Glucose negatively affects Nrf2/SKN-1-mediated innate immunity in C. elegans

High glucose levels negatively affect immune response. However, the underlying mechanisms are not well understood. Upon infection, the round worm C. elegans induces multiple gene transcription programs, including the Nrf2/SKN-1-mediated detoxification program, to activate the innate immunity. In this study, we find that high glucose conditions inhibit the SKN-1-mediated immune response to Salmonella typhimurium, exacerbate the infection and greatly decrease survival. The effect of glucose shows specificity to SKN-1 pathway, as UPRmit and UPRER that are known to be induced by infection, are not affected. Hyper-activation of SKN-1 by wdr-23 RNAi restores partly the immune response and increases the survival rate in response to S. typhimurium. In all, our study reveals a molecular pathway responsible for glucose’s negative effect on innate immunity, which could help to better understand diseases associated with hyperglycemia.


INTRODUCTION
One of the hallmarks of diabetic complication is impaired wound healing [1].This probably is resulted from a negative effect of high glucose on immune response.It is shown that high glucose levels can inhibit the normal function including chemotaxis, phagocytosis, killing of polymorphonuclear cells such as monocytes and macrophages [2].Consistently, in C. elegans, high glucose can cause rapid aging and greatly shorten the lifespan [3][4][5].As pathogenic infections also strongly decrease the lifespan of C. elegans [6,7], glucose may speed up aging by compromising the im-mune response.However, whether this is the case and what are the underlying molecular mechanisms remain poorly understood.
Previously, efforts have been directed to delineate the signaling pathways that mediate the glucose signaling in infection-related diseases.It was found in mice that several key molecular pathways participate in the pathogenic process, such as the protein kinase C pathway, the protein glycation pathway, TNF-α pathway, etc. [8].These pathways are interconnected and dynamically regulate immunity in response to high glucose.Deregulation of these pathways tends to AGING elevate the intracellular levels of reactive oxygen species (ROS), increase oxidative damage and cause abnormal protein modification, which could contribute to the disease state of diabetes [8][9][10].Therefore, antioxidation pathways could play an important role in the regulation of glucose suppression of immune system.
One of the major anti-oxidation pathway is controlled by the nuclear factor erythroid 2 (NFE2)-related factor 2 (Nrf2) [11,12].Nrf2 is a member of the cap 'n' collar (CNC) subfamily of basic region leucine zipper (bZip) transcription factors.Nrf2 protein is highly conserved and its C. elegans homolog SKN-1 functions in similar way to bind promoters of oxidative stress-related genes [12].The regulation of both Nrf2 and SKN-1 is largely through posttranslational modifications and proteasomemediated degradation.Under normal conditions, Nrf2 is sequestered by its associated protein KEAP-1 in the cytoplasm, targeted for ubiquitination and proteasomemediated degradation [13].Upon stress, posttranslational modifications of either Nrf2 or KEAP-1 cause release of the Nrf2 into the nucleus, which then drives transcription of stress responsive genes.Nrf2 has been implicated in many diseases [14,15] and also the aging process [12,16].Similarly in C. elegans, SKN-1 is also targeted by proteasome for degradation upon stress conditions.However, as C. elegans lacks KEAP-1 ortholog [12], WDR-23 has been proposed to be play similar roles in preventing Nrf2 degradation [17].
Aging modulators are commonly targeted by pathogen defense pathways.For example, in C. elegans, SKN-1 isrequired for many lifespan extension events [12], and it is also an essential regulator of innate immunity [18].The well-known aging regulator DAF-16 is required for protection against strains that kill C. elegans slowly by gut colonization [19,20].Recently, DAF-16 is also found to be phosphorylated by MBK-1 [21], a DYRK kinase that promotes resistance to P. aeruginosa [22].In mammals, several studies suggest Nrf2 to be an important modulator of innate immunity.Loss of Nrf2 in mice renders the animal vulnerable to infection by viruses [23].Similarly, disruption of Nrf2 dramatically increases susceptibility to cecal ligation and punctureinduced sepsis [24].On the other hand, activation of Nrf2 reduces the infection by Salmonella typhimurium [25].In addition, mice fed with sulforaphane, a pharmacologic activator of Nrf2, show significant resistance to infection by bacteria such as Pseudomonas aeruginosa and Plasmodium.falciparum [26,27].All these results suggest an important role of Nrf-2/SKN-1 in fighting against infectious disease.However, whether Nrf2/SKN-1 plays a role in high glucose-induced immunity malfunction remains unknown.
We are interested to know the functions of SKN-1 in high glucose conditions.As the physiology of the round worm C. elegans is severely affected by glucose, this animal model is especially suited to study glucose toxicity.In this report, we examine the response of SKN-1 pathway to S. typhimurium in the presence or absence of glucose and found that glucose dampens SKN-1 activity, increases infection and limits survival.Activation of SKN-1 by knocking down wdr-23 expression alleviates the negative effect of glucose on infection and lifespan of C. elegans.This study may help to better understand the pathological pathways underlying human diseases related to high glucose conditions.

