Myocyte enhancer factor‐2 and p300 interact to regulate the expression of homeostatic regulator Pumilio in Drosophila

Abstract Pumilio (Pum), an RNA‐binding protein, is a key component of neuron firing‐rate homeostasis that likely maintains stability of neural circuit activity in all animals, from flies to mammals. While Pum is ubiquitously expressed, we understand little about how synaptic excitation regulates its expression in the CNS. Here, we characterized the Drosophila dpum promoter and identified multiple myocyte enhancer factor‐2 (Mef2)‐binding elements. We cloned 12 dmef2 splice variants and used a luciferase‐based assay to monitor dpum promoter activity. While all 12 dMef2 splice variants enhance dpum promoter activity, exon 10‐containing variants induce greater transactivation. Previous work shows dPum expression increases with synaptic excitation. However, we observe no change in dmef2 transcript in larval CNS, of both sexes, exposed to the proconvulsant picrotoxin. The lack of activity dependence is indicative of additional regulation. We identified p300 as a potential candidate. We show that by binding to dMef2, p300 represses dpum transactivation. Significantly, p300 transcript is downregulated by enhanced synaptic excitation (picrotoxin) which, in turn, increases transcription of dpum through derepression of dMef2. These results advance our understanding of dpum by showing the activity‐dependent expression is regulated by an interaction between p300 and dMef2.

Pumilio binds an eight nucleotide sequence in mRNA (UGUANAUA, where N = A, G, C or U), termed a Pum Response Element (PRE) and, by doing so, induces translational repression (Arvola, Weidmann, Tanaka Hall, & Goldstrohm, 2017;Wharton, Sonoda, Lee, Patterson, & Murata, 1998;Wreden, Verrotti, Schisa, Lieberfarb, & Strickland, 1997). Pum-dependent translational repression requires a number of coregulators, including Nanos (Nos) and brain tumour (Brat), which bind different, but equally characterized, RNA motifs to form a complex with Pum (Arvola et al., 2017). An analysis of 3'UTRs in the Drosophila genome identified 2477 transcripts containing one or more PREs highlighting the possibility that many transcripts undergo Pum-mediated translational regulation. The number of transcripts regulated may, however, be considerably less because specificity is also likely provided by both PRE copynumber and proximity of PRE-, Nos-and Brat-binding motifs within individual transcripts (Arvola et al., 2017).
The number of transcripts expressing PREs underscores the importance of Pum. Despite this, however, our understanding of pum expression and role(s) is limited and, where information is known, is mostly focused on post-transcriptional modification. For example, the dpum transcript is itself regulated through translational repression by the cytoplasmic RNA-binding Fox protein (Rbfox1, aka A2BP1) in order to promote germ cell development (Carreira-Rosario et al., 2016). In mammals, myocyte enhancer factor-2 (Mef2) regulates the expression of miR-134 which, in turn, downregulates pum2 transcript to fine-tune dendrite morphogenesis (Fiore et al., 2009(Fiore et al., , 2014. In mammals, Mef2 is an activitydependent transcription factor that has been implicated to control synapse formation in addition to dendrite morphogenesis (Flavell et al., 2006). Depending on interaction with either positive or negative cofactors, Mef2 can potentiate or repress gene transcription. For example, through an interaction with GATA4, a cardiac-enriched transcription factor, Mef2 activates the Nppa promoter to regulate cardiac development (Morin, Charron, Robitaille, & Nemer, 2000). By contrast, Mef2 forms a complex with class II histone deacetylases (HDACs) to repress gene transcription by deacetylating histones, resulting in chromatin condensation and a reduced accessibility of core transcriptional machinery to promoter regions of target genes (Kao et al., 2001;Lu, McKinsey, Zhang, & Olson, 2000;McKinsey, Zhang, & Olson, 2001).
To identify how transcription of pum is regulated, we cloned the promoter region of dpum and identified putative binding motifs for 114 transcription factors, including multiple dMef2 elements. A luciferase-based reporter, driven by the dpum promoter, shows that dMef2 is sufficient to transactivate the dpum promoter. The magnitude of transactivation varies across the many dMef2 splice variants present in Drosophila CNS. Significantly, we also report that dMef2mediated transactivation of dpum is repressed by p300 (aka Nejire), a histone acetyltransferase (HAT). Unlike dMef2, we show that p300 expression is directly regulated by neuronal activity and, thus, provide a potential route through which membrane depolarization regulates the expression level of dpum.

