NHR-14 loss of function couples intestinal iron uptake with innate immunity in C. elegans through PQM-1 signaling

Iron is essential for survival of most organisms. All organisms have thus developed mechanisms to sense, acquire and sequester iron. In C. elegans, iron uptake and sequestration are regulated by HIF-1. We previously showed that hif-1 mutants are developmentally delayed when grown under iron limitation. Here we identify nhr-14, encoding a nuclear receptor, in a screen conducted for mutations that rescue the developmental delay of hif-1 mutants under iron limitation. nhr-14 loss upregulates the intestinal metal transporter SMF-3 to increase iron uptake in hif-1 mutants. nhr-14 mutants display increased expression of innate immune genes and DAF-16/FoxO-Class II genes, and enhanced resistance to Pseudomonas aeruginosa. These responses are dependent on the transcription factor PQM-1, which localizes to intestinal cell nuclei in nhr-14 mutants. Our data reveal how C. elegans utilizes nuclear receptors to regulate innate immunity and iron availability, and show iron sequestration as a component of the innate immune response.


Introduction
Iron is an essential nutrient required by nearly all organisms for biological processes including DNA replication, mitochondrial respiration, and oxygen transport (Anderson et al., 2012). As a redox active element, iron also has a role in the production of reactive oxygen species (ROS), which can be deleterious to cellular proteins, DNA and membranes. In humans, tight regulation of cellular iron content is essential, as iron deficiency impairs neurological development in infants and causes anemia in adults, while iron overload leads to cardiomyopathy, diabetes mellitus, and neurodegenerative diseases (Fleming and Ponka, 2012). Excess iron also predisposes humans to infectious diseases, and limiting iron availability to microbes is a critical host innate immune response (Lopez and Skaar, 2018;Weiss and Carver, 2018). Given the dual nature of iron in healthy and diseased states, it is critical for organisms to maintain cellular iron concentration in a narrow physiological range.
C. elegans has emerged as a useful model for studying iron metabolism. In addition to its genetic tractability, ease of maintenance and susceptibility to human pathogens, key proteins involved in mammalian iron uptake, storage and efflux are conserved in C. elegans (Anderson and Leibold, 2014). C. elegans express the divalent metal transporter SMF-3 (Au et al., 2009;Bandyopadhyay et al., 2009;Romney et al., 2011), the iron-storage protein ferritin (FTN-1, FTN-2) (Gourley et al., 2003;Kim et al., 2004;Romney et al., 2008;Valentini et al., 2012) and the iron export protein ferroportin (FPN-1.1-3), which are orthologous to human DMT1 (SLC11A2, NRAMP2), ferritin-H and -L subunits (FTH1, FTL) and ferroportin (SLC40A1), respectively. Iron is transported across the apical enterocyte membrane by SMF-3/DMT1 and is used within the enterocyte for the metalation of iron-containing proteins and by mitochondria for Fe-S cluster biosynthesis (Braymer and Lill, 2017;Rouault, 2015). Iron not utilized is stored in ferritin within the enterocyte or exported across the basolateral enterocyte membrane by ferroportin/FPN-1.1. Ferroportin is the sole iron exporter in mammals (Ward and Kaplan, 2012), unlike C. elegans that express three ferroportin orthologs (FPN-1.1-3) that are not well characterized. Like mammals, intestinal iron absorption in C. elegans is upregulated during iron deficiency by HIF-transcriptional activation of smf-3/ DMT1 (C. elegans, HIF-1; human, HIF-2a, EPAS1) (Mastrogiannaki et al., 2009;Romney et al., 2008;Shah et al., 2009). One difference between iron sequestration in C. elegans and vertebrates is that during iron limitation C. elegans ferritin is transcriptionally repressed by HIF-1 Romney et al., 2011), while vertebrate ferritin is translationally repressed by the ironregulatory proteins (IRPs) (Anderson et al., 2012). In addition to iron, ftn-1 is regulated by the insulin/IGF-like (IIS) pathway , the REF-1 transcription factor HLH-29 (Quach et al., 2013), and by the serum/glucocorticoid-regulated kinase (SGK-1) to regulate lipid and iron homeostasis (Wang et al., 2016). The regulation of iron uptake by SMF-3 and iron sequestration by ferritin ensures that cells acquire adequate iron to satisfy their needs while limiting iron toxicity. Precise maintenance of cellular iron homeostasis is underscored by the shortened life span of C. elegans exposed to high iron (Gourley et al., 2003;Valentini et al., 2012).
Iron is required in the establishment of most pathogen infections. Pathogens use a variety of mechanisms to acquire iron from the host, one of which is the synthesis and release of siderophores that bind iron, allowing its uptake by bacteria (Palmer and Skaar, 2016). In response to pathogens, hosts have evolved strategies to limit iron availability that include the production of metal sequestering proteins, the sequestration of iron in ferritin and the regulation of metal transporter function (Lopez and Skaar, 2018;Wessling-Resnick, 2015). C. elegans has provided a valuable genetic model to study host-pathogen interactions because many human pathogens, such as Pseudomonas aeruginosa, Salmonella enterica and Staphylococcus aureus, cause lethal intestinal infections in C. elegans (Cohen and Troemel, 2015). Although C. elegans lack an adaptive immune system, these organisms rely on evolutionarily conserved innate immune systems, such as the p38-mitogen-activated protein kinase (MAPK), insulin and transforming growth factor-b (TGF-b) signaling pathways, for sensing and resolving bacterial infections (Irazoqui et al., 2010;Kim and Ewbank, 2015). Similar to vertebrates, C. elegans use diverse mechanisms to withhold iron and other metals during pathogen infections. FTN-2 provide resistance to S. aureus and E. coli LF82 infections (Simonsen et al., 2011) and SMF-3 provides resistance to S. aureus infections (Bandyopadhyay et al., 2009). In addition, C. elegans respond to pyoverdine, an iron siderophore produced by P. aeruginosa, by upregulation of a mitochondrial surveillance pathway that is triggered by pyoverdine-induced mitochondrial damage (Kang et al., 2018;Kirienko et al., 2015;Tjahjono and Kirienko, 2017). These studies highlight iron limitation as an important component of innate immune response in controlling pathogen infections.
In this report, we identify NHR-14, a nuclear receptor (NR) as a transcriptional repressor of SMF-3-dependent iron uptake and the innate immune response. Loss of nhr-14 promotes the nuclear localization of PQM-1, which binds to GATA-like DAF-16-associated elements (DAEs) in the smf-3 promoter to activate smf-3 transcription. nhr-14 mutants exhibit enhanced resistance to P. aeruginosa infection that in part requires smf-3 as well as other innate immune response genes. Our data implicate NHR-14 as a repressor of iron uptake and the transcriptional innate immune response mediated by PQM-1, showing iron sequestration as an important component of innate immunity in C. elegans.

