An evolutionary perspective of AMPK–TOR signaling in the three domains of life
Valentin Roustan1,*, Arpit Jain2,*, Markus Teige1, Ingo Ebersberger2,3,†,‡ and Wolfram Weckwerth1,4,†,‡
1 Department of Ecogenomics and Systems Biology, University of Vienna, Vienna, Austria2 Department of Applied Bioinformatics, Institute for Cell Biology and Neuroscience, Goethe University, Max-von-Laue Str. 13, D-60438 Frankfurt, Germany3 Senckenberg Biodiversity and Climate Research Centre (BiK-F), Senckenberg Anlage 25, D-60325 Frankfurt, Germany4 Vienna Metabolomics Center (VIME), University of Vienna, Vienna, Austria
* These authors contributed equally to this work.† These authors contributed equally to this work.‡ Correspondence: [email protected] and [email protected]
Received 2 March 2016; Accepted 3 May 2016
Editor: Christine Foyer, Leeds University
Abstract
AMPK and TOR protein kinases are the major control points of energy signaling in eukaryotic cells and organisms. They form the core of a complex regulatory network to co-ordinate metabolic activities in the cytosol with those in the mitochondria and plastids. Despite its relevance, it is still unclear when and how this regulatory pathway was formed during evolution, and to what extent its representations in the major eukaryotic lineages resemble each other. Here we have traced 153 essential proteins forming the human AMPK–TOR pathways across 412 species representing all three domains of life—prokaryotes (bacteria, archaea) and eukaryotes—and reconstructed their evolutionary history. The resulting phylogenetic profiles indicate the presence of primordial core pathways including seven proto-kinases in the last eukaryotic common ancestor. The evolutionary origins of the oldest components of the AMPK pathway, however, extend into the pre-eukaryotic era, and descendants of these ancient proteins can still be found in contem-porary prokaryotes. The TOR complex in turn appears as a eukaryotic invention, possibly to aid in retrograde signal-ing between the mitochondria and the remainder of the cell. Within the eukaryotes, AMPK/TOR showed both a highly conserved core structure and a considerable plasticity. Most notably, KING1, a protein originally assigned as the γ subunit of AMPK in plants, is more closely related to the yeast SDS23 gene family than to the γ subunits in animals or fungi. This suggests its functional difference from a canonical AMPK γ subunit.
Cells require a continuous supply of energy, and con-stantly sense the energy status to modulate cellular metab-olism accordingly. Energy is chemically stored in ATP and is released by hydrolysis into ADP/AMP and phosphate/pyrophosphate, respectively, fueling most of the cellular pro-cesses. Sensing the intracellular energy level and co-ordinating
the organism’s growth in accordance with the energy status is therefore indispensable for cellular life. In animals, the AMPK complex accomplishes the surveillance of the energy status. Under energy stress, it inhibits cell growth and stimu-lates catabolic processes to increase nutrient and energy avail-ability (Halford et al., 2003, 2004; Hardie, 2014). The target
of rapamycin complex (TORC) acts in an antagonistic way. It promotes cell growth and protein synthesis under favora-ble energy conditions (Dunlop and Tee, 2009; Laplante and Sabatini, 2012).
Comparative studies have revealed that the AMPK com-plex is well conserved among eukaryotes, and corresponding complexes have been characterized in yeast (SNF1) and in plants (SnRK1) (Polge and Thomas, 2007). Throughout the rest of the manuscript, we will subsume these evolutionar-ily related and functionally equivalent complexes in different species under the name AMPK. Most eukaryotic AMPK complexes are heterotrimeric. They are formed by the cata-lytic subunit α and the regulatory subunits β and γ. Plants possess an additional subunit called SnRK1βγ which is part of all SnRK1 complexes (Emanuelle et al., 2015). The cor-responding proteins contain, in association with the CBS domain, a KIS domain present normally in β subunits (Gissot et al., 2004). In mammals, AMPK activity requires the phosphorylation of Thr172 located in the activation loop of the α subunit. This phosphorylation site can be regulated in different ways: allosterically by the AMPKγ subunit which senses the ATP:AMP and ATP:ADP ratios (Xiao et al., 2011; Oakhill et al., 2012), or by direct phosphorylation, medi-ated by the serine/threonine kinase 11 (STK11 also called LKB1) (Hawley et al., 2003; Woods et al., 2003; Shaw et al., 2004). However, in plants and fungi, AMPK complexes are not directly regulated by AMP (Crozet et al., 2014). It was shown in plants that phosphorylated sugars, such as glucose-1-phosphate, glucose-6-phosphate, or trehalose-6-phosphate, play a role in the regulation of plant AMPK activity (Nunes et al., 2013a, b; Crozet et al., 2014; Nagele and Weckwerth, 2014; Yadav et al., 2014). Nevertheless, the molecular mecha-nisms by which phosphorylated sugars regulate plant AMPK are unknown. Additionally, alternative STK11-independent modes of activation exist. In response to calcium flux, Thr172 can be directly phosphorylated by CAMKK2 (also known as CAMKKβ) kinase (Hawley et al., 2005; Hurley et al., 2005; Woods et al., 2005; Fogarty et al., 2010). In plants, only the CAMKK proteins GRIK1 and GRIK2 (Geminivirus Rep-interacting kinase 1 or 2) have been found as the activators of the plant AMPK (Shen et al., 2009). In animals and fungi, once AMPK has been activated, its activity is then modulated by the protein phosphatase-2A (PP2A) complex (Park et al., 2013) while in plants it is the protein phosphatase-2-CA that seems to regulate the plant AMPK activity (Rodrigues et al., 2013). It has been shown that hormones such as insulin can also affect AMPK activity (Kovacic et al., 2003; Ning et al., 2011; Valentine et al., 2014).
The AMPK complex influences the metabolic activity of the cell either directly or through downstream signaling cascades. In addition, it can adjust gene expression to cope with low energy availability (Tome et al., 2014). Four major groups of target proteins can be identified. (i) Proteins related to glycolysis, which are either directly or indirectly targeted by the AMPK complex. Examples for a direct regulation are 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase (PFK-2) and glycogen synthase (GYS). In turn, GLUT4, a glucose transporter, is an example of an indirect regulation by AMPK
(Wojtaszewski et al., 2003). (ii) Proteins involved in lipid metabolism and transport, such as hydroxymethylglutaryl-CoA reductase (HMGRC). These are also directly targeted by AMPK (Carling et al., 1991). (iii) Proteins involved in cell cycle control (cyclin A and D) (Huang et al., 2013; Wang et al., 2015). (iv) Proteins involved in translation. This latter group of proteins is particularly relevant as protein synthe-sis is a highly energy-demanding processes. Thus, the cell has to ensure an activation of this process only when sufficient energy supply is available.
