News Article | August 26, 2016
The Colorado Energy Office, Energy Resource Center, and Colorado Springs Utilities announce the installation of a 2kW rooftop solar power array as part of the state's Weatherization Assistance Program.
The Coca-Cola Company has just achieved one of its major environmental goals, five years ahead of schedule. The company announced on Monday that for every drop of water used in its beveraage, it can now give the same amount back to the planet. In 2007, Coca-Cola announced a goal of replenishing the water it uses by the year 2020. Through 248 community water partnerships in 71 countries around the world, the company claims to have already met its goal. An assessment conducted by LimnoTech and Deloitte in conjunction with The Nature Conservancy found that in 2015, Coca-Cola returned 191.9 billion liters of water to nature or human communities — 115 percent of the water it used in its beverages that year. “Now, every time a consumer drinks a Coca-Cola product, they can have confidence that our company and bottling partners are committed to responsible water use today and tomorrow, Muhtar Kent, chairman and chief executive of the Coca-Cola company said in a statement. The milestone comes after years of criticism of Coca-Cola’s water practices from environmental and global justice organizations. The company originally announced its water replenishment goal in 2007 following a campaign by anti-poverty group War on Want claiming that Coca-Cola had exacerbated water shortages and contaminated local water supplies in communities around the world, particularly in India. Just a few years prior, in 2004, officials in the southern Indian state of Kerala had shut down a bottling plant following a protracted legal battle in which local residents argued the plant had overexploited the region’s groundwater. Within the past couple of years, the company has dropped plans for several new facilities in other parts of the country following local protests. And just this year, Coca-Cola shut down a bottling plant in the northern Indian community of Kaladera after local activists claimed the facility was draining groundwater resources. The controversy didn’t end with the company’s pledge to become “water neutral” by 2020, either. Some have claimed that its replenishment efforts still miss the mark. Amit Srivastana, global resistance director at the global justice group India Resource Center, argued in a December blog post that Coca-Cola’s goal of water neutrality could never actually be achieved. “Water issues are local in their impact unlike, for example, climate change,” he wrote. “When Coca-Cola extracts water from a depleted aquifer in Varanasi or Jaipur, the impacts are borne by the local communities and farmers that depend upon it to meet their water needs. Replenishing an aquifer hundreds of miles away from the point of extraction, as Coca-Cola has often done to ‘balance’ their water use, has no bearing on the health of the local aquifer which Coca-Cola depletes through its bottling operations, nor the privations suffered by those who depend upon it.” But the company has argued that “at each of its 863 plants globally, Coca-Cola requires operations to determine the sustainability of the water supply they share with others in terms of quality, quantity, and other issues such as infrastructure to treat and distribute water.” In cases where the sustainability of a water supply is questionable, the company claims, a source water protection plan is implemented to help find solutions for the problem. “While each plant may not replenish all water to its direct source, Coca-Cola’s policy is to require that all plants work to ensure they do not negatively impact water sources and work with the community on longer term solutions,” the company has stated. In his statement, Kent — the Coca-Cola Company CEO — also added, “We are keenly aware that our water stewardship work is unfinished and remain focused on exploring next steps to advance our water programs and performance.”
No statistical methods were used to predetermine sample size. The experiments were not randomized and the investigators were not blinded to allocation during experiments and outcome assessment. For TFs, gateway-compatible entry clones (Invitrogen) containing the open reading frames (ORFs) lacking stop codons were obtained from ref. 11. Drosophila Act5C-promoter driven expression clones were created using the Gateway system. The TF ORFs were shuttled into the GAL4-DNA binding domain (DBD) containing destination vector pAGW-GAL4-DBD (cloned as described below) by mixing 100 ng of TF entry clone, 100 ng of pAGW-GAL4-DBD and 0.7 μl of LR clonase II enzyme mix (Invitrogen). The identities of all TF entry clones have been confirmed by Sanger sequencing using the primers 5′-CCCAGTCACGACGTTG-3′ and 5′-CACAGGAAACAGCTATG-3′. Note that we tested the full-length transcription factors, including their DBDs, as trans-activating and DNA-binding functions might not always reside in entirely separate protein domains. While this implies that the fusion proteins might bind via the TFs’ DBDs in addition to the GAL4-DBD mediated recruitment, this does not influence the results of the assay: the assay itself measures transcriptional activation independently of where TF binding occurs and we expect that the TFs’ DBDs have at most minor effects on binding strengths as the GAL4-DBD binds to DNA already very strongly. For cofactors, we compiled a list of 338 cofactors based on several criteria. We included