Results
Loss of NHR-14 rescues low iron growth of hif-1(ia4) mutants We previously reported that hif-1(ia4) mutant worms are developmentally delayed when grown under iron limiting conditions using the iron chelator, 2,2-bipyridyl (BP), and that this delay was suppressed by elevating intracellular iron by reducing ftn-1 and ftn-2 expression (Romney et al., 2011).
To discover novel genes involved in iron homeostasis, a suppressor screen was performed to identify mutations that rescued the low iron phenotype observed in hif-1(ia4) null animals. The ia4 mutation is a 1,231 bp deletion of the second, third, and fourth exons that deletes much of the bHLH and PAS domains in HIF-1, and likely results in complete loss of hif-1 function (Jiang et al., 2001). This screen led to the isolation of several suppressor mutants, of which nhr-14(qa6909) and nhr-14 (qa6910) were identified as strong suppressors of the hif-1(ia4) low iron phenotype ( Figure 1A). The causative mutation within these lines was determined using whole genome sequencing and Hawaiian strain SNP mapping (see Materials and methods). Both the qa6909 and qa6910 mutations displayed a similar~1 Mb region of the X chromosome that have a Hawaiian/Bristol reads ratio of close to zero ( Figure 1B). Sequence analysis revealed that the qa6909 and qa6910 mutant strains possess distinct mutations within the nuclear-hormone receptor gene nhr-14 (T01B10.4) ( Figure 1C). DNA from qa6909 mutants showed a G > A substitution that changes Trp49 to a STOP codon and DNA from qa6910 mutants showed a G > A substitution that changes Gly33 to Arg ( Figure 1C).
nhr-14 is predicted to encode a protein homologous to the HNF4 family of nuclear receptors, which are transcription factors that regulate gene expression in response to nutritional, metabolic and environmental signals (Taubert et al., 2011). Amino acid sequence alignment between NHR-14 and human HNF4a showed a highly conserved N-terminal zinc finger DNA-binding domain (DBD) and a conserved C-terminal ligand-binding domain (LBD) ( Figure 1C). The qa6909 mutation (Trp49STOP) located within the DBD is predicted to result in a truncated protein that lacks a significant portion of the DBD and the entire LBD. The qa6910 missense mutation (Gly33Arg), which is also located within the DBD, changes non-polar glycine to positively-charged arginine ( Figure 1C). Gly33 is conserved in HNF4 receptors found in vertebrates and Drosophila, and it is likely that this substitution disrupts DNA binding.
To verify that mutations in nhr-14 are responsible for rescuing the low iron developmental delay of hif-1(ia4) mutants, we crossed nhr-14(tm1473) mutants into the hif-1(ia4) background, and grew hif-1(ia4); nhr-14(tm1473) double mutants on low iron (NGM-BP) plates. The tm1473 mutation is a 409 bp deletion of the third exon and intron that results in a premature stop codon and deletion of the ligand-binding domain that likely results in nhr-14 loss of function. hif-1(ia4); nhr-14(tm1473) double mutants developed normally under iron limitation ( Figure 1D), showing that loss of nhr-14 function is responsible for the rescue observed in qa6909 and qa6910 mutant animals.

