Introduction

The symbiosis between cnidarians and dinoflagellate algae in the family Symbiodiniaceae is the basis for coral reef development. Corals can grow in nutrient-poor waters thanks to the support of organic molecules provided by their algal symbionts (Muscatine and Porter 1977). The translocation and exchange of organic molecules that are metabolically energetic has been the focus of many studies aimed at understanding the mechanisms that drive and regulate this symbiosis (Hofmann and Kremer 1981; Fitt and Trench 1983; Wakefield and Kempf 2001; Burriesci et al. 2012; Davy et al. 2012; Kopp et al. 2015). The algal symbionts reside in the host gastrodermis surrounded by several membranes of algal origin, plus an outermost host-derived membrane, known as the symbiosome membrane complex (Fitt and Trench 1983; Wakefield and Kempf 2001), and any exchange of organic molecules must be through this selective barrier. Photosynthetic products translocated to the host, or photosynthates, constitute a major source of energy in this symbiosis, with free sugars like glucose and glycerol being the most important (Muscatine 1967; Lewis and Smith 1971; Schmitz and Kremer 1977; Hofmann and Kremer 1981; Gordon and Leggat 2010; Davy et al. 2012; Kopp et al. 2015). The primary source of nitrogenous compounds to the symbiotic partners is through heterotrophic feeding (Davy et al. 2012).

Transmembrane transporters involved in the distribution and assimilation of nitrogen and carbon have been identified in symbionts of the Symbiodiniaceae family (Aranda et al. 2016). Correspondingly, the animal host must be capable of internalizing photosynthates to cover its basal energetic requirements (Lampert 2016), but these are not thoroughly studied. The molecular exchange of organic compounds could be facilitated by internal membrane proteins located in the cell membranes of each symbiotic partner, and in the symbiosomal interface (Fitt and Trench 1983; Wakefield and Kempf 2001). Some studies have suggested that specialized proteins may be responsible for the transport of organic molecules from symbiont to host, like glucose and glycerol (Sproles et al. 2018). Glucose membrane transporters (GLUT-type) in the Solute Carrier family (SLCA2) are of interest, given their high affinity for glucose and location in internal membranes (Rands et al. 1993; Escher and Rasmuson-Lestander 1999; Wakefield and Kempf 2001; Joost et al. 2002; Wood and Trayhurn 2003; Uldry and Thorens 2004; Wright 2013; Scheepers et al. 2015; Sproles et al. 2018). GLUTs transport glucose down its concentration gradient, hence the name facilitated diffusion (Castrejón et al. 2007). The glucose-sensing function has two components: (1) GLUT-mediated glucose entry into the cell and (2) glucose metabolism through glucokinase phosphorylation (Díaz and Burgos 2002; Castrejón et al. 2007). Glucose selectivity by GLUT transporters is determined by a highly conserved amino acid sequence (QLS) in transmembrane segment seven, the glucose recognition site (Díaz and Burgos 2002). In the symbiotic sea anemone Exaiptasia diaphana, GLUT8 exhibited high levels of mRNA expression under symbiotic conditions in contrast to an aposymbiotic condition (Lehnert et al. 2014). This same transporter was later immunologically detected by Mashini and collaborators (2022) in E. diaphana infected with homologous and heterologous symbionts in the gastrodermis and epidermis of symbiotic and aposymbiotic anemones; however, these authors did not find differences in the immunodetection of this protein according to symbiotic state.

Glycerol is transported through Intrinsic Channel Proteins of the Superfamily of Major Intrinsic Proteins (MIPs) known as Glycerol Facilitators (GFs), channel-like plasma membrane proteins that can take up or exclude glycerol (Castrejón et al. 2007). Glycerol can also be transported through Aquaglyceroporins (GLPs), which are channels that transport small uncharged molecules, such as glycerol, CO2, and urea, but exclude water (Rojek et al. 2008). Once glycerol is transported to the symbiosomal space, it must cross the host cell membrane possibly through a GLP, whose transport mechanism may be by diffusion (Castrejón et al. 2007). GLPs are regulated by phosphorylation or glycosylation (Tamás et al. 1999; Mandal et al. 2012) and by activation of pH-dependent ions or intracellular signals (Von Bülow and Beitz 2015). In aquaglyceroporins (GLPs), conserved motifs (“NPA”) form an aromatic/arginine selectivity filter (ar/RSF) that largely determines solute specificity and substrate transport (Deshmukh et al. 2015). In cnidarians, 16 sequences with homology to GLPs have been identified, all with at least six transmembrane domains, characteristic of these proteins, with amino-terminal and carboxy-terminal ends in the cytoplasm (Agre et al. 2002).

