ARTICLE IN PRESS Molecular and Cellular Endocrinology ■■ (2014) ■■–■■

Contents lists available at ScienceDirect

Molecular and Cellular Endocrinology j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / m c e

Upstream open reading frames regulate cannabinoid receptor 1 expression under baseline conditions and during cellular stress

1 2 3 4

Q2 M. Eggert a,1,*, M. Pfob a,1, V. Jurinovic b, G. Schelling c, O.K. Steinlein a a

5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

b c

Institute of Human Genetics, University Hospital, Ludwig-Maximilians-University Munich, Germany Institute for Medical Informatics, Biometry and Epidemiology, Ludwig-Maximilians-University, Munich, Germany Department of Anesthesiology, University Hospital, Ludwig-Maximilians-University, Munich, Germany

A R T I C L E

I N F O

Article history: Received 17 April 2014 Received in revised form 27 August 2014 Accepted 17 September 2014 Available online Keywords: Endocannabinoid system Cannabinoid receptor 1 CNR1 Upstream open reading frame

A B S T R A C T

The cannabinoid receptor subtype 1 gene CNR1 is not only associated with phenotypes such as cognitive performance, addiction and anxiety, but is also known to be crucially involved in responses to acute and chronic psychological and cellular stress conditions. Functional analysis of the 5’ untranslated regions of the five known mRNA variants of the human CNR1 gene revealed that two of these variants contain upstream open reading frames that are able to modulate gene expression both under baseline condition and conditions of cellular stress including hypoxia, glucose restriction and hyperthermia. The upstream open reading frames might provide a mechanism that enables the cannabinoid 1 receptor to escape the general repression of protein synthesis that is typical for conditions of cellular stress. © 2014 Elsevier Ireland Ltd. All rights reserved.

32 1. Introduction

33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54

The endocannabinoid system (ECS) plays a crucial role in the regulation of a variety of physiological functions, such as learning and memory processing, vegetative control, energy homeostasis, immunity and stress response (Bermudez-Silva et al., 2012; Hillard, 2014; Puighermanal et al., 2012; Strewe et al., 2012). It acts through different endocannabinoids (ECs) which are able to bind to the cannabinoid receptor subtypes 1 and 2 (CB1 and CB2 receptor) (Strewe et al., 2012). The CB1 receptor is mainly located in the central and peripheral nervous system, but also e.g. in the skeletal and gastrointestinal system, whereas the CB2 receptor can be predominantly found in organs and cells involved in immune response but also in the brain (Onaivi et al., 2012). Both subtypes are members of the G-protein coupled receptor family and act mainly through the activation of an associated guanine nucleotide-binding protein, but other, non-G protein mediated effects of endocannabinoids such as extracellular signal-regulated kinase (ERK) 1/2 activation via betaarrestin have also been described (Franklin et al., 2013). In neurons of the brain, endocannabinoids function as retrograde synaptic messengers. They are synthesized and released postsynaptically and travel backwards across the synapses, stimulating CB1 receptors on

55 56 57 58 59 60 61

Q1

* Corresponding author. Institute of Human Genetics, University Hospital, LudwigMaximilians-University Munich, Goethestraße 29, 80336 Munich, Germany. Tel.: +4989/5160 3646; fax: +4989/5160 4468. E-mail address: [email protected] (M. Eggert). 1 Both authors contributed equally to this work.

62 presynaptic axons and suppressing the release of several excit63 atory neurotransmitters (McLaughlin et al., 2014). 64 Particularly the CB1 receptor has been intensively studied as this 65 receptor subtype is known to be linked to several neurological and 66 endocrinological pathways (Vlachou and Panagis, 2014). A growing 67 body of literature indicates a crucial involvement of the CB1 re68 ceptor in acute and chronic conditions of global (psychological) and 69 cellular stress. Cellular stress is the consequence of disturbed ho70 meostasis, and the cannabinoid receptor plays an important role 71 within the functional network that helps the cell to survive such 72 critical situations (Feuerecker et al., 2012; Sanchez and 73 Garcia-Merino, 2012; Zogopoulos et al., 2013b). 74 Yet the important function of the endocannabinoid system as a 75 regulator of homeostasis is not limited to global stress but is also seen on a cellular level [8]. Endocannabinoid signaling through the Q3 76 77 CB1 receptor has been shown to facilitate the survival of stressed 78 neurons as a result of acute brain injury, neuroinflammation and 79 neurodegenerative diseases. Furthermore, the endocannabinoid 80 system plays an important role in neurogenesis and in repair mecha81 nisms after neuronal injury (Zogopoulos et al., 2013a). The molecular 82 mechanisms leading to altered cannabinoid receptor 1 gene (CNR1) 83 expression under acute or chronic cellular stress exposure are not 84 completely understood. It is known that the 5’- and 3’untranslated 85 regions (UTRs) of genes can harbor regulatory elements that are 86 capable of influencing the expression pattern of the main protein 87 coding region. Multiple regulatory elements, like hairpins, protein88 binding sites, internal ribosomal entry sites (IRESs), polyadenylation 89 signals, microRNA binding sites and iron-responsive elements may 90 be found in these regions (Chatterjee and Pal, 2009; Wethmar et al.,

http://dx.doi.org/10.1016/j.mce.2014.09.019 0303-7207/© 2014 Elsevier Ireland Ltd. All rights reserved.