NHR-14 is highly expressed in intestine and head neurons and is not regulated by iron
Given that the mutations in nhr-14 were identified in a screen for mutants that rescued the low iron developmental delay of hif-1(ia4) mutants, we questioned whether nhr-14 mRNA or protein is regulated by changes in iron status. We generated a transgenic line carrying a nhr-14 fosmid transgene that utilizes the endogenous cis-elements and promoter of nhr-14 to drive the expression of NHR-14 tagged with GFP and 3xFLAG. We validated NHR-14::GFP::FLAG as a functional replacement for endogenous NHR-14 by injecting hif-1(ia4); nhr-14(tm1473) double mutants with the nhr-14 fosmid transgene. These transgenic worms displayed the hif-1(ia4) developmental delay phenotype when grown under iron limitation (Figure 2-figure supplement 1). Fluorescent imaging of NHR-14:: GFP::FLAG worms showed that it is highly expressed in intestinal nuclei and in cells in the head (Figure 2A-C). Western blot analysis showed a protein of~80 kDa in worm lysates consistent with the predicted mass of NHR-14::GFP::FLAG ( Figure 2D). No significant change in levels of NHR-14:: GFP::FLAG and nhr-14 mRNA were observed in wild-type N2 worms grown in low iron (NGM-BP) or high iron (NGM-ferric ammonium citrate, FAC) compared to NGM ( Figure 2D-E). Furthermore, nuclear localization of NHR-14::GFP::FLAG was not affected by iron or by BP (Figure 2-figure supplement 2). The enrichment of nhr-14 in intestinal and head cells is consistent with intestinal and neuronal expression of nhr-14 determined by single-cell RNA sequence and SAGE analyses and (Cao et al., 2017;McGhee, 2007).
Given that PQM-1 transcriptionally activates smf-3, we questioned whether PQM-1 nuclear localization is affected by nhr-14 loss. Previous studies showed that a PQM::GFP::FLAG transgene localized to the nucleus of intestinal cells during larval stages and to the cytoplasm in adults (Dowen et al., 2016;O'Brien et al., 2018;Tepper et al., 2013). In agreement with these studies, PQM-1::GFP::FLAG localized to nuclei in intestinal cells in L4 stage worms fed either control RNAi or nhr-14 RNAi, and to the cytoplasm in adults fed control RNAi. PQM-1::GFP::FLAG, however, remained in intestinal nuclei in adult worms fed nhr-14-RNAi ( Figure 4G). Together, these data suggest that PQM-1 activates smf-3 by associating with GATA-like DAE motifs in the smf-3 promoter, and that NHR-14 acts upstream of PQM-1 to inhibit its nuclear translocation and transcriptional activity.
Iron is required for pathogen survival and infectivity, and both pathogens and hosts have evolved specialized mechanisms to control iron availability (Hood and Skaar, 2012). We postulated that P. aeruginosa-mediated NHR-14 repression may provide a strategy to limit iron to the pathogen in the intestinal lumen by increasing iron uptake by SMF-3. If so, iron content would increase in wild-type N2 worms exposed to PA14. As predicted, iron content increased in wild-type N2 worms exposed to PA14 compared to OP50, and was further increased in nhr-14(tm1473) mutants exposed to either OP50 or PA14 ( Figure 8F). The low iron content observed in smf-3(ok1035) single mutants and smf-3(ok1035); nhr-14(tm1473) double mutants was unaffected by PA14 exposure ( Figure 8E and Figure 3B). Iron content was lower in pqm-1(ok485) mutants than nhr-14(tm473) mutants, and was unaffected by PA14. Together, these data show that NHR-14 repression and PQM-1 activation are required for PA14 induction of SMF-3-dependent iron uptake and FTN-1 iron storage, demonstrating iron sequestration as a component of the innate immune response in C. elegans (Figure 9).

