Differential expression of hypoxia pathway genes in honey bee (Apis mellifera L.) caste development

Differential expression of hypoxia pathway genes in honey bee (Apis mellifera L.) caste development

Journal of Insect Physiology 57 (2011) 38–45 Contents lists available at ScienceDirect Journal of Insect Physiology journal homepage: www.elsevier.c...

405KB Sizes 0 Downloads 116 Views

Journal of Insect Physiology 57 (2011) 38–45

Contents lists available at ScienceDirect

Journal of Insect Physiology journal homepage: www.elsevier.com/locate/jinsphys

Differential expression of hypoxia pathway genes in honey bee (Apis mellifera L.) caste development Sergio Vicente Azevedo a, Omar Arvey Martinez Caranton a, Tatiane Lippi de Oliveira b, Klaus Hartfelder b,* a

Departamento de Gene´tica, Faculdade de Medicina de Ribeira˜o Preto, Universidade de Sa˜o Paulo, Av. Bandeirantes 3900, 14049-900 Ribeira˜o Preto, Brazil Departamento de Biologia Celular e Molecular e Bioagentes Patogeˆnicos, Faculdade de Medicina de Ribeira˜o Preto, Universidsade de Sa˜o Paulo, Avenida Bandeirantes 3900, 14049-900 Ribeira˜o Preto, SP Brazil b

A R T I C L E I N F O

A B S T R A C T

Article history: Received 23 July 2010 Received in revised form 21 September 2010 Accepted 22 September 2010

Diphenism in social bees is essentially contingent on nutrient-induced cellular and systemic physiological responses resulting in divergent gene expression patterns. Analyses of juvenile hormone (JH) titers and functional genomics assays of the insulin–insulin-like signaling (IIS) pathway and its associated branch, target-of-rapamycin (TOR), revealed systemic responses underlying honey bee (Apis mellifera) caste development. Nevertheless, little attention has been paid to cellular metabolic responses. Following up earlier investigations showing major caste differences in oxidative metabolism and mitochondrial physiology, we herein identified honey bee homologs of hypoxia signaling factors, HIFa/ Sima, HIFb/Tango and PHD/Fatiga and we investigated their transcript levels throughout critical stages of larval development. Amsima, Amtango and Amfatiga showed correlated transcriptional activity, with two peaks of occurring in both queens and workers, the first one shortly after the last larval molt and the second during the cocoon-spinning phase. Transcript levels for the three genes were consistently higher in workers. As there is no evidence for major microenvironmental differences in oxygen levels within the brood nest area, this appears to be an inherent caste character. Quantitative PCR analyses on worker brain, ovary, and leg imaginal discs showed that these tissues differ in transcript levels. Being a highly conserved pathway and linked to IIS/TOR, the hypoxia gene expression pattern seen in honey bee larvae denotes that the hypoxia pathway has undergone a transformation, at least during larval development, from a response to environmental oxygen concentrations to an endogenous regulatory factor in the diphenic development of honey bee larvae. ß 2010 Elsevier Ltd. All rights reserved.

Key words: Hypoxia inducible factor HIF Insulin signaling Caste development Honeybee

1. Introduction Functional specialization based on differences in morphology was a key step in the evolution of castes in highly social insects. With relatively few exceptions (Schwander et al., 2010) the ontogenetic pathways that generate the different morphologies in highly social Hymenoptera are not contingent on genotype differences but on environmental stimuli that can bias development from a single genotype into two or more phenotypically distinct female castes. While caste functions in their social context have fascinated observers for centuries, the developmental mechanisms underlying the plasticity in developmental pathways are still only barely elucidated, spearheaded by studies on the honey bee, Apis mellifera. In this highly eusocial insect, differential

* Corresponding author. Tel.: +55 16 3602 3063; fax: +55 16 3633 1786. E-mail addresses: [email protected] (S.V. Azevedo), [email protected] (O.A.M. Caranton), [email protected] (T.L. de Oliveira), [email protected] (K. Hartfelder). 0022-1910/$ – see front matter ß 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.jinsphys.2010.09.004

feeding of queen- and worker-destined larvae by nurse bees in the early larval stages triggers a major endocrine response, marked by pronounced differences in the hemolymph juvenile hormone (JH) titer (Rachinsky et al., 1990; Hartfelder and Engels, 1998), as well as in insulin–insulin like signaling (IIS) and its parallel branch, the target-of-rapamycin signaling (TOR) pathway (Wheeler et al., 2006; Patel et al., 2007; Azevedo and Hartfelder, 2008). The composite IIS/TOR pathway is a phylogenetically conserved module for sugar and amino acid sensing which serves to adjust growth and cell proliferation rates to nutrient availability (Brogiolo et al., 2001; Oldham and Hafen, 2003). As an outcome of the honey bee genome annotation (The Honey Bee Genome Sequencing Consortium, 2006; Denison and Raymond-Delpech, 2008), the IIS/TOR pathway has come into focus as a means to understanding differential growth and morphological differentiation in honey bee postembryonic development (Wheeler et al., 2006; Patel et al., 2007; Azevedo and Hartfelder, 2008), as well as reproductive division of labor and lifespan in adult queens and workers (Corona et al., 2007). The analyses on expression patterns and functionality of IIS/TOR pathway genes revealed

