The Pineal Gland Is Required for Optimal Nighttime Immunity to Bacterial Infection in Larval Zebrafish Authors: Gabriel Anasarias‐Dumo, Yevetta Y. Xiang, David Cumin, James F. Cheeseman, Guy R. Warman, Jonathan W. Astin, Christopher J. Hall Published: 2026 ABSTRACT ABSTRACTThe immune response to bacterial infection demonstrates circadian rhythmicity, with heightened antibacterial immunity often observed during the daytime. This circadian‐gated immune response is believed to have evolved to synchronize peak antibacterial immunity with the increased threat of infection when active. Using a larval zebrafish infection model, we have previously shown that light has a positive effect on antibacterial immunity, with enhanced survival and neutrophil recruitment observed when infections are delivered during the light phase. However, the specific contribution of the light‐detecting pineal gland and its principal hormone, melatonin, to this circadian immune response has not been investigated. Here, we examined whether the pineal gland modulates the host response to bacterial infection in a time‐of‐day–dependent manner by targeted ablation of the pineal gland. To generate pineal gland‐deficient larvae we created the transgenic zebrafish line Tg(exorh:YFP‐NTR2.0) that restricts expression of the highly efficient nitroreductase variant NTR2.0 to pinealocytes, enabling metronidazole‐mediated pinealectomy. We also analyzed larvae lacking the brain‐specific homeobox (bsx) gene, that encodes a transcription factor essential for pineal gland development. Both approaches resulted in complete loss of the pineal gland, confirmed by live imaging, the absence of arylalkylamine N‐acetyltransferase 2 (aanat2) expression and diminished melatonin production. Pineal gland‐deficient larvae exhibited normal survival and neutrophil recruitment when infected during the light phase, indicating that daytime immune defenses can be maintained independently of the pineal gland. In contrast, infection of pineal gland‐deficient larvae during the dark phase led to significantly reduced survival, which correlated with impaired neutrophil recruitment to the infection site. These findings demonstrate that the pineal gland plays an important role in supporting nighttime antibacterial immunity in larval zebrafish. Introduction From bacteria to humans, circadian rhythms enable organisms to prepare for recurring daily environmental challenges [1]. The host response to bacterial infection demonstrates a robust circadian rhythm, with optimal immunity often observed during an animal's early active phase [2, 3, 4]. This circadian‐gated antibacterial immunity likely evolved to synchronize optimal immune responses with the increased threat of infection when active. Given that genetic and environmental disruption of our endogenous circadian rhythms can lead to enhanced risk of infections [5], a thorough understanding of circadian‐gated immunity may uncover new approaches to better treat infections. Largely synthesized in the pineal gland, melatonin has been demonstrated to influence the host response to infections [6, 7]. Circulating melatonin is believed to activate innate immune cells (neutrophils and macrophages) that provide the critical first line of defense against bacterial infections [6]. Melatonin also potentiates the chemotactic response of rodent and human neutrophils, the most abundant circulating leukocyte with potent antibacterial functions [8]. The distribution of neutrophils throughout the body and their functional activity also demonstrate circadian rhythmicity, with peak numbers in tissues coinciding with the active phase [9, 10]. Furthermore, the circadian‐gated extravasation of neutrophils into tissues, from the circulation, is dependent upon circadian clock‐controlled expression of the chemokine Cxcl2, and is believed to provide enhanced antibacterial protection during periods of activity [11]. Neutrophils also use their circadian clock machinery to transcriptionally amplify bactericidal activity during the active phase [12]. In addition to their potent bactericidal activity, neutrophils are implicated in the pathogenesis of many diseases, from cancer and inflammatory disorders to metabolic diseases [13]. As such, manipulating circadian‐gated neutrophil activity may offer new and alternative therapeutic possibilities for treating a broad range of diseases [14]. Equipped with a highly conserved innate immune system, larval zebrafish provide a powerful