Please cite this article in press as: M. Eggert, M. Pfob, V. Jurinovic, G. Schelling, O.K. Steinlein, Upstream open reading frames regulate cannabinoid receptor 1 expression under baseline conditions and during cellular stress, Molecular and Cellular Endocrinology (2014), doi: 10.1016/j.mce.2014.09.019

ARTICLE IN PRESS M. Eggert et al./Molecular and Cellular Endocrinology ■■ (2014) ■■–■■

2

1 2 3 4 5 6 7 8 9 10

2010). Furthermore, many genes carry one or more upstream open reading frames (uORFs) in their 5’UTR (Calvo et al., 2009; Iacono et al., 2005; Matsui et al., 2007). Certain uORFs are able to govern protein expression, for example by establishing barrier functions to scanning ribosomes, altering the rate of ribosomal re-initiation, or increasing mRNA instability. In the present study we performed a systematic search for functionally relevant uORFs in the 5’UTRs of the five known CNR1 gene variants. Our experiments revealed that two of these variants possess uORFs that are functional under baseline and/or cellular stress conditions.

11 12 13 14 15 16 17 18 19 20 21 22

2. Material and methods 2.1. In silico analysis of the 5’UTR of CNR1 variants CNR1 variants 1 (NM_016083.4), 2 (NM_033181.3), 3 (NM_001160226.1), 4 (NM_001160258.1) and 5 (NM_001160259.1) were screened for in-frame start and stop codons upstream of the Q4 main start codon using the StarORF Finder (http://star.mit.edu/orf/ runapp.html) (Fig. 1). The putative uORFs contained no validated SNPs (NCBI gene view). Furthermore, none of the SNPs present in CNR1 5’UTRs created or deleted a putative uORF.

23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

For CNR1 variant 3 (NM_001160226.1), possessing two putative uORFs, four constructs were created: one with the wild type mRNA (CNR1-3.1), one with the first uORF switched off (CNR13.2), one with the second uORF switched off (CNR1-3.3), and one construct with both uORFs switched off (CNR1-3.4). CNR1 variant 4 (NM_001160258.1) harbors three putative uORFs. As the third uORF is identical in sequence with the second uORF of variant 3 (and has been analyzed there), constructs were created only for the first two uORFs: one with the wild type mRNA (CNR14.1), one with the first uORF switched off (CNR1-4.2), one with the second uORF switched off (CNR1-4.3), and one construct with both uORFs switched off (CNR1-4.4). For the additional experiments, three more constructs were created, harboring an in-frame stop codon (tga) three bases downstream of the uORF start codon: one with a premature stop codon within the first uORF (CNR1-4.5), one with a premature stop codon within the second uORF (CNR1-4.6), and one construct, in which the putative protein coding sequences of both uORFs were truncated by premature stop codons (CNR1-4.7). For CNR1 variant 5 (NM_001160259.1), harboring one putative uORF, two constructs were created: one with the wild type mRNA (CNR1-5.1) and one with the mutated uORF start codon (CNR1-5.2). An overview of the constructs used in our experiments is shown in the Supplementary Table S1 and S2.

2.2. Plasmid construction Firefly luciferase vector pGL4.10 and renilla luciferase vector pGL4.74 were purchased from Promega (Mannheim, Germany). The TK-Promoter sequence was cut out of pGL4.74 with KpnI and XhoI (Fermentas, St. Leon-Rot, Germany) after generating a XhoI restriction site at position bp 783 using GeneArt site-directed mutagenesis system (Invitrogen, Karlsruhe, Germany). After KpnI and XhoI digestion of pGL4.10, the TK-Promoter sequence was ligated into the multiple cloning site of pGL4.10 upstream of the firefly luciferase coding sequence. The 5’UTR inserts of four different variants of the CNR1 gene were synthesized (MWG Eurofins, Ebersberg, Germany) and cloned into pGL4.10 with XhoI and NcoI directly between the TK-Promoter and the firefly luciferase coding sequence. After cloning the inserts were confirmed by sequencing. To create constructs lacking a certain uORF, the ATG of the uORF was mutated to TTG. For CNR1 variant 1 (NM_016083.4), harboring one putative uORF, two constructs were created: one with the wild type mRNA (CNR11.1) and one with the mutated uORF start codon (CNR1-1.2).