proteins containing Pfam domains typical for transcriptional cofactors (for example, HAT, HDAC, SET, Chromo, Bromo), proteins which are part of chromatin modifying or remodelling complexes or part of complexes associated with RNA polymerases (for example, SAGA, Polycomb, TFIID, Mediator), and Drosophila proteins which are homologues of mammalian chromatin-associated proteins (Supplementary Table 3). We amplified the cofactor ORFs from cDNA using oligonucleotides containing Gateway-compatible attB-sites (5′-GGGGACAAGTTTGTACAAAAAAGCAGGCTTC-3′ and 5′-GGGACCACTTTGTACAAGAAAGCTGGGTC-3′) for subsequent entry clone creation. The primer sequences have been chosen to be as close as possible to an annealing temperature of 60 °C which we calculated using the formula with cA, cT, cG and cC being the number of adenines, thymines, guanines and cytosines, respectively. The full list of resulting primer sequences (lacking the attB sequences) is listed in Supplementary Table 3; for 18 of the cofactors no primer sequences are available because we obtained these entry clones from ref. 11 categorized as TFs but manually re-categorized them as cofactors based on their annotation in FlyBase31 or their protein domain content32. For cDNA generation, RNA was isolated from S2 cells and reverse transcribed as described in ref. 33. For PCR amplification, KOD and KOD XL DNA (Merck Millipore) and KAPA HiFi (KAPA) polymerases were used according to manufacturer’s specifications. We created Gateway entry clones by mixing 1 μl of PCR reaction, 100 ng of pDONR221, and 1 μl of BP clonase II enzyme mix (Invitrogen). The identities and correctness of all entry clones have been ensured using Sanger and next-generation sequencing (see below) and we deposited them at Addgene (http://www.addgene.org/Alexander_Stark/). The cofactor ORFs were then shuttled to the Drosophila Act5C-promoter driven destination vector pAGW-GAL4-DBD as described for TFs. The insert flanks of all obtained cofactor entry clones have been Sanger-sequenced and automatically checked to cover the TSS and TTS of one of the isoforms annotated by FlyBase. All entry clones passing additional manual visual inspection using BLAT and the UCSC genome browser have been subjected to further verification by next-generation sequencing as follows. A pool of 100–300 entry clones corresponding to a total of 5 μg DNA solved in 50 μl TE buffer was sonicated (duty cycle, 20%; intensity, 5; 200 cycles per burst; time, 90 s) to 200–400 bp using a S220 Focused-ultrasonicator (Covaris) as described in ref. 11. The fragmented plasmid pool was then prepared for deep sequencing using the Illumina DNA Sample Prep kit and sequenced using a HiSeq2000 (Illumina) producing 50-nt reads. The resulting reads have been assembled and analysed using PrInSes-C34. All insert sequences not starting with ATG, containing a stop codon or a frameshift were immediately rejected. All sequences with less than five mutations leading to non-synonymous amino acid changes were immediately accepted. The remaining sequences were translated, aligned against the respective protein sequence, and manually decided. The next-generation sequencing reads have been deposited at the NCBI Sequence Read Archive (SRA) under the accession SRS806429; the PrInSes-C-generated full-length transcript sequences are available at http://factors.starklab.org and in Supplementary Data 1, and the cofactor Gateway entry clones from Addgene (http://www.addgene.org/Alexander_Stark/). We cloned a destination vector to conveniently create vectors expressing N-terminally V5- and GAL4-DBD-tagged TFs and cofactors under the control of the Drosophila Act5C promoter using the Gateway cloning system. pAGW-GAL4-DBD was cloned by amplifying the GAL4-DBD from pBPGUw35 using one oligonucleotide containing the V5-tag (peptide sequence MGKPIPNPLLGLDST) 5′-TCTGATATCATGGGGAAGCCAATCCCTAATCCCCTTCTGGGACTCGACTCTACCGGCGGCTCTATGAAGCTACTGTCTTCTATCGAACA-3′ and the oligonucleotide 5′-TATACCGGTGGCCGCCGCCCGACGATACAGTCAACTGTCTTTGAC-3′. Amplification was performed using KOD Polymerase (Merck Millipore) according to the manufacturer’s instructions. The resulting PCR product was digested using EcoRV and AgeI and ligated into pAGW (Drosophila Gateway Vector Collection), which was digested using the same enzymes, thereby replacing eGFP with V5-GAL4-DBD. We created Gateway-compatible (Invitrogen) destination vectors to conveniently clone reporter vectors for different regulatory contexts based on firefly luciferase transcribed from a housekeeping core promoter (hkCP; promoter of ribosomal gene RpS1213) or a developmental core promoter (dCP; Drosophila synthetic core promoter (DSCP) derived from Eve35). We created the destination vector attR_dCP_luc by digesting pGL4.26 (Promega) with FseI and BglII and ligating a fragment containing DSCP and luc+, thereby replacing the minimal promoter and luc2 with DSCP-luc+. We digested the resulting vector with KpnI and BglII and ligated a fragment containing the attR Gateway cassette, yielding attR_dCP_luc. We created two hkCP-driven destination vectors containing a Gateway cassette either upstream (attR_hkCP_luc) or downstream (hkCP_luc_attR) of the luciferase reporter gene by using the plasmid pGL3 (Promega) as a basis and replacing the SV40 promoter with the promoter of RpS12 as described in ref. 13. The resulting vector was digested using either KpnI and BglII (to create attR_hkCP_luc) or AfeI (to create hkCP_luc_attR); in both cases, we amplified a Gateway attR cassette using oligonucleotides