Protein biosynthesis—while inhibited by AMPK—is induced by the target of rapamycin (TOR) pathway, a major regulator of growth in eukaryotes (Albert and Hall, 2015). Like AMPK, the TOR kinase is also conserved in the eukary-otic domain (van Dam et al., 2011). It has been well character-ized in animals (Hall, 2008; Eltschinger and Loewith, 2016), fungi (yeast) (Otsubo and Yamamato, 2008; Wei and Zheng, 2011), and plants (Arabidopsis thaliana) (Xiong and Sheen, 2015). The TOR kinase participates in two distinct heterotri-meric protein complexes, the TOR complex 1 (TORC1) and the TOR complex 2 (TORC2). In mammals, TORC1 com-prises the protein kinase TOR and two regulatory subunits RAPTOR (regulatory-associated protein of TOR) and LST8 (lethal with SEC13 protein 8). TORC2, composed of TOR, RICTOR (rapamycin-insensitive companion of TOR), and LST8 (Wullschleger et al., 2006), is absent in plants due to the lack of RICTOR. TORC1 directly phosphorylates the S6 kinase (S6K) which in turn phosphorylates the S6 subunit of the ribosomal complex (RPS6) and the translational repres-sor 4E-binding proteins (4E-BPs), which then dissociate from eIF4E thereby initiating translation (Ma and Blenis, 2009; Chauvin et al., 2014). TORC1 is activated by the membrane-associated Rheb (Li et al., 2010; Martin et al., 2014). Rheb in turn is repressed by the tuberous sclerosis complex (TSC1/2 complex) (Manning and Cantley, 2003; Saucedo et al., 2003). As well as intracellular signals, such as nutrient availability (Avruch et al., 2009), the TOR pathway also integrates extra-cellular signals (Oldham and Hafen, 2003). One of the best characterized extracellular activation pathways of TORC1 in animals is the insulin pathway. This activates the cas-cade phosphoinositide-3-kinase (PI3K), phosphoinositide-dependent kinase 1 (PDK1), and protein AKT, resulting in an activation of TORC1 via inhibition of the TSC1/2 com-plex (Shaw and Cantley, 2006). AMPK is also known to regu-late TOR activity by direct phosphorylation of the TSC1/2 complex and RAPTOR, a subunit of TORC1 (Inoki et al., 2003; Gwinn et al., 2008).
The mechanistic details of AMPK and TOR signaling in eukaryotes are increasingly well understood. In contrast, their precise evolutionary origins have only been poorly investigated. A more detailed view on the early onset of eukaryotic energy sensing is, however, interesting, especially in the context of eukaryotic organelle emergence. Obviously, the activities of the energy-generating organelles of endos-ymbiotic origin (i.e. mitochondria and chloroplasts) need to be co-ordinated with the energy metabolism in the cyto-sol (metabolic pool of the host). Both the AMPK and the TOR pathways have been implicated in retrograde organellar
3898 | Roustan et al.
Downloaded from https://academic.oup.com/jxb/article-abstract/67/13/3897/1749706by gueston 16 April 2018
signaling (Komeili et al., 2000; Ng et al., 2013). It remains unclear whether the two pathways emerged in the last eukary-otic common ancestor (LECA)—that is, after the formation of organelles—or whether for one or both pathways a pre-LECA version has existed that was later extended to facilitate host–organelle communication.
Eukaryotes probably arose by an endosymbiotic event involving an archaeon and an alpha-proteobacterium (Margulis et al., 2006). Accordingly, the foundations of the primordial eukaryotic gene set should have been mainly deter-mined by the genes encoded in the symbiont’s genomes. Thus, a substantial proportion of metabolic, but also of signaling, networks present nowadays in all eukaryotes can be expected to have their evolutionary roots in either the bacterial or the archaeal lineages. A recent search for higher plant proteins of cyanobacterial origin indicated that the kinase subunit of the higher plant AMPK (AKIN10) and the regulatory γ subunit (SNF4) is older than the eukaryotes (Bayer et al., 2014). Here we extended the evolutionary analysis to the entire AMPK and TOR pathways. We used the components of the human AMPK/TOR KEGG pathways as seeds, and searched for their orthologs in 412 diverse eukaryotic and prokaryotic genomes (Fig. 1). The resulting phylogenetic profiles for 153 proteins thus serve as the basis to reconstruct the evolutionary history of the AMPK/TOR regulatory network from its ori-gin in the pre-eukaryotic era to its representation in contem-porary eukaryotes. Against this background, we discuss the consequences for subcellular energy regulation in eukaryotes.
Methodological basis for exploring the evolutionary origins of the AMPK and TOR pathway components
The AMPK and TOR pathways are described most exten-sively in humans. Therefore, 124 proteins forming the AMPK signaling pathway and 60 proteins accomplishing TOR signaling were selected from the KEGG pathway database (Kanehisa and Goto, 2000). Thirty-one proteins have been assigned to both pathways, leaving a non-redundant set of 153 proteins. Each of these sequences served as the seed in a two-staged search tracing its counterparts in the gene sets of 412 species distributed across the tree of life (Supplementary Fig. S1; Supplementary Table S1 at JXB online). In stage one, we performed a targeted ortholog search for each protein with HaMStR-OneSeq (Ebersberger et al., 2014), an extension of the profile hidden Markov model- (pHMM) based ortholog search tool HaMStR (Ebersberger et al., 2009). The result-ing presence–absence pattern translates into a comprehensive phylogenetic profile across 330 eukaryotes, 26 archaea, and 57 bacteria. To assess whether the identified orthologs could have similar functionalities to the seed protein, their extended domain architectures (feature architecture) were compared using FACT (Koestler et al., 2010). This step is important as it helps in identifying orthologs which may have changed their functionality in the course of evolution, for exam-ple by domain gain or loss (Studer and Robinson-Rechavi, 2009; Gabaldon and Koonin, 2013). The resulting feature
architecture similarity (FAS) score ranges from 0 to 1, where 0 denotes no shared features between the compared proteins and a score of 1 indicates that all features in the seed protein also occur in its ortholog. Based on their FAS, the detected orthologs were grouped into two classes. ‘A’ orthologs have a FAS score >0.5, and all other orthologs are classified as ‘B’. In cases where no ortholog was identified in a given species, the analysis was continued with the more sensitive stage two of our search. Using the seed protein as query, the FAS scores of all proteins in the target species were evaluated and pro-teins with a FAS score >0.5 were retained. Of these, the high-est scoring protein was used to train a pHMM model with ‘hmmscan’ (Finn et al., 2015). Subsequently, we searched the human gene set with ‘hmmsearch’ (Finn et al., 2015) for sequences with a significant similarity to the pHMM. In cases where the seed protein was among the top five highest scor-ing hits, the candidate was kept; otherwise it was discarded. Subsequently, the phylogenetic profile from the ortholog search was complemented with the candidates obtained from the second search stage. To reduce the impact of spurious ortholog assignments for the pathway reconstruction, only orthologs that are present in at least 10% of the species in one clade were considered. Subsequently, the minimum age of the seed protein was assessed as the last common ancestor of all taxa harboring an ortholog of this protein. Using the KEGG mapper tool (Kanehisa and Goto, 2000; Kanehisa et al., 2016), the phyletic distribution of the individual components of AMPK and TOR pathways was visualized for plant, fungi, animals, archaea, and bacteria (Fig. 1). In the same way, the ancestral pathways for the LECA and for the last common ancestor of the opisthokonts were reconstructed by consid-ering only those proteins that could be traced back to these ancestral species (Fig. 1). For phylogenetic tree reconstruc-tion, individual ortholog groups were aligned with Muscle (Edgar, 2004), and the alignments were post-processed by removing alignment columns with >50% gaps. Subsequently, the best fitting model of sequence evolution was selected with ProtTest v.3.4.1 (Abascal et al., 2005), and maximum likelihood trees were reconstructed with RAxML v.8.1.15 performing 100 bootstrap replicates and mapping the branch support values to the maximum likelihood tree (Stamatakis et al., 2008). To reconstruct the evolutionary history of the kinase network at a greater level of resolution, maximum likelihood tree reconstruction of 24 contemporary kinases was integrated with an analysis of their PFAM domain archi-tecture evolution (Supplementary Fig. S2A).