71 2.3. Cell culture The human embryonic kidney cells (HEK) 293 were purchased from Cell Lines Service (Eppelheim, Germany). The HEK cell line was chosen because, unlike neuronal cell lines, it has the advantage of not expressing endogenous CNR1 (information provided by the company ‘Cell Lines’). Culturing of the cells was performed in T25 flasks in monolayer with DMEM (Sigma Aldrich, Hamburg, Germany) containing 4.5 g/l glucose, 1% L-glutamine, 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin (Invitrogen, Karlsruhe, Germany). The cells were maintained in a humidified atmosphere at 37 °C and 5% CO2. For all experiments HEK293 cells were seeded with 3 × 105 cells in 24 well plates (Greiner Bio-One, Frickenhausen, Germany). Cotransfection was performed with the reporter plasmid pGl4.10 + TK containing the 5’UTR of CNR1 variant 1, 3, 4, and 5, respectively, and the control plasmid pGL4.74 at a ratio 20:1 using TransIT®-LT1

43

44 45

46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70

Fig. 1. 5’UTRs of the five cannabinoid 1 receptor (CNR1) variants with putative uORFs. The bp position for each uORF (start/stop) is depicted starting from the main start codon in 5’- direction.

Please cite this article in press as: M. Eggert, M. Pfob, V. Jurinovic, G. Schelling, O.K. Steinlein, Upstream open reading frames regulate cannabinoid receptor 1 expression under baseline conditions and during cellular stress, Molecular and Cellular Endocrinology (2014), doi: 10.1016/j.mce.2014.09.019

72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88

ARTICLE IN PRESS M. Eggert et al./Molecular and Cellular Endocrinology ■■ (2014) ■■–■■

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66

Transfection Reagent (MoBiTec, Göttingen, Germany) with 24 h duration of transfection after seeding. 2.4. Baseline and stress condition tests Baseline test: 24 h after co-transfection culture medium was replaced and the HEK293 cells were kept for 24 h in the incubator at 37 °C and 5% CO2. Hypoxia stress test: 24 h after co-transfection culture medium was replaced and the HEK293 cells were kept in a hypoxia chamber containing 95% nitrogen/5% CO2 for 24 h according to the manufacturer’s protocol (Anaerocult C mini, Merck, Darmstadt, Germany) prior to luciferase assay. No glucose stress test: 24 h after co-transfection culture medium was replaced by DMEM containing no glucose, but FBS, 1% L-glutamine and 1% penicillin/ streptomycin (Invitrogen, Karlsruhe, Germany) for 24 h prior to luciferase assay or qPCR. Hyperthermia test: 24 h after co-transfection culture medium was replaced and the HEK293 cells were kept for 24 h at 41 °C in the incubator prior to luciferase assay. Test conditions were adapted from (Fischer et al., 2008; Lohse et al., 2011). 2.5. Luciferase assay The firefly and renilla luciferase activities were measured with the TRiStar LB941 (Berthold Technologies, Bad Wildbad, Germany) by the Dual-Glow® luciferase assay system (Promega, Mannheim, Germany) according to the manufacturer’s protocol. The ratio of firefly luciferase and renilla luciferase activity was determined and normalized to pGL4.10 + TK. The results for the various 5’UTR constructs are given as normalized firefly activity in percentage. 2.6. Quantitative PCR Quantitative real-time PCR was performed under no glucose conditions for CNR1 variant 4 constructs CNR1-4.1 to CNR1-4.4. Cotransfection with the various constructs was performed as described above. RNA was extracted by using Qiagen RNeasy plus kit, including DNase treatment of 15 min, according to the manufacturer’s protocol (Qiagen, Hilden, Germany). First-strand cDNA synthesis was carried out using 1,6 μg of RNA from each transfection as starting material (QuantiTect reverse transcription kit, Qiagen). Real-time PCR was performed targeting firefly luciferase (target) and renilla luciferase (control) coding sequence using primers firefly-fwd 5’TGCAACACCCCAACATCTTC-3’ and firefly-rev 5’-CCTTTAGGCA CCTCGTCCAC-3’; renilla-fwd 5’-AATGGCTCATATCGCCTCCT-3’ and renilla-rev 5’-CACGACACTCTCAGCATGGA-3’. The reactions were carried out in 20 μl volumes containing 10 μl SsoFast™ EvaGreen® Supermix (Bio-Rad, Munich, Germany), 125 nM of each primer, 2 μl molecular biology grade water and 6 μl of each template after cDNA synthesis. Thermal cycling consisted of denaturation (98 °C for 5 s), annealing and extension (60 °C for renilla; 62 °C for firefly for 5 s), performed in 50 cycle steps. Melting curve analysis ranged from 65 to 95 °C with 0.5 °C intervals. The results for the various 5’UTR constructs are given as firefly mRNA level in percentage normalized to pGL4.10 + TK. 2.7. Statistical analysis All experiments were repeated independently three times with triplicate samples per experiment, which results in a total sample size of nine (3 × 3). The Livak calculation method was applied for real-time PCR data analysis (Livak and Schmittgen, 2001). Twotailed t-test (Tables S3 and S4) was used to compare the values of the target samples and the control samples. For verification of statistical results the general linear model and two-way ANOVA (Table S5) was applied. A p value of p < 0.05 was considered statistically significant.