containing the respective restriction sites, and digested and ligated it into the digested plasmid. All enhancers, motif mutant contexts and other motif or backbone mutant variants were either PCR amplified with primers containing attB Gateway sites or ordered as synthesized fragments (IDT), shuttled into entry clones using TOPO or BP Clonase II (both Invitrogen), and shuttled into the luciferase destination vectors using the LR clonase II enzyme mix (Invitrogen) by mixing 1 μl of PCR product or synthesized DNA solved in TE buffer, 100 ng of destination vector and 0.7 μl of LR clonase II enzyme mix (Invitrogen). We used a modified version of pRL-TK (Promega) to normalize the firefly signal for transfection efficiency and cell number. Ubi-RL has been created by cloning a region upstream of the gene Ubi-p63E (chr3L: 3901760-3902637) upstream of the Renilla luciferase gene in reverse orientation using NheI and BglII. S2 cells, derived from embryos36, were obtained from Life Technologies and grown in Schneider’s Drosophila Medium (Life Technologies 21720-024) supplemented with 10% FBS (Sigma F7524) and 1% penicillin/streptomycin (Life Technologies 15140-122) grown in T75 flasks (ThermoScientific 156499) at 27 °C and passaged every 2–4 days. BG3 neuroblast-like cells, derived from larvae37, were obtained from the Drosophila Genomics Resource Center (DGRC) and grown in Schneider’s Drosophila Medium supplemented with 10% FBS, 1% penicillin/streptomycin, and 10 μg ml−1 Insulin (Sigma-Aldrich I1882) in T75 flasks at 27 °C and passaged every 3–4 days. Kc167 cells, derived from embryos38, were obtained from DGRC and grown in M3/BPYE Medium containing 5% FBS and 1% penicillin/streptomycin in T75 flasks at 27 °C and passaged every 2–3 days. Ovarian somatic cells (OSCs), derived from adult ovaries39, were obtained from the laboratory of J. Brennecke and grown in Shields and Sang M3 Insect Medium (Sigma-Aldrich S8398) supplemented with 10% FBS, 1% insulin, 1% glutathione, 1% fly extract, and 1% penicillin/streptomycin in T75 flasks at 27 °C and passaged every 2–3 days. All cell lines used are regularly checked for mycoplasma contamination. S2 cell transfections were performed using jetPEI (peqlab 13-101-40N). Four hours before transfection, 30,000 cells (30 μl of a 106 cells per ml suspension) were seeded in clear polystyrene 384-well plates (ThermoScientific 164688). For each transfection, we used 30 ng firefly luciferase reporter plasmid, 3 ng Renilla luciferase expressing plasmid Ubi-RL, and 3 ng GAL4-DBD-TF/cofactor or GAL4-DBD-GFP fusion protein expressing plasmid. Beforehand, we assayed the effects of using different amounts of GAL4-DBD fusion protein expressing plasmid and chose 3 ng (Extended Data Fig. 9). The DNA solution containing 36 ng DNA in 5 μl TE buffer was filled up to 15 μl using sterile 150 mM NaCl (polyplus) and prepared in 96-well plates. Transfection reagent (15 μl total: 13.95 μl 150 mM NaCl, 1.05 μl jetPEI) was added to each well of the 96-well plates and mixed rigorously. After 30 min incubation at 25 °C, cells were transfected in quadruplicates by transferring each transfection mix four times (6 μl each) to four adjacent wells of a 384-well plate containing the seeded cells. Luciferase assays were performed after 48 h of growth at 27 °C. Handling the transfection mixes and all subsequent pipetting steps have been performed using a Bravo Automated Liquid Handling Platform (Agilent). Kc167, BG3, and OSC cell transfections were performed using jetPEI in the same way as described above for S2 cells with the exception of transfection reagent composition: 15 μl total containing 14.1 μl 150 mM NaCl and 0.9 μl jetPEI. Human HeLa cells (gift from the laboratory of J. M. Peters) were grown in DMEM medium (Gibco 52100-047) supplemented with 10% heat-inactivated FBS, 1% penicillin/streptomycin and 2 mM L-glutamine (Sigma G7513) in T75 flasks at 37 °C in an atmosphere of 95% air and 5% carbon dioxide. All cell lines used are regularly checked for mycoplasma contamination. We performed HeLa cell transfections using a self-prepared 1 mg ml−1 PEI (25,000 MW, Polysciences 23966) stock solution in PBS (pH adjusted to pH 4.5 and sterile filtered). On the day before transfection we seeded 30 μl of a suspension containing 4,000 HeLa cells in medium (DMEM, 10% FBS, penicillin/streptomycin) into each well of a 384-well plate. Three microlitres of a PEI/DMEM mix (0.24 μl PEI filled to a total of 4.5 μl using DMEM without FBS and penicillin/streptomycin and incubated at room temperature for 5 min) were added to 3 μl of a DNA/DMEM mix (44.5 ng firefly luciferase reporter vector, 4.45 ng TF expression vector (created using pAGW-CMV_ GAL4-DBD, see below) and 4.45 ng pRL-CMV vector for transfection normalization (Promega #E2261) in DMEM without FBS and penicillin/streptomycin. The resulting DNA/PEI mix in DMEM was incubated at room temperature for 30 min and subsequently added to the seeded cells. We performed cell lysis and luciferase assays using the Promega dual-luciferase reporter assay system (Promega E1910) according to the manual. We created the Gateway destination vector pAGW-CMV_GAL4-DBD by replacing the Drosophila Act5C promoter in pAGW-GAL4-DBD with a region containing the CMV enhancer and the T7 promoter amplified from pRL-CMV using the primers 5′-CGACAGATCTTCAATATTGGCCATTAGCCATAT-3′ and 5′-GGTGGCTAGCCTATAGTGAGTCGTATTA-3′. Dual-luciferase assays were performed using self-prepared substrate solutions (D-Luciferin and Coelenterazine have been obtained from GoldBio