| Amplification of dmef2 splice variants and p300
Total RNA was extracted from the third instar CNS of Canton-S (mixed sexes). cDNA synthesis was carried out in a total volume of 20 μl using the manufacturer's protocol | 1729 LIN aNd BaINES (RevertAid First-Strand cDNA Synthesis kit; Thermo Fisher Scientific). Dmef2 PCR was performed by using forward and reverse primers, which introduced a Kpn I and an Xho I site, respectively, and ligated to pAc5.1 expression vector (Thermo Fisher Scientific). Fifty-six plasmids from independent Escherichia coli colonies were isolated and sequenced to identify splice variants of dmef2. The forward and reverse primer sequences are as follows (5' to 3'): ATTAGGTACCGGATAGGAAATCTGTTGCCATGG and ATTACTCGAGCAGCTCGTGCCGGCTATGT. p300 (nejire, CG15319) was PCR amplified with the primer pairs which introduced Kpn I and Xba I sites in the 5' and 3' end of the open reading frame respectively. PCR product was ligated to pAc5.1 expression vector. The forward and reverse primer sequences are as follows (5' to 3'): AATAGGTACCATGATGGCCGATCACTTAGACG and AATATCTAGACTAGAGTCGCTCCACAAACTTG. All clones were checked by sequencing prior to expression analysis.

| Luciferase assay
S2R+ cells (10 5 cells in 100 μl of Schneider's Drosophila Medium, Gibco) were treated with dsRNA (1 μg) in a 96-well plate (Corning ® Costar ® ) for 3 hr, followed by cotransfection (Effectene, QIAGEN) of dpum promoterfirefly construct and renilla-luciferase reporter (100 ng each) for a further 24 hr. The transfection procedure is as described in the manufacturer's instructions (QIAGEN). The luciferase assay was performed using the dual-luciferase reporter assay system (Promega). Briefly, 30 μl of transfected S2R+ cells were transferred to a well of a 96-well white plate (FluoroNunc ™ ) and lysed with 30 μl of passive lysis buffer and then 30 μl Luciferase Assay Reagent II was added to measure firefly luciferase activity. This was followed by 30 μl of Stop & Glo ® to measure renilla-luciferase activity. A GENios plate reader (TECAN) was used to measure luminescence. At least five independent transfections of each experiment were performed. Double-stranded RNA, dmef2 (BKN27383) and p300 (BKN21411), were obtained from the Sheffield RNAi Screening Facility (Sheffield, UK).
Luciferase activities of the third instar larvae CNS (mixed sexes) were measured using the Promega Steady-Glo Luciferase Assay Kit. Briefly, 20 virgin females of dpum promoter-GAL4 line (see below for details) was crossed to five attP24 UAS-luciferase males (Markstein, Pitsouli, Villalta, Celniker, & Perrimon, 2008). Flies carrying the UAS-luciferase transgene alone were used for background controls. Ten third instar larvae CNSs were collected in 100 μl Promega Glo Lysis buffer for each sample, and five independent samples collected for each genotype. CNSs were homogenized, incubated at room temperature for 10 min, centrifuged for 5 min, and supernatant was transferred to a new tube. For luciferase assays, 30 μl of each sample was transferred to a well of a white-walled 96-well plate at room temperature, and 30 μl Promega Luciferase reagent was added to each well and plates were incubated in the dark for 10 min. Luminescence was measured with a GENios plate reader (TECAN). The obtained values were normalized to total protein concentration, measured using the Bradford protein assay (Bio-rad).
Targeted activity-manipulation was achieved using overexpression of UAS-TrpA, and raised temperature to 29°C for 3 hr. Luciferase activity of pumC-GAL4>luc; TrpA was normalized to pumC-GAL4>luc alone and the measurements at 3 hr compared to 0 hr (set at 1).

| Quantitative RT-PCR
Quantitative RT-PCR was performed using a SYBR Green I real-time PCR method (LightCycler ® 480 SYBR Green I Master; Roche, Mannheim, Germany) as described in Lin, He, and Baines (2015). For examining egfp transcript expression of the dpum MiMIC line, 10 third instar larval CNSs were collected in an Eppendorf and treated with or without 10 mM pentylenetetrazol for 1 hr. Total RNA was extracted using the RNeasy microkit (QIAGEN, Hilden, Germany). PCR primers were designed with the aid of LightCycler Probe Design Software 2.0 (v1.0; Roche). Primer sequences (5' to 3') used were: actin-5C (CG4027), CTTCTACAATGAGCTGCGT and GAGAGCACAGCCTGGAT; dmef2, TTCAAATATC ACGCATCACCG and GCTGGCGTACTGGTACA; p300, GTTCTGGACTTCCCACG and TACTGGCTCATTTG CATGTAAC; egfp, ACGGCAACTACAAGACC and GCTTGTCGGCCATGATATAGA (forward and reverse respectively). The relative gene expression was calculated as the 2 −ΔCt , where ΔCt was determined by subtracting the average actin-5C Ct value from that of dmef2, p300 or egfp.