3

3. Results Five partially overlapping variants of the CNR1 5’UTR that are the results of differential splicing and use of alternate exons are annotated (http://www.ncbi.nlm.nih.gov/omim). They are encoding either CNR1 isoform a (variants 1, 3–5) or isoform b (variant 2) that differs from isoform a by its shorter and distinct N-terminus. The relative expression frequencies as well as the individual patterns of expression of the five CNR1 variants in human brain are not known in detail (Zhang et al., 2004). For experimental design we predominately used the comparison of wild type versus mutant (switched off) uORF (Table S3) because changes seen under these conditions can be unequivocally attributed to the respective uORF. 3.1. CNR1 variant 1 CNR1 variant 1 5’UTR (NM_016083.4) contains one putative uORF of 57 nt length (Fig. 1). Luciferase assays were used to analyze the impact the uORF has on luciferase protein expression. Neither the baseline culturing condition nor any of the three stress tests showed a significant difference between the switched off uORF (CNR1-1.2) and the wild type (CNR1-1.1) (Table S3 and Fig. 2A). Likewise, no significant changes were observed when comparing the luciferase protein expression under baseline conditions versus different stress conditions for each construct (Table S4). 3.2. CNR1 variant 2 In silico analysis showed that the 5’UTR of CNR1 variant 2 (NM_033181.3) did not contain putative uORFs (Fig. 1). Therefore, this gene variant was not included in the study. 3.3. CNR1 variant 3 CNR1 variant 3 5’UTR (NM_001160226.1) was found to harbor two overlapping putative uORFs of 147 nt, respectively 93 nt, that share the same stop codon (Fig. 1). The wild type construct (CNR13.1) was compared with constructs in which either the first or second (CNR1-3.2; CNR1-3.3) or both uORFs (CNR1-3.4) were switched off. Under baseline culturing conditions, none of the three constructs showed significant differences in comparison with the wild type construct (Table S3 and Fig. 2B). In the hypoxia test, none of the two uORFs resulted in a significant difference when being switched off alone. However switching off both uORFs at the same time significantly increased luciferase protein expression by 2.1 fold compared with the wild type (Table S3 and Fig. 2B). When comparing the luciferase protein expression under baseline conditions versus hypoxic conditions for each construct, CNR1-3.1 and CNR1-3.2 showed no significant changes in luciferase protein expression. The luciferase protein expression of CNR1-3.3 and CNR1-3.4 resulted in a significant increase by 2.8, respectively 4.2 fold under hypoxia (Table S4). In the no glucose and hyperthermia tests, none of the three constructs containing mutated uORFs showed significant luciferase protein expression changes compared with the wild type (Table S3 and Fig. 2B). Likewise, results were not statistically significant when comparing the luciferase protein expression under baseline conditions versus no glucose or hyperthermic conditions for each construct (Table S4). 3.4. CNR1 variant 4 CNR1 variant 4 5’UTR (NM_ 001160258.1) contains two overlapping putative uORFs of 96 nt, respectively 87 nt in length, that share the same stop codon (Fig. 1). The third uORF of 93 nt is identical with the second uORF of CNR1 variant 3 (see above). Therefore,

Please cite this article in press as: M. Eggert, M. Pfob, V. Jurinovic, G. Schelling, O.K. Steinlein, Upstream open reading frames regulate cannabinoid receptor 1 expression under baseline conditions and during cellular stress, Molecular and Cellular Endocrinology (2014), doi: 10.1016/j.mce.2014.09.019

67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132

ARTICLE IN PRESS 4

1 2 3 4 5 6

M. Eggert et al./Molecular and Cellular Endocrinology ■■ (2014) ■■–■■

Fig. 2. CNR1 variant 1 uORF is not involved in translational control, CNR1 variant 3 harbors a hypoxia-inducible uORF. (A) Luciferase assay of CNR1 variant 1 5’UTR did not show significant results under baseline or stress conditions. Firefly activity normalized to pGl4.10 + TK in percentage is illustrated for the two different constructs: from left to right – CNR1-1.1 (black bars) and CNR1-1.2 (white bars) for each test. Error bars represent ± SEM of 3 x 3 replicates. (B) Luciferase assay of CNR1 variant 3 5’UTR resulted in significant increase in luciferase protein expression under hypoxic condition when switching off both uORFs. Firefly activity normalized to pGl4.10 + TK in percentage is illustrated for the four different constructs: from left to right – CNR1-3.1 (black bars), CNR1-3.2 (white bars), CNR1-3.3 (gray bars), CNR1-3.4 (shaded bars) for each test. Error bars represent ± SEM of 3 × 3 replicates. Asterisks indicate significant differences (p < 0.05).

