Bioelectric regulation of innate immune system function in regenerating and intact Xenopus laevis

Two key inputs that regulate regeneration are the function of the immune system, and spatial gradients of transmembrane potential (V mem). Endogenous bioelectric signaling in somatic tissues during regenerative patterning is beginning to be understood, but its role in the context of immune response has never been investigated. Here, we show that V mem levels modulate innate immunity activity in Xenopus laevis embryos. We developed an assay in which X. laevis embryos are infected with a uropathogenic microorganism, in the presence or absence of reagents that modify V mem, prior to the ontogenesis of the adaptive immune system. General depolarization of the organism’s V mem by pharmacological or molecular genetic (ion channel misexpression) methods increased resistance to infection, while hyperpolarization made the embryos more susceptible to death by infection. Hyperpolarized specimens harbored a higher load of infectious microorganisms when compared to controls. We identified two mechanisms by which V mem mediates immune function: serotonergic signaling involving melanocytes and an increase in the number of primitive myeloid cells. Bioinformatics analysis of genes whose transcription is altered by depolarization revealed a number of immune system targets consistent with mammalian data. Remarkably, amputation of the tail bud potentiates systemic resistance to infection by increasing the number of peripheral myeloid cells, revealing an interplay of regenerative response, innate immunity, and bioelectric regulation. Our study identifies bioelectricity as a new mechanism by which innate immune response can be regulated in the context of infection or regeneration. V mem modulation using drugs already approved for human use could be exploited to improve resistance to infections in clinical settings.

The vertebrate immune system is divided into two categories: innate and adaptive.1 The former provides the first line of defense against pathogens via the actions of surface barriers, secreted antimicrobial peptides, and a subset of blood cell types. The latter is mediated by B and T lymphocytes and relies on the memory of previous exposure to the targeted pathogenic agent. Current vaccination strategies rely on adaptive immunity to create a specific memory that will stimulate immediate immune reaction upon future exposure to the pathogen and as a result, facilitate its elimination. However, with the geographic migration of subtropical diseases, emergence of new pathogenic agents (i.e., exposure to pathogens to which no adaptive memory has developed), clinical interests towards improving immune defenses of lymphocyte-deficient patients, and the need to prevent infection in battlefield and other traumatic injuries, the development of new methods to modulate innate immunity is of utmost importance.

Crucially, the immune system is relevant not only in the context of infection, but is also a major contributor to regenerative response. In general, evolutionary advances in immune defense, especially the development of a fully functional adaptive system, inversely correlate with regenerative capacity.2, 3 The role of immune response in regenerative events is an exciting emerging field.4–6 Thus, the incompletely developed immune system of embryos provides a powerful model to study relationships between regeneration and innate immune response to infection.5, 7 Modulating the pathways involved in embryonic regeneration could lead to potential new avenues for improving innate immune response.

Multiple strategies have been pursued to modulate the effectiveness of innate immunity, including genetic modifications of the host or chemical treatment of blood cells other than lymphocytes.8–10 Here, we explore the role of the host’s bioelectric properties and regenerative response in modulating the effectiveness of the innate immune response. Numerous studies have shown that all cells (not just excitable nerve and muscle) possess endogenous plasma membrane voltage gradients (V mem) that are generated by the ion channels and pumps they express. V mem acts as a crucial regulator of cell differentiation, proliferation, apoptosis, migration, activation, neoplasia, and polarization. Furthermore, changes in the spatial distribution of specific V mem levels across tissues regulates various organ-level functions, including developmental organ patterning and regeneration.11–17 We, therefore, hypothesized that the host’s bioelectric status could modulate the effectiveness of the innate immune response.