Evolutionary conservation of metabolic components in eukaryotic AMPK–TOR pathways
Based on the phyletic distributions of the AMPK and TOR components, the evolution of the AMPK and TOR path-ways in the eukaryotic domain was addressed first (Fig. 1). As energy-sensing and signaling pathways, AMPK and TOR control the cell metabolism in accordance with the energy homeostasis. Therefore, proteins related to glycolysis and
Comparative genomics of AMPK–TOR signaling | 3899
Downloaded from https://academic.oup.com/jxb/article-abstract/67/13/3897/1749706by gueston 16 April 2018
glucose transport, lipid metabolism and mitochondrial activ-ity, carbohydrate storage, and biological processes related to protein translation, autophagy, and the cell cycle are directly under the control of AMPK and/or TOR. A majority of proteins related to these processes were present already in the LECA, while proteins related to G6Pase, GYS, CD36, HNF4α, CRTC2, FOXO, PGC-1α, 4EBP1, HIF1α, PPARγ, and VEGF appeared later in evolution (Fig. 1; Supplementary Table S1). From these genes, GYS and FOXO arose on the opisthokont lineage, while G6Pase, CD36, HNF4α, CRTC2, PGC-1α, 4EBP1, PPARγ, and VEGF resemble metazoan inventions.
Interestingly, evidence also exists for the loss of some key proteins, for example in fungi (e.g. PEPCK, MCD, and cyclinD1) and/or in plants (e.g. PEPCK, FASN, CPT1, and SIRT1). In some instances, this is different from previous studies claiming that the corresponding functionalities are present in these kingdoms (PEPCK, fungi/plants; FASN and CPT1, plants) (Wood et al., 1984; Chen et al., 2002; Hynes et al., 2002; Brown et al., 2006; Li-Beisson et al., 2010). PEPCK may serve as an illustrative example. The protein appears confined to the amoeba and metazoan line-ages, but orthologs were found in both prokaryotic domains (Supplementary Table S1). This implies that PEPCK belongs to the oldest fraction in the data set. In line with their shared evolutionary descent, the human and most of the prokary-otic PEPCK proteins are GTP dependent. In plants and in fungi proteins have been described that exert a PEPCK-like activity, which, however, are not identified as orthologs to the human PEPCK in our analysis, indicating an inde-pendent evolutionary origin. In support of this, the plant PEPCK-like proteins are ATP dependent, as are many of the fungal PEPCK-like proteins (Aich and Delbaere, 2007). Interestingly, despite the different evolutionary origins of PEPCK functionality in plants and fungi, evidence from A. thaliana suggests that the regulation of all these proteins is AMPK dependent (Baena-Gonzalez et al., 2007). Thus, it appears that the same function can be accomplished by dif-ferent and non-homologous proteins.
Conservation of AMPK–TOR signaling components in eukaryotes
The analysis of the phylogenetic profiles (Fig. 1; Supplementary Table S1) for inferring the evolutionary age of the TOR pathway indicates that both TORC1 and TORC2 were present already in the LECA. Similarly, the RAG complex, which regulates TORC1 based on the avail-ability of amino acids, is as old as the eukaryotes (Fig. 1; Supplementary Table S1). The same also applies to the upstream components of signal transduction such as Rheb, TSC2, and the pathways IKBKB, ERK. and AKT, which integrate intracellular energy signals or extracellular sig-nals. Tracing the evolution of this ancestral TOR pathway towards the contemporary species reveals that the TOR pathway has been, with clear exceptions in plants, consid-erably stable. A few modifications include the duplication
of an ancestral TSC2-like gene on the opisthokont line-age to form the contemporary paralogs TSC1 and TSC2, and the invention of REDD1, INS, IGF, and IRS1 in the metazoan ancestor. PRAS40 is an even more recent addi-tion to the pathway that appears confined to the verte-brates. Interestingly, orthologs related to PI3K were not found in plants and fungi, while PI3K catalytic subunits are present in some stramenopiles and Emiliania huxeleyi (Supplementary Table S1). This apparent gene loss in plants is in line with a previous finding that plants lack proteins related to either PIK3CA, PIK3CD, or PIK3CG (Brown and Auger, 2011). Indeed, the situation in plants is even more complex as additional losses seem to have occurred for TSC2, Rheb, RICTOR, and RAG proteins (compare Vernoud et al., 2003; Berkowitz et al., 2008; Rehmann et al., 2008; van Dam et al., 2011).
The TOR complex is directly and indirectly regulated by AMPK in humans (Mihaylova and Shaw, 2011). All kinases (TAK1, LKB1, and CAMKKβ) and phosphatases (PP2A) regulating AMPK activity have been traced back to almost all eukaryotic species. No orthologs were found in fungi and plants for STRADA and STRADB proteins, while proteins related to leptin signaling were present only in metazoans. It is important to highlight that in plants the AMPK com-plex is not directly regulated by AMP but rather by phos-phorylated sugar (Nunes et al., 2013a, b; Crozet et al., 2014; Nagele and Weckwerth, 2014; Yadav et al., 2014). It remains unknown if this mode of regulation is plant specific or rep-resents a more ancient mechanism. AMPK plant specificity is further characterized by the expansion of the plant SnRK family in three subfamilies (Halford and Hey, 2009), suggest-ing that there is a plant-specific adaptation of the AMPK pathway. This opens up novel hypotheses for the analysis of the AMPK/TOR evolution in autotrophic organisms. In the following section, AMPK/TOR protein kinases in the LECA are discussed.
Seven distinct protein kinases involved in AMPK/TOR signaling might have been present in the last eukaryotic common ancestor
Kinases are the key players in the signal transduction and reg-ulative networks of AMPK and TOR. Phylogenetic profiles for the kinases revealed that orthologs are present mostly in bacteria in the prokaryotic domain. This reflects a previous observation that among prokaryotes, mainly bacteria—and here particularly strains forming multicellular aggregates, such as Proteobacteria, Actinobacteria, and Cyanobacteria—harbor eukaryotic-like protein kinases (ELKs) (Perez et al., 2008). Along this line, we found orthologs to the highest num-ber of AMPK/TOR pathway-related kinases in Sorangium cellulosum (Myxococcales, Proteobacteria), which is also among the species with the highest number of ELKs (317) (Perez et al., 2008).
A reconstruction of the evolutionary history of the kinase network (Supplementary Fig. S2A) reveals that the
Comparative genomics of AMPK–TOR signaling | 3901
Downloaded from https://academic.oup.com/jxb/article-abstract/67/13/3897/1749706by gueston 16 April 2018
contemporary kinases are descendants of seven ancient kinases that were already present in the LECA (Fig. 2). Furthermore, most of the PFAM domain architectures seen nowadays in these kinases already existed at that time. This lends support to the hypothesis that the kinase network of AMPK/TOR might have already been functional before the diversification of the eukaryotic domain. The connection to the pre-eukaryotic era remains, however, speculative. There is no phylogenetic evidence that the prokaryotic orthologs that were identified in the course of our analyses are more closely related to any of the eukaryotic kinases. Based on the domain architectures of the pro- and eukaryotic kinases, however, it is tempting to speculate that the LECA kinases 3, 4, 5, and 6, representing the predecessors of the contemporary kinases CAMKK1, STK11, PRKAA, and MAPK1/3 (cf. Fig. 2), are the oldest eukaryotic kinases, and are direct descendants of pre-eukaryotic proto-kinases. They share with most of the
bacterial kinases a simple PFAM domain architecture con-sisting only of a single Pkinase domain (Supplementary Fig. S2B). From this basis, and via the addition of further Pfam domains, LECA 1 (plus Pkinase C-terminal domain), LECA 2 (plus PH domain and Pkinase C-terminal domain), and LECA 7 (plus DUF3543) could have emerged. Note that the addition of a second, C-terminal Pkinase domain in RPS6K-A appears as a metazoan invention. The corre-sponding domain is only very distantly related to the Pkinase domains in all other kinases and its evolutionary origin is unclear (Supplementary Fig. S2B).