Discussion
Using a genetic suppressor screen to rescue the low iron developmental delay phenotype of hif-1 (ia4) mutants, we identified NHR-14 as a transcriptional repressor of SMF-3-dependent iron uptake and innate immunity in C. elegans. Loss of nhr-14 increases SMF-3-mediated iron uptake and induces innate immune response genes that contribute to the resistance of nhr-14(tm1473) mutants to P. aeruginosa. These transcriptional responses are regulated in part by the transcription factor PQM-1. Our data support a model whereby C. elegans respond to pathogen infection by reducing NHR-14 abundance, thus allowing PQM-1 to translocate to the nucleus of intestinal cells to activate genes critical for innate immunity and iron uptake. Several C. elegans nuclear receptors, such as DAF-12 (Liu et al., 2013), NHR-49 (Sim and Hibberd, 2016), NHR-25 (Ward et al., 2014), and NHR-114 and NHR-144 (Yuen and Ausubel, 2018), regulate innate immune response genes; however, there is no evidence to date indicating a role for these nuclear receptors in iron metabolism. NHR-14 may thus be unique among nuclear receptors in coupling innate immunity with iron sequestration as a strategy to limit iron to invading intestinal pathogens.
Originally discovered to be induced in response to paraquat-mediated stress (Tawe et al., 1998), PQM-1 has been shown to be responsible for the activation of the DAF-16/FoxO-Class II subset of IIS genes (Tepper et al., 2013). Under normal conditions, the IIS pathway is upregulated by insulinlike peptides via activation of the DAF-2 tyrosine kinase receptor that culminates in the transcription factor DAF-16/FoxO phosphorylation and its cytoplasmic retention. Under stress, IIS signaling is reduced, DAF-16/FoxO is dephosphorylated and driven into the nucleus to activate Class I genes (stress response) and repress Class II genes (metabolism/growth) (Murphy et al., 2003;Tepper et al., 2013). PQM-1 antagonizes DAF-16/FoxO nuclear localization upon increased insulin signaling, resulting in the activation of Class II genes through binding to the DAE (Tepper et al., 2013). In larval stage worms, PQM-1 localizes to the nucleus of intestinal cells, but in adults, PQM-1 becomes mostly cytoplasmic (Dowen et al., 2016;O'Brien et al., 2018;Tepper et al., 2013). Consistent with these studies, we found PQM-1::GFP localized to the nucleus in L4 stage worms fed either control RNAi or nhr-14 RNAi; however, PQM-1::GFP remained nuclear in adults fed nhr-14 RNAi. These data suggest that PQM-1 nuclear localization caused by loss of nhr-14 in adults may be a mechanism to maintain growth via Class II gene activation as well as defend against stress in aged worms. Other signaling pathways are known to regulate PQM-1 subcellular localization, including mTORC/SGK-1, which promotes PQM-1 cytoplasmic localization in intestinal cells (Dowen et al., 2016), and HSP-90, which promotes PQM-1 nuclear localization in neuronal and intestinal cells (O'Brien et al., 2018). PQM-1 has also been shown to have a role in reproductive aging (Templeman et al., 2018).
The enrichment of Class II and innate immune response genes in nhr-14(tm1473) mutants suggest that these mutants may have a survival advantage when challenged with pathogens. As predicted, nhr-14(tm1473) mutants displayed enhanced resistance to P. aeruginosa that require PQM-1. P. aeruginosa infection reduced NHR-14 abundance in wild-type N2 adults, which led to PQM-1 nuclear localization, activation of smf-3 and specific DAF-16/FoxO-Class II genes and increased iron content. The observation that not all DAF-16/FoxO-Class II genes are regulated by PQM-1 in nhr-14(tm1473) mutants suggests other pathways are involved in their regulation. Our data are consistent with a model whereby increased iron uptake by SMF-3 may be a strategy to limit iron to pathogens in the intestinal lumen. It is also possible that iron helps to maintain iron homeostasis during a pathogen infection as it is required by enzymes involved in critical cellular processes.
We found that pqm-1(ok485) mutants displayed similar sensitivity to P. aeruginosa PA14 as wildtype N2 worms consistent with a recent report (O'Brien et al., 2018), but conflicts with two other studies showing pqm-1 is required for resistance to PA14 (Joshi et al., 2016;Shapira et al., 2006). The reason for the conflicting data is unclear, but may be caused by different background strains or different conditions used for pathogen survival assays.
In vertebrates, natural resistance-associated macrophage protein 1 (NRAMP1), a paralog of DMT1, is found in the phagosome membrane of macrophages, where it transports Fe(II) and Mn(II) out of phagosomes, thus restricting growth of pathogens residing within these compartments by limiting iron availability (Lopez and Skaar, 2018;Wessling-Resnick, 2015). DMT1 is highly expressed in duodenal enterocytes where it is a key regulator of intestinal iron absorption (Fleming et al., 1998;Gunshin et al., 1997). The intestinal mucosa also serves as the first line of defense against pathogens. Upon a pathogen infection in intestine, the innate immune response is triggered by the detection of microorganism-associated molecular patterns (MAMPs) via patternrecognition receptors (PRR) (e.g. Tol/Toll-like and NOD-like receptors) that activate pathways regulating the production of anti-microbial proteins and cytokines, as well as the recruitment of acute inflammatory cells (Lopez and Skaar, 2018;Santaolalla et al., 2011). Several studies have shown that DMT1 is induced in response to TNFa (tumor necrosis factor a) and IFNg (interferon g) in duodenal enterocytes and bronchial epithelial cells (Laftah et al., 2006;Wang et al., 2005). It was speculated in these studies that the increase in DMT1 is a means to sequester iron to prevent microbial access to host iron and limit pathogenesis. Further investigation is needed to determine a role for DMT1-mediated iron uptake in intestinal innate immunity in vertebrates.
How NHR-14 senses pathogens remain to be determined. Unlike vertebrates, C. elegans lack clear pattern-recognition receptors orthologs, and its single Toll-like receptor is not necessary for immune defense (Cohen and Troemel, 2015;Stuart et al., 2013). A surveillance pathway has been identified in C. elegans that detects microbial-derived toxin damage to core cellular processes, such as translation and mitochondrial respiration, and induces signaling pathways controlling innate immune genes (Pukkila-Worley, 2016). The ligand-binding domain of NHR-14 may sense and bind pathogen-derived molecules or endogenous-produced ligands that facilitate NHR-14 stability or coregulator binding to regulate target genes. NHR-14 is also highly expressed in head neurons, suggesting that neuronal NHR-14 may detect small molecules that could modulate the innate immune response in intestine. NHR-14 was reported to function as an estrogen receptor (Mimoto et al., 2007), but conflicts with other studies showing that estrogen and bisphenol A did not affect NHR-14 function (Allard and Colaiácovo, 2010;Fischer et al., 2012).
Of the 284 NRs in C. elegans, 269 are derived from a HNF4-related ancestor (Robinson-Rechavi et al., 2005). This is unlike mammals that express two HNF4 receptors (HNF4a and HNF4g) and Drosophila melanogaster that express a single HNF4 (Taubert et al., 2011). Mammalian HNF4a is expressed in liver, kidney, intestine, and the colon where it activates genes involved in glucose, fatty acid, and xenobiotic metabolism (Gonzalez, 2008). HNF4a has been shown to control iron homeostasis in liver and hepatoma cells through the regulation of transferrin, transferrin receptor-1 and hepcidin genes (Matsuo et al., 2015;Truksa et al., 2009). The expansion of the HNF4-related NRs in C. elegans suggests that they may have evolved dedicated functions that are performed by HNF4 in higher organisms (Taubert et al., 2011). Studies have shown that in the plant parasitic nematode Meloidogyne incognita (Abad et al., 2008) and in the human parasitic nematode Brugia malayi (Ghedin et al., 2007) the number of NRs are reduced to 92 and 27, respectively. Of note, NHR-14 is one of four HNF4-related nuclear receptors that is conserved in both parasites, suggesting that NHR-14 may be important for parasite infectivity.
In summary, the identification of NHR-14 adds a new layer of regulation that links iron metabolism with the innate immune response and provides new insights by which C. elegans responds to pathogen infections.