| Statistics
Statistical significance between group means was assessed using either a Student's t test (where a single experimental group is compared to a single control group and p-values are presented two-sided) or a one-way ANOVA followed by Bonferroni's post hoc test (multiple experimental groups). Data are presented as mean ± standard deviation (SD).

| Identification of the Pumilio promoter region
To analyse transcriptional regulation of dpum, we identified a putative dpum promoter region. A 2-kb region upstream of the transcription start site was targeted as a potential location. Interrogation of transcription factor databases (TRANSFAC model, MAPPER; Marinescu et al., 2005), identified putative binding motifs for 114 transcription factors within the region −2,000 to +1 (transcription initiation marked as +1, motifs listed in Supporting Information Table S1). Putative transcriptional binding motifs include: Sp1, TBP, C/EBP, Oct-1, Mef2, MADS-A/B, Hb, NF-kappaB, TCF, CREB and SRF (Supporting Information Figure S1). To test for function of the dpum promoter region, this 2-kb fragment, (−2,000 to +1, termed pumA) was placed upstream of firefly-luciferase (FF) and transiently transfected in S2R+ cells. After 24 hr, pumA:FF resulted in a 156.5 ± 21.9-fold increase in FF activity compared to transfection of cells with empty vector (set at 1, Figure 1, p = 2.5 × 10 −7 ).
Dmef2 is encoded by a single gene (Lilly, Galewsky, Firulli, Schulz, & Olson, 1994;Nguyen, Bodmer, Abmayr, McDermott, & Spoerel, 1994;Taylor, Beatty, Hunter, & Baylies, 1995) and contains 15 exons. Exons 10 and 14 are alternatively spliced, while exons 9 and 15 contain cryptic splice sites and generate cassettes 9A and 15A respectively ( Figure 2a). To identify the most common splice isoforms of dmef2 transcripts present in the CNS of third instar Drosophila larvae, RT-PCR was used to isolate and clone 56 complete open reading frames (ORFs). Comparison of exon composition of clones revealed nine unique splice variants. Isoforms dmef2(I-IV) were previously identified (Gunthorpe, Beatty, & Taylor, 1999;Taylor et al., 1995), while dmef2(V-VIII) and dmef2(mini) are novel. Analysis of the ORFs showed that dmef2(II) and dmef2(VI) are present at highest frequency ( Figure 2a). Analysis of exon usage across all splice sites show that exon 10 is present at highest abundance followed by exon 14 (Figure 2b).
To gain better understanding of how individual dMef2 splice variants transactivate pumA:FF activity, we constructed another three splice variants (although not found in our CNS analysis, these variants are theoretically possible), dmef2(IX-XI) (Figure 2a). The expression of individual dmef2 variants, in S2R+ cells, resulted in a 1.6 ± 0.1-to 3.3 ± 0.2-fold increase in FF expression compared to control (only intrinsic dmef2 expression, set at 1; Figure 3a). This level of change mirrors previous reports of transgenic MEF2-transactivation of other transcripts (ranging from ~2.5-to 2.9-fold; Lyons, Schwarz, & West, 2012;Wang, Wang, Chen, & Sun, 2017). Sorting splice variants by fold change formed a clear group of dmef2 isoforms that contain exon 10 (Figure 3b). Exon 10 is contained within dmef2 (I, II, III, IV, VI and VII) that, collectively, transactivated the dpum promoter to a greater level than variants lacking F I G U R E 1 Characterization of dpum promoter activity. Activity analysis, by luciferase assays, of constructs bearing defined regions of a putative 2-kb dpum promoter that was placed upstream of firefly-luciferase (FF). These regions are termed pumA, pumB, pumC, pumD and pumE respectively (−2,000, −1,434, −578, −312 and −189 to +1: transcription initiation marked as +1). Constructs were transiently cotransfected, together with a renilla-luciferase (Ren, a loading control, driven by actin promoter), in S2R+ cells for 24 hr and pum:FF to Ren ratio was calculated. PumA, pumB, pumC, pumD and pumE:FF resulted in 156.5 ± 21.9-, 164.8 ± 13.7-, 159.8 ± 17.2-, 103 ± 11.1-and 72.5 ± 13.4-fold increase in FF activity compared to transfection of cells with empty vector (FF to Ren ratio was set at 1; n = 5 independent