Phylogenetic analysis of AMPK core components reveals potential origins in the pre-eukaryotic era for the α and γ subunits
The function of AMPK is conserved across the eukaryotic domain, indicating that it is essential for survival. Here, the evolutionary history of the genes conveying this functional-ity was studied. Several studies have shown that humans and Arabidopsis possess two and three genes, respectively, encod-ing the α subunits (Hardie et al., 2003), whereas fungi have only one. The tree in Supplementary Fig. S4A reveals that the two human genes, PRKAA1 and PRKAA2, are recent paral-ogs that emerged on the vertebrate lineage. Likewise, the three genes in A. thaliana share a common ancestor to the exclusion of other plant genes, indicative of two rounds of gene dupli-cation on the Arabidopsis lineage. Similar but independent duplication events are seen in rice and Selaginella moellen-dorffii. Fungi, as well as invertebrate animals, however, have retained the ancestral state of only one single AMPKα gene. A sequence alignment of the human AMPKα kinase domain with seven bacterial orthologs revealed that the ATP-binding motif, the essential catalytic residues, and Thr172 are con-served (Supplementary Fig. S3).
AMPKβ and AMPKγ display the same pattern of lineage-specific duplications as the α subunit (Fig. 3; Supplementary Fig. 4B). The β subunit is, however, confined to the eukaryotes, while the γ subunit resembles findings of the AMPKα, as its orthologs are also present in the prokar-yotes (Fig. 3). Independent support for the existence of a functionally equivalent AMPKγ subunit in archaea (King et al., 2008) comes from crystallographic structure analyses. The AMPKγ ortholog in the archaeon Methanocaldococcus jannaschii harbors four cystathionine β-synthase domains, resulting in a similar structure to the eukaryotic AMPKγ regulatory subunit (Gomez-Garcia et al., 2010). The tree further reveals one interesting finding with respect to the AMPK in plants. The land plant-specific AMPKβγ subunit is more closely related to the genes encoding the γ subu-nit in animals than to KING1, which was originally identi-fied as the A. thaliana γ subunit (Fig. 3). An independent phylogenetic profiling of KING1 reveals that its orthologs are widespread among eukaryotes and form a distinct clade that is clearly distinct from the AMPKγ clade (Fig. 3). This implies that KING1 has a functionality that is at least not directly linked to AMPK. In line with this observation, it
Fig. 2. The evolutionary history of the protein kinases involved in AMPK–TOR signaling revealed seven proto-LECA protein kinases. A reconstruction of phylogenetic relationships (maximum likelihood tree reconstruction) was integrated with PFAM domain architecture information. In the LECA, seven distinct kinases (marked with 1–7) might have been present. LECA 3–LECA 6 possess only a single Pkinase pfam domain and represent the contemporary human kinases CAMKK1, STK11, PRKAA, and MAPK1/3. The bacterial kinases share with their eukaryotic counterparts only the possession of the Pkinase Pfam domain (PK). (This figure is available in colour at JXB online.)
3902 | Roustan et al.
Downloaded from https://academic.oup.com/jxb/article-abstract/67/13/3897/1749706by gueston 16 April 2018
was previously shown that of the two proteins, KING1 and AMPKβγ, only the latter interacts with the plant AMPK complex (Gissot et al., 2006; Ramon et al., 2013; Emanuelle et al., 2015). Moreover, the AMPKβγ subunit is able to complement a snf4 mutant in yeast, indicating their func-tional equivalency (Bouly et al., 1999; Kleinow et al., 2000; Lumbreras et al., 2001). Integrating both observations with the phylogenetic evidence results in a new view on the evolution of the regulatory γ subunit in plants. It strongly suggests that the original γ subunit in plants underwent a gene fusion to form the present-day βγ subunit (Lumbreras et al., 2001). This fusion gene continued to exert the ances-tral functionality of the AMPKγ subunit. KING1, on the other hand, is related to the fungal SDS23 protein. Still, an indirect link between KING1 and the AMPK complex may exist. In yeast, SDS23 interacts with PP2A and PP2A-related phosphatase complexes. It was proposed that SDS23 inhib-its phosphatase activity during glucose limitation (Hanyu et al., 2009). Additionally, yeast PP2A has been shown to regulate the AMPK activity by dephosphorylating Thr210 (Ruiz et al., 2011). Similar to SDS23 in fungi, KING1 could act as an alternative energy sensor that regulates—for example via intermediate phosphatases, such as PP2A—the activity of AMPK. A corresponding working hypothesis is outlined in Fig. 4. An initial indication that a functional connection between an energy sensor (AMPKγ or KING1) and PP2A may indeed exist comes from the co-occurrence
of both proteins in the phylogenetic profiles. in addition to the energy sensor, only one further protein involved in AMPK regulation was consistently found in bacteria and archaea, the PP2A catalytic subunit.
Consequences of endosymbiosis and subcellular compartmentation on energy sensing and signaling in eukaryotes
The presence of an energy sensor (AMPKγ) and a regu-latory phosphatase (PP2A) suggests that the seed of the AMPK pathway exists in prokaryotes. Based on the phy-logenetic profiles generated in this study, this seed can now be extended with additional functionalities. Orthologs for FBP, PCK, PFKFB, PP2C, GLUT4, RAB11B, ULK1, and EEF2 were found in both archaea and bacteria (Fig. 1; Supplementary Table S1). HMGCR and RPS6 orthologs were seen only in archaea, whereas orthologs for the protein kinases PRKAA (α regulatory subunit of AMPK), STK11, TAK1, MAPK1, RSKs, CAMKK-β, AKT, S6K, PKC, BRAF, and IKBKB appear confined to the bacteria (Fig. 1; Supplementary Table S1). Similarly, some components involved in the eukaryotic energy storage and availability pathways were also detected only in bacteria; for example, PFK-1, a protein involved in glycolysis, and SCD and FASN involved in fatty acid biosynthesis. With respect to the TOR
Fig. 3. The evolutionary relationships of the AMPKγ and KING1 gene families. Maximum likelihood tree of selected orthologs of the human AMPKγ subunits and of arabidopsis KING1. The tree reveals that the plant protein KING1, which was formerly attributed as AMPKγ, belongs to a different gene family that is only distantly related to the AMPKγ subunit. Instead, the plant AMPKβγ subunit represents the ‘true’ AMPKγ in plants, which, however, was modified by a domain addition on the plant lineage. Branch labels denote percentage bootstrap support of the corresponding splits. The gene duplication giving rise to the three human AMPKγ subunits occurred on the vertebrate lineage after the split of Drosophila. Accordingly, all orthologs of the human proteins were subsumed under the name G1-3. (This figure is available in colour at JXB online.)
Comparative genomics of AMPK–TOR signaling | 3903
Downloaded from https://academic.oup.com/jxb/article-abstract/67/13/3897/1749706by gueston 16 April 2018
pathway, apart from AKT, S6K, MAPK1, RSK, ULK pro-tein kinases, and HIF1α, a transcription factor involved in hypoxia response (Wang et al., 1995), no other proteins were found in prokaryotes.
