Reserpine improves Enterobacteriaceae resistance in chicken intestine via neuro-immunometabolic signaling and MEK1/2 activation

Poultry products are the primary vehicle for broad-host, nontyphoidal Salmonella enterica contamination and foodborne disease in the United States^1,2, causing 1.35 million infections and costing approximately $400 million annually3. Although extensive efforts have been made to minimize Salmonella incidence in poultry via antimicrobials, the spread of resistance genes has caused an emergence of Salmonella isolates resistant to essential antibiotics^3,4. Furthermore, live Salmonella vaccines and probiotics are commonly implemented as prophylactics in commercial poultry to reduce Salmonella load, however, their individual efficacies against Salmonella resistance are inconsistent^5–7. Altogether, current methods are insufficient in the reduction of Salmonella in chickens, suggesting that a deeper understanding of biological factors affecting Salmonella colonization is needed to develop more successful treatments.

In chickens, broad host Salmonella serovars induce an immunotolerant state in the chicken intestine via increased regulatory T cells (Tregs), which suppress the inflammatory immune responses necessary to clear Salmonella^8,9. Thus, interfering with Treg activities in the gut may improve antibacterial responses against Salmonella. A largely-understudied field in chicken biology is neuroimmunology, or the interactions between the nervous and immune systems¹⁰. The intestine is highly-innervated with neurons and immune cell populations, which can then interact via neurochemical signaling¹¹. In mammals, Tregs synthesize their own stores of catecholamine neurochemicals like norepinephrine, and disrupting these intracellular stores via reserpine inhibits Treg function¹². However, whether chicken Tregs have similar neurochemical stores and if they too are affected by reserpine have not yet been investigated.

In this report, we found that reserpine causes the release of intracellular norepinephrine stores from chicken ceca explants and intestinal Tregs, driving increased antimicrobial responses against Salmonella. These ex vivo antimicrobial responses were recapitulated in vivo, as birds orally treated with reserpine exhibited reduced gut Enterobacteriaceae and Salmonella post-challenge compared to control birds. Furthermore, we found that reserpine treatment induced T cell activation, reduced CTLA-4 gene expression, and deactivated metabolic pathways like epidermal growth factor receptor (EGFR) signaling and mammalian target of rapamycin (mTOR) signaling, which were linked to antimicrobial responses. Lastly, we found that MEK1/2 activation plays a central role in reserpine-induced antimicrobial activities.

Results

Reserpine treatment induces norepinephrine release from intestinal cells. In an intestinal explant model¹³ (Supplementary Fig. 1), we demonstrated neurochemical release in ceca tissues at 1 h post-reserpine treatment (1 µM) using ultra-high performance liquid chromatography (UHPLC). Culture media from reserpine-treated explants had increased levels of norepinephrine and no changes in serotonergic metabolites compared to controls (Fig. 1a). However, this norepinephrine release did not induce inflammatory damage in the explants, as pathological scores were statistically identical between groups (Supplementary Fig. 2a). Using flow cytometry to sort lymphocyte populations (Fig. 1b) potentially responsible for norepinephrine release in the ceca, Tregs (i.e., CD4⁺CD25⁺) had significantly greater intracellular norepinephrine stores versus naïve T helper (TH) cells (i.e., CD4⁺CD25⁻), and reserpine treatment reduced intracellular norepinephrine levels in Tregs alone (Fig. 1c). However, intracellular stores of serotonergic metabolites were unaffected by reserpine treatment (Supplementary Fig. 2b, c).

