1p or phRG-TK (20 ng/well) combined with different internal deletions (100 ng/well), as indicated on the left side.For example, plasmid pE2.1p-d1850/1750contained a −2.1 kb fragment of ENDOU-1, but deletion of −1850~−1750 nucleotides.The luc activities obtained from control and five deletion clones, as indicated, were quantified both during normal (DMSO) (represented by empty box) and Thapsigargin (TH)-induced stress condition (represented by solid box).HEK-293T cells treated with DMSO served as control group.(B) In vivo system.Histograms presenting the relative luc activity obtained from zebrafish embryos microinjected at the one cell stage with either indicated plasmids (6.9 pg/embryo) and phRG-TK (2.3 pg/embryo) followed by heat-shock treatment at 72 hpf and luminescence detection at 96 hpf.Microinjected zebrafish embryos not exposed to heat shock served as the control group.The relative luc activity was represented by the fold change in the Flu/Rlu ratio over that obtained from the control group normalized as 1.Values representing control and heat-shock-treated embryos were averaged from three independent experiments.A oneway ANOVA, followed by Tukey's multiple comparison test, was used to perform statistical analysis (***, p < 0.001; error bars indicate mean ± SD).

Figure 4 .
Figure 4. Comparison of the promoter activity of ENDOU-1 driven by different internal deletion clones from −2.1 kb of the ENDOU-1 fragment.(A) In vitro system.Histograms presented the relative luc activity obtained from HEK-293T cells cotransfected either phRG-TK (20 ng/well) combined with pE2.1p or phRG-TK (20 ng/well) combined with different internal deletions (100 ng/well), as indicated on the left side.For example, plasmid pE2.1p-d1850/1750contained a −2.1 kb fragment of ENDOU-1, but deletion of −1850~−1750 nucleotides.The luc activities obtained from control and five deletion clones, as indicated, were quantified both during normal (DMSO) (represented by empty box) and Thapsigargin (TH)-induced stress condition (represented by solid box).HEK-293T cells treated with DMSO served as control group.(B) In vivo system.Histograms presenting the relative luc activity obtained from zebrafish embryos microinjected at the one cell stage with either indicated plasmids (6.9 pg/embryo) and phRG-TK (2.3 pg/embryo) followed by heat-shock treatment at 72 hpf and luminescence detection at 96 hpf.Microinjected zebrafish embryos not exposed to heat shock served as the control group.The relative luc activity was represented by the fold change in the Flu/Rlu ratio over that obtained from the control group normalized as 1.Values representing control and heat-shock-treated embryos were averaged from three independent experiments.A one-way ANOVA, followed by Tukey's multiple comparison test, was used to perform statistical analysis (***, p < 0.001; error bars indicate mean ± SD).