system to directly observe the host response to microbial infection [15, 16]. By 2 days postfertilization (dpf), larval zebrafish are populated with innate immune cells that functionally resemble those in mammals. Supported by transgenic reporter lines marking neutrophils and macrophages, transparent larval zebrafish provide a powerful system to directly observe the innate immune response to microbial infection, at the single‐cell level [17, 18, 19, 20]. As a diurnal animal possessing a highly conserved set of circadian clock genes, larval zebrafish are also a popular model in which to study circadian clock biology [21, 22], and they possess a light‐responsive pineal gland that rhythmically produces melatonin from 26 h post fertilization (hpf) [23]. We and others have leveraged off the conserved immune system and circadian biology of larval zebrafish to study the influence of light and circadian rhythms on host immune responses [12, 24, 25, 26, 27, 28, 29, 30]. When examining the host response to bacterial infection in larval zebrafish, our previous work has shown that light is protective, with better survival outcomes when infections are delivered during the daytime [24]. We have also shown that this positive impact of light is largely dependent upon the light‐induced clock gene per2, the expression of which is enriched within the pineal gland during the daytime [12, 31]. In this study, we wanted to investigate whether the light‐detecting pineal gland contributes to the host response to infection. By studying the antibacterial immune response within pineal gland‐deficient larvae, under different lighting conditions, we uncover an unexpected role for the pineal gland in helping direct neutrophil recruitment to sites of infection during the night. Materials and Methods Zebrafish and Lighting Conditions Zebrafish research was conducted with the approval of the University of Auckland Animal Ethics Committee (approval number AEC22563). Adult zebrafish (Danio rerio) were housed in the University of Auckland's Zebrafish Facility on a 14‐ to 10‐h light–dark (LD) cycle. Larval zebrafish were generated by natural spawning and raised at 28°C in standard E3 medium (0.33 mM CaCl2, 0.33 mM MgSO4, 0.17 mM KCl, and 5 mM NaCl). When required, E3 medium was supplemented with 0.003% phenylthiourea (PTU) to inhibit pigmentation. For LD treatment, embryos were harvested and raised under 14‐ to 10‐h LD conditions within environmental incubators (maintained at 28°C). For the reversed DL condition, embryos were harvested within 2 h of spawning and raised in environmental chambers in 10‐ to 14‐h DL conditions. WT AB zebrafish were obtained from the Zebrafish International Resource Centre (ZIRC). The neutrophil specific Tg(lyz:DsRED2) nz50 reporter line was used in this study [19]. Generation of Transgenic Lines To create the exorh:EGFP transgene construct, p5E‐exorh [32] (Addgene ID 195949) was gateway‐cloned into pDestTol2pA with pME‐EGFP and p3E‐polyA using the Tol2 kit [33]. To create the exorh:YFP‐NTR2.0 transgene construct, p5E‐exorh was similarly gateway‐cloned into pDestTol2pA with pENTR‐TagYFP‐P2A‐NfsB_VvF70A/F108Y [34] (herein referred to as pME‐YFP‐NTR2.0) and p3E‐polyA. The exorh:EGFP and exorh:YFP‐NTR2.0 constructs were injected into one cell‐stage Tg(lyz:DsRED2) embryos, along with Tol2 transposase mRNA, to create the stable lines Tg(exorh:EGFP) nz40 ;Tg(lyz:DsRED2) and Tg(exorh:YFP‐NTR2.0) nz41 ;Tg(lyz:DsRED2), respectively. Metronidazole Treatment Tg(exorh:YFP‐NTR2.0) larvae were immersed in metronidazole (MTZ) (Sigma‐Aldrich) at the indicated concentrations (dissolved in E3 medium) and times, as previously described [35]. For all treatments, larvae were continuously immersed in MTZ throughout the experiment. Generation of bsx Crispants To generate bsx crispants, injection mixes of 5 μM bsx sgRNAs 1 and 2 (sourced from IDT using their pre‐designed library, design ID: Dr.Cas9.BSX.1.AA and Dr.Cas9.BSX.1.AE, respectively, and resuspended in sterile H2O) + Cas9 protein (Alt‐R Streptococcus pyogenes Cas9 nuclease, v.3, IDT) were incubated at 37°C for 5 min, as previously described [36]. One nanoliter was then injected into the cytoplasm of one cell‐stage embryos. Genomic DNA was isolated at 2 dpf using the HotShot method [37] to assess for targeted editing and to quantify editing efficiency using the T7 endonuclease I assay and next‐generation sequencing (NGS). As a control for the crispant experiments, a tyrosinase‐targeting