Reserpine treatment increases Salmonella resistance in ex vivo and in vivo conditions. In ceca explants from 21-day-old birds, supernatant from the reserpine-treated group had higher killing ability against Salmonella compared to that of control explants regardless of strains tested, e.g., Salmonella Typhimurium and Salmonella Kentucky (Fig. 1d). However, reserpine itself was not bactericidal (Supplementary Fig. 2d), confirming that Salmonella killing was mediated by host factors. To test in vivo reserpine-induced antimicrobial responses, we orally treated chickens with 0, 0.5, or 5 mg reserpine/kg body weight from 1 to 3 days post-hatch (dph). Reserpine treatment at either concentration did not affect the chicken weight gain at pre- (Supplementary Fig. 3a) nor post-Salmonella challenge (Supplementary Fig. 3b), nor did oral reserpine treatment induce the significant release of any neurochemicals systemically (Supplementary Fig. 3c). Given that reserpine induced antimicrobial responses ex vivo, we predicted reserpine may affect the commensal gut microbiota. However, 16S rRNA sequencing showed that reserpine treatment did not affect the levels of the majority of commensal bacteria in the ceca (Fig. 2a and Supplementary Figs. 4, 5). Nevertheless, antimicrobial responses were clearly observed after birds were challenged with Salmonella Typhimurium UK-1. At two days post-Salmonella challenge, fecal shedding of total Enterobacteriaceae and Salmonella was significantly reduced by reserpine treatment regardless of concentration (Fig. 2b). Similarly, total Enterobacteriaceae and Salmonella CFUs in ceca content were reduced by reserpine treatment at four days post-challenge (Fig. 2c). In addition to colonizing the chicken intestine, broad host Salmonella strains like UK-1 have the capacity to invade internal organs in young birds¹⁴. Here, Salmonella Typhimurium UK-1 was detected in ceca, spleen, and bursa but not in the liver of challenged birds. Although Salmonella levels were statistically identical between groups in the bursa, reserpine treatment significantly reduced Salmonella levels in the spleen (Supplementary Fig. 6). Furthermore, reserpine treatment did not induce pathological inflammation at any concentration in the small intestine nor ceca (Supplementary Fig. 7), and ceca goblet cell numbers were significantly increased by reserpine treatment (Fig. 3). This is in line with a previous study demonstrating that, in mammals, reserpine treatment increases the production of intestinal mucus^15,16, which is synthesized by goblet cells in the epithelium¹⁷.

Reserpine treatment increases antimicrobial peptide expression while decreasing CTLA-4 expression. To determine underlying mechanisms responsible for improved antimicrobial responses upon reserpine treatment, we measured genes expression through transcriptional changes via RT-qPCR. Expression of the regulatory cytokine IL-10¹⁸ was unchanged (Fig. 4a); however, the expression of CTLA-4, a surface-bound protein associated with Tregs that downregulates immune responses¹⁹, was downregulated in reserpine-treated explants versus controls (Fig. 4A). In line with this downregulated immunosuppressive factor, reserpine treatment increased gene expression of antimicrobial peptides (AMPs) like beta defensin 12 (BD-12), BD-14, and fowlicidin 1 (Fowl-1) versus controls (Fig. 4a). Furthermore, the expression of IL-2, a cytokine released by activated T cells^20,21, was also increased by reserpine treatment versus control (Fig. 4a).

Reserpine-treated explants undergo large immunometabolic shifts. To determine what global immunometabolic pathways were affected by reserpine, we used a chicken-specific kinome peptide array, which measures changes in phosphorylation activities within several signaling pathways²². Overall, reserpine treatment altered several immunological and metabolic pathways (Table 1). In total, 414 proteins from the top 25 KEGG pathways were differentially phosphorylated upon reserpine treatment (Table 1). Within these pathways, several were involved in the EGFR signaling pathway and T cell receptor (TCR) signaling pathway, and these pathways were further analyzed. EGFR was dephosphorylated at the Tyr869 residue (Table 2). Furthermore, in the EGFR signaling pathway, mTOR was phosphorylated at Ser2448 and Thr2446 but was dephosphorylated at Ser2481 (Table 2). Uniquely, mitogen-activated protein (MAP) kinase 2 (MEK2), a component of the MEK1/2 signaling pathway²³, was phosphorylated at the Ser306 residue (Table 2), important for MEK2 activation²⁴. Similarly, MEK2 is also involved in the TCR signaling pathway, in which CD28, a T cell co-receptor crucial for T cell activation²⁵, was phosphorylated (Table 2).

Reserpine improves Enterobacteriaceae resistance in chicken intestine via neuro-immunometabolic signaling and MEK1/2 activation - Image 1

Reserpine-induced antimicrobial responses are dependent on norepinephrine and metabolic signaling. Given that reserpine treatment (1) increased intracellular norepinephrine release and (2) induced changes in EGFR and mTOR phosphorylation, we investigated the roles of these pathways in antimicrobial responses. Explants treated with norepinephrine alone similarly induced antibacterial responses in a dose-dependent manner (Fig. 4b), which was blocked by inhibiting beta-adrenergic receptors 2 and 3 (Fig. 4c). Treatment of explants with recombinant EGF alone prevented reserpine-induced antimicrobial responses (Fig. 4d). However, treatment with EGFR inhibitor AG1478 alone did not trigger any antimicrobial responses (Fig. 4d). Additionally, treatment of explants with rapamycin, an inhibitor of the mTOR pathway, increased bactericidal responses (Fig. 4e). Overall, these findings demonstrate that reserpine treatment induces antimicrobial responses through multiple signaling pathways.