S.V. Azevedo et al. / Journal of Insect Physiology 57 (2011) 38–45

surprising differences when compared to insect model systems such as Drosophila melanogaster. Whereas the knockdown of TOR function resulted, as expected, in a significant developmental bias towards the worker caste (Patel et al., 2007), expression levels for genes encoding insulin receptor (InR) and insulin-like peptides (ILP) were in concordance with expectations only for the earliest larval stages (Wheeler et al., 2006), when larval growth rates are still similar for prospective queens and workers (Wang, 1965; Thrasyvoulou and Benton, 1982). In contrast, in the late larval stages when queen larvae by far surpass workers in terms of growth rates, the expression of insulin receptor genes appeared markedly down-regulated in queens and, furthermore, higher transcript levels for AmILP-2, the predominantly expressed honey bee ILP gene, were found in fifth instar worker larvae rather than in queens (Azevedo and Hartfelder, 2008). This unexpected temporal and caste-specific expression profile of IIS pathway genes prompted us to broaden the view and look for signaling pathways that might interact with IIS/TOR. Based on evidence from vertebrates and invertebrates, the hypoxia signaling pathway was seen as such a candidate. Befitting with the key role of oxygen in organismic functions, sensory systems adjusting a cell’s physiology to environmental oxygen levels converge on a highly conserved signaling pathway, this consisting of two hypoxia-inducible transcription factor subunits (HIFa and HIFb) and a modifying enzyme (PHD). HIFa and HIFb are members of the basic Helix-Loop-Helix-Period-ArntSingle-minded (bHLH–PAS) family of transcription factors. The modifying enzyme, a 2-oxoglutarate and iron-dependent dioxygenase PHD (prolyl hydroxylase domains), hydroxylates specific prolyl residues in the oxygen-dependent degradation domain (ODD) of the HIFa subunit (Bruick and McKnight, 2001; Epstein et al., 2001), enabling its ubiquitination and proteasomal degradation (Ivan et al., 2001; Jaakkola et al., 2001; Maxwell et al., 2001). Under normoxic conditions, the oxygen-dependent catalytic PHD activity constantly hydroxylates two prolyl residues in HIFa, whereas under hypoxic condition this activity is downregulated, thus stabilizing HIFa and allowing its translocation to the nucleus where it forms a functional dimer with HIFb that then can bind to HIF-responsive elements (HREs) in the control region of hypoxia-regulated genes (Carver et al., 1994; Wang et al., 1995). The resulting adjustments in cellular physiology are an adaptive stress response that allows organisms to deal with unfavorable conditions, both at the level of system functions, by stimulating blood vessel growth in mammals (Maxwell et al., 2001) or tracheal branching in insects (Centanin et al., 2008, 2010), or at the organismic level, by modulating general metabolic functions, these frequently being associated with longevity (Fuchs et al., 2010). In insects, the hypoxia signaling pathway is best studied in D. melanogaster, where the core components are Sima and Tango (the HIFa and HIFb homologs), and Fatiga (the PHD homolog) (Nagao et al., 1996; Lavista-Llanos et al., 2002; Gorr et al., 2004). The functionality of the pathway has been elucidated in embryonic and larval tracheolar trees, where local oxygen concentrations drive branching morphogenesis through controlling the expression levels of the FGF receptor homolog Breathless and the FGF family member Branchless, the latter acting as a chemo-attractant (Centanin et al., 2010). Through its homeostasis function for all oxygen-dependent physiological processes, the hypoxia pathway is not only evolutionarily conserved, but also tightly integrated with other signaling pathways controlling metabolic rates. In Caenorhabditis elegans, hypoxia resistance is directly coupled to the functionality of the daf-2 gene which encodes the nematode insulin/insulin-like receptor (Mabon et al., 2009). In mouse myoblasts, the decision of whether to enter the differentiation pathway or remain in a proliferative state is also directly influenced by oxygen tension,

39

with HIF-1 overexpression inhibiting Akt-mTOR-dependent myogenesis (Ren et al., 2010). In the fruit fly, several reports strongly argue for an intricate interaction between the IIS/TOR and hypoxia signaling pathways, denoting that (i) nuclear localization of Sima is promoted by PI3K-AKT and TOR (Dekanty et al., 2005), (ii) that Sima is a negative regulator of IIS/TOR-mediated growth (Romero et al., 2007), and (iii) that the genes encoding the tuberous sclerosis complex proteins Tsc1 and Tsc2 emerged as major effect genes downregulating translation, as observed in a genome wide RNAi screen designed to reveal hypoxia effects (Lee et al., 2008). Although honey bee larvae probably do not experience noteworthy hypoxia conditions while being fed by nurse bees, differences in oxidative metabolism have long been recognized as setting apart queens and workers. Meticulous analyses of O2 consumption and CO2 production by individual larvae showed that queens by far exceed workers during the phase of intense growth and that the respiratory quotient of worker larvae is generally divergent from that of queens, thus indicating metabolic differences in the conversion of nutrients provided in worker diet and royal jelly (Melampy and Willis, 1939). Subsequently, measurements of mitochondrial protein mass and spectra analyses of mitochondrial respiratory chain proteins revealed that queen larvae have relatively more mitochondria than workers and, especially so, higher cytochrome oxydase activity and cytochrome c contents (Osanai and Rembold, 1968). Upon integrating respiration rates and cytochrome c content into a calculation of respiratory values, queen larvae were seen to exceed workers by a factor six (Eder et al., 1983). Either because these results seemingly confirmed the obvious, viz. higher oxidative metabolism in queen larvae, or because mechanisms that could link oxidative metabolism to signaling pathways underlying caste development were not evidenced at the time, these studies received little attention in subsequent investigations which, starting in the seventies, became focused on the regulation of caste development by the morphogenetic hormones JH and ecdysteroids. With a sequenced genome at hand it now became feasible to conduct expression analyses of candidate genes integrating signaling pathways, and in this manner, the apparent paradox arising from gene expression analyses of IIS pathway genes (Azevedo and Hartfelder, 2008) prompted us to revisit the early results on oxidative metabolism under a new perspective. In the present study we investigated transcript levels of the three core genes of the hypoxia signaling pathway in the critical stages of honey bee caste development by means of quantitative RT-PCR assays. As a first step we assayed whole body expression levels of the Amsima, Amtango and Amfatiga genes. Next, we investigated the respective transcript levels in tissues and organs of specific interest for caste functionality, such as the larval brain, leg imaginal discs, and the developing ovary. 2. Material and methods 2.1. Rearing of honey bee larvae and RNA extraction Larvae of the third (L3), fourth (L4) and fifth instar (L5) were obtained from experimental hives of Africanized honey bees, A. mellifera. As in previous studies (Rachinsky et al., 1990), the fifth instar was subdivided into nine substages, these being three feeding stages (L5F1, L5F2 and L5F3, defined by weight), three cocoon-spinning phases (L5S1, L5S2 and L5S3, distinguishable by position of the larva in the brood cell and degree of gut voidance), and three prepupal stages (PP1, PP2 and PP3, divided by metathoracic leg size). Worker larvae were retrieved directly from brood combs. Queen larvae were reared by standard apicultural techniques through transfer of first instar larvae into queen cups that were subsequently introduced into the queenless upper story of Langstroth hives.