7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46

only the first two uORFs were tested by luciferase assays. The wild type construct (CNR1-4.1) was compared with constructs in which either one or both uORFs were switched off. Under baseline and hypoxic conditions, a significant increase in luciferase protein expression was observed when the second uORF or both uORFs were switched off (Table S3 and Fig. 3A). In the no glucose test, all three constructs containing mutated uORFs showed highly significant luciferase protein expression changes compared with the wild type. The construct with the first uORF switched off resulted in a considerable downregulation of luciferase protein expression. However, the constructs with either the second uORF switched off or both uORF switched off showed a significant upregulation of luciferase protein expression compared to the wild type (Table S3 and Fig. 3A). The hyperthermia stress test showed a significant decrease of main ORF translation when comparing the construct with the first uORF switched off (CNR1-4.2) to the wild type. The results for CNR14.3 (second uORF switched off) were marginally not significant, but a significant increase was seen for the variant with both uORFs switched off (Table S3 and Fig. 3A). When comparing for each construct the luciferase protein expression under baseline conditions versus the various stress conditions, the results showed a trend toward significance but did not reach the required level (Table S4). The most constantly significant findings were obtained for CNR1 variant 4 in the no glucose test. We therefore conducted additional experiments in which the putative uORF proteins were truncated by the introduction of premature stop codons to investigate if the observed changes were due to functional uORF peptides. The comparison of the wild type (CNR1-4.1) with the constructs with stop codons inserted into the first, second or both uORF sequences (CNR14.5; CNR1-4.6; CNR1-4.7) showed significant differences (Table S3 and Fig. 3B). Likewise a significant difference was seen when comparing the construct with the first start codon mutated (CNR14.2) to the one carrying a premature stop codon in the first uORF (CNR1-4.5) (p = 0.001) (Fig. 3B). The comparison of the construct with the second start codon mutated (CNR1-4.3) with the one harboring a premature stop codon in the second uORF (CNR1-4.6) did not result in any significant difference (p = 0.943) (Fig. 3B).

Additionally, we performed relative qPCR for the constructs CNR14.2 to CNR1-4.4 and compared results with the wild type (CNR14.1) in order to ascertain whether the observed no glucose-induced changes in luciferase protein expression originated at the transcriptional level. For all CNR1 variant 4 constructs mRNA expression levels determined at no glucose levels were non-significant (Table S3 and Fig. 3C).

47 48 49 50 51 52 53 54

3.5. CNR1 variant 5 In silico analysis showed that the 5’UTR of CNR1 variant 5 (NM_001160259.1) harbors one putative uORF with 102 nt (Fig. 1). When comparing the wild type construct with the uORF-switchedoff construct in the luciferase assay, results were not statistically significant for any of the four tests (Table S3 and Fig. 3D). When for each of the two constructs the luciferase protein expression was compared under baseline conditions versus the various stress conditions, results showed a trend toward significance but did not reach the level of significance (Table S4).

55 56 57 58 59 60 61 62 63 64 65 66

4. Discussion and conclusions 4.1. Discussion The regulation of gene expression by uORFs has been previously described for other genes involved in cellular stress response. Their uORFs seem to play a role in several feedback loops that ensure adequate gene expression and enable the organism to adapt to different types of internal or environmental stress. Some of them display rather complicated regulatory mechanisms, such as ATF4 that contains a uORF that is usually inhibitory but nevertheless able to induce translation under stress conditions (Mueller and Hinnebusch, 1986). Another example is GADD34 which interacts with protein phosphatase 1, causing dephosphorylation of the eukaryotic initiation factor 2α (eIF2α) and thereby relaxing the stress induced inhibition of general protein synthesis. GADD34 contains a uORF that directs translation of the main ORF during eIF2α phosphorylation and is therefore directly involved in stress response (Lee et al., 2009).

Please cite this article in press as: M. Eggert, M. Pfob, V. Jurinovic, G. Schelling, O.K. Steinlein, Upstream open reading frames regulate cannabinoid receptor 1 expression under baseline conditions and during cellular stress, Molecular and Cellular Endocrinology (2014), doi: 10.1016/j.mce.2014.09.019

67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84

ARTICLE IN PRESS M. Eggert et al./Molecular and Cellular Endocrinology ■■ (2014) ■■–■■