Based on the above evidence, a proto-LECA pathway was reconstructed that would have resulted from a bacte-rial/archaeal endosymbiosis event to form the primordial eukaryote (Fig. 1). Subsequently, this proto-LECA pathway layout was compared with the one that results when integrat-ing the components that could be traced back to the LECA, but which have no counterparts in the prokaryotic domains (Fig. 1).
There are three major changes when comparing the proto-LECA pathways with the LECA versions. First, the β subu-nit complementing the AMPK complex was added. This subunit has been proposed to facilitate substrate specificity and subcellular localization of the AMPK complex (Pierre et al., 2007; Polge et al., 2008). Secondly, genes associated with mitochondrial metabolism, such as ACC1 and 2, MCD, and CPT1, arose in line with the emergence of mitochondria prior to the diversification of eukaryotes. Thirdly, the TOR pathway, as it is seen nowadays in eukaryotes, was formed. This includes the emergence of TSC2, Rheb, and the TORC1 and 2 complexes. Also, PIPK3 proteins and the downstream pathway controlling the TOR activity emerged in the LECA. PIP3 is often associated with insulin signaling in multicellular
metazoans. However, this molecule has also been found to play an important role in chemotaxis and cell motility in unicellular eukaryotes (Devreotes and Janetopoulos, 2003; Gamba et al., 2005). The presence of the PIPK3 signaling components in the LECA suggests that the latter functional-ity is ancestral.
Our comparison of the primordial AMPK/TOR pathways with the LECA versions suggests a concomitant appearance of mitochondria and the TOR signaling pathway. In line with the presence of the mitochondria, the TORC complex/path-way acts downstream of the glycolysis-mitochondrial electron transfer chain (ETC) in mammals (Kwon et al., 2011) and plants (Xiong et al., 2013; Xiong and Sheen, 2015). Therefore, it can be speculated that the TOR pathway emerged to form a crossroads between PIK3/AKT signaling and mitochondria, and further to mediate retrograde signaling (Komeili et al., 2000). In animals it was proposed that TOR signaling links the status of mitochondria and energy production to transla-tional activity and growth by controlling both mitochondrial activity and biogenesis through 4E-BP-dependent transla-tional regulation (Morita et al., 2013, 2015).
Conclusion
Energy sensing and signaling is an essential process forming the basis of organismal life. Based on a comprehensive phy-logenetic profiling of the AMPK and TOR pathway compo-nents paired with maximum likelihood tree reconstruction, a fine-grained scenario reconstructing the evolutionary history of the corresponding gene interaction network was created. It was shown that the AMPK and TOR pathway comprises a core set of proteins, which are widely conserved across eukary-otes. This underlines the ubiquitous relevance of this function-ality in eukaryotes. In line with this notion is the evolutionary age of the kinase network within AMPK/TOR. The presence of seven LECA kinases spanning almost the full contempo-rary diversity of contemporary domain architectures indicates that kinase regulation could have been fully active prior to the diversification of eukaryotes. Notably, evidence is provided that KING1, which has hitherto been assigned as AMPKγ, repre-sents a distinct evolutionary lineage and is more closely related to SDS23 in yeast. On this basis, it is proposed that KING1 is as an alternative energy sensor, which could—similarly to the situation in yeast—act in conjunction with an intermediate phosphatase, such as PP2A, in the regulation of downstream processes. While the kinases have their evolutionary origins in the bacterial domain, the energy sensor, AMPKγ, is well con-served between eukaryotes and archaea, indicating that both domains might share a related system for energy sensing. This suggests a scenario where the formation of the eukaryotic AMPK pathway emerged as a synergy from the endosymbio-sis between a bacterium providing the regulatory network and an archaeon providing the energy sensor. This basic system was then complemented by a substantial number of proteins involved in metabolic activity whose evolutionary ages are older than the eukaryotes. At the eukaryotic transition, a diversifica-tion of the pathways seems to have occurred. In particular, the
Fig. 4. Working hypothesis on the functional role of plant KING1. It is hypothesized, here, that plant KING1 could have a similar function to SDS23 in fungi (Hanyu et al., 2009). We therefore propose that plant KING1 could regulate its potential interactors based on the energy level. Additionally, it was proposed that the TOR pathway in plants promotes Tap46 expression (Ahn et al., 2011). Tap46 is a regulatory subunit of plant PP2A, which promote TOR signaling (protects S6K from PP2A) (Ahn et al., 2015). Since in vivo regulation of plant AMPK (SnRK1) dephosphorylation by PP2C is debated (Rodrigues et al., 2013; Emanuelle et al., 2015), it would be reasonable to investigate further the PP2A and PP2A-related phosphatase activity on plant AMPK phosphorylation as shown for yeast (Ruiz et al., 2011). Question marks correspond to so far missing links. LES, Low Energy Syndrome; PP2C, protein phosphatase 2C; PP2A, type 2A protein phosphatase; PPase, PP2A-related protein phosphatase; AMPKK, this corresponds to the upstream kinases which phosphorylate plant AMPK; in Arabidopsis it is SnaK1 and SnaK2; Tap46, regulatory subunit of plant PP2A complex. (This figure is available in colour at JXB online.)
3904 | Roustan et al.
Downloaded from https://academic.oup.com/jxb/article-abstract/67/13/3897/1749706by gueston 16 April 2018
TOR complex appeared with mitochondrial development. It is proposed that the TOR complex and downstream proteins could play an important role in retrograde signaling.
Supplementary data
Supplementary data are available at JXB online.Table S1. Combined matrix between the ortholog search
and feature architecture similarity search.Table S2. Correspondence between gene names in
Supplementary Table S1 and Figure 1.Figure S1. Workflow of the feature aware phylogenetic
profiling.Figure S2. Phylogenetic analysis of the kinases involved in
AMPK/TOR pathways.Figure S3. Structure-based alignments of the AMPKα2
subunit with seven bacterial orthologs.Figure S4. Phylogenetic analyses of the AMPK α and β
subunits.
AcknowledgmentsVR and AJ were financed by the Marie Curie ITN project CALIPSO (GA ITN-2013 607 607). We thank Alexander Kirchmair, Ella Nukarinen, Andrea Mair, and Bernhard Wurzinger for helpful discussions.
ReferencesAbascal F, Zardoya R, Posada D. 2005. ProtTest: selection of best-fit models of protein evolution. Bioinformatics 21, 2104–2105.
Ahn CS, Ahn HK, Pai HS. 2015. Overexpression of the PP2A regulatory subunit Tap46 leads to enhanced plant growth through stimulation of the TOR signalling pathway. Journal of Experimental Botany 66, 827–840.
Ahn CS, Han JA, Lee HS, Lee S, Pai HS. 2011. The PP2A regulatory subunit Tap46, a component of the TOR signaling pathway, modulates growth and metabolism in plants. The Plant Cell 23, 185–209.
Aich S, Delbaere LT. 2007. Phylogenetic study of the evolution of PEP-carboxykinase. Evolutionary Boinformatics Online 3, 333–340.
Albert V, Hall MN. 2015. mTOR signaling in cellular and organismal energetics. Current Opinion in Cell Biology 33, 55–66.
Avruch J, Long X, Ortiz-Vega S, Rapley J, Papageorgiou A, Dai N. 2009. Amino acid regulation of TOR complex 1. American Journal of Physiology. Endocrinology and Metabolism 296, E592–E602.