MEK1/2 signaling plays a central role in reserpine-induced antimicrobial responses. In our kinome analyses, we found that these immunometabolic signaling changes were associated with MEK2 phosphorylation, suggesting MEK1/2 signaling plays a vital role in these responses. Using the MEK1/2 signaling inhibitor U0126, MEK1/2 signaling inhibition reversed the antimicrobial response induced by reserpine (Fig. 4d). Similarly, antimicrobial responses in rapamycin-treated explants were partially reversed upon MEK1/2 inhibition (Fig. 4e). Finally, antimicrobial responses in norepinephrine-treated explants were reversed upon MEK1/2 inhibition (Fig. 4f). Overall, these data demonstrate a central role for MEK1/2 signaling in the antimicrobial response induced by reserpine and other neuroimmunometabolic signaling pathways.

Discussion

Chicken products like meat and eggs are primary vehicles for salmonellosis^1,2. Reducing Salmonella colonization in the chicken intestine is paramount to mitigating salmonellosis in humans. In this study, we demonstrate that reserpine treatment releases intracellular stores of norepinephrine and induces significant changes in chicken ceca immunometabolism, resulting in increased antibacterial responses against Salmonella. The ex vivo explant model used in this study allows for preserving the totality of intestinal cell populations present in vivo while maintaining spatial organization, which provides a more accurate representation of in vivo conditions¹³. In support of the utility of this model, we found that reserpine treatment induces antimicrobial responses against Salmonella ex vivo and in vivo. In our study, reserpine treatment increased the expression of several AMPs, including beta-defensins 12 and 14 as well as fowlicidin-1. Betadefensins are crucial to regulating the gut microbiota and homeostasis²⁶. Thus, strategies that increase host beta-defensin production are viable replacements for antibiotic treatment²⁷. Although these molecules are directly bactericidal, they have additional functions as well. For example, fowlicidin-1 can neutralize bacterial lipopolysaccharide (LPS)²⁸, a microbe-associated molecular pattern that potently induces inflammation²⁹. Furthermore, beta-defensins reduce intestinal apoptotic signals in LPS-treated animals³⁰. Thus, improving the production of these AMPs may both increase resistance against bacterial pathogens, as well as mitigate host damage induced by these antibacterial responses. In support of this, we found no differences in pathological scores between groups despite a clear elevation in immunological responses in reserpine-treated explants. However, the transcriptional factors responsible for reserpine-induced antimicrobial peptide production are unclear at this time. Activation of the transcription factor c-FOS increases antimicrobial responses in macrophages³¹ while suppressing excessive inflammatory responses^32–34. Given these findings were reflected in our study, reserpine-induced c-FOS activation may be driving these antimicrobial responses, although this remains to be determined.

Reserpine improves Enterobacteriaceae resistance in chicken intestine via neuro-immunometabolic signaling and MEK1/2 activation - Image 2

Reserpine improves Enterobacteriaceae resistance in chicken intestine via neuro-immunometabolic signaling and MEK1/2 activation - Image 3

Reserpine improves Enterobacteriaceae resistance in chicken intestine via neuro-immunometabolic signaling and MEK1/2 activation - Image 4

Reserpine improves Enterobacteriaceae resistance in chicken intestine via neuro-immunometabolic signaling and MEK1/2 activation - Image 5