LUCK-250 and pjk-Gmbh 102111) and lysis buffer as described in ref. 40. For cell lysis, the supernatant was removed and 30 μl of lysis buffer added and incubated gently shaking for 30 min. Ten microlitres of the cell lysates were transferred to black 384-well plates for luminescence assays (Nunc MaxiSorp, Sigma-Aldrich P6491-1CS). All pipetting steps have been performed using a Bravo Automated Liquid Handling Platform (Agilent). Luminescence was measured after adding 20 μl of each substrate, for firefly and Renilla luciferase respectively, using a Biotek Synergy H1 plate reader coupled to a plate stacker. We normalized all firefly luciferase signals to the signal of Renilla luciferase to control for transfection efficiency and cell number (the relative luciferase signal). We then further normalized all relative luciferase signals for TF- and cofactor-GAL4-DBD transfections to relative luciferase signals obtained for GAL4-DBD-GFP transfections (fold-change over GFP). We assessed statistical significance by two-sided unpaired t-tests on the two sets of quadruplicate relative luciferase signals (GAL4-DBD-TF/COF versus GAL4-DBD-GFP). Throughout the paper, ‘activation’ was defined as a fold-change ≥1.5 (P < 0.05), and ‘repression’ was defined as a fold-change ≤1/1.5 (P < 0.05), both compared to the signal for GAL4-DBD-GFP. We corrected the P values for multiple testing using the Benjamini and Hochberg method as implemented in R (p.adjust with method ‘BH’ or its alias ‘fdr’). All statistical calculations and graphical displays, if not stated otherwise, have been performed using version 2.15.3 of the R software suite41. Enrichment analyses have been performed for each of the 15 clusters and for 6 types of features. To first obtain a coarse functional characterization of the clusters, we assessed the enrichments and depletions of TFs which are able to activate or repress a developmental (dCP) or housekeeping (hkCP) core promoter on their own (≥1.5-fold activation or repression (P < 0.05), both compared to the signal for GAL4-DBD-GFP when tested on a context comprised of UAS sites upstream of a developmental core promoter (4×UAS-dCP) or a housekeeping core promoter hkCP (4×UAS upstream hkCP)). Homopolymeric amino acid repeat motifs have been de novo discovered using MEME42 (version 4.8.1, q-value threshold of 1 × 10−5) in TFs that activated or repressed on their own outside enhancer contexts (tested in the 4×UAS dCP context; ≥1.5-fold; P < 0.05). Pfam domain32 signature matches in the Drosophila proteome have been generated using hmmer43 (version 3.0b3, e-value threshold of 0.01). Eukaryotic Linear Motifs44 (ELM; version 08/2014) were matched to the amino acid sequences of the tested TF protein isoforms, after masking the TFs’ Pfam. Additionally, Gene Ontology45 (GO) annotations, and gene expression patterns in the Drosophila embryo as annotated by ref. 46 (IMAGO) have been subjected to enrichment and depletion analyses. To control for multiple testing, we empirically determined false-discovery rates (FDRs) for the different hypergeometric P values. For this, we repeated the feature enrichment analyses 1,000 times, each after randomly shuffling the TF-to-cluster assignments, and recorded the best (that is, most significant) P values. We then adjusted the original P values such that only 10% of the 1,000 random controls reached the P values of the original data (FDR < 10%). Following this protocol, we separately adjusted the FDR cut-off for each cluster (15) and feature type (ELM, MEME, Pfam, GO, IMAGO). To assess if tethering via the GAL4-DBD reflects the different TFs’ regulatory functions when bound to their endogenous motifs, we selected two sets of TFs, three TFs that preferentially activated the CGCG- versus the GATA-context (Fig. 1e) and four TFs that preferentially activated the hormone-receptor contexts; Fig. 2c). We replaced each UAS site in the enhancer mutant contexts S2-1 CGCG, S2-1 GATA, and Nhe2 EcR3, 12 (which also corresponds to an endogenous TF motif in the wild-type enhancers, for example, the EcR motif for the hormone contexts) with a sequence corresponding to the consensus motif of the respective TF as reported in refs 47, 48. (Dfd: CTTAATGA, Hey: CAGCCGACACGTGCCCC, Ets21C: ATTTCCGGT, Ato: AACAGGTGG, Ets96B: ACCGGAAGTAC, Gl: ATTTCAAGAATA, HLH4C: AAAAACACCTGCGCC). The enhancer rescue constructs were synthesized by IDT, shuttled into the luciferase reporter vector attR_dCP_luc using the Gateway system and tested in luciferase assays in S2 cells exactly as described above. To assess potential functional associations of assigned TFs and cofactors, we followed the strategy from ref. 30, recruiting TFs via GAL4-DBD and providing untagged cofactors. For this, we chose contexts in which the different TFs (Clk of cluster 8, Bsh of cluster 10, and CG17186 of cluster 14) were active (4×UAS-dCP for Clk and 4×UAS-upstream-hkCP for Bsh and CG17186). We prepared DNA mixes to be transfected containing 29 ng firefly luciferase reporter plasmid, 3 ng Renilla luciferase expressing plasmid Ubi-RL, 1 ng (Bsh and CG17186) or 0.5 ng (Clk) of GAL4-DBD–TF fusion protein expressing plasmid and an increasing series of untagged cofactor expressing plasmid (0 ng, 0.003 ng, 0.006 ng, 0.012 ng, 0.023 ng, 0.047 ng, 0.094 ng, 0.188 ng, 0.375 ng, 0.75 ng, 1.5 ng, 3 ng). We kept the total amount of transfected plasmid DNA constant at 36 ng for all experiments using a GFP-expressing plasmid. To clone the expression plasmids for the untagged cofactors and GFP, we used the