Baena-Gonzalez E, Rolland F, Thevelein JM, Sheen J. 2007. A central integrator of transcription networks in plant stress and energy signalling. Nature 448, 938–942.
Bayer RG, Kostler T, Jain A, Stael S, Ebersberger I, Teige M. 2014. Higher plant proteins of cyanobacterial origin: are they or are they not preferentially targeted to chloroplasts? Molecular Plant 7, 1797–1800.
Berkowitz O, Jost R, Pollmann S, Masle J. 2008. Characterization of TCTP, the translationally controlled tumor protein, from Arabidopsis thaliana. The Plant Cell 20, 3430–3447.
Bouly JP, Gissot L, Lessard P, Kreis M, Thomas M. 1999. Arabidopsis thaliana proteins related to the yeast SIP and SNF4 interact with AKINalpha1, an SNF1-like protein kinase. The Plant Journal 18, 541–550.
Brown AP, Affleck V, Fawcett T, Slabas AR. 2006. Tandem affinity purification tagging of fatty acid biosynthetic enzymes in Synechocystis sp. PCC6803 and Arabidopsis thaliana. Journal of Experimental Botany 57, 1563–1571.
Brown JR, Auger KR. 2011. Phylogenomics of phosphoinositide lipid kinases: perspectives on the evolution of second messenger signaling and drug discovery. BMC Evolutionary Biology 11, 4.
Carling D, Clarke PR, Hardie DG. 1991. Adenosine monophosphate-activated protein kinase: hydroxymethylglutaryl-CoA reductase kinase. Methods in Enzymology 200, 362–371.
Chauvin C, Koka V, Nouschi A, et al. 2014. Ribosomal protein S6 kinase activity controls the ribosome biogenesis transcriptional program. Oncogene 33, 474–483.
Chen ZH, Walker RP, Acheson RM, Leegood RC. 2002. Phosphoenolpyruvate carboxykinase assayed at physiological concentrations of metal ions has a high affinity for CO2. Plant Physiology 128, 160–164.
Crozet P, Margalha L, Confraria A, Rodrigues A, Martinho C, Adamo M, Elias CA, Baena-Gonzalez E. 2014. Mechanisms of regulation of SNF1/AMPK/SnRK1 protein kinases. Frontiers in Plant Science 5, 190.
Devreotes P, Janetopoulos C. 2003. Eukaryotic chemotaxis: distinctions between directional sensing and polarization. Journal of Biological Chemistry 278, 20445–20448.
Dunlop EA, Tee AR. 2009. Mammalian target of rapamycin complex 1: signalling inputs, substrates and feedback mechanisms. Cellular Signalling 21, 827–835.
Ebersberger I, Simm S, Leisegang MS, Schmitzberger P, Mirus O, von Haeseler A, Bohnsack MT, Schleiff E. 2014. The evolution of the ribosome biogenesis pathway from a yeast perspective. Nucleic Acids Research 42, 1509–1523.
Ebersberger I, Strauss S, von Haeseler A. 2009. HaMStR: profile hidden markov model based search for orthologs in ESTs. BMC Evolutionary Biology 9, 157.
Edgar RC. 2004. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Research 32, 1792–1797.
Eltschinger S, Loewith R. 2016. TOR complexes and the maintenance of cellular homeostasis. Trends in Cell Biology 26, 148–159.
Emanuelle S, Hossain MI, Moller IE, et al. 2015. SnRK1 from Arabidopsis thaliana is an atypical AMPK. The Plant Journal 82, 183–192.
Finn RD, Clements J, Arndt W, Miller BL, Wheeler TJ, Schreiber F, Bateman A, Eddy SR. 2015. HMMER web server: 2015 update. Nucleic Acids Research 43, W30–W38.
Fogarty S, Hawley SA, Green KA, Saner N, Mustard KJ, Hardie DG. 2010. Calmodulin-dependent protein kinase kinase-beta activates AMPK without forming a stable complex: synergistic effects of Ca2+ and AMP. Biochemical Journal 426, 109–118.
Gabaldon T, Koonin EV. 2013. Functional and evolutionary implications of gene orthology. Nature Reviews Genetics 14, 360–366.
Gamba A, de Candia A, Di Talia S, Coniglio A, Bussolino F, Serini G. 2005. Diffusion-limited phase separation in eukaryotic chemotaxis. Proceedings of the National Academy of Sciences, USA 102, 16927–16932.
Gissot L, Polge C, Bouly JP, Lemaitre T, Kreis M, Thomas M. 2004. AKINbeta3, a plant specific SnRK1 protein, is lacking domains present in yeast and mammals non-catalytic beta-subunits. Plant Molecular Biology 56, 747–759.
Gissot L, Polge C, Jossier M, Girin T, Bouly JP, Kreis M, Thomas M. 2006. AKINbetagamma contributes to SnRK1 heterotrimeric complexes and interacts with two proteins implicated in plant pathogen resistance through its KIS/GBD sequence. Plant Physiology 142, 931–944.
Gomez-Garcia I, Oyenarte I, Martinez-Cruz LA. 2010. The crystal structure of protein MJ1225 from Methanocaldococcus jannaschii shows strong conservation of key structural features seen in the eukaryal gamma-AMPK. Journal of Molecular Biology 399, 53–70.
Gwinn DM, Shackelford DB, Egan DF, Mihaylova MM, Mery A, Vasquez DS, Turk BE, Shaw RJ. 2008. AMPK phosphorylation of raptor mediates a metabolic checkpoint. Molecular Cell 30, 214–226.
Halford NG, Hey S, Jhurreea D, Laurie S, McKibbin RS, Paul M, Zhang Y. 2003. Metabolic signalling and carbon partitioning: role of Snf1-related (SnRK1) protein kinase. Journal of Experimental Botany 54, 467–475.
Halford NG, Hey S, Jhurreea D, Laurie S, McKibbin RS, Zhang Y, Paul MJ. 2004. Highly conserved protein kinases involved in the regulation of carbon and amino acid metabolism. Journal of Experimental Botany 55, 35–42.
Halford NG, Hey SJ. 2009. Snf1-related protein kinases (SnRKs) act within an intricate network that links metabolic and stress signalling in plants. Biochemical Journal 419, 247–259.
Comparative genomics of AMPK–TOR signaling | 3905
Downloaded from https://academic.oup.com/jxb/article-abstract/67/13/3897/1749706by gueston 16 April 2018
Hall MN. 2008. mTOR—what does it do? Transplantation Proceedings 40, S5–S8.
Hanyu Y, Imai KK, Kawasaki Y, et al. 2009. Schizosaccharomyces pombe cell division cycle under limited glucose requires Ssp1 kinase, the putative CaMKK, and Sds23, a PP2A-related phosphatase inhibitor. Genes to Cells 14, 539–554.
Hardie DG. 2014. AMPK—sensing energy while talking to other signaling pathways. Cell Metabolism 20, 939–952.
Hardie DG, Scott JW, Pan DA, Hudson ER. 2003. Management of cellular energy by the AMP-activated protein kinase system. FEBS Letters 546, 113–120.
Hawley SA, Boudeau J, Reid JL, Mustard KJ, Udd L, Makela TP, Alessi DR, Hardie DG. 2003. Complexes between the LKB1 tumor suppressor, STRAD alpha/beta and MO25 alpha/beta are upstream kinases in the AMP-activated protein kinase cascade. Journal of Biology 2, 28.