40

S.V. Azevedo et al. / Journal of Insect Physiology 57 (2011) 38–45

When in the desired developmental stage, larvae were homogenized in 1 ml of Trizol (Invitrogen) for RNA extraction. Whole-body extracts from L3 samples consisted of five larvae each, whereas from the fourth instar on each sample consisted of a single individual. Four biological replicates were prepared for each developmental stage. Specific tissue samples were prepared by dissection of worker larvae in the L4, L5F1 and L5F3 developmental stages, these being the ones when growth is exponential (L4 and L5F1) or when growth rates decline and reach a plateau (L5F3). From each larva we dissected the complete set of leg imaginal discs, the head capsule containing the brain, and the pair of ovaries. Each sample for homogenization in 250 ml of Trizol consisted of tissue dissected from three larvae. Three biological replicates were prepared for each developmental stage. RNA quality and quantity were checked spectrophotometrically (Nanovue, GE Healthcare).

endogenous control gene for honey bees has been validated (Lourenc¸o et al., 2008). Each sample was analyzed in triplicate (technical replicates). In whole-body transcript analyses, four biological samples were analyzed for each larval stage and caste, whereas in tissue-specific expression assaying we performed triplicate analyses. Relative expression levels were calculated by the comparative DD Ct method and are expressed as 2 Ct (Pfaffl, 2001). All values were calibrated against a single sample, this being the mean of the technical replicates for Amtango of a third instar worker sample. Data were analyzed by Kruskal–Wallis tests and two-way ANOVA using GraphPad Prism v. 5.0 software (Jandel Scientific), considering P  0.05 as statistically significant.

2.2. Annotation of honey bee hypoxia genes and design of RT-PCR primers

3.1. Annotation of hypoxia pathway genes in the honey bee

The core of the hypoxia signaling pathway is made up by three genes that are highly conserved across metazoan phyla. We identified homologs of these in the honey bee genome by mutual best-hit BLASTP searches against the Glean3 predicted Official Gene Set (The Honey Bee Genome Sequencing Consortium, 2006). The D. melanogaster genes encoding Sima (CG7951), Tango (CG11987) and Fatiga (CG31543) and the corresponding mammalian protein sequences HIF1a, HIF1b and PHD (GenBank accession numbers NP_001512.1, NP_848514.1, and NP_071334.1, respectively) served as queries. The retrieved honey bee protein sequences, were then manually annotated using the Artemis v11 platform (Rutherford et al., 2000). Putative functional domains were evidenced by domain-specific BLAST analyses (http:// www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi). Gene-specific primers designed by means of the Primer3 (http:// frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi) and Gene Runner v. 3.05 softwares were: Amsima (F 50 TGAACGACAGCATGAACGACAGCATGGCCGA 30 ; R 50 CGTTCGTTGCTCCTTCTCCG 30 ), Amtango (F 50 ATGAAACAACACAATCGCCTAT30 ; R 50 TGTCTTCCTGTCTTCCAATAGCCCATGC 30 ), Amfatiga (F 50 GTAGTGATCAAATAGTAGTGATCAAATAACGTGGC 30 ; R 50 CCTTTGTCCTTCCATTGATTGT 30 ). For Amsima they bridged exons 12 and 13, for Amtango they were within exon 12, and for Amfatiga they were in exon 1. Primer specificity and efficiencies were checked by dissociation curve analysis and 1:10 step serial dilutions, evidencing amplification efficiencies of 2.0–2.1 and R2 = 0.99 for all three primer sets. The amplification fragments generated by these primers were sequenced and checked by BLASTN alignments against the honey bee genome and non-redundant databases to confirm gene identity. 2.3. Quantitative RT-PCR analysis of hypoxia gene transcript levels in honey bee larvae 1 mg RNA of each sample was treated with 1 U DNase I (Invitrogen) to remove DNA contaminants. This DNase-treated RNA was reverse transcribed using Superscript II (Invitrogen) enzyme and oligo(dT)18 primer (Fermentas Life Sciences), generating first-strand cDNA profiles for the RNA samples. Each quantitative RT-PCR (RT-qPCR) analysis was performed with 1 ml cDNA (diluted 1:10), 8 ml of Power SYBR Green (Applied Biosystems), 0.5 ml of each primer (10 pmol/ml) and 6 ml of deiononized water (MilliQ, Millipore). Reaction cycles were run in a Real-Time PCR 7500 system (Applied Biosystems). PCR conditions were 50 8C for 2 min and 95 8C for 10 min, followed by cycles of 95 8C for 15 s and 60 8C for 1 min. For normalization of cDNA levels we performed assays using rp49 primers. rp49 encodes a ribosomal protein and its utility as