This reserpine-driven increase in AMP production was associated with increased IL-2 expression and reduced CTLA-4 expression. Upon activation of naïve T cells, IL-2 production is increased, which induces further T cell proliferation, promotes CD4⁺ differentiation, and facilitates effector and memory CD8⁺ T cell formation²⁰. This activation process is dependent on the interaction between costimulatory ligand CD28, expressed on naïve T cells, and CD80/86, expressed on antigen-presenting cells (APCs)³⁵. However, Tregs can interfere with this interaction via CTLA-4, which outcompetes CD28 for CD80/86 binding, inhibiting IL-2 accumulation and thus preventing T cell activation^25,36. One of the mechanisms in which Salmonella persists in the chicken gut is by increasing intestinal Tregs, which prevents the inflammatory responses necessary to clear Salmonella⁹. Thus, we hypothesized that reserpine treatment could inactivate chicken Tregs as shown in human Tregs¹², which would permit anti-Salmonella responses in the gut. As expected, reserpine decreased CTLA-4 expression, which is constitutively expressed on Tregs³⁷. We found that CD28 was phosphorylated in reserpine-treated explants, suggesting that CD28 activation and IL-2 production were occurring due to reduced CTLA-4 levels. Furthermore, NFATC1 (but not NFATC2) was phosphorylated upon reserpine treatment. Activation of these transcription factors has been linked to IL-2 production in memory CD4⁺ T cells³⁸, suggesting that reserpine is increasing IL-2 gene expression through NFATC1 activation.

One notable observation is that reserpine treatment in vivo did not dramatically change the resident gut microbiota in young birds. The gut microbiota is crucial to proper animal development, driving immune and physiological maturation^39,40. Furthermore, antibiotic treatment in young animals causes dramatic changes in their gut microbiota⁴¹, which can predispose them to bacterial infection and physiological dysfunction later in life by depleting populations of gut microbes crucial for normal development^42,43. Thus, oral reserpine treatment in day-old birds is a feasible way to promote resistance against Salmonella without negatively affecting the developing gut microbiota, although the long-term effects of early-age reserpine treatment on the gut microbiota are unclear. Although changes in Fusobacteria, Lactobacillaceae, and Erysipelotrichaceae were induced by reserpine treatment, these changes were not consistent between reserpinetreated groups and did not appear to be associated with any biological parameter measured in this study (antimicrobial responses, inflammation, mucus production, etc). Thus, the biological impact of these specific shifts in the commensal microbiota is unclear and does not contribute to the host responses investigated in this study. Still, this lack of antimicrobial activity may appear to contrast the reserpine-induced antimicrobial responses seen in our ex vivo explant model. In birds, innatelyproduced gallinacins are the primary AMPs produced in the intestine at post-hatch, peaking at days 1–3 post-hatch and begin to drop by day 4 post-hatch, in which AMPs controlled through the adaptive immune system become dominant in the chicken intestine⁴⁴. In our study, explants from 21-day-old birds were used to assess reserpine efficacy, in which these intestinal explants would have a more-mature adaptive immune system. Thus, reserpine-induced antimicrobial responses appear to be dependent on the adaptive immune system. This is supported by our finding that reserpine induces norepinephrine release in chicken intestinal Tregs, which coincidentally migrate from the thymus to the chicken intestine around day four post-hatch^45–47. Dhawan and colleagues (2016) determined that specific subsets of intestinal Tregs are crucial for regulating AMP responses⁴⁸, although the subset of Tregs responsible for this mechanism in chickens is still unclear and warrant further investigation. In humans, reserpine inhibits intracellular vesicle storage of catecholamines such as norepinephrine, which induce autocrine/ paracrine signaling loops that suppress Treg function and stimulate immune activation¹². In this study on chickens, reserpine treatment increased norepinephrine release from both explants and intestinal Tregs. Thus, Tregs at least partially contribute to the total pool of norepinephrine released by intestinal cells. However, in our hands and due to limited reagents and methods for primary chicken cell cultures, we could not culture chicken intestinal Tregs for longer than six hours, preventing any direct examination of reserpine on Treg immunosuppressive function. However, we did find that treatment with norepinephrine alone at the physiological concentration released after one hour of reserpine treatment could stimulate antibacterial responses, which was dependent on beta-adrenergic receptors. Norepinephrine is a well-known mediator of neuroimmunological responses, inducing cytokine production, cell proliferation, and antibody secretion by lymphocytes^49,50 and has been demonstrated to improve antibacterial responses via cross-talk between sympathetic ganglia and resident tissue macrophages⁵¹. Overall, the intracellular release of norepinephrine drives antimicrobial responses via autocrine/paracrine signaling of intestinal cell populations. Future work should determine which specific cellular populations (i.e., enterocytes, enteric neurons, APCs) interact with the regulatory T cells involved in this mechanism.