Gateway-compatible vector pAW (Drosophila Gateway Vector Collection). The remaining experimental procedure and analysis was performed as described above. We clustered the 474 TFs based on the log -transformed fold-change values (TF over GFP) from all 24 contexts. First, we standardized all contexts and constructed a k-nearest-neighbour graph (k = 15). We used the Euclidean distance as distance measure as it reflects both the variation of the enhancer activity profile across contexts and the effect sizes within each context; that is, it is able to discriminate between strong and weak activators and repressors even if they vary similarly across the 24 contexts. Next, we took a symmetrized (A + AT) adjacency matrix of this graph and solved multiclass spectral clustering as described in ref. 49 and implemented in the Python package scikit-learn50. In order to decide about the number of clusters and to assess the clustering validity, we analysed the clustering stability upon bootstrapping the data set51. In order to visualize the data, we mapped the data onto a plane by a specialized nonlinear dimensionality reduction technique (t-SNE)52. The algorithm provides the visualization by mapping data points close in the original space to nearby locations in the plane, preserving the local structure. We extended the k-nearest-neighbour graph to include cofactors by comparing the log -transformed fold-change values (cofactor over GFP) of cofactors and TFs (k = 5, Euclidean distance). The locations of the cofactors in the visualization were obtained from spring layout. We know that UAS sites in the enhancer mutant contexts most probably replace binding sites that are functional3 but we do not know which TFs bind them in vivo. In order to check whether we recover these positive controls in the enhancer mutants, we took all the TFs expressed in S2 cells (RPKM > 1) (ref. 53) for which motifs are known3. We scanned the wild-type enhancer sequences (S2-1-wt, S2-2-wt, S2-3-wt, Ubi-1-wt, Ubi-2-wt, Ubi-3-wt) for motif matches with P < 9.76 × 10−4 (1/4,096) using an in-house motif-detection program. For each mutant context, we considered only those TFs for which any of its motif matches had at least 5 mutated base pairs. In the resulting set of TFs (Extended Data Table 1) there is at least one TF per each of the enhancer mutant contexts that activated the respective context when recruited via the GAL4-DBD (≥1.5-fold activation compared to GFP; P < 0.05). We tested a subset of the original 472 TFs in four different cell types (S2, Kc167, BG3 and OSC). This subset consists of 171 TFs covering all the 15 clusters by 9–17 TFs, including all the TFs mentioned in the main text. In each cell type, we computed Euclidean distances after standardizing the log -transformed fold-change values in each context. Then we compared the distances of intra-cluster TF–TF pairs (both TFs belong to the same cluster) to inter-cluster TF–TF pairs (each of the TFs belongs to a different cluster). In order to test whether the medians of these two groups of distances are significantly different, we determined empirical P values as follows. We randomly shuffled TF-to-cluster assignments 106 times and each time computed the medians of the distances for both groups. We mark the P values P < 1 × 10−6 as we never obtained a difference between the medians of intra- and inter-cluster distances as large as for the actual data for any of the cell types.
No statistical methods were used to determine sample size. The experiments were not randomized and the investigators were not blinded during experiments and outcome assessment. DvPdf-GAL4 was provided by J. H. Park; Clk4.1M-GAL4 was from P. Hardin; UAS-dTrpA1 (2nd) was from P. Garrity; UAS-CaLexA was from J. Wang41; UAS-TNT and UAS-Tet were from H. Amrein; Pdf-GAL80 and CRY-GAL80 are described by Stoleru et al.12; LexA-P2X2 and Clk856-GAL4 were from O. Shafer11, 29. UAS-CD4::spGFP1-10 and LexAop-CD4::spGFP11 were from K. Scott; Clk4.1M-lexA was from A. Sehgal5, LexAop-LUC was generated by X. Gao and L. Luo48. LexAop-dTrpA1 was from G. M. Rubin. UAS-VGLUT RNAi 1 (VDRC 104324), UAS-mGluRA RNAi 1 (VDRC 103736), UAS-mGluRA RNAi 2 (VDRC 1793) were from the Vienna Drosophila Resource Center (VDRC). The following lines were ordered from the Bloomington Stock Center: Pdfr (R18H11)-GAL4 (48832), Pdfr (R18H11)-LexA (52535),UAS-CsChrimson (55136), UAS-eNPHR3.0 (36350), UAS-Denmark (33064), UAS-ArcLight (51056), UAS-GCaMP6f (42747),UAS-syt-GFP (33064), UAS-VGLUT RNAi 2 (40845, 40927), VGlutMI04979-GAL4 (60312). Flies were reared on standard cornmeal/agar medium supplemented with yeast. The adult flies were entrained in 12:12 light-dark cycles at 25 °C. The flies carrying GAL4 and UAS-dTrpA1 were maintained at 21 °C to inhibit dTrpA1 activity. Locomotor activity of individual male flies (aged 3–7 days) was measured with Trikinetics Drosophila Activity Monitors or video recording system under 12:12 light:dark conditions. The activity and sleep analysis was performed with a signal-processing toolbox implemented in MATLAB (MathWorks). Group activity was also generated and analysed with MATLAB. For dTrpA1-induced neuronal firing experiments (Fig. 3 and Extended Data Fig. 9), flies were entrained in light:dark for 3–4 days at 21 °C, transferred to 27 °C for two days, followed by 2 subsequent days at 21 °C. The evening activity index (Extended Data Fig. 9) was calculated by dividing the average activity from ZT8–12 by the average activity from ZT 0–12. The behaviour experiments involving RNAi expression (Extended Data Fig. 10b) were done at 27 °C to enhance knockdown efficiency. All statistical analyses were conducted using IBM SPSS software. The sample size was chosen based on the pilot studies to ensure >80% statistical power to detect significant difference between different groups. Animals within the same genotype were randomly allocated to experimental groups and then processed. We were not blind to the group allocation as the experimental design required specific genotypes for experimental and control groups. However, the data analyser was blinded when assessing the outcome. The Wilks–Shapiro test was used to determine normality of data. Normally distributed data were analysed with two-tailed, unpaired Student’s t-tests, one-way analysis of variance (ANOVA) followed by a Tukey–Kramer HSD test as the post-hoc test or two-way analysis of variance (ANOVA) with post-hoc Bonferroni multiple comparisons. Nonparametrically distributed data were assessed using the Kruskal–Wallis test. Data were presented as mean behavioural responses, and error bars represent the standard error of the mean (s.e.m.). Differences between groups were considered significant if the probability of error was less than 0.05 (P < 0.05). Experiments were repeated at least three times and representative data was shown in figures. For mechanical stimulation, individual flies from different groups were loaded into 96-well plates and placed close to a small push–pull solenoid. The tap frequency of the solenoid was directly driven by an Arduino UNO board (Smart Projects). One tap was used as a modest stimulus and ten taps (1 Hz) was used as a strong stimulus. Arousal threshold was measured during the middle of the day (ZT6) and evening (ZT10) with different intensities. The movement of flies before and after the stimulus was monitored by the web camera and the recording videos (1fps) were processed by the MTrack2 plugin in Fiji ImageJ software to convert the videos into binary images and to calculate the trajectory and moving area as well as the percentage of aroused flies. All trans-retinal (ATR) powder (Sigma) was dissolved in alcohol to prepare a 100 mM stock solution for CsChrimson experiments23. 100 μl of this stock solution was diluted in 25 ml of 5% sucrose and 1% agar medium to prepare 400 μM of ATR food. Newly eclosed flies were transferred to ATR food for at least 2 days before optogenetic experiments. The behavioural setup for the optogenetics and video recording system is schematized in Supplementary Fig. 1. Briefly, flies were loaded into white 96-well Microfluor 2 plates (Fisher) containing 5% sucrose and 1% agar food with or without 400 μM ATR. Back lighting for night vision was supplied by an 850 nm LED board (LUXEON) located under the plate. Two sets of high power LEDs (627 nm) mounted on heat sinks (four LEDs per heat sink) were symmetrically placed above the plate to provide light stimulation. The angle and height of the LEDs were adjusted to ensure uniform illumination. The voltage and frequency of red light pulses were controlled by an Arduino UNO board (Smart Projects). The whole circuit is described in ref. 25. The flat surface and compact wells of the 96-well plate allow uniform illumination, which was difficult to achieve in Trikinetics tubes. We used 627 nm red light pulses at 10 Hz (0.08 mW mm−2) to irradiate flies expressing the red-shifted channelrhodopsin CsChrimson within the DN1s23. (The CsChrimson illumination protocol had no effect on halorhodopsin eNpHR3.0). Fly behaviour was recorded by a web camera (Logistic C910) without an infrared filter. We used time-lapse software to capture snapshots at 10 s intervals. The light:dark cycle and temperature was controlled by the incubator, and the light intensity was maintained in a region that allowed entrainment of flies without activating CsChrimson. Fly movement was calculated by Pysolo software and transformed into a MATLAB readable file14. 5 pixels per second (50% of the Full Body Length) was defined as a minimum movement threshold15, 16. The activity and sleep analyses were performed with a signal-processing toolbox implemented in MATLAB (MathWorks) as described above. The design of the invention has been filed for patent. To monitor bioluminescence activity in living flies, we used previously described protocols49. White 96-well Microfluor 2 plates (Fisher) were loaded with 5% sucrose and 1% agar food containing 20 mM d-luciferin potassium salt (GOLDBIO). 250 μl of food was added to each well. Individual male or female flies expressing CaLexA–LUC were first anaesthetized with CO and then transferred to the wells. We used an adhesive transparent seal (TopSeal-A PLUS, Perkin Elmer) to cover the plate and poked 2–3 holes in the seal over each well for air exchange. Plates were loaded into the stacker of a TopCount NXT luminescence counter (Perkin Elmer). Assays were carried out in an incubator under light:dark conditions. Luminescence counts were collected for 5–7 days. For temperature shift experiments (Fig. 4b), the incubator temperature was set to 21 °C for 3 days and then increased to 30 °C at ZT 0 of the 4th day. Other experiments were performed at 25 °C. Three different modes were used in our experiments: (1) To record CaLexA–LUC activity only, 9 plates were placed in a stacker, and each plate was sequentially transferred to the TopCount machine for luminescence reading. Every cycle took about 1 h, and the recording was continued for several days. (2) To combine optogenetic stimulation with the luciferase assays (Fig. 4a and Extended Data Fig. 6a), we replaced the stacker with a chamber of our own design (Fig. 4a). 