Hawley SA, Pan DA, Mustard KJ, Ross L, Bain J, Edelman AM, Frenguelli BG, Hardie DG. 2005. Calmodulin-dependent protein kinase kinase-beta is an alternative upstream kinase for AMP-activated protein kinase. Cell Metabolism 2, 9–19.
Huang WW, Tsai SC, Peng SF, Lin MW, Chiang JH, Chiu YJ, Fushiya S, Tseng MT, Yang JS. 2013. Kaempferol induces autophagy through AMPK and AKT signaling molecules and causes G2/M arrest via downregulation of CDK1/cyclin B in SK-HEP-1 human hepatic cancer cells. International Journal of Oncology 42, 2069–2077.
Hurley RL, Anderson KA, Franzone JM, Kemp BE, Means AR, Witters LA. 2005. The Ca2+/calmodulin-dependent protein kinase kinases are AMP-activated protein kinase kinases. Journal of Biological Chemistry 280, 29060–29066.
Hynes MJ, Draht OW, Davis MA. 2002. Regulation of the acuF gene, encoding phosphoenolpyruvate carboxykinase in the filamentous fungus Aspergillus nidulans. Journal of Bacteriology 184, 183–190.
Inoki K, Zhu T, Guan KL. 2003. TSC2 mediates cellular energy response to control cell growth and survival. Cell 115, 577–590.
Kanehisa M, Goto S. 2000. KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic Acids Research 28, 27–30.
Kanehisa M, Sato Y, Kawashima M, Furumichi M, Tanabe M. 2016. KEGG as a reference resource for gene and protein annotation. Nucleic Acids Research 44, D457–D462.
King NP, Lee TM, Sawaya MR, Cascio D, Yeates TO. 2008. Structures and functional implications of an AMP-binding cystathionine beta-synthase domain protein from a hyperthermophilic archaeon. Journal of Molecular Biology 380, 181–192.
Kleinow T, Bhalerao R, Breuer F, Umeda M, Salchert K, Koncz C. 2000. Functional identification of an Arabidopsis snf4 ortholog by screening for heterologous multicopy suppressors of snf4 deficiency in yeast. The Plant Journal 23, 115–122.
Koestler T, von Haeseler A, Ebersberger I. 2010. FACT: functional annotation transfer between proteins with similar feature architectures. BMC Bioinformatics 11, 417.
Komeili A, Wedaman KP, O’Shea EK, Powers T. 2000. Mechanism of metabolic control. Target of rapamycin signaling links nitrogen quality to the activity of the Rtg1 and Rtg3 transcription factors. Journal of Cell Biology 151, 863–878.
Kovacic S, Soltys CL, Barr AJ, Shiojima I, Walsh K, Dyck JR. 2003. Akt activity negatively regulates phosphorylation of AMP-activated protein kinase in the heart. Journal of Biological Chemistry 278, 39422–39427.
Kwon KY, Viollet B, Yoo OJ. 2011. CCCP induces autophagy in an AMPK-independent manner. Biochemical and Biophysical Research Communications 416, 343–348.
Laplante M, Sabatini DM. 2012. mTOR signaling in growth control and disease. Cell 149, 274–293.
Li L, Kim E, Yuan H, Inoki K, Goraksha-Hicks P, Schiesher RL, Neufeld TP, Guan KL. 2010. Regulation of mTORC1 by the Rab and Arf GTPases. Journal of Biological Chemistry 285, 19705–19709.
Li-Beisson Y, Shorrosh B, Beisson F, et al. 2010. Acyl-lipid metabolism. Arabidopsis Book 8, e0133.
Lumbreras V, Alba MM, Kleinow T, Koncz C, Pages M. 2001. Domain fusion between SNF1-related kinase subunits during plant evolution. EMBO Reports 2, 55–60.
Ma XM, Blenis J. 2009. Molecular mechanisms of mTOR-mediated translational control. Nature Reviews Molecular Cell Biology 10, 307–318.
Manning BD, Cantley LC. 2003. Rheb fills a GAP between TSC and TOR. Trends in Biochemical Sciences 28, 573–576.
Margulis L, Chapman M, Guerrero R, Hall J. 2006. The last eukaryotic common ancestor (LECA): acquisition of cytoskeletal motility from aerotolerant spirochetes in the Proterozoic Eon. Proceedings of the National Academy of Sciences, USA 103, 13080–13085.
Martin TD, Chen XW, Kaplan REW, Saltiel AR, Walker CL, Reiner DJ, Der CJ. 2014. Ral and Rheb GTPase activating proteins integrate mTOR and GTPase signaling in aging, autophagy, and tumor cell invasion. Molecular Cell 53, 209–220.
Mihaylova MM, Shaw RJ. 2011. The AMPK signalling pathway coordinates cell growth, autophagy and metabolism. Nature Cell Biology 13, 1016–1023.
Morita M, Gravel SP, Chenard V, et al. 2013. mTORC1 controls mitochondrial activity and biogenesis through 4E-BP-dependent translational regulation. Cell Metabolism 18, 698–711.
Morita M, Gravel SP, Hulea L, Larsson O, Pollak M, St-Pierre J, Topisirovic I. 2015. mTOR coordinates protein synthesis, mitochondrial activity and proliferation. Cell Cycle 14, 473–480.
Nagele T, Weckwerth W. 2014. Mathematical modeling reveals that metabolic feedback regulation of SnRK1 and hexokinase is sufficient to control sugar homeostasis from energy depletion to full recovery. Frontiers in Plant Science 5, 365.
Ng S, Giraud E, Duncan O, et al. 2013. Cyclin-dependent kinase E1 (CDKE1) provides a cellular switch in plants between growth and stress responses. Journal of Biological Chemistry 288, 3449–3459.
Ning J, Xi G, Clemmons DR. 2011. Suppression of AMPK activation via S485 phosphorylation by IGF-I during hyperglycemia is mediated by AKT activation in vascular smooth muscle cells. Endocrinology 152, 3143–3154.
Nunes C, O’Hara LE, Primavesi LF, Delatte TL, Schluepmann H, Somsen GW, Silva AB, Fevereiro PS, Wingler A, Paul MJ. 2013a. The trehalose 6-phosphate/SnRK1 signaling pathway primes growth recovery following relief of sink limitation. Plant Physiology 162, 1720–1732.
Nunes C, Primavesi LF, Patel MK, Martinez-Barajas E, Powers SJ, Sagar R, Fevereiro PS, Davis BG, Paul MJ. 2013b. Inhibition of SnRK1 by metabolites: tissue-dependent effects and cooperative inhibition by glucose 1-phosphate in combination with trehalose 6-phosphate. Plant Physiology and Biochemistry 63, 89–98.
Oakhill JS, Scott JW, Kemp BE. 2012. AMPK functions as an adenylate charge-regulated protein kinase. Trends in Endocrinology and Metabolism 23, 125–132.
Oldham S, Hafen E. 2003. Insulin/IGF and target of rapamycin signaling: a TOR de force in growth control. Trends in Cell Biology 13, 79–85.
Otsubo Y, Yamamato M. 2008. TOR signaling in fission yeast. Critical Reviews in Biochemistry and Molecular Biology 43, 277–283.
Park S, Scheffler TL, Rossie SS, Gerrard DE. 2013. AMPK activity is regulated by calcium-mediated protein phosphatase 2A activity. Cell Calcium 53, 217–223.