To test this hypothesis, we exploited the high amenability of Xenopus laevis embryos to biophysical and molecular-genetic perturbations,7, 18, 19 and the high level of conservation between amphibian and mammalian immunity. A significant advantage of this model organism with respect to the current study is that during X. laevis development, there are no detectable mature B or T cells until 12 days post-fertilization, providing a significant window of investigation in which innate immunity is the only defense the organism has against pathogens.7, 20 Moreover, X. laevis embryos constitute a robust model for regeneration,21, 22 allowing us to investigate how the innate immune mechanisms recruited by the regenerative response23, 24 are affecting the innate immune response to infection.

In this study, we used pharmacological and genetic methods to modulate the bioelectric properties of X. laevis embryo tissues after infection with bacteria that expressed the green fluorescent protein (GFP). We found that V mem depolarization of the host increases its resistance to infection and that its hyperpolarization had the opposite effect. This effect is mediated by a serotonergic signaling pathway at the organismal level and involves migration of embryonic myeloid cells. We also show that X. laevis embryos undergoing tail regeneration/repair show increased resistance to infection. This suggests that bioelectric modulators, including many small molecule drugs already approved for human use, as well as signaling molecules activated by regeneration pathways, represent promising new approaches for manipulating innate immunity responses following exposure to new pathogens.

Our initial goal was to identify pathogenic bacteria that would infect X. laevis embryos and could be easily detected and quantified using fluorescence. Up to developmental stage 48, innate defenses are the only immune resistance active in these transparent embryos, making X. laevis embryos an ideal model for studying modulation of innate immunity.7, 20 We selected a subset of uropathogenic E. coli strains (generously provided by Dr.Matthew A. Mulvey, University of Utah25) based on their ability to infect evolutionarily distant model species (adult mouse and zebrafish embryos) and eliciting immune responses. Moreover, these strains are traceable due to their harboring of a plasmid encoding an unstable form of the green fluorescent protein (GFP-LVA mutant) that allows the experimenters to visualize the progression of the infection, and whose very short half-life circumvents fluorescent noise left by dead bacteria. We first tested the potential of the bacteria to infect embryos at stages 8–9 (blastula) or 11–12 (gastrula). No propagation of the non-pathogenic E. coli strain (K12 lab strain) was detected in embryos infected with it at either stage, while the pathogenic strains showed clear propagation throughout many anatomical regions in the majority of infected embryos (Fig. 1a). These data show that the uropathogenic strains have the capacity to colonize X. laevis embryos at early stages of embryonic development.

We then determined the survival rate of embryos daily for 6 days after infection. When compared to non-infected embryos, no lethality was found among the embryos infected with the non-pathogenic strain. However, only 30–50% of infected embryos survived when the uropathogenic strains were used, with most of the lethality occurring between 24 and 72 h following infection (example shown in Fig. 1b). Thus, since the adaptive immune is not yet developed at these developmental stages, we established that survival rates of X. laevis embryos after infection with human-relevant bacteria could be used as a proxy for the degree of activation of the innate immune system.

Many physiological and developmental processes have been shown to be regulated by bioelectrical signaling.14, 26–30 We, therefore, investigated whether the organism’s bioelectric state could modulate innate immunity activity in vivo. It is important to note that in this study, the terms bioelectric depolarization or hyperpolarization are used in the sense of electrophysiology (referring to plasma membrane resting potential), not “polarization” in the sense of cell activation by chemical chemoattractants or cytokines as described in immunology literature. First, we targeted chloride channels with IVM, which selectively opens glycine receptor chloride (GlyCl) channels in the cell membrane. Since the concentration of Cl− is higher in the intracellular environment (40–60 mM) than in the extracellular one (10 mM), treating the cell with IVM leads to its depolarization, as the negatively charged chloride ions exit the cell through the open GlyCl channels. This strategy has recently been used to efficiently depolarize host tissues and exploited to test the role of tissue resting potentials in guiding innervation,31 muscle patterning,32 and metastasis.33, 34 Treatment of the embryos with 1 μM IVM, a dose that does not inhibit bacterial growth,35, 36 one hour after infection led to a significant increase in the ratio of surviving embryos (31.7 +/− 12.3%; Fig. 1c) when compared with untreated infected ones. Interestingly, by using a medium with higher concentrations of Cl− ions (70 mM NMDG-chloride), which leads to hyperpolarization,31, 34 we observed the inverse effect: there was a significant decrease in the survival ratio (−32.0 +/− 16.0%; Fig. 1c).

To determine whether this effect was chloride-specific or due to depolarization of V mem per se, we tested another depolarizing compound, barium chloride, which is a well-known potassium channel blocker.37–39 We found that it had an effect similar to IVM, significantly increasing the survival ratio (32.3 +/− 7.0%; Fig. 1c). This suggests that the previously observed effect on infection-resistance was not specific to one compound or a type of ion, but rather due to the reagents’ depolarizing action. Membrane voltage readings on embryos treated with barium chloride showed increased depolarization when compared with untreated embryos (Supplementary Fig. 1), indicating the effectiveness of the treatment in depolarizing embryos. We also assayed a medium that was highly concentrated in potassium ions (60 mM), which depolarizes embryos by preventing the exit of K+ ions through potassium channels, by supplementing with potassium gluconate (as in prior studies).40–42 Potassium gluconate (K-gluc) must be used in such bioelectricity studies instead of the commonly-used KCl because the excess chloride ions can confound the results. Surprisingly, no significant increase in survival ratios was observed (−20 +/− 19.3%; Fig. 1c). We hypothesized that this was due to gluconate serving as a nutritious agent for E. coli: if true, any increase in the host resistance to infection would be masked by an increase in bacterial proliferation independently from the host’s bioelectric state. To test this hypothesis, we grew the pathogenic strain on its own in suspension in media supplemented with K-gluc and observed a pronounced increase in bacterial proliferation compared to controls (Supplementary Fig. 2). In addition, we showed that IVM and barium chloride do not have any negative effect on bacteria growth.

Taken together, our data suggest that depolarization of host tissue leads to an increase in its resistance to infection, while hyperpolarization decreases it.

Interestingly, bacteria also utilize bioelectric signaling at their membranes.43, 44 To confirm that the aforementioned effects were not simply due to the compounds’ impact on bacteria survival/proliferation (or their own bioelectric state), we injected fertilized eggs with mRNAs encoding ion channels that reliably lead to depolarization or hyperpolarization in Xenopus 45—this specifically targets the resting potential of host cells, by expressing these channel proteins in the larval tissues. We first used the EXP1 channel, a constitutively conductive cation channel from the nematode C. elegans that acts as a depolarizing agent under these conditions.46 Embryos were injected at the one-cell stage with the highest dose of mRNA that did not cause an increase in spontaneous lethality in non-infected embryos and avoided developmental defects or toxicity. We found that the ratio of survival to infection was significantly increased by 11.0 +/− 2.8% in EXP1-injected embryos (Fig. 1c), reinforcing our previous finding that the observed effects with chemical depolarization were not simply caused by the compounds’ direct action on bacteria and were not due to off-target effects of drug compounds, but rather were a specific consequence of bioelectric potential change in the host cells.

We then used forced expression of hyperpolarizing ion translocators to determine the effect of genetic hyperpolarization of the host on resistance to infection. Overexpression of two hyperpolarizing potassium channels, Kv1.5 and an overactive Kir2.1 mutant,47–49 led to respective significant decreases of 35.3 +/− 17.4 and 39.0 +/− 2.0% in survival after infection (Fig. 1c), confirming at the molecular-genetic level the outcome previously obtained with chemical hyperpolarization. We also hyperpolarized the embryos by injecting the PMA1.2 mRNA, which encodes a plasma membrane-localized proton pump previously used to hyperpolarize Xenopus cells.50, 51 Similar to the effects of hyperpolarizing potassium channels, we observed a significant decrease of 20.0 +/− 10.2% in survival after infection, suggesting that the effect of genetic hyperpolarization was due to the general hyperpolarized state of the embryos, and not specific effects of individual cation types (Fig. 1c).

Taken together, our data from the chemical and genetic modification of the embryos bioelectric state using multiple different ion families, pharmacological agents, and genetic translocators, indicate that depolarization of the host increases resistance to infection, while hyperpolarization decreases it.

In order to probe the mechanism(s) by which the increase/decrease in infection-resistance linked to the host’s bioelectric state, we studied the mobilization of leukocytes (a key mediator of innate immunity) following infection in the multiple bioelectric states described above. We followed leukocyte localization within the embryo by immunofluorescence using XL2, an antibody that specifically recognizes Xenopus embryonic leukocytes.52 We first found that embryos surviving infection 96 h after inoculation were showing peripheral mobilization of embryonic leukocytes: the non-infected controls had XL2-labeled leukocytes concentrated in the middle part of the body (enriched along the borders of the trunk tissues), while the resistant embryos showed a more peripheral distribution of these cells into the fin (Fig. 1d), indicating an infection-triggered migration. This clearly demonstrated that the survivors were exposed to the infectious agents and not simply surviving as a result of experimental inefficiency in the infection protocol (as all specimens were exposed).

Supplementary Figure 3A shows the mobilization of XL2-labeled cells in the periphery of an infected tadpole 4 days post-infection: the density of labeled leukocytes was higher in the ventral fin (right panel) when compared to the non-infected tadpole (left panel). We quantified the density of XL2-labeled cells in the post-anal portion of the ventral fin of non-infected and infected tadpoles for all the previously studied bioelectric states (via the automated counting of XL2 spots in single-color images using ImageJ). In non-infected embryos, only chemical hyperpolarization with high extracellular Cl− concentrations led to a significant increase in relative leukocyte mobilization (Supplementary Fig. 3B, panel iii). Following infection, all conditions showed significant increases in relative leukocytes concentrations in the ventral fin when compared to non-infected tadpoles, but no infected group in a modified bioelectric state showed a significant difference with infected controls (Supplementary Fig. 3B, panels i–iv). This suggests that modification of the organism’s bioelectric state does not result in significant variations in leukocyte mobilization.

To gain additional insight into the events leading up to the effect of V mem on animal survival end points, we characterized bacterial load profiles at intermediate time-points. To quantify bacteria at various times following infection, we adapted a protocol for the quantification of GFP (as the bacteria we are using harbor a GFP-expressing plasmid) activity in embryo lysates.25 Preliminary measurements indicated that the level of infection was sufficient starting at 48 h post-infection to be detectable by spectrofluorometry in non-denaturing embryo lysates (data not shown). We compared the levels of infection at the 72 h post-infection time point (Supplementary Fig. 4) and found that, when compared to controls, the average, median, and third quartile values were all decreased in the two depolarizing conditions leading to an increased survival rate. However, these differences were not statistically significant (p = 0.195 and 0.127 for IVM and barium chloride treatments, respectively).

Interestingly, the hyperpolarizing condition (high chloride) led to a significant increase in bacteria-derived GFP activity (p = 0.008). We also noted that the presence of potassium gluconate leads to a significant increase in bacteria-derived GFP signal (p = 0.019), which correlates with this compound not increasing the survival rate of infected embryos. The lower averages and quartiles observed in the IVM and barium chloride conditions are due to the higher number of embryos showing no detectable level of bacteria-derived fluorescence: these embryos will eventually survive and, although some had visible GFP levels 24 h post-infection, the levels were too low to be detected by spectrofluorometry on whole embryo lysates. Given that all embryos showing visible GFP activity at 48 and 72 h die (Supplementary Fig. 5), we suggest that the critical time for triggering an immune reaction sufficient to eliminate and resist the infection occurs within the first 24 h following infection and that, once a certain bacterial threshold is reached within the host, the progression of infection becomes irreversible and its rate of progression is impacted by the host’s bioelectric status.

Previous studies have shown that bioelectric changes are often transduced into downstream pathways and changes in cell behavior by voltage regulation of serotonergic signaling.31, 33, 34, 53–55 We asked whether a similar signaling pathway was involved in the increased resistance of depolarized embryos to infection, since IVM treatment leads to both hyperpigmentation and increased survival following infection. We exploited a suppression/rescue strategy that targets the serotonin transporter SERT, a key player in the regulation of serotonin movement by resting potential. We repeated the infection experiments with embryos depolarized using IVM or barium chloride, and observed a similar average increase in survival (37.5 +/− 22.5% and 33.5 +/− 16.5% for IVM and barium chloride, respectively; Fig. 2a) as reported above. When we combined these treatments with fluoxetine, a specific inhibitor of serotonin transport,56, 57 the depolarization-induced increases in survival were nullified, suggesting the involvement of a serotonergic pathway in bioelectrically-induced resistance to infection. Similarly, when treated with SHU9119, an MSH agonist leading to a hyperpigmented phenotype acting downstream from serotonergic signaling following treatment with IVM,33 we likewise observed a significant increase in resistance to infection (37.5 +/− 4.5%; Fig. 2a), indicating that hyperpigmentation was sufficient for this increase, and/or that additional pro-resolution functions of MSH signaling58, 59 directly affected immune system functions. Taken together, these results identify signaling pathways involving serotonin and/or MSH in the resistance to infection.

We then molecularly validated the role serotonergic signaling plays in depolarization-induced increased resistance to infection by substituting fluoxetine treatment with the injection of rSert-D98G, an mRNA encoding a dominant negative mutant form of the rat serotonin transporter.55, 60 As observed with fluoxetine, expression of the dominant negative serotonin transporter significantly reduced (for ivermectin treatment) or nullified (for BaCl2 treatment) the depolarization-induced increase in the resistance to infection (Fig. 2b). Together with the results obtained with fluoxetine treatments, these data confirm that a signaling pathway mediated by serotonin modulates depolarization-induced resistance to infection.

The increase in resistance to infection observed following treatment with an MSH agonist (SHU9119; Fig. 2a) suggests that melanocytes could have an active role in mediating innate immunity. This possibility is further supported by the fact that melanocytes have been shown in zebrafish to migrate to wound sites and then fragment releasing their contents.61 We found this to also be the case in Xenopus, in muscle punctures and tail amputations (Fig. 3a). To test the role of melanocytes in immune response, we treated infected albino embryos, whose melanocytes do not synthesize melanin, with IVM and quantified the surviving embryos. We found that the IVM-induced increase in resistance to infection was still present in albinos (31.5 +/− 23.1 % in albinos vs. 15.3 +/− 10.4 % in wild-type; Fig. 3b), suggesting that the melanocyte-dependent effect observed was not dependent on the synthesis/secretion of melanin.

We then tested the effect of factors secreted by melanocytes as well as of their intracellular content on sensitivity to infection. Two subgroups of cultured human melanocytes were evaluated: low density, in which the melanocytes are proliferating and do not show pigmentation, and high density, in which the melanocytes are confluent, quiescent, and pigmented. We found that both the supernatant and cell lysate from actively proliferating melanocytes significantly (p < 0.05) increased the resistance to infection compared to control infected animals (35.0 +/− 7.0% and 159.5 +/− 37.5% for supernatant and lysate, respectively; Figs. 3c–d), while no significant difference was observed with extracts derived from quiescent, melanin-producing melanocytes (−30.0 +/− 42.4% and −21.5 +/− 30.4% for supernatant and lysate, respectively; Figs. 3c–d). These data suggest that non-quiescent melanocytes secrete and harbor factors other than melanin that modulate the innate immune response to increase survival following infection.

Since depolarization of instructor cells by IVM led to increased resistance to infection, we performed a transcriptomic analysis on stage 45 tadpoles to identify likely transcriptional mediators of this effect. Stage 45 was chosen due to its advanced stage of development, which yielded a higher number of innate immunity-related candidates when compared to earlier stages (analyzing at st. 15, we found only 19 transcripts to have been altered, with no known links to immune system function; data not shown). We compared embryos treated with IVM with untreated embryos, as described in (ref. 33). This analysis identified 517 differentially expressed probes following IVM treatment (Appendix 1). Several of these genes correspond to human homologs involved in immune pathways and were associated with immune-related physiological phenomena such as infection (Fig. 4a). Transcripts associated with immunity and those downregulated more than two-fold included eukaryotic translation initiation factor 2, subunit 1 alpha, cathepsin L1, and albumin while those upregulated by more than two-fold included platelet-activating factor receptor, thioredoxin, complement component 4A (Rodgers blood group), complement component 6, gene 1, cathepsin S, and immunoresponsive gene 1.

Using a second, independent transcriptome dataset in Xenopus, we characterized transcript changes that occurred via glycine receptor activation40 and identified 520 differentially expressed probes (Appendix 2). Despite the fact that there was only 5% overlap of common probes within each data set, processes related to inflammation and infection were also significantly represented by these transcripts (Fig. 4b). Network enrichment analysis for expression targets revealed that transcripts regulated by glycine receptor agonism are regulated by molecules such as IL1B, IL11, NF-kappaB, interleukins, and tumor necrosis factor. Moreover, gene network analysis for cell processes with glycine receptor activation40 revealed that many differentially expressed probes were related to inflammation, infection, lymphocyte infiltration, and wound healing. Transcripts in this network included solute carrier family 20 (phosphate transporter), member 2, immunoresponsive 1 homolog (mouse), proteasome (prosome, macropain) subunit, beta type, 8 (large multifunctional peptidase 7), and annexin A1 were upregulated more than 20-fold while protease, serine, 3, CD48 molecule, solute carrier family 26, member 4, and solute carrier family 5 (sodium iodide symporter), member 5 were downregulated more than 20-fold. Other genes in this network associated with immunity and infection were toll-interleukin 1 receptor (TIR) domain containing adaptor protein, CD38 molecule, CD48 molecule, CD81 molecule, inducible T-cell co-stimulator ligand. Both these transcriptomic studies support the idea that improved resistance to infection after depolarization may be mediated in part through the differential expression of transcripts, conserved between frog and mammals, whose products play a role in the immune system.

In order to identify additional cell-level mechanisms that mediate the increase in resistance observed following barium chloride-mediated depolarization, we examined the expression of an early primitive myeloid marker, spib-a,62 and a gene coding for a protein mediating extracellular matrix remodeling (and, therefore, allowing migration) by macrophages, mmp7 (ref. 63). Shortly after hatching (stage 26), the number of spib-a-positive cells was similar in animals from all conditions, but the number of mmp7-positive cells was significantly increased in barium chloride-treated specimen (283 +/− 13 vs. 193 +/− 34 per specimen for barium chloride and control conditions, respectively; Figs. 5a–b). These data suggest that the depolarizing treatment does not affect primitive myeloid cells’ commitment, but acts to increase the proportion of committed myeloid cells acquiring the capacity to migrate throughout the embryos.

At stages 33/34, the majority of spib-a-expressing cells localized within the yolk region and a sizable portion resided in the posterior fin region (Fig. 5c(i)). In embryos showing detectable levels of bacteria-derived GFP activity (qualified as infected), spib-a expression intensified in the ventral yolk region and was absent from the periphery (Fig. 5c(ii)), contrary to resistant embryos (no detectable GFP) in which spib-a expression was still present in the periphery and closer to control levels in the yolk region (Fig. 5c(iii)). For mmp7, expression was detected in cells distributed all over the embryo (Fig. 5c(iv)), becoming concentrated in the ventral yolk region in infected specimens (Fig. 5c(v)), while resistant embryos showed a pattern of expression closer to that of controls (Fig. 5c(vi)). These data indicate that the infection status influences the localization of both spib-a- and mmp7-positive myeloid cells, and suggests that their presence at the embryo periphery is crucial for survival after infection. The increase in mmp7-positive cells (and, therefore, migratory immune cells) shown in embryos depolarized with barium chloride could explain the observed increase in infection-resistance.

Tail amputation has been shown to depolarize the X. laevis embryos at the amputation site,21, 42 and regeneration is now known to involve the recruitment of innate immunity mediators to the site of injury in order to form a blastema and lead to successful development of the ablated tissues.23, 24 Bioelectric signaling is important in both regenerative response64–66 and immune modulation, but no prior studies have examined bioelectrics, immunity, and regeneration in the same context. We thus investigated the effect of tailbud amputation of the host on survival following infection. When tailbuds from infected embryos were amputated (as schematized in Fig. 6a), we found that their survival percentage was significantly increased (21.3 +/− 5.9% vs. 36.8 +/− 5.8% for non-amputated and tail-amputated, respectively; Fig. 6b).

We then determined the localization of mmp7-expressing cells by in situ hybridization of amputated embryos and found that some specimens mobilized a subpopulation a myeloid cells at the site of injury as soon as two hours after amputation (Fig. 6c, top panel), a localization that correlates with that of a foci of depolarized cells at the site of injury as visualized with a voltage dye (Fig. 6c, lower panel). Quantification of mmp7-expressing cells in whole populations of embryos amputated and/or infected showed that, at stage 28 (2 h after amputation), infection resulted in a decreased amount of mmp7-expressing myeloid cells. However, tail amputation led to a significant increase in mmp7-expressing cells in both non-infected and infected specimens when compared to their non-amputated counterpart (416.3 +/− 95.5 vs. 348.8 +/− 72.5, p = 0.049 for non-infected amputated and non-amputated, respectively; 177.0 +/− 67.2 vs. 123.4 +/− 38.0, p = 0.004 for infected amputated vs. non-amputated, respectively; Fig. 6d). This shows that an intervention that depolarizes a portion at the periphery of the embryo42, 67 and leads to the recruitment of myeloid cells to the site of injury can successfully increase the resistance of embryos to infection.

In this study, we demonstrated that uropathogenic strains of E. coli are able to colonize and infect Xenopus laevis embryos. Under the described method, the majority of infected embryos succumb to infection within 4 days, while the surviving minority shows peripheral mobilization of leukocytes indicative of an activated immune response. Considering functional lymphocytes only reach detectable levels later during development,7, 20 the observed immune response is achieved via innate immunity. Crucially, the rate of survival is modulated by the bioelectric status of the host: when chemical or genetic depolarization of the infected embryos occurs at levels that do not affect the viability of uninfected controls, resistance to infection increases, and the survival rates are higher. Conversely, forced hyperpolarization of embryos leads to the opposite effect. Most importantly, triggering regenerative pathways enhances immune response efficiency in eliminating the pathogen.

We characterized the response of animals to bioelectric modulation both, at the protein level (via serotonergic signaling) and at the transcriptional level (microarray analysis); as in other contexts, this system has both a physiological component involving signaling molecule transporters (SERT) and an mRNA component that regulates numerous target genes. These results show that X. laevis embryos are a robust model for studying or screening modulators of innate immune response and for characterizing the cellular and molecular components of this response, as well as their relationship to regeneration, with respect to both molecular-genetic and biophysical parameters. Moreover, the ease in obtaining large quantities of fertilized embryos and the infection method (direct injection) makes it a relatively high-throughput screening method, which could become even more effective with the future development of robotic automated injection.