transfections). Four putative Mef2 binding sites (Mef2-a, -b, -c and -d) with a consensus sequence (C/T)T(A/T)(A/T)AAATA(A/G) were identified (black arrows) and each binding sequence is shown in the inset (identical nucleotides shown in grey boxes). White arrows indicate two potential p300 binding sites. The location of putative transcriptional binding motifs, Sp1, TBP, C/EBP, Oct-1, Mef2, MADS-A/B, Hb, NF-kappaB, TCF, CREB and SRF within the dpum 2-kb promoter region and the initiation of each dpum promoter fragment, pumA-E, are indicated in Supporting Information Figure S1. The detailed binding motifs identified within the 2-kb promoter region of dpum, mouse pum2 and human pum2 are listed in Supporting Information Tables S1-S3 respectively. Data are presented as mean ± SD ****p ≤ 0.0001 (ANOVA with Bonferroni's post hoc)  , p = 2.9 × 10 −5 , 0.005, 1.2 × 10 −6 , 0.0002 and 2.5 × 10 −5 , t test, respectively) confirms that inclusion of this exon results in further significant transactivation.

T T T A A A T A T T T A A A T A T T T A T A T A T T T T A A C A T A T A T A A A T A
To confirm dMef2 transactivation of dpum, we compared FF activity of each dpum promoter construct (pumA to E) following dmef2(VII) overexpression (the variant showing the strongest transactivation). Promoter fragments pumA-E contain 4, 3, 3, 2 and 0 predicted Mef2 binding elements respectively ( Figure 1). Overexpression of dmef2 resulted in 1.9 ± 0.3-, 1.7 ± 0.1-, 1.4 ± 0.3-and 1.4 ± 0.3-fold increase in promoter activity (pumA-D, p = 1.2 × 10 −6 , 4.1 × 10 −6 , 0.004, 0.05, respectively) compared to control (no dmef2 expression, set at 1) while, as predicted, pumE (lacking an Mef2 binding motif) showed no change (1.2 ± 0.1, p > 0.05; Figure 4a).

| dMef2 regulates Pumilio expression level
To confirm observations of transcriptional regulation of dpum by dMef2, we used a previously developed dPum protein activity monitor to determine effect to dPum protein level (Lin et al., 2017). Essentially, we constructed an actin promoter driven firefly-luciferase reporter gene (FF-PRE), containing PREs (Pumilio Response Elements), in the 3' UTR. Increased dPum is sufficient, through binding the PREs and inhibiting translation, to reduce FF activity. An identical renilla-luciferase (Ren) reporter, lacking the PRE sites (and thus not affected by dPum), is coexpressed to allow ratiometric determination of activity (to compensate for batch differences between construct expression). Although this is an indirect measurement, it is currently the best monitor of dPum protein activity, because of the poor performance of commercially available anti-Pum antibodies for Western Blotting in Drosophila (our unpublished data). Overexpression of dmef2(VI) or dmef2(VII) isoforms in S2R+ cells, that express both reporters, significantly reduces the ratio of FF-PRE/Ren to 0.7 ± 0.04 and 0.7 ± 0.21, respectively (control, no dmef2 expression, set as 1, p = 0.00013 and 0.03 respectively, Figure 4b). We conclude that increasing dmef2 is sufficient to increase dPum protein activity, in addition to transcript.

Pumilio transactivation
Our prior work has shown levels of dPum and rPum2 (in fly and rat, respectively) are sensitive to neuronal activity: increasing as levels of synaptic excitation increase (Driscoll et al., 2013;Mee, Pym, Moffat, & Baines, 2004). It was expected, therefore, that dmef2 would show activity-dependent   transcription. Similar to mammals, ingestion of the proconvulsant PTX by larvae is sufficient to increase synaptic excitation and induce a seizure-like state (Stilwell, Saraswati, Littleton, & Chouinard, 2006). Therefore, we performed RT-qPCR to examine dmef2 transcript expression in the CNS taken from PTX-fed larvae. We did not, however, observe a significant fold-change (0.97 ± 0.06, n = 5, p > 0.05) compared to vehicle control (set at 1). This lack of effect is indicative that the expression of dmef2, in Drosophila, is not activity dependent. We are not, however, able to rule out activity-dependent post-transcriptional and/or post-translational modifications of dMef2 which may, in turn, influence the expression or activity of dPum.