Glucose renders C. elegans vulnerable to infection by S. typhimurium
To test if glucose would affect the innate immunity in C. elegans, we infected the animals with Salmonella typhimurium (S. typhimurium) by exposing L4 stage worms to the pathogen for 48 hours.The number of live bacteria inside the worms is commonly used to evaluate the extent of infection in C. elegans [28].To determine the degree of infection, we washed day-5 adult worms extensively to get rid of the bacteria on the outer cuticle.We then lysed the worms and plate the lysate in order to count the number of live bacteria by colony-forming assay.As shown in Fig. 1A, glucose supplemented in culture medium already robustly increased the infection, with 0.5% reaching the highest.We therefore used 0.5% of glucose for the rest experiments in this study if not otherwise stated.
As infection is highly associated with survival rate, we determined if glucose would affect survival of animals after infection by pathogen S. typhimurium.After infection, we transferred the animals to normal NGM plate with non-pathogenic OP-50 bacteria and recorded survival every other day until the worms were all dead.Consistent with the enhanced infection in Fig. 1A, glucose supplementation decreased survival rate of worms infected by S. typhimurium.Glucose also shortened the lifespan of C. elegans in the nonpathogenic bacteria such as OP-50 (Fig. 1B), as has been reported by several studies before [3][4][5].Both S. typhimurium and glucose shortened lifespan of C. elegans in a non-additive manner (Fig. 1C).

Glucose prevents S. typhimurium from activating SKN-1
There are several transcriptional programs that mediate the innate immune response in C. elegans.Next, we wanted to know if SKN-1-mediated transcriptional program would be negatively affected by high glucose conditions.To this end, we first asked if promoter activity of gst-4 gene, a direct transcriptional target of SKN-1, could be changed by feeding glucose.This is revealed by examining directly GFP expression from the gst-4 promoter (Pgst-4::gfp) [29].gst-4::gfp can be activated by many stressors, but in response to pathogen, it was robustly activated in a SKN-1dependent manner [18].Therefore, we used gst-4::gfp as a proxy for SKN-1 activity in response to pathogen.As a result, infection by S. typhimurium robustly increased the gst-4::gfp expression, indicating a strong SKN-1 activation (Fig. 2A).However, when raised on NG medium supplemented with glucose, SKN-1 activation was greatly reduced (Fig. 2A and 2B).Consistent with previous studies [18], the Pgst-4::gfp reporter was not activated by non-pathogenic OP-50 (Fig. S1); adding glucose suppressed S. typhimuriuminduced GFP, but did not change the GFP levels in OP-50 medium (Fig. S1).We have observed similar results in RNAi experiment using HT115 bacteria (Fig. 4A and  4B).We conclude that the glucose effect we observed is specific to conditions of infection by S. typhimurium.
Second, we tested the expression of several Nrf-2 target genes (gst-4, gcs-1 and gst-10) by qRT-PCR.Consistently, induction of these SKN-1 target genes by S. typhimurium was suppressed when glucose was pre-sented in the culture medium (Fig. 2C).Without induction by S. typhimurium infection, gst-4 expression was low and not affected by glucose (Fig. S2), consistent with the gst-4::gfp reporter in Fig. S1.Expression of genes encoding skn-1 and upstream kinase such as sek-1, pmk-1 and gsk-3 were not affected by glucose (Fig. 2C), suggesting that the mRNA levels of upstream regulators were not affected (Fig. 2C).Third, we examined the subcellular localization of SKN-1/Nrf2 protein under high glucose conditions.Transcription of SKN-1/Nrf2 target genes requires the accumulation of SKN-1/Nrf2 in the nucleus upon infection [18].We found that glucose strongly attenuated the accumulation of SKN-1::GFP in the intestinal cells of C. elegans (Fig. 2D and 2E).Therefore, glucose targets SKN-1/Nrf2 to impair immune response to S. typhimurium in C. elegans.

Glucose specifically inhibits SKN-1-mediated immune response to S. typhimurium
We also tested if other pathogen defense pathways were affected by glucose.The mitochondrial unfolded protein response (UPR mit ) and the endoplasmic reticulum unfolded protein response (UPR ER ) are key stress responses that are also induced upon certain pathogen infection [30,31].We asked if S. typhimurium could induce similar responses and how they were affected by glucose.To this end, we examined several downstream genes in these two pathways by qRT-PCR.The expression of hsp-6 and hsp-60 genes are well established markers for UPR mit [32], while hsp-3, hsp-4, pek-1 and atf-6 are key genes involved in UPR ER [33].We found that hsp-6, hsp-60, hsp-3 and hsp-4 were significantly increased in expression upon infection by S. typhimurium.However, for all these genes, we detected no significant difference between worms cultured with and without glucose (Fig. 3A and 3B).These data suggest that the glucose specifically targets SKN-1 pathway to affect immune response in C. elegans.
Loss of SKN-1 shortens lifespan.Glucose can also greatly shorten lifespan.If glucose shortens lifespan through inhibiting SKN-1, one would expect that, knocking down skn-1 expression can no longer shorten lifespan in the presence of glucose.Indeed, we confirmed that wild-type N2 worms cultured with SKN-1 RNAi bacteria were much shorter lived than the control.However, when the worms were cultured on the same RNAi bacteria plate supplemented with 0.5% of glucose, lifespan was no longer shortened (Fig. 3C).

SKN-1 hyper-activation diminishes the negative effect of glucose on immunity
Since glucose suppress the innate immunity through inhibiting SKN-1, it is possible that hyper-activation of SKN-1 could reverse the negative effect of glucose.To test this idea, we activated the SKN-1 activity through RNAi knocking down the expression of wdr-23 gene.WDR-23 is analog of mammalian KEAP-1, which binds directly to SKN-1 to promote its proteasome-mediated degradation [34].Knocking down wdr-23 can accumulate SKN-1 rapidly in the nucleus and activate transcription of SKN-1 target genes such as gst-4 [17].Confirming the previous results, we found that the promoter activity of gst-4 was highly induced by wdr-23 RNAi knockdowns, as revealed by gst-4::gfp reporter (Fig. 4A and 4B).Interestingly, when wdr-23 was knocked down, gst-4::gfp expression was no longer affected by glucose (Fig. 4A), suggesting that WDR-23 is functioning in the downstream or in parallel to glucose signaling.
We then evaluated the biological significance of wdr-23 knockdown by examining directly the infection degree through colony assay.Our results showed that the negative effect of glucose on infection was significantly mitigated by wdr-23 knockdowns (Fig. 4B).Furthermore, upon wdr-23 knockdowns, glucose's negative effect on survival in the presence of S. typhimurium was prevented (Fig. 4C and Fig. S3), suggesting that forced activation of SKN-1 could be sufficient to reverse hyperglycemia -induced infection.In the absence of Salmonella.typhimurium, however, wdr-23 RNAi showed no interaction with glucose (Fig. S3).These results suggest that glucose enhances the pathogen killing effect through WDR-23.

DISCUSSION
It has been known for a long time that high glucose condition can exacerbate pathogenic infection in human [2,8].But the underlying mechanism remains incomepletely understood.By using the round worm C. elegans as a model, we aimed to better understand the mechanisms by which glucose modulates the innate immune response.Our studies reveal SKN-1 as a key mediator of glucose's negative effect on infection by Salmonella typhimurium.SKN-1 is homologous to human Nrf2, which has been implicated in many human diseases such as cancers, neurodegenerative disease and diabetic Nephropathy [35][36][37].Our study therefore may have significant relevance to the human physiology and pathology.Quantitative RT-PCR was conducted using established primer sets (Table S4).P values were obtained by student's t-test.**, P<0.001; ns, not significant.(B) Genes known to be induced by endoplasmic reticulum unfolded protein response (UPR ER ) were not affected by glucose.Animals raised on medium with and without glucose from L1 to L4/young adult stage were infected by S. typhimurium for 2 days.mRNA were extracted and reverse transcribed to cDNA.Quantitative RT-PCR was conducted using established primer sets (Table S4).Two independent experiments shows similar results and one of them are shown.P values were obtained by student's t-test.ns, not significant.(C) Glucose and skn-1 RNAi knockdown is not additive in decreasing C. elegans' lifespan.Lifespan and infection were carried out at 20 ºC.Animals raised on medium with and without glucose from L1 to L4/young adult stage were infected by S. typhimurium for 2 days, then transferred back to non-infected OP-50 bacteria plate for the rest of life.Survival were recorded every other day until all died.Data were collected from two independent experiments with number of worms >100.See Table S2 for details.

AGING
Recent efforts to understand the underlying mechanisms that lead to compromised immune response by hyperglycemia have implicated several pathways, such as PKC, ROS and polyol pathway [8].It is interesting to note that dysregulation of these pathways are prone to increase intracellular ROS levels [8,9].As ROS can activate Nrf2/SKN-1, our finding of SKN-1 in mediating glucose's effect on immunity is in line with abovementioned results [8,9].Interestingly, because high glucose metabolism is generally known to accelerate the production of ROS, one would expect that SKN-1 would be activated, instead of suppressed by glucose as shown in this study.One possibility for the increased ROS in glucose metabolism is that it is a result of SKN-1 inhibition, which will be in line with the data presented.Another possible explanation could be that the prominent mechanism for glucose to suppress Nrf2/SKN-1 is not mediated by ROS, but through other unknown molecules.One can imagine that there could be many possible mechanisms that target transcription, translation and posttranslational modification of certain components in the Nrf2/SKN-1 pathway.Gaining insight into the underlying mechanism requires further efforts and expertise from multiple fields of study.S3 for details.
Another very interesting possibility that deserves special attention is that high glucose condition may suppress the generation of ROS, which then prevent the activation of SKN-1.Nrf2/SKN-1 is a key regulator of the innate immunity [18,24].It is known that lipopolysaccharides from the invading microbes trigger a pathway that lead to the generation of superoxide.The superoxide then disrupts the function of KEAP-1/WDR-23, a negative regulator of Nrf2/SKN-1 [38,39], therefore activating Nrf2/SKN-1.High glucose may act on these molecules in the upstream of KEAP-1/WDR-23.This is consistent with our result that by RNAi knocking down of wdr-23, glucose's negative effect on SKN-1 target genes expression and survival upon infection is alleviated.Recently, several groups found that ROS, at low levels, can serve as signal to activate protective programs in order to counteract further accumulation of deleterious ROS [29,40].It is found that high glucose can suppress the increase of such ROS [4], negatively affecting key biological processes that determine survival.Similar mechanisms could exercise in the interface of glucose and infection, resulting in inhibition of innate immunity in C. elegans upon S. typhimurium infection.Our results may suggest a conserved and novel pathway underlying the negative effect of hyperglycemia on human immune response.It will be of great interest to examine in other animal models, especially mammalian model such as mice to see if the current finding is conserved in regulation of hyperglycemia-mediated diseases, such as diabetes.
Interestingly, although opposing in terms of gst-4::gfp induction, both S. typhimurium and glucose shortened lifespan of C. elegans in a non-additive manner (Fig. 1C).Consistent with our study, another recent report also shows that glucose has no additive effect on killing, sometimes even extends the survival of C. elegans that are infected with certain pathogens [41].These results suggested that S. typhimurium and glucose may shorten lifespan through a common pathway.Alternatively, as lifespan/survival is a downstream phenotype that can be complicated by many factors, another possibility could be that the lifespan is affected by other pathways such as mitochondrial and/or ER unfolded protein responses, which are known to contribute to lifespan extension in C. elegans [30,42].

Pathogen infection and survival measurement
S. typhimurium infections were done as described in [28].Briefly, pathogenic bacteria were cultured overnight and plated on NGM agar plates overnight.L4 and young adult worms were transferred to S. typhimurium-containing plate without glucose and raised for 48 hours, then transferred back to plates seeded with non-pathogenic OP-50 or HT115(DE3) when doing RNAi knockdown.To determine the degree of infection, 20 day-5 worms were washed extensively then mechanically disrupted by homogenizer in 0.5 ml of M9 buffer.The homogenized solution was serial diluted and plated on S. typhimurium-selective XLD agar plates (Xylose Lysine Desoxycholate, EMD Chemical Inc.) to count the titer.
To determine survival rate, infected animals were examined for death every other day and percentage of death were plotted using Graphpad Prism 5.

RNAi treatment
RNAi clones were from a collection initially generated in Julie Ahringer's laboratory [43], with the same bacteria strain (HT115) containing the empty vector L4440 as control.RNAi experiments was conducted by feeding worms on agar plates with bacteria expressing double-stranded RNA (dsRNA) corresponding to genes to be knocked down.Specifically, bacteria bearing a wdr-23 DNA fragment were cultured to log phase and seeded on NG plates containing 50 ug/mL Carbenicillin and 1 mM Isopropyl β-D-1-thiogalactopyranoside (IPTG) for at least 24 hours to induce dsRNA expression.L1 stage worms were then transferred to and maintained on the RNAi plate throughout life, gene knockdowns were confirmed by activation of skn-1 reporter gst-4::gfp.

Quantitative real time polymerase chain reaction (qRT-PCR)
The qRT-PCR was done similarly as described in [44].Briefly, worms were washed from agar plates with icecold M9 buffer, total mRNA were then extracted by Trizol.mRNA was reverse-transcribed to cDNA using AGING QIAGEN One-Step RT-PCR Kit.Quantitative PCR was performed using SYBR Green 2X Mater Mix (Applied Biosystems).Gene expression levels were normalized to actin (ACT1) and expressed as fold changes to that of the wild-type.Primers are published before [45], which are also listed in Table S4.

Microscopic imaging
Worms were paralyzed in 1mM levamisole solution and mounted on 3% of agarose gel pad, covered with cover slide and subject to immediate examination by fluorescent microscope.Worms expressing gst-4::gfp were imaged with stereo microscope (Leica Microsystem).Worms expressing SKN-1::GFP were imaged through confocal microscope (Perkin Elmer UltraView Vox Spinning Disk Confocal).Signals from individual animals were quantified with Image J software and plotted as dots.

Statistical analysis
Survival curves and associated data including mean lifespan, standard errors and P values were generated by bioinformatics software Graphpad Prism.P values for bar data are based on student's t-test and survival curves are based on log rank test.P < 0.01 was considered statistically significant.

Figure 1 .
Figure 1.Glucose medium exacerbates S. typhimurium infection and shortens survival.(A) Glucose increases S. typhimurium infection in C. elegans.Animals cultured in the presence and absence of various concentrations of glucose were infected with S. typhimurium at L4 or young adult stage for 2 days.The numbers of infected pathogen were determined by lysing 20 worms and colony forming assay of live S. typhimurium inside the worms.Colony forming unit (CFU) was plotted using Log2.Data from two independent experiments were pooled and plotted.Error Bars stands for standard error of the mean (SEM).P values were obtained by student's t-test.*, P<0.01, **, P<0.001.(B) Glucose decreases lifespan of infected animals.Lifespan and infection were carried out at 20 ºC.Animals were infected with S. typhimurium at L4 or young adult stage for 2 days then transferred to normal NGM plates.Survival of control and infected animals were recorded every other day.Data were collected from two independent experiments with number of worms >100.See Table S1 for details.(C) Comparison of killing effect in the presence and absence of glucose.Lifespan of animals (n>100) were plotted in Whiskers box.P values were obtained by Log-rank test.***, P<0.0001.

Figure 2 .
Figure 2. Glucose medium decreases SKN-1 activity.(A) SKN-1 reporter gst-4::gfp is suppressed by high glucose medium.Animals expressing the SKN-1 reporter gst-4::gfp were cultured in medium supplemented with and without 0.5% glucose from L1 stage to L4/young adult stag, then transferred to infection plate without glucose for 2 days before imaging.Shown are representative images of at least 4 independent experiments (20 animals each).Scale bars are 200 µm.(B) Quantification of experimental results in Fig. 2A by measuring the signal intensity of 10 animals from 1 experiment by ImageJ software.P values were obtained by student's t-test.***, P<0.0001.Error bars indicates standard error of the mean (SEM).(C) SKN-1 target genes (gst-4, gst-10, gcs-1) but not skn-1 and upstream kinase genes (sek-1, pmk-1, gsk-3) are affected by glucose.Animals raised on medium with and without glucose from L1 to L4/young adult stage were infected by S. typhimurium for 2 days.mRNA were extracted and reverse transcribed to cDNA.Quantitative RT-PCR was conducted using established primer sets and protocols.Shown are representative data from 1 of 2 independent experiments.Error bars indicate standard error of the mean (SEM) of 3 replicates.P values were obtained by student's t-test.**, P<0.001; ***, P<0.0001; ns, not significant.(D) Glucose inhibits SKN-1 nuclear localization upon infection.Transgenic C. elegans expressing skn-1::gfp were raised on medium with and without glucose from L1 stage to L4/young adult stag, then infected with S. typhimurium for 2 days before imaging.Shown are representative image of 2 independent experiments."n" marks above the nucleus of intestinal cells.Scale bars are 40 µm.The punctate signals in the intestine are non-specific signals as also shown in Fig. S4.(E)Quantification of experimental results in Fig. 2D by counting the SKN-1::GFP positive nuclei/worms of about 20 worms.Shown are representative data from 1 of the 2 independent experiments.Error bars stands for standard error of the mean (SEM).P values were obtained by student's t-test.***, P<0.0001.

Figure 3 .
Figure 3. Glucose does not affect expression of marker genes of UPR mt or UPR ER .(A) SKN-1 target gene (gst-4) but not marker genes of mitochondrial unfolded protein response (UPR mt ) were affected by glucose.Animals raised on medium with and without glucose from L1 to L4/young adult stage were infected by S. typhimurium for 2 days.mRNA were extracted and reverse transcribed to cDNA.Quantitative RT-PCR was conducted using established primer sets (TableS4).P values were obtained by student's t-test.**, P<0.001; ns, not significant.(B) Genes known to be induced by endoplasmic reticulum unfolded protein response (UPR ER ) were not affected by glucose.Animals raised on medium with and without glucose from L1 to L4/young adult stage were infected by S. typhimurium for 2 days.mRNA were extracted and reverse transcribed to cDNA.Quantitative RT-PCR was conducted using established primer sets (TableS4).Two independent experiments shows similar results and one of them are shown.P values were obtained by student's t-test.ns, not significant.(C) Glucose and skn-1 RNAi knockdown is not additive in decreasing C. elegans' lifespan.Lifespan and infection were carried out at 20 ºC.Animals raised on medium with and without glucose from L1 to L4/young adult stage were infected by S. typhimurium for 2 days, then transferred back to non-infected OP-50 bacteria plate for the rest of life.Survival were recorded every other day until all died.Data were collected from two independent experiments with number of worms >100.See TableS2for details.