1 2 3 4 5 6 7 8 9 10 11 12 13

5

Fig. 3. CNR1 variant 4 contains two functional uORFs. CNR1 variant 5 uORF is not involved in translational control. (A) Luciferase assay of CNR1 variant 4 5’UTR resulted in significant decrease in luciferase protein expression with only the more 3’ located uORF switched on. A significant increase in luciferase protein expression was observed with only the more 5’ located uORF switched on or both uORFs switched off. Firefly activity normalized to pGl4.10 + TK in percentage is illustrated for the four different constructs: from left to right – CNR1-4.1 (black bars); CNR1-4.2 (white bars); CNR1-4.3 (gray bars); CNR1-4.4 (shaded bars) for each test. Error bars represent ± SEM of 3 × 3 replicates. Asterisks indicate significant differences (p < 0.05). (B) Luciferase assay of CNR1 variant 4 5’UTR with inserted stop codons to destroy the uORF sequence significantly reversed the decrease in luciferase protein expression previously observed with the more 3’ located uORF switched on under no glucose condition. This effect was not seen when destroying the uORF sequence of the more 5’ located uORF. Firefly activity normalized to pGl4.10 + TK in percentage is illustrated for the different constructs: from left to right – CNR1-4.1 (black bar); CNR1-4.2 (white bar); CNR1-4.3 (gray bar); CNR1-4.5 (horizontally shaded bar); CNR1-4.6 (vertically shaded bar); CNR1-4.7 (dotted bar). Error bars represent ± SEM of 3 × 3 replicates. Asterisks indicate significant differences (p < 0.05). (C) qPCR results ruled out that these findings were due to a transcriptional effect. Firefly mRNA level normalized to pGl4.10 + TK in percentage is illustrated for the different constructs: from left to right – CNR1-4.1 (black bar); CNR14.2 (white bar); CNR1-4.3 (gray bar); CNR1-4.4 (shaded bar). Error bars represent ± SEM of 3 × 3 replicates. (D) Luciferase assay of CNR1 variant 5 5’UTR did not show significant results under baseline or stress conditions. Firefly activity normalized to pGl4.10 + TK in percentage is illustrated for the two different constructs: from left to right – CNR15.1 (black bar) and CNR1-5.2 (white bar) for each test. Error bars represent ± SEM of 3 × 3 replicates.

14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

The results presented here show that two of the five 5’UTR variants of CNR1 contain uORFs that are able to govern expression of a downstream main ORF and might have a role in cellular stress responses. For CNR1 variant 3 these effects were significant only under conditions of hypoxia and only if both uORFs were removed, implicating a weak to moderate effect restricted to certain stressful situations. The results obtained for CNR1 variant 4 suggest a more complex regulatory mechanism in which the uORFs are functional both at baseline conditions and under different types of cellular stress. Our qPCR experiments under no glucose conditions showed no differences in mRNA concentration levels between cells expressing variant 4 constructs containing intact or mutated uORFs, indicating that the observed effects originated at the level of translation rather than transcription. Interestingly, under conditions of cellular stress the two uORFs within CNR1 variant 4 showed opposite effects. Mutagenesis of the

start codon of the more 5’ located uORF significantly decreased translation of the main ORF. The apparently contrary effect, i.e. increased protein expression, was observed after mutagenesis of the more 3’ located uORF or after mutagenesis of both uORFs. At a first glance the decrease in main protein caused by the loss of the more upstream uORF seems unusual because most functional uORFs are known to be repressors of gene expression (Calvo et al., 2009). However, recent studies have shown that the regulation of gene expression by uORFs can be rather complex, especially in situations when stress conditions elicit specific cellular responses designed to help minimize the potential damage unfavorable conditions might inflict. Cellular stress is known to propagate phosphorylation of the initiation factor eIF2α, which in turn represses the translation of most cellular mRNAs in order to save resources for important cellular functions or later attempts to perform cellular repair. Only a few selected mRNAs exist that escape this general repression of

Please cite this article in press as: M. Eggert, M. Pfob, V. Jurinovic, G. Schelling, O.K. Steinlein, Upstream open reading frames regulate cannabinoid receptor 1 expression under baseline conditions and during cellular stress, Molecular and Cellular Endocrinology (2014), doi: 10.1016/j.mce.2014.09.019