| p300 represses Pumilio promoter activity in vivo
To measure dpum promoter activity in vivo, we generated a transgenic pumC-GAL4 fly and mated it with UAS-luciferase (UAS-luc; Markstein et al., 2008). The resultant third instar larval CNS, expressing pumC-GAL4>luc showed a significant increase (2.6 ± 0.6-fold) in luc activity compared to control (UAS-luc line, set at 1, p = 0.0005, t test, n = 5). To test how the pumC promoter responds to increased synaptic excitation, we established a stable line of pumC-GAL4>luc and raised larvae on food containing PTX (1 μg/ml). This resulted in a 1.9 ± 0.4-fold increase in luc compared to the vehicle control (i.e. no PTX, set at 1, p = 0.009, t test, n = 5).
To restrict activity-manipulation to only dpum expressing cells, we crossed pumC-GAL4>luc with UAS-TrpA and raised the ambient temperature to 29°C (activating the TrpA channel). This also resulted in a 2.7 ± 1.3-fold increase in luc expression compared to a 25°C control (set at 1, p = 0.004, t test, n = 15). To further confirm activity dependence of the dpum promoter, we exploited an egfp-inserted dpum MiMIC-RMCE (Minos-mediated integration cassette-recombination mediated cassette exchange) line. This line has egfp inserted within the dpum locus. Exposure of isolated third instar larval CNS to the proconvulsant pentylenetetrazole (PTZ; 10 mM) for 1 hr resulted in a 2.2 ± 1.3-fold increase in egfp transcript expression compared to untreated larvae (set at 1, p = 0.05, t test, n = 8). For this acute treatment on isolated CNS, we used PTZ (a water-soluble GABA A receptor inhibitor) rather than PTX (required to be dissolved in DMSO) to avoid a confounding effect of the vehicle when using isolated CNS (but not observed with whole larvae). These results not only confirm our previous observations that dpum expression is regulated by membrane depolarization (Mee et al., 2004), but transfers our capability to measure dpum promoter activity to in vivo. Overexpression of dmef2, by crossing pumC-GAL4>luc with UAS-dmef2, resulted in a 1.6 ± 0.2-fold increase in luc expression compared to pumC-GAL4>luc crossed with control UAS-GFP (set at 1, p = 0.003 n = 5, Figure 6). This confirms that the pumC promoter is transactivated by dMef2 in vivo. To validate our observation that p300 represses dpum expression, we also crossed pumC-GAL4>luc with UAS-p300 or UAS-p300 RNAi . Overexpression of F I G U R E 4 dMef2 is sufficient to modify dpum promoter transactivation and dPum protein activity. (a) dpum promoter:fireflyluciferase (pum:FF) constructs, containing different numbers of dMef2 binding motifs (pumA-E, 4, 3, 3, 2 and 0, respectively), were cotransfected with dmef2(VII) and a renilla-luciferase gene (Ren, loading control, driven by actin promoter) in S2R+ cells. Expression of dmef2 resulted in 1.9 ± 0.3-, 1.7 ± 0.1-, 1.4 ± 0.3-, 1.4 ± 0.3-and 1.2 ± 0.1-fold increase in promoter activity (pumA-E, respectively) compared to control (set at 1: each construct expressed in the absence of dmef2(VII)) (n = 5 independent transfections). (b) dPum protein activity can be measured using an actin promoter driven fireflyluciferase (FF) reporter gene containing Pumilio Response Elements (PRE) in the 3' UTR (FF-PRE). Increased dPum is sufficient, through binding the PREs and inhibiting translation, to reduce FF activity. An identical renilla-luciferase (Ren) reporter, but lacking the PRE sites (and thus not affected by dPum), is coexpressed to allow ratiometric determination of activity. Expression of dmef2(VI) or dmef2(VII) isoforms in S2R+ cells, that express both reporters, significantly reduces the ratio of FF-PRE/Ren to 0.7 ± 0.04 and 0.7 ± 0.21 respectively (control [CTRL], no dmef2 expression, set as 1; n = 5 independent transfections). Data are presented as mean ± SD.

| DISCUSSION
Pumilio is a well characterized RNA-binding protein that, among its many reported functions, is a key regulator of F I G U R E 5 p300 represses dMef2-mediated transactivation of the dpum promoter. (a) The pumA promoter:firefly-luciferase (pumA:FF) reporter was cotransfected with p300 and a renillaluciferase gene (Ren, loading control, driven by actin promoter) in S2R+ cells. Expression of p300 reduced pumA promoter activity to 0.76 ± 0.07, (control [CTRL], pumA:FF/Ren alone set at 1). By comparison, the expression of dmef2(VII) resulted in a 2.14 ± 0.18fold increase in luciferase activity compared to control. The activity of dmef2(VII) is abolished when coexpressed with p300 (0.75 ± 0.04). Pre-treatment of S2R+ cells with p300 dsRNA enhanced pumA promoter activity to 2.12 ± 0.22. Overexpression of dmef2, in the presence of p300 dsRNA, further increased pumA promoter activity to 4.41 ± 0.74 (n = 5 independent transfections). (b) Co-transfection of pumA:FF with increasing doses of p300 showed a clear dosedependent suppression of dMef2(VII)-mediated transactivation (2.79 ± 0.22, 1.8 ± 0.17, 1.53 ± 0.03, 1.19 ± 0.05 and 0.76 ± 0.07, p300 plasmid: 0, 10, 20, 50 and 100 ng, respectively, pumA:FF/Ren alone set at 1; n = 5 independent transfections). (c) Co-transfection of pumC:FF (which lacks consensus p300 binding sequences) with dmef2(VII) resulted in a 2.41 ± 0.2-fold increase in luciferase activity compared to control (CTRL, pumC:FF/Ren alone set at 1). This enhancement is abolished when coexpressed with p300 (1.03 ± 0.06, n = 5 independent transfections). (d) Yeast two-hybrid assay shows a protein-protein interaction between dMef2 and p300. The bait plasmid pGBKT7-p300 (BD-p300) and the prey pGADT7-dmef2 (AD-dmef2(VI) or AD-dmef2(VII)) were cotransformed into a Y2HGold yeast strain. Coexpression of 1. BD-p300/AD-dMef2(VI) or 2. BD-p300/AD-dMef2(VII) resulted in activation of all four GAL4responsive markers, HIS3, ADE2, AUR1-C and MEL1 reporters in Y2HGold. The negative control groups, 3. BD-lam/AD-dMef2(VI), 4. BD-lam/AD-dMef2(VII) and 5. BD-p300/AD-T, were not able to activate reporter gene expression in Y2HGold. 6. BD-p53/AD-T was used as a positive control. AbA: neuronal firing-rate homeostasis that maintains stability of neuronal circuits Mee et al., 2004;Muraro et al., 2008). Despite this critical role, little is known concerning how Pum levels are regulated. Its involvement with neuronal homeostasis suggests that pum transcription will be governed by an activity-dependent process (Mee et al., 2004). In this study, we characterized the dpum promoter region and identified both Mef2 and p300 to be part of an activity-dependent regulatory mechanism. The complete regulatory mechanism is, however, likely to be more complicated based on our identification of binding motifs for 114 putative transcription factors within a 2-kb region upstream of the dpum transcription start site. A series of deletion constructs show that pumA, pumB and pumC promoters (composed of 2,000, 1,434 and 578 nucleotides, respectively) exhibit promotor activity, while pumD and pumE (consist of 312 and 189 nucleotides, respectively) result in reduced, but still significant, activity. The pumE promoter, the shortest construct we tested, contains binding motifs for 15 transcription factors (e.g., Pbx-1, BR-C Z4, Zen and Lhx3a) and is, as we show, still capable of driving FF expression in S2R+ cells. Further truncations will be required to identify the minimal promoter region for dpum. We identify four dMef2 binding motifs within a 2-kb putative dpum promoter region and, moreover, show that manipulating dmef2 expression is sufficient to influence promoter activity. Comparison with genes known to be transactivated by dMef2 (e.g., myoD, inflated, mir-1, tubulin60D and others) identifies, on average, 5 ± 1.6 Mef2 binding motifs (analysis of eight genes ± SD). By comparison, analysis of a similar number of genes with no reported regulation by Mef2 (e.g., tailup, cry, lim3, eve, shaker and three others) identifies 3.2 ± 1.4 binding motifs per gene (p = 0.009, unpaired t test, Lin and Baines, unpublished data). This difference strongly suggests that dpum expression is regulated, at least in part, by dMef2.
Dmef2, which is encoded by a single gene, contains 15 exons. Of these exons, four are subject to alternative splicing and generate at least 12 splice variants (dmef2(I-XI) and dmef2(mini)). All splice isoforms transactivate the dpum promoter and, notably, exon 10-containing isoforms result in greater activation compared to variants lacking this exon. This differential activity may be indicative that the level of dpum transactivation can be fine-tuned through altering of dMef2 isoform expression, a possibility that will require future investigation. In muscle, by contrast, it is the expression level of dMef2, rather than isoform expression, which is seemingly more important for cellular differentiation (Gunthorpe et al., 1999). This conclusion was reached, however, without consideration of exon 10 lacking isoforms, because splicing of exon 10 was not previously observed (Gunthorpe et al., 1999;Taylor et al., 1995). Our bioinformatics shows that Mef2 binding sites are conserved in the mammalian (e.g. human and mouse) pum2 promoter region. Interrogation using the Harmonizome search engine followed by ChIP-X enrichment analysis (ChEA) of transcription factor targets database, identifies 23 transcription factors, including p300 and Mef2A, associate with the pum2 promoter region (Lachmann et al., 2010;Rouillard et al., 2016). A genome-wide tiling array (ChIP-Chip) analysis similarly identifies dpum as a target of dMef2 (Sivachenko, Li, Abruzzi, & Rosbash, 2013). Our analysis of the promoter region (−2,000 to +1, set transcription initiation at +1), in both mouse and human pum2, identifies 5 and 6 Mef2 elements, respectively (Supporting Information Tables S1 and S2). Thus, it seems likely that Mef2 is a direct regulator of pum transcription in both insects and mammals. This extends the activity of this transcription factor in addition to its reported inhibitory control of pum2 transcript abundance via upregulation of miR-134 (Fiore et al., 2009(Fiore et al., , 2014. It will be important to understand the relative efficacy of the different Mef2 splice variants for their ability to regulate the expression level of miR-134. However, validation of the interaction of Mef2 with its putative binding sites will require confirmation by additional approaches (e.g. chromatin immunoprecipitation or EMSA).
A lack of effect on dmef2 transcript expression level following exposure to PTX is indicative that this factor does not, at least in Drosophila, form a primary link between neuron membrane depolarization and altered expression of dpum. However, we have been unable to source a usable dMef2 antibody and thus dMef2 protein levels were not determined. Our studies instead spotlight p300. p300 contains HAT activity and is also an accessory protein that interacts with transcription factors to function as either coactivator or repressor. For F I G U R E 6 dMef2 and p300 regulate dpum promoter activity in vivo. Expression of dmef2, p300 or p300 F2161A , respectively, resulted in a 1.6 ± 0.2, 0.6 ± 0.1 and 0.6 ± 0.2-fold change in luciferase (luc) activity compared to control (pumC-GAL4>luc cross with UAS-GFP (GFP), set at 1). Knockdown of p300 expression, achieved by crossing UAS-p300 RNAi (p300 RNAi ), resulted in a 1.4 ± 0.3-fold increase in luc activity (n = 5 independent sample collections). Data are presented as mean ± SD. *p ≤ 0.05, **p ≤ 0.01 and ***p ≤ 0.001 (ANOVA with Bonferroni's post hoc)  Akimaru, Chen et al., 1997;Waltzer & Bienz, 1999). By contrast, T-cell factor (TCF)-mediated Wnt/ Wingless signalling is repressed by p300 acetylation (Waltzer & Bienz, 1998). In mammals, p300 has been reported to bridge the complex of thyroid hormone receptor-retinoid X receptor-Mef2A to abrogate transactivation of α-myosin heavy chain gene promoter activity when an inhibitor, adenovirus E1A for example, is recruited (De Luca et al., 2003).
Here, we show p300 acts as a repressor of dMef2-mediated transactivation of dpum. We further show that this effect is likely achieved through direct binding of p300 to dMef2. This result mirrors the reported physical interaction between a TAZ2 domain (within the cysteine-histidine-rich region 3 (CH3)) of p300 and Mef2A/Mef2C (through MASD/Mef2 domains) in mammals (De Luca et al., 2003;He et al., 2011;Sartorelli et al., 1997). Lacking reliable anti-dMef2 and anti-p300 antibodies for Drosophila, we were not able to examine dMef2 and p300 protein-protein interaction by coimmunoprecipitation. However, a comparison of these two domains in human and Drosophila show 85% (TZA2 domains; Lin and Baines unpublished observation) and 86%-89% (MASD/ Mef2 domains) amino acid identify (Lilly et al., 1994;Nguyen et al., 1994). These similarities support our yeast two-hybrid data showing interaction between p300 and Mef2.