C. elegans strains and culture
Strains obtained from the CGC and those generated for this study are listed in Supplementary file 1 and Key Resources Table. nhr-14(tm1473) mutants were backcrossed 6x to wild-type N2 worms. These worms displayed normal development and brood size. hif-1(ia4); nhr-14(tm1473) and smf-3 (ok1035); nhr-14(tm1473)double mutants, and smf-3(ok1035); hif-1(ia4); nhr-14(tm1473) triple mutants were constructed and verified using standard genetic and molecular methods (Brenner, 1974). Worms were grown on nematode growth medium (NGM) agar plates seeded with Escherichia coli OP50 bacteria at 22˚C unless otherwise noted. For iron experiments, worms synchronized by hypochlorite treatment at the L4 stage to young adult stage were grown overnight on NGM plates and then transferred to NGM plates supplemented with 6.6 mg/ml ferric ammonium citrate (FAC) or 25 uM 2,2'-bipyridyl (BP, iron chelator) for rescue experiments and 100 uM BP for RNAi experiments for times indicated in each experiment. The pH of FAC-NGM agar was adjusted to pH 7.0. For synchronizing worms, eggs were obtained by treating gravid adults with alkaline hypochlorite. Eggs were allowed to hatch overnight in Egg Buffer (118 mM NaCl, 48 mM KCl, 2 mM CaCl 2 , 2 mM MgCl 2 , and 25 mM HEPES) and arrest in the L1 stage. Synchronized L1 stage worms were grown on NGM plates seeded with OP50 until reaching the L4 stage.