3. Results

Using protein sequences of the mammalian and fruit fly genes HIF1a/Sima, HIF1b/Tango and PHD/Fatiga as entries for BLASTP searches against the honey bee gene set we retrieved putative homologs for all three core genes of the hypoxia pathway. The respective honey bee protein sequences were: AmSima [GB16786, E = e82 to Drosophila Sima (CG7951) and E = e83 to human HIF1a (NP_001521.1)], AmTango [GB17763, E = 0.0 to Drosophila Tango (CG11987) and E = e137 to human HIF-1b (NP_848514.1)], and AmFatiga [GB18380, E = e73 to Drosophila Fatiga (CG31543) and E = e74 to human PHD (NP_071334.1)]. These were then manually mapped to the honey bee genome sequence using the Artemis v11 platform. The Amsima gene model is represented by a CDS of 2331 bp. It maps to genomic scaffold Group5.25 on chromosome 5, where it spans a genomic region of approximately 26 kb. The coding sequence (CDS) consists of 14 exons which are separated into two clusters by an intron of approximately 14 kb situated between exons 8 and 9 (Fig. 1A). The search for functional domains indicated the presence of an HLH DNA-binding domain and of two PAS domains which are considered as attributing specificity to DNA binding by this transcription factor. Amtango is represented by a CDS of 2805 bp. It maps to genomic scaffold Group7.9 on chromosome 7, where it spans a genomic region of approximately 44 kb. Its CDS also consists of 14 exons and these are also separated into two clusters by an intron, this being of approximately 17 kb and located between exons 11 and 12 (Fig. 1B). Like its counterpart AmSima, the AmTango amino acid sequence also contains a predicted HLH and two PAS domains. Finally, Amfatiga is represented by a CDS of 768 bp. It maps to genomic scaffold Group 5.7, where it spans a genomic region of approximately 13 kb. Its CDS consists of only 3 exons, with exon 1 separated from exons 2 and 3 by an intron of 10.5 kb (Fig. 1C). The search for conserved domains in the predicted AmFatiga protein returned a single 20G-FeII_Oxy domain, in accordance with its predicted role in the oxygen-sensitive hydroxylation of prolyl residues in AmSima. 3.2. Transcriptional activity of hypoxia genes in honey bee queen and worker larvae As a first step towards an understanding of the role of hypoxia signaling in honey bee caste development we investigated the transcriptional activity of the three core genes during the critical phase of caste development by means of RT-qPCR analysis. Starting with the third larval instar (L3), when queen and worker larvae start to be fed different diets, we covered the phase of exponential growth during the fourth (L4) and the feeding phase of the fifth instar (L5F1–L5F3), as well as the remainder of this instar, which

[()TD$FIG]

S.V. Azevedo et al. / Journal of Insect Physiology 57 (2011) 38–45

41

Fig. 1. Schematic genomic structure of honey bee hypoxia genes. (A) Amsima (GB16786-RA) maps to genomic scaffold Group5.25, with a coding sequence (CDS) consisting of 14 exons; (B) Amtango (GB17763-RA) maps to genomic scaffold Group7.9, with a CDS also made up of 14 exons; (C) Amfatiga (GB18380-RA) maps to genomic scaffold Group5.7, with a CDS consisting of 3 exons. Exons are depicted as boxes with indication of respective nucleotide positions, starting with the translation initiation site as 1. Also shown is the position of the respective primers used in quantitative RT-PCR analyses (Amsima-F and Amsima-R; Amtgo-F and AmtgoR; Amfatiga-F and Amfatiga-R).

comprises the larval-pupal transition accompanied by castespecific differentiation processes shaping internal organs and external morphology (L5S1-PP3). The relative expression levels of the three genes (Fig. 2) exhibited two major tendencies, first, a pronounced divergence between the two castes initiating with the nutritional switch and being sustained until shortly before the pupal molt, and second, apparently correlated expression patterns for the three genes for each of the two castes (Fig. 3). In general terms, means of relative expression levels were higher for all three genes in worker larvae, and these differences were statistically significant for most developmental stages (twoway ANOVA, Bonferroni post hoc tests, considering P  0.05 as statistically significant). A second general trend was the continuous increase of their transcript levels in both castes, accompanying exponential larval growth from the third to the early fifth instar. As soon as the larvae entered the fifth instar, this trend however, reverted in queens, resulting in a drop to near basal levels, whereas expression was sustained at elevated levels in workers. In a pairwise comparison of caste-specific relative expression levels, the two HIF gene orthologs, Amsima and Amtango, showed a marked parallelism in their drop of expression levels in both castes, this occurring towards the end of the feeding phase and marking the transition to the spinning phase (Fig. 2A and B). Even though such a drop in expression levels also occurred in the PHD gene homolog Amfatiga of queens, it was not observed in workers, where this gene continued to be expressed at elevated levels until and throughout prepupal development (Fig. 2C). 3.3. Tissue-specific expression analysis of hypoxia genes in honey bee worker development Upon finding that relative expression levels of the three core genes of the hypoxia signaling pathway are significantly elevated in worker larvae we decided to analyze transcript levels in the larval ovaries, the brain, and leg imaginal discs, all these being structures that exhibit marked caste differences in the adult phenotypes. In this analysis we focused on the fourth and early fifth instar, thus covering the developmental phase when whole body expression levels for the two genes start to diverge between the two castes, with higher levels in workers as compared to queens. The results shown in Fig. 4 illustrate that tissue-specific transcript levels for the three genes do not necessarily reflect the whole-body expression levels (shown for comparison as the

last column in all graphs). A general trend that could be extracted from this analysis was that tissue-specific and whole body transcript levels for the three genes were still similar in fourth instar worker larvae, but diverged as soon as the larvae molted to the fifth instar. Even though a statistically significant difference (P < 0.05; Kruskal–Wallis with Dunn’s post hoc tests, testing tissue sample against whole body transcript level separately for each gene and developmental stage) could only be detected for Amsima expression in L5F1 leg discs (Fig. 4B), mean transcript levels were also lower in early fifth-instar brain and leg discs for all three genes analyzed (Fig. 4B, C, E, F, H and I). Furthermore, leg discs were the structures where mean expression levels most consistently appeared to deviate from whole body levels (Fig. 4 B, C, E, F, H and I), always taking into account the caveat that statistical significance was established for L5F1 leg discs only. Nevertheless, as all these samples were composed of tissue dissected from three larvae each, and three different biological samples were analyzed for each stage, we are confident that the differences in mean transcript levels between tissues and whole body are meaningful and reflect the actual scenario of gene expression levels of the hypoxia signaling pathway genes in these tissues. 4. Discussion The high degree of conservation in core genes of the hypoxia signaling pathway made it possible to retrieve mutual best-hit candidates to fruit fly and mammalian Sima/HIFa, Tango/HIFb and Fatiga/PHD orthologs from the A. mellifera gene set (The Honey Bee Genome Sequencing Consortium, 2006). Both Amsima and Amfatiga, which correspond to the alpha subunit of HIF and its regulatory protein, genomically mapped to scaffold groups of chromosome 5, whereas the Amtango gene was localized on chromosome 7. The similarity of AmTango with its fruitfly and human orthologs was stunningly high, with an overall identity of 60% of its amino acid residues with the Drosophila protein. In the initial identification of the genes encoding the two HIF transcription factor subunits our searches were based on HLH and PAS functional domains, and this approach also retrieved other genes of the HLH–PAS family, such as putative homologs of the Drosophila genes period (per), cycle (cyc), clock (Clk) trachealess (trh), single-minded (sim), spineless (ss) and dysfusion (dys). But in these genes, similarities to the hypoxia pathway genes were limited to the HLH and PAS domains, respectively (data not shown), thus making possible the design of gene-specific primers for Amsima