By phylogenetic analysis, Sproles and collaborators (2018) identified putative GLUT transporter proteins in cnidarians; these transporters clustered into two classes, regarded as highly conserved human homologs. These same authors identified the GLUT8 transcript in the sea anemone Aiptasia pallida (accepted name: Exaiptasia diaphana) as being probably located in internal membranes and, for non-symbiotic species, in the plasma membrane, highlighting the importance of these proteins in the exchange of potential photosynthates in the symbiosis. Also, aquaglyceroporins considered by these same authors (Sproles et al. 2018) shared similarities to human GLP3 and GLP9 and suggested the need for experimental evidence to confirm their role in cnidarian transport. Mashini et al. (2022) later reported elevated levels of AQP3 protein in aposymbiotic Exaiptasia, suggesting that this was a response to a reduced supply of symbiont-derived organic carbon.

Model organisms, such as the medusa Cassiopea xamachana, are an attractive option for studying symbiosis. Unlike corals, the lack of a carbonate skeleton facilitates its handling and cultivation under laboratory conditions (Ohdera et al. 2018). The adult medusa depends on the energy transferred from its symbionts to cover up to 70% of its basal needs (Lampert 2016). About 3 weeks after the symbionts are acquired by asexual polyps, depending upon symbiont species, temperature, and food intake, metamorphosis of the newly symbiotic polyp into a medusa larva occurs (Colley and Trench 1983; Fitt and Costley 1998). In the medusa Cassiopea andromeda, glucose and glycerol are the only two free carbon molecules that acquire an isotopic label in short-term studies (under 90 s), suggesting that glycerol may fuel the synthesis of lipids, being rapidly metabolized (Hofmann and Kremer 1981). Further, previous studies have shown that adult medusae lose mass (size and weight) when symbiont density declines or in the absence of light, even when food is supplemented (Lampert 2016). Apparently, when medusae shrink under low light levels, some polyunsaturated fatty acids seem to be transferred from symbiont to host (Mortillaro et al. 2009). We hypothesized that if glucose and glycerol are important carbon molecules for this symbiosis, the expression of coding genes for their transport would differ in medusae with and without symbionts. This work aimed to evaluate the expression levels of genes coding for glucose transporter proteins GLUT3-like and GLUT8-like, and two genes for glycerol transport through GLP9-like, in the model organism C. xamachana, comparing the conditions of symbiosis and bleaching.

Materials and methods

Experimental design

C. xamachana medusae were donated by Xcaret Park, collected at 20° 34′ 24.59″ N; -87° 07′ 5.40″ W from NE Quintana Roo, Mexico in October 2021. They were acclimated in an open system pond (100 L seawater) under natural light and temperature conditions for over a year. Adult medusae with a similar size (6 cm in diameter) were individually transferred to 1 L beakers containing 500 mL of natural seawater (SW) with bubbling air and placed under a roof with indirect natural sunlight. Medusae were not fed during the experimental phase; the SW was replaced daily. Control (n = 3) and bleached (n = 4) medusae were maintained for 21 days at room temperature (28.3 ± 2.27 ºC; SAMMO 2021). Samples from a tentacle of each medusa were taken weekly, starting on day 0. The tentacle fragments were flash frozen in liquid N2 and stored at -80 °C until further processing. After 14 days, the addition of sugars to the bleached medusae was discontinued (see below) maintaining them for one more week and sampled. At the end of the treatments, medusae were returned to the open system pond. The adults of this medusa associate with Symbiodinium type A1 in Florida, Symbiodinium microadriaticum (Thornhill et al. 2006). We presume that the same symbiont occurs in the adult medusae we used, as they are from the same geographic area.

Artificial bleaching of medusae

The medusae were artificially bleached by the daily addition of a monosaccharide mix following Pogoreutz et al. (2017), but with a sugar mixture and final concentration of 0.3 mg L−1 according to Carabantes et al. (2022). Bleaching of the medusae was evaluated by weekly sampling of a tentacle fragment from each medusa extracted with a Dounce homogenizer. The symbionts were collected by centrifugation (13,000 RPM, 5 min), washed twice with milli-Q water, and resuspended in 500 µL of sterile seawater, adding Lugol (30%) to aid in symbiont counting. Symbiont density was quantified with a hemocytometer on an optical microscope (3 replicate counts per sample), and the data normalized to the wet weight (g). Further, we photographed a portion of a tentacle of each medusa with a fluorescence microscope (Axioskop 40 with aim 20X Tex Red Fs 15) at the beginning and at the end of the 21-day treatment.