Reserpine improves Enterobacteriaceae resistance in chicken intestine via neuro-immunometabolic signaling and MEK1/2 activation - Image 6

Given the clear immunological stimulation induced by reserpine treatment, we hypothesized that several metabolic pathways might also be affected due to the interplay between host metabolism and the immune system¹⁰. To this end, we used the chicken kinome peptide array, which measures immunometabolic signaling at the post-translational level²² and thus enables a more accurate evaluation of which processes are affected by reserpine. EGFR signaling is crucial for goblet cell-associated antigen passage (GAP) formation in the mammalian intestine⁵², and inhibiting EGFR increases beta-defensin production in intestinal cells in vitro⁵³. In this study, we found that EGFR was dephosphorylated in reserpine-treated explants and using recombinant EGF reversed reserpine-induced antimicrobial responses in vitro, demonstrating the importance of EGFR signaling in this system. However, EGFR inhibition alone did not trigger antimicrobial responses, suggesting that EGFR signaling alone is not sufficient to induce antimicrobial responses. Additionally, the mTOR pathway is conserved among eukaryotic organisms and has received vast attention due to its diverse involvement in nutrient sensing, immunity, and aging in animals⁵⁴. Rapamycin, originally derived from the soil bacterium Streptomyces hygroscopicus, is commonly used as an mTOR inhibitor⁴⁰. In this study, reserpine induced differential mTOR phosphorylation at multiple sites upon reserpine treatment. Phosphorylation of S2448 and T2446 is carried out by the kinase S6K⁵⁵, and pS2448 drives mTORC1 activation⁵⁶. In this study, mTORC1 may have been activated upon reserpine treatment, as these two mTOR sites, S6K, and raptor (i.e., RPTOR) were all phosphorylated. However, mTORC1 activation does not play a role in these antimicrobial responses, as deactivating mTOR via rapamycin treatment induced similar antimicrobial responses as reserpine treatment. Although these mTOR sites were phosphorylated, S2481 was uniquely dephosphorylated upon reserpine treatment. The sole site for mTOR autophosphorylation⁵⁷, S2481 has been the only site determined to regulate intrinsic mTOR activities^58,59. Thus, S2481 dephosphorylation deactivates mTOR function, and our study finds that mTOR inhibition increases antimicrobial responses in this ceca explant model. This finding is supported by previous work demonstrating rapamycin treatment increases anti-Campylobacter responses in the murine intestine and directly stimulates antimicrobial responses in splenocytes⁶⁰. Thus, in addition to inducing norepinephrine signaling, reserpine also deactivates EGFR and mTOR, and all three of these pathways contribute to antimicrobial responses in chickens. Given that numerous mTOR sites were phosphorylated and dephosphorylated by reserpine treatment, future studies should look at the individual roles of these sites in antimicrobial responses, which could serve as drug targets to promote bacterial resistance.

Although we identified several pathways that differed in phosphorylation patterns, MEK1/2 signaling is well-established as an essential component of beta-defensin production at mucosal barriers^53,61,62. However, MEK1/2 signaling has never been previously described to be involved in reserpine activity. Here, upon reserpine treatment, MEK2 was phosphorylated at S306. Using the inhibitor U0126, we found that inhibiting MEK1/2 signaling reversed reserpine induced antimicrobial responses, as well as those induced by norepinephrine and rapamycin treatment alone, suggesting that pS306 is a central component of this signaling pathway induced by reserpine and is critical to achieving an antimicrobial response.

In summary, we found that reserpine increases AMP production and immune activation in the chicken intestine by inducing norepinephrine release and beta-adrenergic receptor activation. These changes are correlated with reduced CTLA-4 expression, as well as EGFR and mTOR deactivation, and these antimicrobial responses were dependent on MEK1/2 activation. Thus, we propose that targeting the neuroimmunological axis via oral reserpine treatment could be a viable strategy for increasing Salmonella resistance in poultry animals. Furthermore, since oral reserpine treatment also increased resistance against total Enterobacteriaceae populations, this treatment may also increase resistance against other bacterial pathogens.

This article was originally published in Communications Biology (2021) 4:1359. https://doi.org/10.1038/s42003-021-02888-3. www.nature.com/commsbio. This is an Open Access article licensed under a Creative Commons Attribution 4.0 International License.