Perez J, Castaneda-Garcia A, Jenke-Kodama H, Muller R, Munoz-Dorado J. 2008. Eukaryotic-like protein kinases in the prokaryotes and the myxobacterial kinome. Proceedings of the National Academy of Sciences, USA 105, 15950–15955.
Pierre M, Traverso JA, Boisson B, Domenichini S, Bouchez D, Giglione C, Meinnel T. 2007. N-myristoylation regulates the SnRK1 pathway in Arabidopsis. The Plant Cell 19, 2804–2821.
Polge C, Jossier M, Crozet P, Gissot L, Thomas M. 2008. Beta-subunits of the SnRK1 complexes share a common ancestral function together with expression and function specificities; physical interaction with nitrate reductase specifically occurs via AKINbeta1-subunit. Plant Physiology 148, 1570–1582.
Polge C, Thomas M. 2007. SNF1/AMPK/SnRK1 kinases, global regulators at the heart of energy control? Trends in Plant Sciences 12, 20–28.
Ramon M, Ruelens P, Li Y, Sheen J, Geuten K, Rolland F. 2013. The hybrid four-CBS-domain KINbetagamma subunit functions as the canonical gamma subunit of the plant energy sensor SnRK1. The Plant Journal 75, 11–25.
3906 | Roustan et al.
Downloaded from https://academic.oup.com/jxb/article-abstract/67/13/3897/1749706by gueston 16 April 2018
Rehmann H, Bruning M, Berghaus C, Schwarten M, Kohler K, Stocker H, Stoll R, Zwartkruis FJ, Wittinghofer A. 2008. Biochemical characterisation of TCTP questions its function as a guanine nucleotide exchange factor for Rheb. FEBS Letters 582, 3005–3010.
Rodrigues A, Adamo M, Crozet P, et al. 2013. ABI1 and PP2CA phosphatases are negative regulators of Snf1-related protein kinase1 signaling in Arabidopsis. The Plant Cell 25, 3871–3884.
Ruiz A, Xu X, Carlson M. 2011. Roles of two protein phosphatases, Reg1-Glc7 and Sit4, and glycogen synthesis in regulation of SNF1 protein kinase. Proceedings of the National Academy of Sciences, USA 108, 6349–6354.
Saucedo LJ, Gao X, Chiarelli DA, Li L, Pan D, Edgar BA. 2003. Rheb promotes cell growth as a component of the insulin/TOR signalling network. Nature Cell Biology 5, 566–571.
Shaw RJ, Bardeesy N, Manning BD, Lopez L, Kosmatka M, DePinho RA, Cantley LC. 2004. The LKB1 tumor suppressor negatively regulates mTOR signaling. Cancer Cell 6, 91–99.
Shen W, Reyes MI, Hanley-Bowdoin L. 2009. Arabidopsis protein kinases GRIK1 and GRIK2 specifically activate SnRK1 by phosphorylating its activation loop. Plant Physiology 150, 996–1005.
Stamatakis A, Hoover P, Rougemont J. 2008. A rapid bootstrap algorithm for the RAxML Web servers. Systematic Biology 57, 758–771.
Studer RA, Robinson-Rechavi M. 2009. How confident can we be that orthologs are similar, but paralogs differ? Trends in Genetics 25, 210–216.
Tome F, Nagele T, Adamo M, et al. 2014. The low energy signaling network. Frontiers in Plant Science 5, 353.
Valentine RJ, Coughlan KA, Ruderman NB, Saha AK. 2014. Insulin inhibits AMPK activity and phosphorylates AMPK Ser(4)(8)(5)/(4)(9)(1) through Akt in hepatocytes, myotubes and incubated rat skeletal muscle. Archives of Biochemistry and Biophysics 562, 62–69.
van Dam TJ, Zwartkruis FJ, Bos JL, Snel B. 2011. Evolution of the TOR pathway. Journal of Molecular Evolution 73, 209–220.
Vernoud V, Horton AC, Yang ZB, Nielsen E. 2003. Analysis of the small GTPase gene superfamily of Arabidopsis. Plant Physiology 131, 1191–1208.
Wang GL, Jiang BH, Rue EA, Semenza GL. 1995. Hypoxia-inducible factor-1 is a basic-helix–loop–helix-PAS heterodimer regulated by cellular
O-2 tension. Proceedings of the National Academy of Sciences, USA 92, 5510–5514.
Wang Y, Kuramitsu Y, Baron B, Kitagawa T, Tokuda K, Akada J, Nakamura K. 2015. CGK733-induced LC3 II formation is positively associated with the expression of cyclin-dependent kinase inhibitor p21Waf1/Cip1 through modulation of the AMPK and PERK/CHOP signaling pathways. Oncotarget 6, 39692–39701.
Wei Y, Zheng XF. 2011. Nutritional control of cell growth via TOR signaling in budding yeast. Methods in Molecular Biology 759, 307–319.
Wojtaszewski JF, Nielsen JN, Jorgensen SB, Frosig C, Birk JB, Richter EA. 2003. Transgenic models—a scientific tool to understand exercise-induced metabolism: the regulatory role of AMPK (5'-AMP-activated protein kinase) in glucose transport and glycogen synthase activity in skeletal muscle. Biochemical Society Transactions 31, 1290–1294.
Wood C, Jalil MNH, Mclaren I, Yong BCS, Ariffin A, Mcneil PH, Burgess N, Thomas DR. 1984. Carnitine long-chain acyltransferase and oxidation of palmitate, palmitoyl coenzyme-a and palmitoylcarnitine by pea mitochondria preparations. Planta 161, 255–260.
Woods A, Dickerson K, Heath R, Hong SP, Momcilovic M, Johnstone SR, Carlson M, Carling D. 2005. Ca2+/calmodulin-dependent protein kinase kinase-beta acts upstream of AMP-activated protein kinase in mammalian cells. Cell Metabolism 2, 21–33.
Woods A, Johnstone SR, Dickerson K, Leiper FC, Fryer LG, Neumann D, Schlattner U, Wallimann T, Carlson M, Carling D. 2003. LKB1 is the upstream kinase in the AMP-activated protein kinase cascade. Current Biology 13, 2004–2008.
Wullschleger S, Loewith R, Hall MN. 2006. TOR signaling in growth and metabolism. Cell 124, 471–484.
Xiao B, Sanders MJ, Underwood E, et al. 2011. Structure of mammalian AMPK and its regulation by ADP. Nature 472, 230–233.
Xiong Y, McCormack M, Li L, Hall Q, Xiang C, Sheen J. 2013. Glucose–TOR signalling reprograms the transcriptome and activates meristems. Nature 496, 181–186.
Xiong Y, Sheen J. 2015. Novel links in the plant TOR kinase signaling network. Current Opinion in Plant Biology 28, 83–91.
Yadav UP, Ivakov A, Feil R, et al. 2014. The sucrosetrehalose 6-phosphate (Tre6P) nexus: specificity and mechanisms of sucrose signalling by Tre6P. Journal of Experimental Botany 65, 1051–1068.
Comparative genomics of AMPK–TOR signaling | 3907
Downloaded from https://academic.oup.com/jxb/article-abstract/67/13/3897/1749706by gueston 16 April 2018
![Priority Down-regulation of AMPK signaling pathway rescues ......related hearing loss in the transgenic Tg-B1 mouse model of mitochondrial deafness [12]. We here explored tissue-specific](https://static.documents.pub/doc/80x56/60dbdd3cd32a7244666d85cb/priority-down-regulation-of-ampk-signaling-pathway-rescues-related-hearing.jpg)