RNA extraction

RNA was extracted from a tentacle of each medusa on the initial day, and on days 14 and 21, following Pawlowski et al. (1999) with some modifications. Briefly, frozen tentacle pieces were macerated in liquid N2 and added as powder to a preheated (95 °C) acidic phenol–buffer mixture and vortexed for two minutes, adding RNase Out. After a 25 min centrifugation at 4 °C, the upper phase was chloroform extracted twice and the RNA precipitated with 8 M LiCl. RNA samples (n = 21) were treated with DNase. RNA integrity was evaluated by electrophoresis on 1% agarose gels. RNA concentration and quality were assessed with a BioSpectrometer (Eppendorf). The extracted RNA (500 ng) was copied to cDNA using the ImProm-II TM Reverse Transcriptase (Promega) with 4 μl of buffer (250 mM Tris–HCl pH 8.3, 375 mM KCl, 50 mM DTT, 25 mM MgCl2), 1 μl dNTPs, 0.5 μl ribonuclease inhibitor, and 1 μl of the RT enzyme. cDNAs were synthesized in a BioRad T100 Thermal Cycler under the following conditions: 25 °C (5 min), 42 °C (1 h 3 min) and 70 °C (5 min). For the corroboration of cDNA synthesis, each gene fragment was RT-PCR amplified using the primers described in Table 1.

Table 1 Specific primer sequences for fragment amplification of genes coding for GLUT3-like, GLUT8-like, GLP9-like (genes 1 and 2), and FE1α
Full size table

Transcripts encoding GLUT and GLP transporter proteins in C. xamachana

Nucleotide sequences reported in the C. xamachana genome (https://cassiopeabase.org/resources/c-xamachana-genome/) indicated genes that potentially code for glucose and glycerol transporters. We identified ten transcripts with homologies for GLUT1, GLUT10, four transcripts for GLUT3 and GLUT8, and three transcripts from GLP9 in the genome. From these sequences, a BLAST alignment (https://blast.ncbi.nlm.nih. gov/) with sequences for GLUT and GLP transporters from other cnidarians, with identities > 50%, was identified. We selected one sequence for GLUT3 (> TRINITY_DN109051), one sequence for GLUT8 [Transcript (3204 bp)/CDS Sequence (1545 bp)], and two sequences for the GLP9 gene (Transcript (807 bp)/CDS Sequence (807 bp; Transcript (945 bp) /CDS Sequence (945 bp)). Conserved motifs were identified in these translated sequences after an alignment made in the Clustal Omega program (https:/ /www.ebi.ac.uk/Tools/msa/clustalo/). The presence of motifs that fulfill the function of sugar transport in GLUT (GR, GK, GRR, and PETK) (Joost and Thorens 2001), and glycerol transport in GLP (two NPA motifs and a D residue) (Agre et al. 2002) were corroborated (Supplementary Fig. S1).

The identity of the amplified fragments corresponding to the GLUT3, GLUT8, and two GLP9 genes was confirmed by endpoint PCR amplification using the primers designed (Table 1). Further, the amplification products were purified using the Wizard® SV Gel and PCR Clean-Up System and sequenced at the Institute of Cellular Physiology at UNAM. Through a BLAST search (https://blast.ncbi.nlm.nih.gov/), we corroborated the identity of all the genes (Supplementary Fig S2).

qRT-PCR assays

Quantitative RT-PCR assays were performed in the StepOnePlus thermocycler (Applied Biosystems). Samples of 20 ng of cDNA from tentacles of C. xamachana were used in each amplification reaction, and the expression of fragments corresponding to genes GLUT3, GLUT8, GLP9-1, and GLP9-2 were evaluated in control (n = 3) and bleached (n = 4) medusae at times 0, 14, and 21 days. A fragment of the gene encoding Elongation Factor 1 alpha (EF1α) from C. xamachana was used as reference (Nicot et al. 2005; Cabrales-Arellano et al. 2017). The amplification reactions contained 2X the reaction buffer with intercalating agent (QuantiTect® SYBR®-Green, BioRad), 5 μM forward primer, 5 μM reverse primer, water, and 20 ng of cDNA as template. The amplification conditions were: one cycle at 95 °C for 10 min, 40 cycles: 95 °C for 15 s, and 55 °C for 1 min, and 60 °C for 1 min; and for the melting curve 95 °C for 15 s, 60 °C for 1 min, and 95 °C for 15 s. To determine cycle quantification (CT) values for each gene fragment amplified, the dissociation curve analysis method was used, setting the fluorescence umbral to 0.05. CT values ​​for each gene were averaged and normalized to the FE1α reference gene. The 2-∆∆CT relative quantification method (Livak and Schmittgen 2001) was performed to determine the fold change in expression. The stability for the expression of the reference gene was calculated using free access programs (GenNorm, ∆-CT, and BestKeeper). Means and standard deviations were determined for each sample.

Statistical analysis

A T-Student test was used for comparing the density of symbionts and umbrella size for the sampling times on each experimental treatment. One-way Analysis of Variance was used to assess differences in the expression levels of the selected genes, between the conditions of symbiotic and bleached medusae, with an alpha of 0.05. Pairwise multiple comparisons were run with a Tukey test (Supplementary Material SM1).

Results

Cassiopea xamachana experimental bleaching

Symbiont density in the bleached treated medusae diminished significantly after 7 days of treatment, going from 1.53 × 107 cells g−1 at day 0 to 0.033 × 107 cells g−1 at day 14 (t-value 0vs7d = 4.73). After 14 and 21 days, symbiont density remained low, showing 84–99% fewer symbionts from the initial time; symbiont density was also significantly different (t-value 7vs14d = 7.90; t-value 14vs21d = 8.80). Control medusae conserved their symbionts after 21 days (Fig. 1a); the data collected at each time point were not different from the initial symbiont density (t-value 0vs7d = -1.95, t-value 7vs14d = -2.08, t-value 14vs21d = -2.01). We observed a decrease in umbrella size in bleached medusae from 6.5 cm on day 0 to 4.5 cm on day 21; however, control medusae also reduced their size from 6.5 cm at time 0 to 5.7 cm on day 21 (Fig. 1b). Paired t-test comparisons, however, indicated no significant differences in umbrella size between control and bleached medusae on any of the sampling days (t-value 0vs7d = 0.00, t-value 7vs14d = 4.68, t-value 14vs21d = 0.52) The phenotypic characteristics showed the absence of brownish color due to the diminishing of symbionts in bleached medusae, corroborated by fluorescence microscopy (Supplementary Fig. S3).

Fig. 1
figure 1

Cassiopea xamachana condition after experimental treatments. (a) Symbiont density (1 × 107 cells g.−1) and (b) umbrella size (diameter in cm) in control (filled triangles) and experimentally bleached (filled circles) medusae, during 21 days of treatment (mean values ​​ ± sem; n = 3 for control, n = 4 for bleached medusae)

Full size image

Expression levels of glucose and glycerol putative transporters in symbiotic and bleached medusae

The expression levels for the gene fragments GLUT3, GLUT8, GLP9-1, and GLP9-2 in bleached medusae and the controls, at day 0, were set at 1 ± 0, normalized to the reference gene (Supplementary Table S2). Results showed that the expression of GLUT transporters was not significantly different after 14 days of treatment in control and bleached medusae; however, control medusae showed a tendency to diminish their expression, while bleached medusae increased it (Fig. 2a). The comparative CT values (2-∆∆Ct) after 21 days indicated a decrease in the expression levels of GLUT3 (X ± SE = 0.519 ± 0.395, n = 4) and GLUT8 (X ± SE = 0.225 ± 0.210, n = 4), with significantly different values compared to the symbiotic control (ANOVA: F(1,17) = 5.263, p = 0.009, for GLUT3, and F(1,17) = 5.382, p = 0.008, for GLUT8) (Supplementary Fig S3, Supplementary Material SM1).

Fig. 2
figure 2

Gene expression levels for glucose and glycerol transport genes in symbiotic and bleached medusae. Gene expression levels for (a) glucose transport genes GLUT3 (dark blue) and GLUT8 (clear blue) and (b) glycerol transport genes GLP9-1 (dark peach) and GLP9-2 (clear peach) measured in symbiotic (Control, n = 3) and bleached medusae (n = 4), sampled at day 0 and on days 14 and 21 of treatment. Sugar addition in bleached medusae was discontinued after 14 days. *Denotes significant differences (p < 0.001) in expression levels for each gene from its control

Full size image

Gene expression values​​ for the GLP9-1 gene were significantly lower in the controls at day 14 (ANOVA: F(1,17) = 142.570, p < 0.001) (Fig. 2b) with a lower symbiont density in bleached medusae (0.056 × 107 cells g−1) (Fig. 1a); however, both GLP9 genes changed their expression in bleached medusae (X ± SE = 1.332 ± 0.0816, n = 4 for GLP9-1, and X ± SE = 1.517 ± 0.668, n = 4 for GLP9-2) (Fig. 2b, Supplementary Tables S1 and S2). The sugar treatment was discontinued after 14 days, maintaining the medusae for seven further days in a bleached condition without food. At this time (21 days of treatment), symbiont density remained low with 0.004 × 107 cells g−1 in contrast to 1.269 × 107 cells g−1 in the control (Fig. 1a). An increase in the expression levels of GLP9-1 (X ± SE = 3.045 ± 0.287, n = 4) and GLP9-2 (X ± SE = 2.718 ± 0.654, n = 4) was observed, with expression levels significantly different from the symbiotic control on day 21 (ANOVA: F(1,17) = 15.170, p < 0.001) (Fig. 2b, Supplementary Tables S1 and S2, Supplementary Material SM1).

Discussion

Our knowledge of the symbiosis between algae and cnidarians is still incomplete. We have a limited understanding of the cellular and molecular mechanisms for the biosynthesis and translocation of carbon from symbiont to host. In the present study, we measured the expression of selected genes that code for glucose and glycerol transporters in Cassiopea xamachana, identified as main carbon compounds potentially translocated to the medusa host (Hoffman and Kremer 1981). The addition of sugars successfully diminished the symbiont density in the experimental bleaching, as was demonstrated by Carabantes et al. (2022). GLUT transporters only showed a significant variation in their expression after 21 days of treatment in both conditions, although control medusae increased them, while bleached medusae decreased them. These results suggest that control medusae, being at low light, were probably increasing their chance of acquiring any organic carbon that the symbionts might transfer. Also, in Exaiptasia diaphana, it was documented that anemones with symbionts showed significantly higher levels of GLUT8 than aposymbiotic ones (Mashini et al. 2022). According to these same authors, the GLUT8 transporter localized to the symbiosome membrane complex of E. diaphana, although not exclusively. This observation, along with our results, suggests that this GLUT8 protein could be involved in the transport of symbiont-derived glucose in the medusa as well.

In symbiotic medusae, the expression of glycerol transporters diminished with time, perhaps from being under stressful conditions in the vessels (e.g., low circulation and reduced water volume). However, it is also possible that the aquaglyceroporins we assayed are not responsible for transporting glycerol from symbiont to host. In bleached medusae, the diminishing density of symbionts affected the levels of mRNA transcripts for the selected glycerol genes. After 14 days, bleached medusae significantly increased the number of transcripts for the two aquaglyceroporins compared to the control. Previous studies have reported an increase in the levels of glycerol of approximately threefold in bleached anemones (Molina et al. 2017), suggesting that host animals may degrade lipids to compensate for the loss of symbionts, as has also been documented to occur in the face of nutrient deficiency (Fitt and Pardy 1981). Moreover, after discontinuing the addition of sugars, these same transporters showed further increases in expression at 21 days, consistent with our interpretation. The catabolism of carbohydrate and lipid reserves, when compared between symbiotic and aposymbiotic states, increased in E. diaphana anemones in symbiosis with a heterologous symbiont, Symbiodinium trenchii, that conferred metabolic and transcriptomic profiles in between the symbiotic and aposymbiotic states (Matthews et al. 2017). It seems that with less appropriate symbionts, or in their absence, the animal host may utilize reserves to compensate for the loss of symbiont-derived organic carbon; this has also been observed in medusae, even when food is supplemented (Lampert 2016).

In our results, the high expression levels of the two genes encoding glycerol transporters suggest that bleached medusae were degrading lipids and transporting glycerol through a GLP-type protein to compensate for the absence of symbionts, given the lack of glucose on which they depend as an essential metabolite. Since glycerol is produced by lipid degradation and is used for gluconeogenesis, its transport within the cnidarian host via a GLP9 transporter could also be modulating fat metabolism (Hibuse et al. 2006; Madeira et al. 2015). In the study by Molina et al. (2017), the reduction of the glucose reserve in the anemone Exaiptasia pallida (E. diaphana) due to a diminished symbiont density led to an increase in glycerol-3 phosphate (G3P), as well as a decrease in the specific activity of glycerol-3 phosphate dehydrogenase, suggesting that no lipids were being synthesized. These results were consistent with a steady specific activity of the enzyme Glucose-6-phosphate dehydrogenase (G6PDH), which participates in the pentose pathway where NADPH is regenerated for anabolic metabolism and in fatty acid biosynthesis (Park et al. 2015). Further, in the coral Acropora aspera after 13C labeling experiments, fatty acids and lipogenesis intermediates acquired the label in symbiotic specimens (Hillyer et al. 2017); however, after 9 days under high-temperature conditions, the breakdown of energy stores was detected, diminishing such compounds. It would be worthwhile coupling studies of the expression of transporters with metabolic profiles in this medusa, to achieve a better understanding of the dynamics involved.

Our results suggest that lipolysis might be involved in the cellular metabolism of energy reserves in bleached medusae. In this process, triglycerides (TAG) and diglycerides (DAG), which are chief energy reserves in symbiotic cnidarians (Imbs et al. 2010; Garret et al. 2013; Hillyer et al. 2017; Imbs et al. 2021), are hydrolyzed into fatty acids and glycerol by the action of enzymes known as lipases, found in the adipose tissue of animals (Li et al. 2012). The acyl-CoA produced in the fatty acid degrading process is transported into the mitochondria as acyl-carnitine, where β-oxidation occurs. In this oxidative process, the resulting acetyl-CoA is a substrate for the citric acid cycle that will later generate energy in the form of ATP. Our results regarding the increased expression in GLP9 transporters suggest that bleached medusae could be shifting their metabolism toward the catabolism of lipid stores, explaining the higher expression of glycerol transporters.

Other studies suggest that the loss of symbionts can alter the nutrient cycle in cnidarian symbioses: the cessation in the transfer of photosynthates “gradually reduces the translocation of carbon by the algal symbionts”, leading the symbiotic system to a carbon-limited state (Rädecker et al. 2021), but this would depend on lipid stores as suggested by our results. Further, glucose is an important energy molecule for this medusa, evidenced by the decrease in GLUT3 and GLUT8 transcript levels after a significant reduction of symbionts. By contrast, E. diaphana anemones were found to have the same intensity in the immunodetection of GLUT8 in symbiotic and aposymbiotic organisms (Mashini et al. 2020), suggesting a tighter relationship of the medusa with its symbionts. Interestingly, increased levels of GLUT8 transporter were found in symbiotic anemones harboring heterologous symbionts like Symbiodinium microadriaticum, with higher photosynthetic rates, suggesting an increased effort by the host to acquire photosynthetic products (Mashini et al. 2020). Alternatively, the synthesis of glucose from the glycerol produced by lipid degradation through gluconeogenesis would be unfavorable, since it is an energy-consuming pathway (Molina et al. 2017). Finally, the present work provided experimental evidence for glucose transport in C. xamachana jellyfish involving proteins GLUT3 and GLUT8, as previously proposed by phylogenetic analysis (Sproles et al. 2018). However, evidence for their location in inner membranes and other transporter proteins, such as SGLTs, also needs experimental support for their role in carbon exchange in the cnidarian–dinoflagellate symbiosis. Finally, even though the bleaching of medusae appears to have been homogeneous, we worked with tentacle fragments which could have introduced variations in the quantification of transcripts unrelated to gene expression. Therefore, our results should be taken with caution.

In conclusion, the loss of symbiont cells in C. xamachana decreased the levels of transcripts for the glucose transporters. Our results showed higher expression of glucose transporters GLUT3 and GLUT8 in control compared to bleached medusae, in agreement with glucose being a mobile molecule. There is no significant glucose transport in the absence of symbionts, as the decreased expression of GLUT3 and GLUT8 in bleached medusae suggests. Previous studies suggested the function of GLUT8 as a hexose transporter in intracellular membranes and in the symbiosomal membrane; our results consistently showed transcript levels for GLUT8 decreasing significantly on day 21 in bleached medusae. The results also suggest that symbiont-derived glucose was probably replaced in bleached medusae by lipid catabolism, which increased GLP9 expression, associated with glycerol transport probably within the host tissues. We also observed that sugar addition alone was insufficient to cover the host's energetic needs after 14 days of treatment, leading to a steady loss of umbrella size in bleached medusae. The greatest size reduction was for the bleached medusae at day 21, coincident with the decrease in the expression of glucose transporters GLUT3 and GLUT8.