Reporter constructs
Psmf-3::GFP-H2B was generated by PCR amplification of sequences 1,500 bp upstream from the initiation ATG of smf-3 (Y69A2AR.4) using primers containing SalI and NheI restrictions sites, cloned into TOPO Zeroblunt (Invitrogen) followed by digestion and insertion into Sal1 and NheI sites of pAP.10 as previously described (Romney et al., 2011). The GATA-like DAE 1-3 mutants were generated by mutation of GATA >CATA in Psmf-3::GFP-H2B using the QuikChange Site-Directed Mutagenesis kit (Stratagene). Transgenic strains were generated using a standard microinjection protocol. Each DAE construct (20 ng/ml), a selection plasmid pBX-1 (100 ng/ml) and salmon sperm DNA (25 ng/ul) were coinjected into pha-1 animals. Transgenic animals were obtained after growth at 25˚C. Generation of NHR-14::FLAG::GFP transgenic worms The expression pattern of NHR-14 was determined using transgenic animals expressing a fosmid (40 kb) encoding a NHR-14::3xFLAG::GFP transgene (TransgeneOme) (Sarov et al., 2006). This fosmid contains the entire nhr-14 locus as well as large amounts of flanking DNA critical for endogenous nhr-14 expression. The fosmid was injected into ttTi5606; unc-119 worms at 250 ng/ul and three independent lines were established. Line three was used for experiments in Figure 2A-D.

EMS screen and SNP mapping
Ethyl methanesulfonate (EMS) mutagenesis was performed as previously described (Brenner, 1974).
Briefly, L4 stage hif-1(ia4) worms (Jiang et al., 2001) were treated with 50 mM EMS in M9 buffer (22 mM KH 2 PO 4 , 42 mM Na 2 HPO 4 , 86 mM NaCl) for 4 hr. Mutagenized worms were washed five times to remove excess EMS, plated on NGM plates and grown until gravid. Gravid F1s were then bleached, and F2 eggs plated on NGM plates. Gravid F2 worms were transferred to 25 uM BP plates and removed after laying F3 eggs. F3 progeny that rescued the low iron phenotype of hif-1(ia4) mutants (determined by body length measurements using ImageJ) were scored as suppressors. We identified 16 suppressors from 15,000 haploid genomes using EMS, six that fully and 10 that partially rescued the hif-1(ia4) low iron phenotype. We focused on two mutations, qa6909 and qa6910, that strongly suppressed the hif-1(ia4) low iron phenotype. The other four strong suppressors were sequenced and were not mutations in nhr-14.
Whole genome sequencing-single nucleotide polymorphism (WGS-SNP) mapping was used to identify the causative mutations in qa6909 and qa6910 (Doitsidou et al., 2010). hif-1(ia4) mutants were crossed into the polymorphic Hawaiian strain CB4856 six times. Hawaiian hif-1(ia4) males were crossed with qa6909 and qa6910 mutants and F1 progeny were moved to fresh NGM plates. Gravid F1 adults were allowed to lay eggs overnight on NGM plates containing 25 uM BP. Adults were removed the next day and~50 F2 recombinants that reached adulthood within five days were singled to NGM plates and DNA from their F3 and F4 progeny was purified using the PureLink Genomic DNA Mini kit (Invitrogen) and pooled. WGS was performed using an Illumina HiSeq sequencer. Hawaiian SNP positions were mapped on the genomes of suppressor mutants qa6909 and qa6910 using the single-step SNP-mapping pipeline CloudMap (Minevich et al., 2012).

Rescue experiments
Five gravid adults from each strain were picked to either NGM or NGM-25 uM BP plates and allowed to lay eggs overnight. Adults were removed the following day and progeny were incubated for five days at 22˚C after which images were collected using a Leica M205 FA microscope and Leica DFC 310 FX camera (Leica Microsystems). Progeny were scored based on length using Image J.