sgRNA (tyr sgRNA) was used (sourced from IDT using their pre‐designed library, design ID: Dr.Cas9.TYR.1.AA). T7 Endonuclease I Assay For initial testing of bsx sgRNAs 1 and 2 editing, the T7 endonuclease I assay was used, as previously described [38] using the following primers: forward 5′‐GCTGGGTGCTAATCTACTTCCTG‐3′; reverse 5′‐GACATTACATTTGAGGCAAATAGGCC‐3′. Quantification of CRISPR‐Cas9 Editing by Next‐Generation Sequencing To quantify the editing efficiencies of bsx sgRNAs 1 and 2, NGS was performed on nested PCR amplicons. Primary PCRs were performed with the same primer pairs used for the T7 endonuclease I assay using Platinum SuperFi II DNA polymerase (Life Technologies, Thermo Fisher Scientific). The primary PCR was then used as a template for a secondary PCR using primer pairs with incorporated Illumina adapter sequences (forward 5′‐CCTACACGACGCTCTTCCGATCTATGCATCAAAAGACGCACGG‐3′; reverse 5′‐ GGAGTTCAGACGTGTGCTCTTCCGATCTCAGGTTTGTGTAACGGATGATGC‐3′, Illumina adapter sequence underlined). Following a final PCR with indexing primers, products were cleaned using AMPure XP Beads (Beckman Coulter Life Sciences), as per the manufacturer's protocol, and sequenced on an Illumina MiSeq system. Indels were quantified from FASTQ sequencing files using the online sequence analysis tool CRISPResso2, with a quantification window set at 17 nucleotides on either side of the predicted cleavage position. The editing efficiency of tyr sgRNA was performed as previously described [12]. Live Fluorescence Microscopy Larvae were anesthetized with 4% tricaine and arrayed in 3% methylcellulose in E3. Images were taken using a Zeiss Axio Zoom V16 stereo microscope (with enhanced GFP and monomeric red fluorescent protein filter sets) equipped with Zeiss Axiocam 506 mono and 208 color cameras. Whole‐Mount In Situ Hybridization Whole‐mount in situ hybridization for aanat2 expression was performed as previously described [24, 39]. Staining was performed using BM Purple (Roche) as the substrate. After staining, larvae were fixed in 4% paraformaldehyde (PFA) and stored in 100% glycerol at 4°C. Melatonin Quantification Melatonin extraction and ELISA were performed as described [23, 40]. In brief, pools of 100 larvae were collected at 76 hpf (zeitgeber time [ZT]16 DL condition) and homogenized in 0.5 mL of 0.1 N NaOH. One mL methylene chloride was added to each sample, followed by vortexing for 30 s and centrifugation at 4250 × g for 3 h at 4°C. The organic phase was transferred to a fresh tube, mixed with 0.5 mL of 0.1 N NaOH, vortexed for 30 s, and centrifuged for 20 min at 4°C. The organic phase was again transferred to a fresh tube, supplemented with 0.15 g Na2SO4, vortexed for 30 s, and centrifuged for 20 min at 4°C. One mL of the final organic phase was collected from each sample, and methylene chloride was evaporated at room temperature under vacuum using a rotary evaporator. Dried extracts were reconstituted in 200 µL PBS with 0.1% gelatin (w/v) (Sigma‐Aldrich). Melatonin concentrations were quantified in experimental duplicate using a commercial ELISA kit (IBL International) according to the manufacturer's instructions. Infections and Survival Analysis Infections and survival analyses were performed as previously described [15]. In brief, larvae were anaesthetized in tricaine and arrayed in 3% methyl cellulose (in E3 medium) and infected with 1000 CFU of GFP‐expressing S. enterica serovar Typhimurium (herein referred to as Sal‐GFP) by hindbrain microinjection at 52 hpf. For infections in the dark phase (DL condition), larvae were arrayed and infected under dim light to minimize light exposure, as previously described [12]. Infected larvae were monitored daily for death (cardiac arrest). Dead larvae were removed, and any displaying excessive Sal‐GFP dissemination (as detected by fluorescence microscopy) were transferred to a separate Petri dish to prevent bacterial spread through the E3 medium, in the event of death and larvae dissociation between monitoring periods. Larvae were monitored for death during the light period for both the LD and DL conditions. Bacterial Growth Assay To assess the effect of MTZ on bacterial growth, Sal‐GFP cultures were established in 4 mL Luria–Bertani (LB) broth supplemented with either 100 μM MTZ, 100 μM DMSO, or no treatment. Each culture was inoculated with approximately 4 × 103 CFU Salmonella and incubated at 37°C with shaking at 200 rpm. At 1, 3, 6, 9, and 24 h post‐inoculation, aliquots were collected