[()TD$FIG]

42

[()TD$FIG]

S.V. Azevedo et al. / Journal of Insect Physiology 57 (2011) 38–45

Fig. 3. Relative quantification of Amsima, Amtango and Amfatiga transcript levels in third (L3), fourth (L4) and fifth instar (F5F1-PP3) honey bee larvae. The samples are the same as those shown in Fig. 2, but here focusing on coordinated expression patterns in the three hypoxia genes. (A) RNA of queen and (B) of worker larvae assayed by SYBR Green RT-qPCR. Relative expression levels were calculated as DD 2 Ct (Pfaffl, 2001) with rp49 serving as endogenous control, followed by calibration against an L3 worker sample for Amtango. Amsima, squares and full line; Amtango, rhomboids and gray dotted line line; Amfatiga, triangles and broken line. Shown are means  S.E.M. of four biological samples, each analyzed in triplicate, per developmental stage and caste. Fig. 2. Relative quantification of (A) Amsima, (B) Amtango and (C) Amfatiga transcript levels in queen and worker larvae of the honey bee. RNA from third (L3), fourth (L4) and fifth instar larvae (F5F1-PP3) was extracted and used in SYBR Green DD RT-qPCR assays. Relative expression levels were calculated as 2 Ct (Pfaffl, 2001), with rp49 serving as endogenous control, followed by calibration against an L3 worker sample for Amtango. Shown are means  S.E.M. of four biological samples, each analyzed in triplicate, per developmental stage and caste. Workers are represented by black open circles and continuous lines, queens by gray closed rhomboids and dotted lines. Asterisks indicate statistically significant differences in transcript levels between queen and worker larvae (two-way ANOVA, Bonferroni post hoc tests; *P  0.05, **P  0.01, ***P  0.001).

and Amtango outside of these conserved domains. A further important consensus motiv, LXXLAP, which characterizes the oxygen-dependent degradation (ODD) domain (Zhao et al., 2002), was found in Amsima but not in any other member of the HLH–PAS family, this being relevant as only this HIF subunit is subject to degradation under normoxia conditions. When assaying the expression levels of the three hypoxia signaling genes in honey bee larvae by quantitative RT-PCR we could show that the mean transcript levels of all three genes were consistently higher in worker larva than in queens (Fig. 2). In both castes, the two HIF subunit genes showed a marked peak of expression in the early feeding phase of the fifth instar, with transcript levels markedly declining as the larvae diminished their feeding activity and started to prepare for metamorphosis. Thereafter, Amsima and Amtango showed a second peak of transcriptional activity, shortly before the larvae entered the prepupal stage. Changing the point of view and looking at the joint expression of Amsima, Amtango and Amfatiga in each of the two castes (Fig. 3),

a high degree of apparent coregulation was evidenced, especially so in queen larvae. Furthermore, in both queens and workers, it was Amsima that exhibited the highest mean transcript levels among the three genes. This is of interest because under normoxia conditions this HIF subunit is targeted for degradation by the PHD enzyme activity which acts as an oxygen sensor. The HIFa subunit, thus, apparently needs to be continuously replenished by translation from its mRNA. Most mammalian body tissues experience oxygen concentrations well below the 21% atmospheric ones, and in some tissues such concentrations can be as low as 3–6% without eliciting a marked hypoxia response (Giaccia et al., 2004). This is also in accordance with findings for the atlas moth and probably most other insects where low intratracheal oxygen levels are maintained irrespective of ambient levels, probably as a protection against the toxicity of higher (ambient) oxygen concentrations (Hetz and Bradley, 2005). Nevertheless, variation in intraorganismal oxygen levels may diversely affect oxygen-dependent metabolic processes in different tissues and as a first approximation to addressing this question in honey bees we investigated the transcript levels of Amsima, Amtango and Amfatiga in three tissue types that undergo caste differentiation processes via different cellular mechanisms. Whereas programmed cell death is the prominent factor in the ovary of worker larvae, epithelial proliferation is eminent in imaginal discs of the legs of both castes, and neuroblast proliferation and neuronal differentiation can occur side by side in brain tissue. When compared to whole body levels of hypoxia gene expression, a major transition towards tissue-specific transcriptional responses became evident at the

[()TD$FIG]

S.V. Azevedo et al. / Journal of Insect Physiology 57 (2011) 38–45

43

Fig. 4. Relative quantification of Amsima (A)–(C), Amtango (D)–(F) and Amfatiga (G)–(I) in ovary, brain and leg imaginal discs of honey bee workers dissected from fourth instar (L4), early fifth instar (L5F1) and late feeding phase fifth instar (L5F3) larvae. RNA extracted from these tissues was used in SYBR Green RT-qPCR assays. Shown are means  S.E.M. of relative expression levels (with rp49 serving as endogenous control, followed by calibration against the same L3 worker sample for Amtango as used in Fig. 2) of three biological samples, each analyzed in triplicate, per developmental stage and caste. The values for whole body samples shown for comparison as the last (grey) bar in each graph were taken from Fig. 2. Asterisks indicate statistically significant differences in transcript level comparing tissue type against whole-body transcript level separately for each gene and developmental stage (Kruskal–Wallis with Dunn’s post hoc tests; *P  0.05).