31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46

ARTICLE IN PRESS 6

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66

M. Eggert et al./Molecular and Cellular Endocrinology ■■ (2014) ■■–■■

protein synthesis. Most of these are required either for the stress response itself or to aid the cells recovery from stress (Barbosa et al., 2013). It has been shown that regulation of translation by uORFs constitutes one of the major escape mechanisms that allow selected mRNAs to avoid the general shut-down during cellular stress response. This mechanism has already been studied in examples such as the yeast transcription factor GCN4, a gene containing four uORFs in its 5’UTR (Hood et al., 2009; Mueller and Hinnebusch, 1986). Under baseline conditions translation is initiated at each succeeding uORF-ATG, and the inhibitory features exhibited by these uORFs decrease the chances that the main ORF will be translated. Under conditions of stress the first uORF is still translated. However, due to shortage of eIF2α-formed ternary complex the 40S ribosomal subunit that remains associated with the mRNA is less likely to be able to recruit ternary complex before reaching downstream uORFs. This reduced reinitiation efficiency increases the chances that ternary complex is available once the 40S ribosomal subunit has travelled the whole length of the 5’UTR and reached the main ORF (Barbosa et al., 2013; Hood et al., 2009; Mueller and Hinnebusch, 1986). Overall, this mechanism helps the main GCN4-ORF to evade global repression of translation during stress. The very same mechanism would help to increase the impact the cannabinoid receptor is known to exert in situations of global or cellular stress situations. The effects observed after mutagenesis of the two uORFs within CNR1 variant 4 would fit well with such a hypothesis of differential upregulation of gene expression. In such a scenario the uORFs would show a complex pattern of regulation during conditions of stress. After mutagenesis of the first uORF the 40S ribosomal subunit would be free to move until it reaches the ATG of the second uORF, which, compared with the first uORF, is located considerably closer to the start codon of the main ORF. With restricted availability of ternary complex during conditions of stress this would lower the chances of the 40S ribosomal subunit to recruit the latter in time before reaching the main ORF. This would explain the reduced amount of main protein observed after mutagenesis of the upstream uORF under stress but not baseline conditions. The increased translation of the main ORF noted after mutagenesis of the second uORF would be due to the fact that the second uORF presents an obstacle between the first uORF and the main ORF, and that its removal increases the chances of the scanning 40S ribosomal subunit to recruit the ternary factor needed for the translation of the main ORF. This could mean that in situations of stress the apparently conflicting effects the two CNR1 variant 4 uORFs show with regard to protein expression are needed as a kind of finely tuned regulatory mechanism. It is at this point not possible to predict the effect such a complex pattern of regulation would have in vivo, especially since not much is known about the expression patterns of the CB1 receptor variants in brain and other organs. It is, however, tempting to speculate that the complexity of the variant 4 uORF regulatory mechanism might be especially helpful in critical situations in which the cannabinoid receptor is needed to support the survival of healthy cells but should not prevent the elimination of damaged ones. It is also worth mentioning that the two uORFs of variant 4 seem to use different mechanisms while modulating the translation of the main ORF. The insertion of a premature stop codon within the uORF sequences affected the function of both uORFs under no glucose conditions, strongly suggesting that the proteins encoded in the uORF sequences are functional. However, a significant difference was also seen for the first uORF when comparing the construct carrying a mutated start codon with the one that carried an artificial premature stop codon. This suggests that the first uORF mainly acts in a peptide-dependent manner while the second uORF uses additional mechanisms. Likely candidates would be ribosome stalling or nonsense-mediated decay (Morris and Geballe, 2000), mechanisms that are not dependent on an intact open reading frame and would therefore remain functional after the stop codon insertion.

4.2. Conclusions

67 68 69 70 71 72 73 74 75 76 77 78

The data presented here suggest that CNR1 not only is involved in an organism’s answer to global stressors but that it is likely to also play a role in stress responses on a more basic, cellular level. Our results suggest that for at least two of the CNR1 variants both the baseline expression and the stress responses are modulated by the presence of uORFs within their 5’UTRs. It is not possible yet to predict the impact these uORFs might have on CNR1 function at in vivo conditions, but it will be interesting to find out how they fit within the complex system of general and cellular stress management.

79 Funding

80 81 82 83 84 85

This work was supported by the Deutsche Forschungsgemeinschaft [STE16511-3]. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

86 Acknowledgments

87 88 89

We thank Franz Jansen for excellent technical assistance.

90 Appendix: Supplementary Material

91 92 93 94

Supplementary data to this article can be found online at doi:10.1016/j.mce.2014.09.019.

95 References Barbosa, C., Peixeiro, I., Romao, L., 2013. Gene expression regulation by upstream open reading frames and human disease. PLoS Genet. 9, e1003529. Bermudez-Silva, F.J., Cardinal, P., Cota, D., 2012. The role of the endocannabinoid system in the neuroendocrine regulation of energy balance. J. Psychopharmacol. 26, 114–124. Calvo, S.E., Pagliarini, D.J., Mootha, V.K., 2009. Upstream open reading frames cause widespread reduction of protein expression and are polymorphic among humans. Proc. Natl. Acad. Sci. U.S.A. 106, 7507–7512. Chatterjee, S., Pal, J.K., 2009. Role of 5’- and 3’-untranslated regions of mRNAs in human diseases. Biol. Cell 101, 251–262. Feuerecker, M., Hauer, D., Toth, R., Demetz, F., Holzl, J., Thiel, M., et al., 2012. Effects of exercise stress on the endocannabinoid system in humans under field conditions. Eur. J. Appl. Physiol. 112, 2777–2781. Fischer, U., Radermacher, J., Mayer, J., Mehraein, Y., Meese, E., 2008. Tumor hypoxia: impact on gene amplification in glioblastoma. Int. J. Oncol. 33, 509–515. Franklin, J.M., Vasiljevik, T., Prisinzano, T.E., Carrasco, G.A., 2013. Cannabinoid agonists increase the interaction between beta-Arrestin 2 and ERK1/2 and upregulate beta-Arrestin 2 and 5-HT(2A) receptors. Pharmacol. Res. 68, 46–58. Hillard, C.J., 2014. Stress regulates endocannabinoid-CB1 receptor signaling. Semin. Immunol. Hood, H.M., Neafsey, D.E., Galagan, J., Sachs, M.S., 2009. Evolutionary roles of upstream open reading frames in mediating gene regulation in fungi. Annu. Rev. Microbiol. 63, 385–409. Iacono, M., Mignone, F., Pesole, G., 2005. uAUG and uORFs in human and rodent 5’untranslated mRNAs. Gene 349, 97–105. Lee, Y.Y., Cevallos, R.C., Jan, E., 2009. An upstream open reading frame regulates translation of GADD34 during cellular stresses that induce eIF2alpha phosphorylation. J. Biol. Chem. 284, 6661–6673. Livak, K.J., Schmittgen, T.D., 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25, 402–408. Lohse, I., Reilly, P., Zaugg, K., 2011. The CPT1C 5’UTR contains a repressing upstream open reading frame that is regulated by cellular energy availability and AMPK. PLoS ONE 6, e21486. Matsui, M., Yachie, N., Okada, Y., Saito, R., Tomita, M., 2007. Bioinformatic analysis of post-transcriptional regulation by uORF in human and mouse. FEBS Lett. 581, 4184–4188. McLaughlin, R.J., Hill, M.N., Gorzalka, B.B., 2014. A critical role for prefrontocortical endocannabinoid signaling in the regulation of stress and emotional behavior. Neurosci. Biobehav. Rev. 42, 116–131. Morris, D.R., Geballe, A.P., 2000. Upstream open reading frames as regulators of mRNA translation. Mol. Cell. Biol. 20, 8635–8642.