RNAi
RNAi clones nhr-14 (T01B10) and pqm-1 (F40F8.7) were obtained from the Ahringer RNAi feeding library . Empty vector (L4440) was used as a control. All RNAi clones were verified by sequencing. Bacterial strains (HT115 DE3) were grown overnight in Luria-Bertani medium containing 50 mg/ml ampicillin and seeded onto NGM or NGM-BP plates containing 100 mg/ml ampicillin and 1 mM isopropyl b-D-1-thiogalactopyranoside (IPTG) to induce the gene of interest. Bacteria were then incubated at room temperature overnight to induce dsRNA. Five hif-1(ia4) gravid adults were allowed to lay eggs overnight on NGM or NGM-25 uM BP RNAi plates at 20˚C. Adults were removed the following day and progeny were scored for rescue against BP as described above.
For qPCR experiments, 20-30 gravid adults were placed on RNAi plates for 48-72 hr. The progeny that reached L3 and L4 stages were harvested for RNA isolation.
For nhr-14(tm1473); pqm-1 RNAi experiments, L1 stage synchronized worms were seeded onto RNAi plates and incubated at 22˚C to obtain adult hermaphrodites (P0). Gravid hermaphrodites from P0 were used to obtain a L1 stage synchronized population (F1), which was seeded onto pqm-1 RNAi plates for a second round. Gravid F1 hermaphrodites were synchronized to obtain a L1 stage population (F2) and were seeded on pqm-1 RNAi plates for a third round. Worms were allowed to reach L4 stage and then harvested for qPCR experiments.

qPCR
For indicated conditions and time-points, worms were washed off the plate with M9 buffer followed by three further washes to remove external bacteria, and rocked for 0.5 hr in M9 buffer to clear bacteria from the gut. RNA was extracted using TRIzol Reagent (Invitrogen) according to the manufacturer's protocol. cDNA synthesis was performed using SuperScript III First-Strand Synthesis SuperMix (Invitrogen) or qScript XLT cDNA SuperMix (QuantaBio). After reverse transcription, qPCR was performed using TaqMan Gene Expression Assays (Applied Biosystems) and analyzed using an Applied Biosystems 7900 HT qPCR instrument. The cycle threshold (Ct) value for each transcript was normalized to either act-1, act-2b or nhr-23. For pathogen experiments, all values were normalized using control genes nhr-23 or tba-1. The comparative Ct method was used to quantify transcript abundance. At least three biological replicates with duplicate technical replicates were used for qPCR experiments. TaqMan Assays are listed in Key Resources Table. RNA-seq analysis Wild-type N2 and nhr-14(tm1473) L1 stage worms were grown on NGM plates seeded with OP50 and harvested at L4 stage (48 hr at 22˚C). Three biological replicates for each strain were used for RNA-seq analysis. Worms were collected by centrifugation and washed with M9 buffer to remove, and RNA was extracted using a modified TRIzol procedure. Briefly, the aqueous phase was mixed with an equal volume of 70% EtOH and added to RNeasy spin columns (Qiagen). RNA concentration and quality were determined using a NanoDrop spectrophotometer followed by further quality control with the Bioanalyzer RNA 9000 Nano Chip. Library construction was performed using the Illumina TruSeq RNA Sample Preparation Kit v2 (RS-122-2001 andRS-122-2002) and was sequenced with the Illumina HiSeq (50 cycle, single-end). Sequenced reads were aligned to the Ce10 (Worm-Base Build 220) transcriptome index using Novoalign, and differential expression was determined using Useq 8.7.4 (Nix et al., 2008). Transcripts with a log2 ratio of 1 or higher at a false discovery rate (FDR) 0.05 were considered differentially expressed. Gene enrichment analysis was performed using GOrilla (Eden et al., 2009), which determines enriched GO terms at the top of a ranked list of genes. BioProspector (Liu et al., 2001) was used for de novo motif finding with the width parameter set to 16. Consistent motifs were identified using 1000 bp upstream of upregulated and downregulated genes.

Pathogen infection and lifespan analysis
P. aeruginosa strain PA14 slow-kill (SK) assays were performed as previously described (Tan et al., 1999). PA14 was cultured in Luria broth and seeded on slow-kill plates. Eggs from gravid adults were obtained by alkaline hypochlorite treatment, allowed to hatch overnight in S-basal medium, and arrested in the L1 stage. Synchronized L1 stage worms were grown on NGM plates seeded with OP50 until reaching the L4 stage. Each strain of L4 stage worms were transferred to NGM or SK plates containing 75 uM 5-fluoro-2-deoxyuridine (FUdR) to prevent growth of progeny and incubated at 22˚C. C. elegans viability was scored every day using a dissecting microscope to detect dead worms (Figure 8-source data 1). For iron content quantification, worms were harvested after 24 h PA14 exposure and processed for ICP-MS analysis.
Lifespan analysis of nhr-14(tm1473) and wild-type N2 worms grown under low iron conditions was performed by transferring L4 stage worms to NGM and NGM-25 uM BP plates containing 75 uM FUdR to prevent growth of progeny. Assays were carried out at 22˚C and worms were scored every day. For pathogen and lifespan analysis, worms were scored as dead if they did not move, pump, or respond to gentle prodding. Survival analyses were analyzed by Kaplan-Meir method, and statistical significance was assessed using the Mantel-Cox log-rank test (GraphPad Prism).