627 nm LEDs mounted to a pair of heat sinks were symmetrically positioned in the chamber to ensure uniform illumination of the 96-well plate (0.08 mW mm−2 for CsChrimson stimulation and 1 mW mm−2 for eNPHR3.0 stimulation). Flies pre-fed with ATR were loaded into a plate. Single plates stayed in the LED chamber for 8 min and then automatically transferred to the TopCount for luminescence reading for 2 min. (3) To assay fly movement in 96-well plates and CaLexA–LUC activity at the same time, single plates were recorded using a web camera attached to the top of chamber (Fig. 4a). During each hour, the plate sat in the video chamber for 58 min and then was automatically transferred to the TopCount machine for a 2 min luminescence reading. The raw data were analysed in MATLAB and in Microsoft Excel. All experiments were repeated at least three times. Immunostaining was performed as described50. Fly heads were removed and fixed in PBS with 4% paraformaldehyde and 0.008% Triton X-100 for 45–50 min at 4 °C. Fixed heads were washed in PBS with 0.5% Triton X-100 and dissected in PBS. The brains were blocked in 10% goat serum (Jackson Immunoresearch) and subsequently incubated with primary antibodies at 4 °C overnight or longer. For VGLUT and GFP co-staining, a rabbit anti-DVGlut (1:10,000) and a mouse anti-GFP antibody (Invitrogen; 1:1,000) antibody were used as primary antibodies. For GRASP staining, a mouse anti-GFP monoclonal antibody (Invitrogen; 1:1,000) and a rabbit anti-GFP antibody (Roche; 1:200) were used. After washing with 0.5% PBST three times, the brains were incubated with Alexa Fluor 633 conjugated anti-rabbit and Alexa Fluor 488 conjugated anti-mouse (Molecular Probes) at 1:500 dilution. The brains were washed three more times before being mounted in Vectashield Mounting Medium (Vector Laboratories) and viewed sequentially in 1.1 μm sections on a Leica SP5 confocal microscope. To compare the fluorescence signals from different conditions, the laser intensity and other settings were set at the same level during each experiment. Fluorescence signals were quantified by ImageJ as described. mRNA profiling from E cells and DN1s was performed as previously described34. DN1s and E cells were purified from Clk4.1M-GAL4, UAS-EGFP flies (DN1s) and Dv-Pdf-GAL4, UAS-EGFP, PDF-RFP flies, (E cells; GFP+RFP− cells), respectively. Flies were entrained for 3 days and then collected every 4 h for a total of six time points. Two replicates of six time points were performed for each cell type. Sequencing data were aligned to the Drosophila genome using TopHat51. Gene expression was quantified using the End Sequencing Analysis Toolkit (ESAT; publicly available at http://garberlab.umassmed.edu/software/esat/). ESAT quantifies gene expression only using information from the 3′-end of the genes. Imaging experiments were performed as previous described52. Adult male fly brains were dissected in ice-cold haemolymph-like saline (AHL) (108 mM NaCl, 5 mM KCl, 2 mM CaCl2, 8.2 mM MgCl , 4 mM NaHCO , 1 mM NaH PO -H O, 5 mM trehalose, 10 mM sucrose, 5 mM HEPES; pH 7.5). Brains were then pinned to a layer of Sylgard (Dow Corning) silicone under a small bath of AHL contained within a recording/perfusion chamber (Warner Instruments) and bathed with room temperature AHL. Brains expressing GCaMP6f and Arclight were exposed to fluorescent light for approximately 30 s before imaging to allow for baseline fluorescence stabilization. Perfusion flow was established over the brain with a gravity-fed ValveLink perfusion system (Automate Scientific). ATP or glutamate was delivered by switching the perfusion flow from the main AHL line to another channel containing diluted compound after 30 s of baseline recording for the desired durations followed by a return to AHL flow. For the mGluRA antagonist imaging experiments, 700 nM LY341495 (Tocris Bioscience) was used to block the glutamate-induced inhibition. Imaging was performed using an Olympus BX51WI fluorescence microscope (Olympus) under an Olympus ×40 (0.80 W, LUMPlanFl) or ×60 (0.90W, LUMPlanFI) water-immersion objective, and all recordings were captured using a charge-coupled device camera (Hamamatsu ORCA C472-80-12AG). For GCaMP6f and Arclight imaging, the following filter sets were used (Chroma Technology): excitation, HQ470/×40; dichroic, Q495LP; emission, HQ525/50 m. Frames were captured at 2 Hz with 4 × binning for either 2 min or 4 min using μManager acquisition software52. Neutral density filters (Chroma Technology) were used for all experiments to reduce light intensity and to limit photobleaching. For recordings using GCaMP6f and Arclight, ROIs were analysed using custom software developed in ImageJ52 and National Institute of Health. The fluorescence change was calculated by using the formula: ΔF/F = (F – F )/F × 100%, where F is the fluorescence at time point n, and F is the fluorescence at time 0. The fluorescence was calibrated by subtracting the background fluorescence value. To compare the fluorescence change between neurons in the same brain, fluorescence activities from different neurons were normalized to the highest fluorescence level during the recording time window.