transition from the fourth to the fifth instar, when proliferating tissues, such as the brain and imaginal discs showed relatively low mean transcript levels. The lowest hypoxia gene expression levels were detected in imaginal discs, which are ectodermal structures situated externally right beneath the cuticle. Furthermore, they have an inner network of tracheoles and are primarily composed of proliferating diploid cells which probably have low overall metabolic activity. While intriguing, the caste differences in hypoxia gene expression levels cannot easily be explained. Although not directly measured herein, there is no reason to expect oxygen levels to vary within the brood nest of a colony to an extent that worker larvae but not neighboring queens would experience different ambient hypoxia conditions. Two hypotheses can be proposed to explain the apparent caste differences, first, queen and worker larvae may have intrinsically distinct reaction thresholds to low oxygen supply, with workers being more sensitive than queens. Second, as a consequence of a better food supply, queen larvae may have a cellular physiology that better sustains a high oxidative metabolism in the phase of major growth. Evidence in favor of the second hypothesis comes from studies on respiratory metabolism in honey bee caste development, all pointing towards intrinsic physiological hypoxia conditions in worker larvae when compared to queens (Melampy and Willis, 1939; Osanai and Rembold, 1968; Eder et al., 1983). As these earlier physiological/biochemical investigations have largely been disregarded in favor of subsequent, essentially hormone action-based models on honey bee caste development, it is worthy to cite their main results. On investigating oxygen consumption and carbon dioxide production, Melampy and Willis (1939) showed that the oxidative metabolism

in queens is maintained at consistently higher levels than in workers, both when at maximal levels during the phase of most pronounced larval growth, as well as at low levels during pupal development. Measurements of mitochondria numbers revealed that these organelles were three times more numerous in queens (Osanai and Rembold, 1968), a finding subsequently confirmed by cytochrome c immunoassay quantification as an indicator of mitochondrial mass (Eder et al., 1983), indicating that respiratory rates in larvae are regulated by adaptive changes in quantities of respiratory chain enzymes. Furthermore, a mitochondrial gene expression analysis (Corona et al., 1999) showed that mitochondrial genes are expressed at higher levels in fifth instar queen larvae when compared to workers. Our results now reveal the other side of this physiological caste difference in oxidative metabolism, reflected in an apparently higher activation of the hypoxia signaling pathway in worker larvae. The major emergent question will now be on the potential crosstalk between hypoxia signaling and the other two pathways, viz. juvenile hormone/ecdysteroids and IIS/TOR signaling, that, so far, have been shown to play a role in differentiation processes leading to phenotypic differences between queens and workers. Such information is currently only available for regulatory interactions between the hypoxia and IIS/TOR pathways. In mammals, hypoxia provokes a reversible hypophosphorylation of mTOR and its effectors, the eIF4E-binding proteins (4E-BPs) and ribosomal protein S6 kinase (S6K), amongst others, thus downregulating translation on a broad scale (Arsham et al., 2003). In Drosophila, hypoxia signaling plays a similar role in fine-tuning the IIS/TOR response, whereby the hypoxia-induced gene products Scylla and Charybdis inhibit growth by up-regulating the tuberose

44

S.V. Azevedo et al. / Journal of Insect Physiology 57 (2011) 38–45

sclerosis complex (TSC), which is a negative regulator of TOR, thus down-regulating S6K activity (Reiling and Hafen, 2004). Furthermore, Tsc1 and Tsc2 which together form TSC, as well as the gene encoding tyrosine phosphatase 61F have been revealed in a genome-wide RNAi screen as genes down-regulating TOR activity under hypoxia conditions (Lee et al., 2008). Since insulin mediates its effects on the HIF-dependent transcriptional response via its downstream components PI3K-AKT, which affect TOR activity (Dekanty et al., 2005), the regulation of TOR thus becomes a key factor in integrating the two signaling pathways. In fruit fly and mammalian cells, this regulation occurs at the Sima protein level and its nuclear localization (Lavista-Llanos et al., 2002; Romero et al., 2007), without alteration in sima mRNA levels (Dekanty et al., 2005). Although at this point we are still lacking information on AmSima protein levels and its intracellular localization, the developmental profile of the honey bee Sima ortholog is indicative of a concerted regulatory control that includes the transcriptional level, and this is so not only for Amsima, but also the other two genes of the hypoxia signaling pathway. Another open question in the honey bee system is the occurrence of splice variants, as reported for the Drosophila sima gene (Gorr et al., 2004), where the longer variant is regulated by hypoxia, whereas the shorter one lacks both the oxygen-dependent degradation and the nuclear localization domains, indicating alternative functions for Sima and its dimerization partner Tango. While these points still require investigation in the honey bee, the observed developmental modulation and the clear caste differences in hypoxia gene transcript levels primarily need to be understood in relation to their integration with JH and IIS/TOR signaling. Whereas TOR function (Patel et al., 2007), transcript levels of insulin-like peptides and insulin receptors (Wheeler et al., 2006; Azevedo and Hartfelder, 2008), and the JH titer (Rembold, 1987; Rachinsky et al., 1990) showed major caste differences primarily in the early larval stages, the transcript levels of the hypoxia pathway genes mainly differed in the fourth and especially the early fifth instar, when larval growth rates are highest. The hypoxia transcriptional responses might thus represent a negative (hypoxia) feedback circuit on JH and IIS/ TOR-mediated tissue growth and differentiation. Such hypoxia pathway action on IGF-mediated growth versus differentiation has been evidenced in mammalian myoblasts (Ren et al., 2010) and, thus, could represent a mechanistic explanation for the largescale transcriptional differences in honey bee caste development, with the queen larval transcriptome being characterized by an overexpression of metabolic genes, whereas in the worker transcriptome differentiation-related genes are overrepresented (Barchuk et al., 2007). In conclusion, we could show that (i) the core genes of the hypoxia signaling module are represented in the honey bee genome, (ii) expression appears to be modulated in a concerted fashion during larval development, and (iii) their transcript levels are higher in worker larvae. These results do not only shed new light on prior, largely neglected results on caste differences in mitochondrial activity found in honey bee larvae, but also point towards the hypoxia signaling pathway as a possible modulator of IIS/TOR and JH-regulated processes underlying the formation of alternative phenotypes in this highly eusocial insect. Acknowledgments We thank, Luiz Roberto Aguiar and Marcela Bezerra Laure for technical assistance with honey bee rearing and tissue dissections and Angela Kaysel Cruz for making available the Real-Time PCR system. Financial support was by the Fundac¸a˜o de Amparo a` Pesquisa do Estado de Sa˜o Paulo (FAPESP grants 2007/04859-5 and