Figure 4 .
Figure 4. Activation of SKN-1 pathway mitigates the negative effect of glucose on immune response to S. typhimurium.(A) Knocking down wdr-23 bypasses glucose to activate gst-4::gfp.Animals expressing the SKN-1 reporter gst-4::gfp were fed bacteria expressing double-stranded RNA of wdr-23 from L1 stage to L4/young adult stage on medium with and without glucose, then infected with S. typhimurium for 2 days before imaging.Two independent trials gave similar results and data from one of them were shown.Scale bars are 600 µm.(B) Quantifications of gst-4::gfp intensity of 10 animals in images shown in A by ImageJ software.Ctrl, control; gluc, 0.5% glucose; S, S. typhimurium.P values were obtained by student's t-test.**, P<0.001.ns, not significant.(C) Knocking down wdr-23 alleviates glucose's negative effect on infection.Animals fed bacteria expressing double-stranded RNA of wdr-23 from L1 stage to L4/young adult stage on medium with and without glucose were infected with S. typhimurium for 2 days.The numbers of infected pathogen were determined colony forming assay of live S. typhimurium inside the worms.Colony forming unit (CFU) was plotted using Log2.P values were obtained by student's t-test.**, P<0.001.(D) Knocking down wdr-23 prevents glucose from shortening lifespan of infected C. elegans.Lifespan and infection were carried out at 20 ºC.Animals fed bacteria expressing double-stranded RNA of wdr-23 from L1 stage to L4/young adult stage on medium with and without glucose were infected with S. typhimurium for 2 days.Worms were then transferred to non-infected RNAi bacterial plates.Survival was recorded every other day until all worms died.Data were collected from two independent experiments with number of worms >100.See TableS3for details.

Figure S2 .
Figure S2.(Related to Figure 2C) glucose specifically affect S. typhimurium activation of SKN-1 target genes.Animals raised on medium with and without glucose from L1 to L4/young adult stage were infected with and without S. typhimurium for 2 days.mRNA were extracted and reverse transcribed to cDNA.Quantitative RT-PCR of SKN-1 target genes (gst-4, gst-10, gcs-1) was conducted using established primer sets and protocols.Shown are representative data from 1 of 2 independent experiments.Error bars indicate standard error of the mean (SEM) of 3 replicates.P values were obtained by student's t-test.**, P<0.001; ***, P<0.0001; ns, not significant.

Figure S3 .
Figure S3.(Related to Figure 4D) wdr-23 RNAi cannot attenuate glucose's negative effect on lifespan of non-infected C. elegans.Lifespan and infection were carried out at 20 ºC.Animals fed bacteria expressing double-stranded RNA of wdr-23 from L1 stage to L4/young adult stage on medium with and without glucose were infected with S. typhimurium for 2 days.Worms were then transferred to non-infected RNAi bacterial plates.Survival was recorded every other day until all worms died.number of worms >100 for each sample.See TableS5for details. ns, not significant, * P<0.01.

Figure S4 .
Figure S4.(Related to Figure 2D) The punctate signals are auto-florescence in the intestine of C. elegans.Animals without (left) and with SKN-1::GFP transgenes (right) were raised on medium with and without glucose from L1 stage to L4/young adult stag, then infected with S. typhimurium for 2 days before imaging."n" marks above the nucleus of intestinal cells.Scale bars are 40 µm.