Western blot analysis
At indicated time points, worms were washed off the plate with ddH 2 O and washed three times to remove bacteria. Worms were resuspended in lysis buffer (20 mM HEPES pH 7.5, 25 mM KCl, 0.5% NP-40) and disrupted by two 5 s pulses using an Ultrasonic Processor (Sonics) at 50% amplitude. Protein concentrations were determined using Coomassie Plus Protein Assay Reagent (Thermo Scientific). Protein samples were incubated for 10 min at 95˚C in NuPage LDS Sample Buffer (Invitrogen) containing 10 mM DTT then subjected to electrophoresis using NuPage 4-12% Bis-Tris gels (Invitrogen) and transferred to Amersham Hybond ECL nitrocellulose (GE Healthcare). Blots were incubated with mouse anti-FLAG antibody (Sigma cat# F3165) along with monoclonal b-tubulin antibody (MP Biomedicals LLC cat# 08691261) in 1x Tris-buffered saline (0.01% Tween-20) containing 5% non-fat dry milk. Horseradish peroxidase-conjugated secondary antibodies (Jackson Laboratories) were visualized using Western Lighting Plus-ECL Chemiluminescence Substrate (Perkin Elmer). Three biological replicates were performed and NHR-14::GFP::FLAG was quantified by densitometry analysis using Image J (NIH).

Iron content quantification
Synchronized L1 stage worms were grown on NGM plates seeded with OP50 until reaching the L4 stage. Worms were washed extensively with M9 buffer and incubated in M9 buffer at room temperature for 2 hr to allow for purging of the gut followed by three rinses with ddH 2 O. Worms were pelleted and dried, and metal analysis determined by inductively-coupled plasma-optical emission spectroscopy (ICP-OES) (Children's Hospital Oakland Research Institute (CHORI) Elemental Analysis Facility). Empty tubes were run in parallel to serve as controls. Iron content was normalized to sulfur content. For pathogen experiments, worms were harvested in water, washed and homogenized for metal analysis. Protein concentration in lysates was measured using the BCA Protein Assay (Thermo Fisher Scientific) and 100 mg of lysate was digested overnight in 5:1 HNO 3 :H 2 O 2 , dried, and the pellet resuspended in 2% HNO 3 for metal analysis by inductively-coupled plasma mass spectrometry (ICP-MS) (University of Utah, Center for Iron and Hematology Disorders, Iron and Heme Core). At least 3-5 biological experiments were performed for each strain with triplicate technical replicates. Calibration standard solutions for determination of Fe, Zn, Cu and Mn were prepared from Agilent multi-element calibration standard-2A. Iron content was normalized to protein.

COPAS biosort
Animals expressing Psmf-3::GFP-H2B containing 1,500 bp of smf-3 promoter sequences (Romney et al., 2011), Psmf-3(mDAE1)::GFP-H2B, Psmf-3(mDAE2)::GFP-H2B and Psmf-3(mDAE3):: GFP-H2B reporters were synchronized by hypochlorite treatment, and arrested L1 stage larvae were grown on 10 cm NGM plates containing nhr-14 RNAi or control (empty vector) RNAi for 48 hr described above. Worms were washed in M9 buffer and transferred to nhr-14 RNAi or control RNAi plates supplemented with either 100 mM BP or 6.6 mg/mL FAC and grown overnight. Worms were collected by centrifugation and washed several times in M9 buffer to remove bacteria and debris. Worms were analyzed using COPAS Biosort where L4 stage to young adult worms were gated based on extinction and time of flight parameters (TOF). Extinction and TOF parameters were held constant for subsequent GFP fluorescence acquisition throughout all conditions. GFP fluorescence for 1000 worms was analyzed using FlowJo.

Microscopy
Transgenic worms carrying NHR-14::GFP::FLAG fosmid were imaged at 40x and 60x using an Olympus FV1000 confocal microscope and camera using Olympus FV10-ASW 3.1 confocal imaging software. PQM-1::GFP::FLAG worms were imaged at either 20x or 40x using a Leica DM6000B microscope. Following acquisition, images were rotated, cropped and sized using Adobe Photoshop.

Statistical analysis
Statistical analyses were performed using GraphPad software Prism 6 or 7.04 and Excel. Results are expressed as the mean ± SEM. Data were analyzed by an unpaired two-tailed Student's t-test or two-way ANOVA with Tukey's multiple comparison test. Survival comparisons were performed using the Mantel-Cox log-rank test. All data were evaluated at the significance level p 0.05. Biological replicates reflect different sources of material and/or experiments performed on different days. Statistical details for experiments are indicated in the figure legends. Data availability RNA-seq data have been deposited in GEO under accession code GSE89783. In addition, raw RNAseq data are reported in the source data files.
The following dataset was generated: