ABSTRACT
The insect gut, which plays a role in ion and water balance, has been shown to leak solutes in the cold. Cold stress can also activate insect immune systems, but it is unknown whether the leak of the gut microbiome is a possible immune trigger in the cold. We developed a novel feeding protocol to load the gut of locusts (Locusta migratoria) with fluorescent bacteria before exposing them to −2°C for up to 48 h. No bacteria were recovered from the hemolymph of cold-exposed locusts, regardless of exposure duration. To examine this further, we used an ex vivo gut sac preparation to re-test cold-induced fluorescent FITC-dextran leak across the gut and found no increased rate of leak. These results question not only the validity of FITC-dextran as a marker of paracellular barrier permeability in the gut, but also to what extent the insect gut becomes leaky in the cold.
INTRODUCTION
The majority of insects are chill-susceptible, meaning they suffer negative effects of chilling (chilling injuries) at low temperatures well above the freezing point of their body fluids (Overgaard and MacMillan, 2017). As temperatures drop below an insect's critical thermal minimum (CTmin), they lose coordinated motor control. Continued cold exposure eventually leads to the onset of chill coma, a state characterized by a complete but reversible paralysis (Andersen et al., 2015; Hazell and Bale, 2011; MacMillan and Sinclair, 2011a,b; Rodgers et al., 2010). Prolonged exposure to low temperatures leads to tissue damage, and these chilling injuries can accumulate and increase in severity if temperatures remain low. Chilling injuries are thought to be largely driven by cell death resulting from a loss of ion homeostasis (Andersen et al., 2017a; Bayley et al., 2018; Carrington et al., 2020; Koštál et al., 2004; MacMillan and Sinclair, 2011b; Overgaard et al., 2021). Water and ion balance play a key role in maintaining neuromuscular function, but at low temperatures, active transport rates of solutes are slowed to a point at which they cannot counterbalance the passive leak of solutes and water (Overgaard et al., 2021).
The insect gut is structurally divided into three regions – foregut, midgut and hindgut – and plays a major role in maintaining this osmotic and ionic balance under benign conditions (Overgaard et al., 2021). Mechanical breakdown of food occurs in the foregut, the bulk of digestion and nutrient absorption happens in the midgut, and any water remaining is reabsorbed in the hindgut before the digested bolus passes into the rectum and is excreted (Linser and Dinglasan, 2014; Phillips et al., 1987). Additionally, the hindgut and specialized diverticulae known as Malpighian tubules (somewhat analogous to human kidneys) work together to maintain renal function (MacMillan and Sinclair, 2011b; MacMillan et al., 2017; Overgaard et al., 2021; Yerushalmi et al., 2018). Water and ions can move across renal epithelia in two primary ways: transcellularly through aquaporins, ion transporters or channels, or paracellularly through structures called septate junctions (Izumi and Furuse, 2014; Jonusaite et al., 2016, 2017a,b). These junctions are ladder-like protein complexes located between gut epithelial cells that regulate the passive movement of solutes and water (MacMillan et al., 2017; O'Donnell, 2008; Phillips et al., 1987).
Under optimal environmental conditions, transport and leak rates are balanced such that hemolymph water volume and [Na+] remain high, whereas [K+] remains low (D'Silva et al., 2017; Harvey et al., 1983; MacMillan and Sinclair, 2011b; MacMillan et al., 2015b; Overgaard and MacMillan, 2017). At low temperatures, ion and water homeostasis become disrupted when a net leak of Na+ and water into the gut lumen occurs, reducing hemolymph volume (MacMillan and Sinclair, 2011b). Recent evidence has suggested that gut epithelial barriers of both Drosophila melanogaster and Locusta migratoria become disrupted in the cold, and this failure of barrier function has been hypothesized to contribute to ion balance disruption in the cold by allowing water and/or solutes to leak down their electrochemical gradients (Andersen et al., 2017b; Brzezinski and MacMillan, 2020; MacMillan et al., 2017). In both locusts and D. melanogaster, this leak was observed in vivo using the fluorescently labelled dextran (FITC-dextran, 3–5 kD), a molecule used frequently in epithelial barrier research because it is too large to move transcellularly and cannot be metabolized by animals (Andersen et al., 2017b; Brzezinski and MacMillan, 2020; Jensen-Jarolim et al., 1998; MacMillan et al., 2017; Woting and Blaut, 2018). Brzezinski and MacMillan (2020) found that dextran leak in the cold occurs unidirectionally from the gut to the hemocoel when fed to locusts, but not in the opposite direction when injected into the hemocoel, implying that gut contents may be particularly likely to leak into the hemocoel of insects during cold stress.
In addition to regulating the flow of water and ions, the gut also houses an abundant microbial community that is primarily composed of bacteria and yeasts (Dillon and Dillon, 2004; Padilla, 2016; Wong et al., 2011). Recent studies have suggested that gut bacteria and yeasts may affect an insect's survival at low temperatures. For example, D. melanogaster with a healthy gut flora exhibit significantly increased cold tolerance (Henry and Colinet, 2018; Moghadam et al., 2018; Padilla, 2016). However, immune activation, typically associated with bacterial pathogens, has also been reported in adult and larval D. melanogaster following cold stress. Specifically, the cold stress response is characterized by increased expression of genes in the Toll, immune deficiency (Imd) and/or Janus kinase (JAK)-signal transducer and activator of transcription (STAT) pathways (Salehipour-shirazi et al., 2017; Sinclair et al., 2013; Štětina et al., 2019).
These repeated reports of immune activation in the cold could putatively be explained by the presence of bacteria in the hemolymph that originate from the ‘leaky’ gut, but no studies so far have directly tested whether gut bacteria leak into the hemocoel of insects after a cold exposure. Given that previous studies suggest that small molecules leak across the gut and extensive cell death occurs in the insect gut during a cold stress, bacteria may also leak, through either disrupted cell barriers or lesions in the gut wall. Here, we used in vivo and ex vivo experiments with migratory locusts to test the hypothesis that immune activation, previously observed in cold-stressed insects, is a direct response to cold-induced bacterial leak from the gut.
MATERIALS AND METHODS
Locust rearing
Locusts [Locusta migratoria (Linnaeus 1758)] used in the experiments were derived from a colony maintained at Carleton University, Ottawa, ON, Canada. The colony was reared on a 16 h:8 h light:dark cycle at 28°C at 60% relative humidity, under crowded conditions. All locusts were provided with a dry mixture of oats, wheat germ, wheat bran and dry milk powder, as well as fresh wheat clippings, 3 days a week ad libitum. All locusts used in experiments were sexed and equal numbers of males and females were used in all experiments.
Chill coma recovery time and chilling injury scores
Cold tolerance of locusts was quantified using chill coma recovery time (CCRT) and degree of chilling injuries (injury score). The methodology used here was slightly modified from that of Brzezinski and MacMillan (2020). Locusts from the colony were collected on the day of the experiment and placed individually in 50 ml polypropylene Falcon tubes (n=10 for each group). Holes were made in the lids of the tubes, which provided locusts access to air for the duration of the experiment. Control locusts were placed in an incubator (Isotemp BOD Refrigerated Incubator 3720A, Thermo Fisher Scientific, ON, Canada) with dry oat mixture and fresh wheat clippings for 48 h at 25°C. Locusts undergoing cold stresses were suspended in a circulating cooling bath (model AP28R-30, VWR International, Radnor, PA, USA) using a Styrofoam rig. The bath was filled with 100% ethylene glycol, pre-set to 25°C, and cooled to −2°C at a rate of −0.25°C min−1. Locusts were then left at −2°C for 12, 24, 36 or 48 h. Temperature was monitored and confirmed throughout the duration of the cold exposure using type-K thermocouples (TC-08 Data Logger, Pico Technology, Tyler, TX, USA) in the glycol and in the hemocoel of an additional locust not used in the experiments. After each time point, locusts, in their comatose state, were removed from their tubes and placed on their side on a clean bench at room temperature (24.0–25.0°C). To measure CCRT, each locust was closely monitored for the time taken to regain neuromuscular function and stand on all six legs, which was recorded as that locust's CCRT. A 90 min cut-off point was used, after which any locust that had failed to stand on all legs was marked as ‘unrecovered’.
After 90 min, locusts (regardless of their state) were placed in clean 50 ml polypropylene tubes with dry oat mixture and fresh wheat clippings and left to recover in the incubator at 25°C. After 24 h at 25°C, an injury assessment was done by a single assessor (M.I.E.-S.) using a five-point scale adapted from MacMillan et al. (2014). Scores were defined as follows: (0) no movement observed (dead); (1) limb movement (leg and/or head twitching); (2) greater limb movement (leg extension and retraction, and whole body twitching), but unable to stand; (3) able to stand, but unable or unwilling to walk or jump; (4) able to stand, walk and/or jump, but lacks coordination; and (5) movement restored similar to pre-exposure levels of coordination. After scoring injury, locusts were returned to their respective tubes with replenished oats and wheat clippings and placed in the incubator. The injury assessment was repeated three more times for each locust – 48, 72 and 96 h post-cold exposure.
Development of fluorescent bacteria feeding protocol
To be confident that any bacteria present in the hemolymph originated from the gut lumen, we developed a novel feeding and bacterial leak assay. Ingestion of a fluorescent-tagged bacterial species was previously validated in moth larvae (Spodoptera littoralis) by Teh et al. (2016). Our protocol used a mutant fluorescent strain of Escherichia coli, GFPmut3 (λmax excitation: 500 nm, λmax emission: 513 nm), with a green fluorescent gene on a plasmid alongside an ampicillin-resistant gene (Chalova et al., 2008) and preliminary trials were conducted to optimize the assay. We point out that owing to overlapping fluorescence spectra between GFPmut3 and FITC-dextran, we opted to test for in vivo leak of bacteria and dextran in separate experiments with separate locusts (refer to Supplementary Materials and Methods). In the final protocol, an overnight culture of GFPmut3 E. coli was grown in Luria–Bertani (LB) broth containing ampicillin (100 μg ampicillin 1 ml−1 medium) in an incubator at 37°C (model MIR-154, PHC Corporation, Wood Dale, IL, USA). Then, 125 ml of medium was centrifuged at 8000 rpm (9730 g) for 15 min (Sorvall RC 6 Plus, Thermo Fisher Scientific, Waltham, MA, USA). The supernatant was discarded, and the bacterial pellet was resuspended in 1.25 ml of distilled water in a 2 ml centrifuge tube, effectively concentrating the bacterial solution 100-fold. Ten 3 cm strands of freshly cut wheat were added to the tube and allowed to soak for 24 h in an incubator at 37°C. Simultaneously, locusts that were to be used in the experiments were moved to a separate cage for 24 h with no wheat or oats to fast (which ensured they would eat the soaked wheat when presented with it). Each locust was then placed in separate plastic containers (36.3×25.1×59.9 cm) with holes in the lid, along with their own bacteria-soaked wheat, where they fed for 24 h in an incubator at 25°C. Further details on the validation of this protocol are included in the Supplementary Materials and Methods, with results of validation experiments provided in Figs S1–S3.
Investigating cold-induced bacterial leak from the gut
To test whether bacteria in the gut leak into the hemolymph during cold stress, locusts were suspended in a cooling bath at −2°C for 12, 24, 36 or 48 h (n=6, n=10 for 48 h group) following 24 h of feeding on the wheat soaked in the fluorescent bacteria solution. Control locusts were kept in the incubator at 25°C for 48 h with wheat and oats provided ad libitum. To collect hemolymph, locusts were pricked dorsally at the head–thorax junction. Hemolymph was then collected using techniques adapted from Findsen et al. (2013). A 50 μl capillary tube was used to collect hemolymph via capillary action at the site of injury. By inserting a pipette tip at the end of the capillary tube, 10 μl of hemolymph was drawn and pipetted into 190 μl of sterile locust saline (in mmol l−1: 140 NaCl, 8 KCl, 2.3 CaCl2 dihydrate, 0.93 MgCl2 hexahydrate, 1 NaH2PO4, 90 sucrose, 5 glucose, 5 trehalose, 1 proline and 10 HEPES, pH 7.2) in a centrifuge tube, and this process was completed twice to generate two hemolymph samples from each animal. After briefly vortex mixing, one of the 200 μl solutions was spread on a Petri dish containing LB broth with ampicillin, and the other solution was spread on an LB agar plate without ampicillin. Both plates were then incubated at 37°C. The plates were checked for colony growth every day for 4 days. Colony-forming units (CFU µl−1) in extracted hemolymph samples were then determined using serial dilution plating on LB agar plates containing ampicillin.
Bacterial leak could plausibly occur following, rather than during, a cold stress, so we performed a follow-up experiment to test for bacterial leak following a 6 h rewarming period after the cold stress. Cold exposures were performed in an identical manner as described above following bacterial feeding. In this case, however, locusts (n=6 in each group) were placed in small plastic deli containers after the cold stress, with freshly cut wheat (not soaked in the bacterial solution) and dry oat mixture in excess. The containers with the locusts were then placed in the incubator and left to recover at 25°C for 6 h.
To ensure that fluorescent bacteria that are present in the hemolymph of cold-stressed locusts could be recovered using our extraction method, we included a positive control. Locusts (n=4) were suspended in a cooling bath and were left undisturbed while the bath ramped down to −2°C. Individuals were then removed and were injected dorsally at the head–thorax junction with 10 µl of a 1.46×108 CFU ml−1 solution of GFPmut3 E. coli in sterile locust saline. The injection site was then sealed with high vacuum grease (Dow Corning, Etobicoke, ON, Canada) before returning the locusts back to the cooling bath for 1 h at −2°C. Hemolymph samples were then collected as above.
Determining the role of gut bacteria in cold-induced paracellular barrier disruption
The leak of FITC-dextran in L. migratoria from the gut to the hemocoel in vivo was first validated using methods adapted from Brzezinski and MacMillan (2020). For 24 h in the incubator (set to 25°C), locusts (n=6–7 per group) were left to feed on a dry oat mixture (oats, wheat bran, wheat germ and skim milk powder) saturated with a 1 mmol l−1 solution of FITC-dextran in water. Locusts were then either exposed to −2°C or left in the incubator, with fresh food, for 48 h (see above for cold stress protocol). Upon removal from the cold or incubator, a 2 μl aliquot of hemolymph was collected and diluted 50-fold in a 96-well black/clear-bottomed plate (Corning Falcon Imaging Microplate, black/clear bottom) for imaging via fluorescence spectrophotometry (λmax excitation: 485 nm, λmax emission: 528 nm; BioTek Cytation 5 Imaging Reader, Winooski, VT, USA). FITC-dextran concentrations (in μg ml−1) were determined by reference to a standard curve.
After confirming in vivo leak of FITC-dextran into the hemolymph (see Results), we proceeded to better investigate cold-induced paracellular leak across the gut ex vivo and in the absence of the natural gut microbiome. We tested for leak of FITC-dextran (FD4; 3–5 kDa, Sigma-Aldrich, St Louis, MO, USA) from isolated gut segments of L. migratoria using a modified gut sac approach (Gerber and Overgaard, 2018; Hanrahan et al., 1984). We modified the preparation by (1) not everting the gut segments, and (2) inserting two separate pieces of polyethylene tubing on the anterior and posterior side of each segment. On the day of the experiment, locusts were decapitated and prepared for dissection by removing all appendages. The thorax and abdomen were placed in a Sylgaard-lined Petri dish and locust saline was used to keep tissues moist during dissections. An incision was made in the anterior to posterior direction along the ventral side to pin open the body cavity. All structures aside from the gut tract were cleared away before the segments were isolated. Portions of the gut were isolated in a manner convenient to the method being applied rather than by anatomical definition. Briefly, the segments were described as follows: anterior, from the anterior-most portion of the esophagus to the midgut caecae; central, from posterior of the midgut caecae to the midgut–hindgut junction, where the Malpighian tubules connect; and posterior, from the anterior-most portion of the ileum to the posterior end of the rectum.
To suspend each isolated gut segment within our system, a heat flared polyethylene (PE) tube (VWR ID×OD: 0.023×0.038 inches) was inserted and tied into the anterior margin of the section. Once secure, standard locust saline was injected through the PE tube to thoroughly rinse out the gut contents, including the vast majority of the gut microbiome. A second heat-flared PE tube was then inserted and tied into the posterior margin of the segment. Preparations were kept in a Petri dish containing continuously oxygenated (95% O2, 5% CO2, Praxair, Danbury, CT, USA) saline at room temperature (23°C) until all three segments had been prepared for suspension. Once complete, a 1 mmol l−1 FITC-dextran solution was injected via one of the PE tubes into each preparation until it had filled both PE tubes, ensuring the lumen was filled with the saline containing the FITC-dextran. Each preparation was suspended in a beaker containing 25 ml of continuously oxygenated locust saline, which acted as our serosal environment. After a 30 min rest period at room temperature to allow for tissue stabilization, the beaker was moved into the cooling bath (preset at 0°C) and monitored for 5 h (Gerber and Overgaard, 2018). Preparations were then removed from the cooling bath and monitored for an additional 2 h at room temperature to account for any effect of rewarming on the rate of leak.
Throughout the experiment, 90 μl aliquots were collected directly from the beakers once every hour for the duration of the experiment and transferred to a 96-well black/clear bottom plate for fluorescence spectrophotometry. Concentrations of FITC-dextran in the samples were determined by reference to a standard curve of FITC-dextran in locust saline. The results obtained from fluorescent analyses were then plotted to obtain leak rates per cm2 of gut tissue for all preparations. Briefly, the slope [FITC-dextran concentrations (μmol) against time (h)] of each gut sac sample and measurements of tissue length and width (treated as cylindrical surface area) were used to calculate leak rates per cm2 of gut tissue.
Data analysis
All collected data were analyzed in RStudio version 3.5.3 (https://posit.co/products/open-source/rstudio/). The distribution and variance of residuals were assessed using Shapiro–Wilk tests and Q–Q plots, which supported the use of non-parametric tests. The effect of cold duration on CCRT was analyzed with a Kruskal–Wallis (KW) test followed by pairwise Wilcoxon tests with the Benjamini–Hochberg (BH) correction. Because the assumption of normality was not met, a generalized linear model (GLM) with a Poisson distribution was used to test for the effect of cold duration and assessment day on injury scores. A KW test followed by pairwise Wilcoxon tests (with BH correction) were used to test for significant differences in injury scores on the first and fourth assessment days following the cold exposures. Again, because of the non-normal distribution of the data, a GLM with a Poisson distribution was used to examine the effect of bacterial feeding on chilling injuries. Cold exposure duration was held as a fixed categorical variable, whereas assessment day was held as a continuous variable (in their respective analyses). For in vivo measurements of FITC-dextran leak, a Wilcoxon rank sum test was used to test for significant differences in FITC-dextran concentrations in the hemolymph. For the ex vivo gut sac experiment, FITC-dextran leak per cm2 of tissue in the cold and post-cold was analyzed using the lmer() function (lme4 and lmerTest packages for R). Time was held as a continuous factor, gut segment as a fixed effect and each individual locust as a random effect. Finally, paired t-tests were performed to compare the rates of FITC-dextran leak per cm2 of gut tissue in the cold and post-cold. Log10-transformed ex vivo gut data were used for statistical analyses. Values presented on all graphs are shown as means±s.d. with the α-level being 0.05 for all statistical tests. For the gut bacterial leak assay, no statistics were used to analyze growth on plates owing to lack of any colonies observed from the cold-stressed locusts. All data are available in Dataset 1.
RESULTS AND DISCUSSION
Chill coma recovery time and survival score following cold exposures
CCRTs significantly increased from 19.3±7.07 min after 12 h at −2°C to 80.4±2.95 min after 48 h at −2°C (for those locusts that recovered within the 90 min cut-off period; KW, χ2=17.67, P<0.001). Longer cold exposures also resulted in fewer locusts recovering before the 90 min cut-off. Exposure to −2°C for 12 or 24 h yielded a 100% recovery rate, which decreased to 75% at 36 h and 37.5% at 48 h (Fig. 1A).
Chilling injuries were quantified using injury scores that clearly demonstrated how longer cold exposures led to higher degrees of injury (lower injury scores; GLM, F4,140=44.62, P<0.001) (Fig. 1B,C). Over the 4 days following the cold stresses, chilling injuries worsened (GLM, F3,140=4.30, P=0.006; Fig. 1D). This effect, however, was only significant when we included the 48 h cold exposure group in the statistical model, suggesting that particularly severe injuries get progressively worse after the cold stress (compare Fig. 1B and C).
As previously found in the same species (e.g. Andersen et al., 2017a; Brzezinski and MacMillan, 2020; MacMillan et al., 2014), increasing durations of exposure to low temperature (–2°C) resulted in higher degrees of injury (Fig. 1). Latent chilling injuries manifested in locusts exposed to −2°C for 48 h, quantified by decreasing injury scores (more severe chilling injuries) over a 3-day period after the first survival assessment. A similar pattern was found in D. melanogaster exposed to a relatively long cold stress (25 h) at 0°C, where lower injury scores 1 day after the cold stress led to significantly more deaths the following 3 days after removal from the cold (El-Saadi et al., 2020). These results from D. melanogaster and L. migratoria suggest that chilling injuries do not fully heal and may even continue to worsen in the days following removal of the insect from a severe cold stress (see fig. 2A in El-Saadi et al., 2020).
No bacterial leak across the gut following cold exposure
If the continuous decline in survival is a result of bacterial infection or an adverse immune response (Sadd and Siva-Jothy, 2006), then this could be explained by bacteria leaking from the gut lumen into the hemocoel during or following a cold stress. To test this, we fed locusts a fluorescent strain of E. coli before exposing them to the cold. Immediately following a cold exposure, hemolymph samples were taken from locusts and spread on LB agar plates with or without ampicillin to look for GFPmut3 colony growth exclusively or any colony growth, respectively. Although 48 h of exposure to −2°C led to significantly higher dextran leak in vivo (Fig. 2A), no bacterial colonies were seen on any plates (LB agar plates with or without ampicillin) containing hemolymph from the locusts, regardless of cold exposure duration (Table 1). This finding was also confirmed in a follow-up experiment to examine bacterial leak following a period of rewarming, in which locusts were left to recover with food at benign temperature for 6 h following the cold exposure. Similar to the acute experiments, no bacterial colonies were observed on any plates regardless of cold exposure duration (Table 1).
In another orthopteran species, the spring field cricket (Gryllus veletis), the ability of the immune system to clear bacteria from the hemolymph is significantly reduced at low temperatures (Ferguson et al., 2016). This is also seen in some other orders of insects (Ferguson and Sinclair, 2017). As hemolymph was plated immediately after the cold stresses in one of our experiments, it is unlikely that the absence of bacteria was a result of immune-related bacterial clearance. In this case, there are two probable scenarios: (1) gut barriers maintain their integrity well enough to prevent septicemia, despite damage to the gut epithelia and leak of small solutes, or (2) gut epithelia retain their barrier properties in the cold. From our data (see below) and those of Gerber and Overgaard (2018), the degree of FITC-dextran leak in the cold is insufficient to support a purported increase in gut barrier permeability that would have allowed bacteria to cross over in our experiments. It is possible, however, that the size of E. coli (∼1×3 μmol l−1; Reshes et al., 2008) may prevent it from crossing the septate junctions, even when epithelial cells in the gut might be damaged by the cold, because intercellular spaces are only approximately 20 nm wide when healthy (Jonusaite et al., 2016) and may exclusively allow smaller molecules such as FITC-dextran to pass. Hence, the second explanation is more likely.
One possible explanation for cold-induced immune activation involves the leak of immunogenic components of bacteria such as lipopolysaccharide (LPS) or peptidoglycan (Charles and Killian, 2015), which are far smaller than whole bacteria. Although purified LPS has been shown to not activate the Imd pathway in Drosophila (Kaneko et al., 2004), it does induce overexpression of antimicrobial peptide genes in silkworms (Tanaka et al., 2009). Peptidoglycan, a major constituent of the bacterial cell wall, signals immunity through the Toll or Imd pathway (Leulier et al., 2003). Although whole bacteria do not leak across the gut epithelia in the cold, these small molecular components of the bacterial membrane or wall (or other small metabolites, ions or toxins) may still leak between cells in the cold and contribute to downstream physiological impairments.
Another possibility is that cellular damage or death in the midgut, for example (MacMillan et al., 2017), can lead to the release of actin into the hemolymph of insects (Dominguez and Holmes, 2011). Srinivasan et al. (2016) clearly show that actin in the hemolymph of D. melanogaster elicits an immune response via JAK/STAT pathway activation. However, no studies to date have clearly measured actin release in vivo following cold stress, and so this possibility of sterile immune activation remains an unexplored avenue.
FITC-dextran leak across epithelia of isolated gut segments
After confirming in vivo FITC-dextran leak in cold-stressed locusts (Wilcoxon rank sum test; P=0.002; Fig. 2A), we used a modified gut sac technique to examine ex vivo leak of FITC-dextran in the absence of the vast majority of the natural gut microbiota. At 0°C, no significant differences between FITC-dextran leak rates were found between the segments (LME, F2,15=1.03, P=0.379; Fig. 2B). The same was true for segments held at room temperature over the course of experiments (LME, F2,10=0.989, P=0.406). When comparing between both temperature treatments (–2°C or 23°C), FITC-dextran leak rates did not significantly differ between segments (LME, F2,30=0.882, P=0.424) or between treatments (LME, F1,30=1.04, P=0.315). Furthermore, there was no statistically significant interaction between the gut segment and type of treatment received when analyzing FITC-dextran leak rates (LME, F2,30=1.10, P=0.346; Fig. 2B). Finally, no significant differences were found between leak rates after 5 h of cold stress and after 2 h at 23°C following cold stress in any of the three segments (foregut: two-tailed t5=0.668, P=0.534; midgut: two-tailed t5=1.19, P=0.288; hindgut: two-tailed t5=1.27, P=0.261).
When we exposed isolated gut segments to the cold ex vivo, we found no significant increase in the rate of FITC-dextran leak from the lumen to the surrounding solution compared with isolated gut sacs at benign temperature (Fig. 2B). This agrees with a previous study on the same species using everted rectal sacs ex vivo, where no change was found in the mucosal-to-serosal clearance of FITC-dextran in the cold (Gerber and Overgaard, 2018). These data now lead us to question the validity of FITC-dextran as a marker of paracellular permeability, as a common factor in the studies that reported FITC-dextran leak in the mucosal to serosal direction at low temperatures was that the FITC-dextran was administered to the insects orally. The ingestion of FITC-dextran means that it would be concentrated in the gut initially, where there also exists an abundant microbial community. Some bacterial species produce dextranases, which cleave larger dextran molecules into smaller fragments (Khalikova et al., 2005). If these bacterial species are present in the gut of healthy locusts, then it would explain the leak of FITC-dextran at benign temperature, which would most likely arise from the fragmentation of the large FITC-dextran molecule into smaller polysaccharides.
The findings from the present study may shift our understanding of barrier failure in cold-stressed insects and raise questions as to what may trigger an insect's immune system in the cold. Considering that we did not find any evidence of bacterial leak from the gut, we propose that immune activation in the cold may arise from sterile causes, possibly as a result of cell damage from chilling injuries or relatively small immunogenic molecules from microbes, such as LPS, that can more readily cross paracellular barriers. From this standpoint, we suggest that future studies should attempt to examine cold-induced immune activation by looking at damage- or pathogen-associated molecular patterns such as actin and LPS in the hemocoel following cold stresses.
Acknowledgements
We thank Jeffery Dawson and Marshall Ritchie for assistance with locust rearing and colony maintenance, Amanda Carroll for help with procuring bacterial stocks and early troubleshooting with microbiological techniques, and Laura Ferguson for valuable discussion in the early stages of the study's design.
Footnotes
Author contributions
Conceptualization: M.I.E., L.G., H.A.M.; Methodology: M.I.E., A.H., L.P., A.W., L.G., J.O., H.A.M.; Formal analysis: M.I.E., K.B.; Resources: A.W., H.A.M.; Data curation: M.I.E., K.B.; Writing - original draft: M.I.E.; Writing - review & editing: M.I.E., K.B., A.H., L.P., A.W., L.G., J.O., H.A.M.; Visualization: M.I.E.; Supervision: H.A.M.; Project administration: H.A.M.; Funding acquisition: H.A.M.
Funding
This work was supported by a Discovery Grant from the Natural Sciences and Engineering Research Council of Canada (RGPIN-2018-05322) and Ontario Early Researcher Award (ER19-15-080) to H.A.M. Equipment used in this study was purchased through a Canada Foundation for Innovation John R. Evans Leaders Fund (37721) and Ontario Research Fund Award to H.A.M.
Data availability
All relevant data can be found within the article and its supplementary information.
References
Competing interests
The authors declare no competing or financial interests.