2008/07231-00) and Conselho Nacional de Desenvolvimento Cientı´fico e Tecnolo´gico (CNPq grant 470335/2009-2). References Arsham, A.M., Howell, J.J., Simon, M.C., 2003. A novel hypoxia-inducible factorindependent hypoxic response regulating mammalian target of rapamycin and its targets. Journal of Biological Chemistry 278, 29655–29660. Azevedo, S.V., Hartfelder, K., 2008. The insulin signaling pathway in honey bee (Apis mellifera) caste development – differential expression of insulin-like peptides and insulin receptors in queen and worker larvae. Journal of Insect Physiology 54, 1064–1071. Barchuk, A.R., Cristino, A.S., Kucharski, R., Costa, L.F., Simoes, Z.L.P., Maleszka, R., 2007. Molecular determinants of caste differentiation in the highly eusocial honeybee Apis mellifera. BMC Developmental Biology 7, e70. Brogiolo, W., Stocker, H., Ikeya, T., Rintelen, F., Fernandez, R., Hafen, E., 2001. An evolutionarily conserved function of the Drosophila insulin receptor and insulin-like peptides in growth control. Current Biology 11, 213–221. Bruick, R.K., McKnight, S.L., 2001. A conserved family of prolyl-4-hydroxylases that modify HIF. Science 294, 1337–1340. Carver, L.A., Hogenesch, J.B., Bradfield, C.A., 1994. Tissue-specific expression of the rat AH-receptor and ARNT messenger-RNAs. Nucleic Acids Research 22, 3038– 3044. Centanin, L., Dekanty, A., Romero, N., Irisarri, M., Gorr, T.A., Wappner, P., 2008. Cell autonomy of HIF effects in Drosophila: tracheal cells sense hypoxia and induce terminal branch sprouting. Developmental Cell 14, 547–558. Centanin, L., Gorr, T.A., Wappner, P., 2010. Tracheal remodelling in response to hypoxia. Journal of Insect Physiology 56, 447–454. Corona, M., Estrada, E., Zurita, M., 1999. Differential expression of mitochondrial genes between queens and workers during caste determination in the honeybee Apis mellifera. Journal of Experimental Biology 202, 929–938. Corona, M., Velarde, R.A., Remolina, S., Moran-Lauter, A., Wang, Y., Hughes, K.A., Robinson, G.E., 2007. Vitellogenin, juvenile hormone, insulin signaling, and queen honey bee longevity. Proceedings of the National Academy of Sciences of the United States of America 104, 7128–7133. Dekanty, A., Lavista-Llanos, S., Irisarri, M., Oldham, S., Wappner, P., 2005. The insulin-PI3K/TOR pathway induces a HIF-dependent transcriptional response in Drosophila by promoting nuclear localization of HIF-alpha/Sima. Journal of Cell Science 118, 5431–5441. Denison, R., Raymond-Delpech, V., 2008. Insights into the molecular basis of social behaviour from studies on the honeybee, Apis mellifera. Invertebrate Neuroscience 8, 1–9. Eder, J., Kremer, J.P., Rembold, H., 1983. Correlation of cytochrome c titer and respiration in Apis mellifera: adaptative response to caste determination defines workers, intercastes and queens. Comparative Biochemistry and Physiology 76B, 703–716. Epstein, A.C.R., Gleadle, J.M., McNeill, L.A., Hewitson, K.S., O’Rourke, J., Mole, D.R., Mukherji, M., Metzen, E., Wilson, M.I., Dhanda, A., Tian, Y.M., Masson, N., Hamilton, D.L., Jaakkola, P., Barstead, R., Hodgkin, J., Maxwell, P.H., Pugh, C.W., Schofield, C.J., Ratcliffe, P.J., 2001. C. elegans EGL-9 and mammalian homologs define a family of dioxygenases that regulate HIF by prolyl hydroxylation. Cell 107, 43–54. Fuchs, S., Bundy, J.G., Davies, S.K., Viney, J.M., Swire, J.S., Leroi, A.M., 2010. A metabolic signature of long life in Caenorhabditis elegans. BMC Biology 8, e14. Giaccia, A.J., Simon, M.C., Johnson, R., 2004. The biology of hypoxia: the role of oxygen sensing in development, normal function, and disease. Genes & Development 18, 2183–2194. Gorr, T.A., Tomita, T., Wappner, P., Bunn, H.F., 2004. Regulation of Drosophila hypoxia-inducible factor (HIF) activity in SL2 cells – identification of a hypoxia-induced variant isoform of the HIF alpha homolog gene similar. Journal of Biological Chemistry 279, 36048–36058. Hartfelder, K., Engels, W., 1998. Social insect polymorphism: hormonal regulation of plasticity in development and reproduction in the honeybee. Current Topics in Developmental Biology 40, 45–77. Hetz, S.K., Bradley, T.J., 2005. Insects breathe discontinuously to avoid oxygen toxicity. Nature 433, 516–519. Ivan, M., Kondo, K., Yang, H.F., Kim, W., Valiando, J., Ohh, M., Salic, A., Asara, J.M., Lane, W.S., Kaelin, W.G., 2001. HIF alpha targeted for VHL-mediated destruction by proline hydroxylation: implications for O2 sensing. Science 292, 464– 468. Jaakkola, P., Mole, D.R., Tian, Y.M., Wilson, M.I., Gielbert, J., Gaskell, S.J., von Kriegsheim, A., Hebestreit, H.F., Mukherji, M., Schofield, C.J., Maxwell, P.H., Pugh, C.W., Ratcliffe, P.J., 2001. Targeting of HIF-alpha to the von Hippel-Lindau ubiquitylation complex by O2-regulated prolyl hydroxylation. Science 292, 468–472. Lavista-Llanos, S., Centanin, L., Irisarri, M., Russo, D.M., Gleadle, J.M., Bocca, S.N., Muzzopappa, M., Ratcliffe, P.J., Wappner, P., 2002. Control of the hypoxic response in Drosophila melanogaster by the basic helix-loop-helix PAS protein similar. Molecular and Cellular Biology 22, 6842–6853. Lee, S.J., Feldman, R., O’Farrell, P.H., 2008. An RNA interference screen identifies a novel regulator of target-of-rapamycin that mediates hypoxia suppression of translation in Drosophila S2 cells. Molecular Biology of the Cell 19, 4051–4061. Lourenc¸o, A.P., Makert, A., Cristino, A.S., Simoes, Z.L.P., 2008. Validation of reference genes for gene expression studies in the honey bee, Apis mellifera, by quantitative real-time RT-PCR. Apidologie 39, 372–385.