Please cite this article in press as: M. Eggert, M. Pfob, V. Jurinovic, G. Schelling, O.K. Steinlein, Upstream open reading frames regulate cannabinoid receptor 1 expression under baseline conditions and during cellular stress, Molecular and Cellular Endocrinology (2014), doi: 10.1016/j.mce.2014.09.019

Q5

96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139

ARTICLE IN PRESS M. Eggert et al./Molecular and Cellular Endocrinology ■■ (2014) ■■–■■

1 2 3 4 5 6 7 8 9 10 11 12 13

Mueller, P.P., Hinnebusch, A.G., 1986. Multiple upstream AUG codons mediate translational control of GCN4. Cell 45, 201–207. Onaivi, E.S., Ishiguro, H., Gu, S., Liu, Q.R., 2012. CNS effects of CB2 cannabinoid receptors: beyond neuro-immuno-cannabinoid activity. J. Psychopharmacol. 26, 92–103. Puighermanal, E., Busquets-Garcia, A., Maldonado, R., Ozaita, A., 2012. Cellular and intracellular mechanisms involved in the cognitive impairment of cannabinoids. Philos. Trans. R. Soc. Lond. B. Biol Sci. 367, 3254–3263. Sanchez, A.J., Garcia-Merino, A., 2012. Neuroprotective agents: cannabinoids. Clin. Immunol. 142, 57–67. Strewe, C., Feuerecker, M., Nichiporuk, I., Kaufmann, I., Hauer, D., Morukov, B., et al., 2012. Effects of parabolic flight and spaceflight on the endocannabinoid system in humans. Rev. Neurosci. 0, 1–8.

7

Vlachou, S., Panagis, G., 2014. Regulation of brain reward by the endocannabinoid system: a critical review of behavioral studies in animals. Curr. Pharm. Des. 20, 2072–2088. Wethmar, K., Smink, J.J., Leutz, A., 2010. Upstream open reading frames: molecular switches in (patho)physiology. Bioessays 32, 885–893. Zhang, P.W., Ishiguro, H., Ohtsuki, T., Hess, J., Carillo, F., Walther, D., et al., 2004. Human cannabinoid receptor 1: 5’ exons, candidate regulatory regions, polymorphisms, haplotypes and association with polysubstance abuse. Mol. Psychiatry 9, 916–931. Zogopoulos, P., Vasileiou, I., Patsouris, E., Theocharis, S., 2013a. The neuroprotective role of endocannabinoids against chemical-induced injury and other adverse effects. J. Appl. Toxicol. 33, 246–264. Zogopoulos, P., Vasileiou, I., Patsouris, E., Theocharis, S.E., 2013b. The role of endocannabinoids in pain modulation. Fundam. Clin. Pharmacol. 27, 64–80.

Please cite this article in press as: M. Eggert, M. Pfob, V. Jurinovic, G. Schelling, O.K. Steinlein, Upstream open reading frames regulate cannabinoid receptor 1 expression under baseline conditions and during cellular stress, Molecular and Cellular Endocrinology (2014), doi: 10.1016/j.mce.2014.09.019

14 15 16 17 18 19 20 21 22 23 24 25 26

Upstream open reading frames regulate cannabinoid receptor 1 expression under baseline conditions and during cellular stress.

The cannabinoid receptor subtype 1 gene CNR1 is not only associated with phenotypes such as cognitive performance, addiction and anxiety, but is also ...
911KB Sizes 1 Downloads 5 Views