Michael Rout, head of the Laboratory of Cellular and Structural Biology, and Brian Chait, head of the Laboratory of Mass Spectrometry and Gaseous Ion Chemistry, together with their colleagues, have been working to understand the nuclear pore complex for nearly two decades. They first isolated and described the components of the yeast nuclear pore complex in 2000 and then released a first draft of its structure in 2007. In yeast, as well as in humans, the nuclear pore complex is composed of an inner ring sandwiched between two outer rings: one facing in and one facing out. Attached to these outer rings are classes of proteins that are compatible with the unique chemistry of either the nucleus or cytoplasm found in the rest of the cell. This asymmetric distribution of nuclear pore components is important for helping to establish the directionality of transport through the pore. The complex is so fundamental that its architecture was thought to be shared by all eukaryotes. However, Samson Obado, a research associate in the Rout lab, has found a eukaryotic species that has pores with a structure unlike any that has yet been studied, a finding which implies that the pore's evolution is probably more involved than had previously been assumed. Obado's results were published in February in PLoS Biology. "We were driven by curiosity about the evolution of the nuclear pore complex," says Obado. Almost everything we know about the complex is from yeast and human, which although quite far apart in evolutionary terms are nevertheless closely related when compared to plants and many other single-celled eukaryotes. "But what did the original pores look like, and how have they developed in different eukaryotes?" Obado wondered. He was the perfect candidate to investigate this, as his graduate studies focused on trypanosomes; these parasites, responsible for serious diseases such as sleeping sickness and Chagas disease, diverged from the families that developed into yeast, humans, plants, and other eukaryotes roughly one and a half billion years ago - close to the time of the last common ancestor of all eukaryotes. "Compared with many other eukaryotes, trypanosomes have an unusual and quirky molecular biology. A key example is how they transcribe their genes into messenger RNAs for translation into proteins, which is very different from textbook models. Furthermore, because they are so divergent, you can't just search for gene sequences similar to those in yeast or humans," Obado says. Rout, Chait, and colleagues worked together with Mark Field of Dundee University in Scotland to identify putative nuclear pore components from trypanosomes. To do so, the team walked protein by protein through the nuclear pore complex, purifying each one along with other proximal proteins. These other proteins, they figured, could be part of the nuclear pore complex as well. Obado then purified the new proteins to determine whether they were also part of the complex, along with their proximal proteins, and so on until he had a complete survey of all the proteins involved. Finally, they worked with the university's Electron Microscopy Resource Center to determine where key components are located with respect to each other. The result was a complete first picture of the entire trypanosome pore structure. The teams' results show that the architecture of the inner ring of the nuclear pore complex is fundamentally similar in trypanosomes, yeast, plants, and vertebrates, suggesting an ancient origin for this common feature. In contrast, the trypanosome nuclear pore complex possesses a unique mechanism tethering it to the surrounding nuclear envelope, and the outer ring is less well "conserved" by evolution, with additional never-before-seen components. However, the most notable difference between trypanosome nuclear pore complexes and others that have been examined to date, is that the trypanosome's nuclear pore complex exhibits a near-complete symmetry of its components, and lacks almost all the proteins in yeast and humans that are required to establish an asymmetric assembly within the pore. Instead, it seems that soluble proteins in the nucleus and cytoplasm are responsible for setting transport's directionality in these parasites for both proteins and RNA. The fact that the trypanosome pore has such dissimilarities with the pore of humans may provide an opportunity, says Obado. "These differences may offer something we can target therapeutically without risking harm to our own transport mechanisms." These discoveries also lend further credence to an idea in evolutionary theory first suggested by Rout and his colleagues. The theory, known as the "protocoatomer hypothesis," suggest that the nuclear pore and nuclear envelope share a common ancestor with structures that form coats on other membrane-bound structures in the cell, like those in the Golgi apparatus and endoplasmic reticulum. Indeed, the trypanosome nuclear pore components, despite their dissimilarities from other eukaryotes, still retain the kinds of structures and organization found in these other coats, the scientists say. Moreover, since the inner core of the trypanosome nuclear pore complex are structurally well preserved and similar to those from other eukaryotes, Rout and his team believe this core represents the original units of a simpler nuclear pore complex that pre-dates the last ancestor of all eukaryotes, and thus may provide new clues as to how the nucleus originally evolved. Explore further: Research suggests core nuclear pore elements shared by all eukaryotes More information: Samson O. Obado et al. Interactome Mapping Reveals the Evolutionary History of the Nuclear Pore Complex, PLOS Biology (2016). DOI: 10.1371/journal.pbio.1002365