Our results are consistent with increasing neuronal depolarization negatively regulating the expression of p300 which, in turn, allows increased dMef2-mediated transactivation of dpum. The expression of p300 is known to be activity dependent and is similarly downregulated in pilocarpine (increased acetylcholine signalling) treated mouse hippocampus (Hansen, Sakamoto, Pelz, Impey, & Obrietan, 2014). The expression of mef2 has also been reported to be regulated by increased synaptic excitation in mammals (Mao, Bonni, Xia, Nadal-Vicens, & Greenberg, 1999), an observation we could not validate in Drosophila. We cannot rule out, however, that additional post-transcriptional and/or post-translational modifications, including alternative splicing and/ or phosphorylation of dmef2 might influence its activity. In this regard, a report of Mef2 activation by Ca 2+ -activated dephosphorylation, via calcineurin is particularly attractive (Flavell et al., 2006). This is because Ca 2+ entry across the neuronal membrane is widely regarded as an initial reporter of neuronal activity in homeostatic mechanisms (Cudmore & Turrigiano, 2004;Gunay & Prinz, 2010;O'Leary, van Rossum, & Wyllie, 2010). Alternative post-transcriptional modifications of Mef2 activity have also been reported. For example, p38 mitogen-activated protein kinase (p38-MAPK) induced phosphorylation of Mef2C is critical for activation of Mef2 target genes (Han, Jiang, Li, Kravchenko, & Ulevitch, 1997;Mao et al., 1999).
The upregulation of dpum transcript expression in late stage 17 Drosophila embryos, due to increased synaptic excitation, was previously reported (Mee et al., 2004). Intriguingly, the analysis of transcript expression in the third instar CNS between wild type and wild type raised on food F I G U R E 7 Transcription of dpum is coregulated by dMef2 and p300. (a) Under "normal" synaptic excitation, p300 is more abundant and binds to dMef2 to inhibit the transactivation of dpum. (b) Increasing exposure to synaptic excitation is sufficient to downregulate the expression of p300 resulting in the release of dMef2 from inhibition. This facilitates transactivation of dpum. Increased dPum protein can, in turn, translationally repress paralytic mRNA (para; voltage-gated sodium channel) to achieve a homeostatic reduction in action potential firing promoter dpum containing PTX revealed a significant reduction in dpum transcript expression (Lin et al., 2017). These paradoxical results might be a result of dpum autoregulation. The dpum transcript contains multiple PRE motifs in its 3'UTR region (Chen et al., 2008;Gerber, Luschnig, Krasnow, Brown, & Herschlag, 2006). Similarly, human PUM1 and PUM2 are also potential targets of PUM protein (Bohn et al., 2018).
In this study, we used a dpum promoter to drive firefly-luc expression, which lacks PRE motifs, and showed enhanced dpum promoter activity in third instar larvae raised on food containing PTX. This result validates that the dpum promoter is responsive to levels of synaptic activity and also provides additional evidence to suggest that Pum regulates its own expression through negative feedback.
In summary, we show that regulation of dpum expression is mediated by an interaction between dMef2 and p300, the latter being an activity-dependent negative regulator. Under "normal" synaptic excitation, p300 is more abundant and binds to dMef2 to inhibit transactivation of dpum. Increased synaptic excitation reduces p300 expression which, in turn, releases dMef2 from inhibition. The increase in dPum protein translationally represses paralytic (voltage-gated sodium channel) mRNA to achieve a homeostatic reduction in action potential firing (a schematic mechanism is shown in Figure 7). It follows, therefore, that inhibition of p300 would be predicted to be anticonvulsant (mirroring increased dPum activity). Indeed, treatments (genetic or pharmacological) that elevate dPum activity are potently anticonvulsive in Drosophila seizure mutants (Lin et al., 2017). Moreover, seizures in flies, rodents and human are associated with decreased dPum or Pum activity respectively (Follwaczny et al., 2017;Lin et al., 2017;Siemen et al., 2011;Wu et al., 2015). Thus, while drug interventions that directly activate Pum may be difficult to achieve in the clinic, inhibition of p300 may represent a more achievable route to better control epilepsy.