S.V. Azevedo et al. / Journal of Insect Physiology 57 (2011) 38–45 Mabon, M.E., Scott, B.A., Crowder, C.M., 2009. Divergent mechanisms controlling hypoxia sensitivity and lifespan by the DAF-2/Insulin/IGF-Receptor pathway. PLoS One 4, e7939. Maxwell, P.H., Pugh, C.W., Ratcliffe, P.J., 2001. Activation of the HIF pathway in cancer. Current Opinion in Genetics & Development 11, 293–299. Melampy, R.M., Willis, E.R., 1939. Respiratory metabolism during larval and pupal development of the female honeybee (Apis mellifica L.). Physiological Zoology 12, 302–311. Nagao, M., Ebert, B.L., Ratcliffe, P.J., Pugh, C.W., 1996. Drosophila melanogaster SL2 cells contain a hypoxically inducible DNA binding complex which recognizes mammalian HIF-1 binding sites. FEBS Letters 387, 161–166. 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. Osanai, M., Rembold, H., 1968. Entwicklungsabha¨ngige Enzymaktivita¨ten in den Mitochondrien von Ko¨niginnen und Arbeiterinnen der Honigbiene. Biochimica et Biophysica Acta 162, 22–31. Patel, A., Fondrk, M.K., Kaftanoglu, O., Emore, C., Hunt, G., Frederick, K., Amdam, G.V., 2007. The making of a queen: TOR pathway is a key player in diphenic caste development. PLoS One 2, e509. Pfaffl, M.W., 2001. A new mathematical model for relative quantification in realtime RT-PCR. Nucleic Acids Research 29, e45. Rachinsky, A., Strambi, C., Strambi, A., Hartfelder, K., 1990. Caste and metamorphosis – hemolymph titers of juvenile hormone and ecdysteroids in last instar honeybee larvae. General and Comparative Endocrinology 79, 31–38. Reiling, J.H., Hafen, E., 2004. The hypoxia-induced paralogs Scylla and Charybdis inhibit growth by down-regulating S6K activity upstream of TSC in Drosophila. Genes & Development 18, 2879–2892. Rembold, H., 1987. Caste-specific modulation of juvenile hormone titers in Apis mellifera. Insect Biochemistry 17, 1003–1006.

45

Ren, H.X., Accili, D., Duan, C.M., 2010. Hypoxia converts the myogenic action of insulin-like growth factor into mitogenic action by differentially regulating signaling pathways. Proceedings of the National Academy of Sciences of the United States of America 107, 5857–5862. Romero, N.M., Dekanty, A., Wappner, P., 2007. Cellular and developmental adaptations to hypoxia: a Drosophila perspective. Methods in Enzymology 435, 123– 144. Rutherford, K.M., Parkhill, J., Crook, J., Horsnell, T., Rice, P., Rajandream, M.-A., Barrell, B., 2000. Artemis: sequence visualisation and annotation. Bioinformatics 16, 944–945. Schwander, T., Lo, N., Beekman, M., Oldroyd, B.P., Keller, L., 2010. Nature versus nurture in social insect caste differentiation. Trends in Ecology & Evolution 25, 275–282. The Honey Bee Genome Sequencing Consortium, 2006. Insights into social insects from the genome of the honeybee Apis mellifera. Nature 443, 931–949. Thrasyvoulou, A.T., Benton, A.W., 1982. Rates of growth of honeybee larvae. Journal of Apicultural Research 21, 189–192. Wang, G.L., Jiang, B.H., Rue, E.A., Semenza, G.L., 1995. Hypoxia-inducible factor-1 is a basic-helix-loop-helix-PAS heterodimer regulated by cellular O2 tension. Proceedings of the National Academy of Sciences of the United States of America 92, 5510–5514. Wang, I.D., 1965. Growth rates of young queen and worker honeybee larvae. Journal of Apicultural Research 4, 3–5. Wheeler, D.E., Buck, N., Evans, J.D., 2006. Expression of insulin pathway genes during the period of caste determination in the honey bee, Apis mellifera. Insect Molecular Biology 15, 597–602. Zhao, J.H.Q., Mooney, S.M., Lee, F.S., 2002. Sequence determinants in hypoxiainducible factor-1alpha for hydroxylation by the prolyl hydroxylases PHD1, PHD2, and PHD3. Journal of Biological Chemistry 277, 39792–39800.