from each culture, serially diluted in sterile PBS, and plated onto LB agar plates for colony enumeration. For each time point, serial dilutions were plated in technical triplicate. Plates were incubated overnight at 37°C, and colonies were subsequently counted to determine the total CFU in the LB broth at each time point. Immunofluorescence and Confocal Imaging Immunofluorescence detection of neutrophils within Tg(lyz:DsRED2) larvae was performed as previously described [41]. In brief, larvae were fixed overnight at 4°C in 4% PFA containing 4% (w/v) sucrose, followed by dehydration into 100% methanol. Samples were then rehydrated through a graded methanol series back to PBST (PBS supplemented with 0.1% (v/v) Tween‐20) and permeabilized with 100% acetone at −20°C for 7 min. After incubation in blocking solution (PBST supplemented with 10% (v/v) normal goat serum) for 2 h at room temperature, samples were incubated sequentially overnight in primary and secondary antibodies. The primary and secondary antibodies used for the detection of DsRED2 and their dilutions were as follows: rabbit anti‐DsRED (1:500; Takara Bio, catalog no. 632496) and goat anti‐rabbit Alexa Fluor 546 (1:500; Thermo Fisher Scientific, catalog no. A‐11035). To quantify neutrophil recruitment, larvae were mounted in 1% (w/v in PBS) UltraPure Low Melting Point Agarose and imaged using a Nikon C1 Eclipse confocal microscope equipped with a 20X objective, focusing on the hindbrain ventricle infection site. Z‐stacks were collected at 3 µm intervals over a total depth of 120 µm, using a scan size of 512 × 512 pixels, scan speed of 0.25, and 2.5X digital zoom. All image acquisition settings were kept identical within, and across, imaging experiments. Statistical Analysis Statistical significance was determined using GraphPad Prism 10. Multiple comparison analysis was carried out by ordinary one‐way analysis of variance (ANOVA) with Dunnett's multiple comparisons test. Pairwise comparisons were carried out by Student's unpaired t‐test. Kaplan–Meier survival plot analysis was carried out using the Gehan–Breslow–Wilcoxon method. Statistical significance was defined by p‐values < 0.05. Results Tg(exorh:YFP‐NTR2.0) Larvae Facilitate Ablation of Pineal Photoreceptor Cells With Very Low Doses of Metronidazole To genetically ablate photoreceptor cells within the pineal gland we generated the Tg(exorh:YFP‐NTR2.0);Tg(lyz:DsRED2) line that directs expression of a rationally‐engineered nitroreductase (NTR) enzyme variant, called NTR2.0, to the pineal gland using the pineal gland‐specific exorh promoter (Figure 1A,B) [32, 34]. Similar transgenic lines expressing NTR2.0 in specific cell compartments, to sensitize those cells to metronidazole (MTZ)‐mediated ablation, have been used to efficiently ablate cells with MTZ doses 100‐fold less than conventional NTR [34, 35]. Metronidazole‐mediated ablation of pineal gland photoreceptor cells in Tg(exorh:YFP‐NTR2.0) larvae. (A) Schematic illustrating the exorh:YFP‐NTR2.0 transgenic construct used to generate the Tg(exorh:YFP‐NTR2.0);Tg(lyz:DsRED2) line. (B) Live imaging of Tg(exorh:YFP‐NTR2.0);Tg(lyz:DsRED2) larvae at 2 dpf (lateral view, anterior to left). Arrow marks YFP‐NTR2.0‐expressing pineal gland. (C) Schematic illustrating conditions used for experiment in (D). (D) Percent of Tg(exorh:YFP‐NTR2.0);Tg(lyz:DsRED2) larvae without pineal gland photoreceptor cells following treatment with 1, 10, 50, 100, 200, 300 μM doses of metronidazole (MTZ) and DMSO (control) and 100 μM MTZ‐treated Tg(exorh:EGFP) controls (n = 25 larvae/treatment). (E) Live imaging of Tg(exorh:YFP‐NTR2.0);Tg(lyz:DsRED2) larvae at 52 hpf, following 100 μM MTZ treatment (dorsal views, anterior to left). Numbers in parentheses represent proportion of larvae with displayed phenotype. Scale bars, 250 μm in (B) and 125 μm in (E). ZT, zeitgeber time. To evaluate pineal photoreceptor ablation, Tg(exorh:YFP‐NTR2.0);Tg(lyz:DsRED2) embryos with YFP‐expressing pineal glands were first selected and then exposed to a range of MTZ doses (1–300 μM), from 32 hpf (ZT8, under normal LD condition). Pineal photoceptor cell ablation was then assessed 20 h later (52 hpf, ZT4), by live fluorescence microscopy (Figure 1C). This analysis revealed that MTZ doses from 100 μM effected cell ablation in all treated larvae, as assessed by the absence of YFP expression (Figure 1D,E). This contrasted with DMSO‐treated Tg(exorh:YFP‐NTR2.0);Tg(lyz:DsRED2) and 100 μM‐treated Tg(exorh:EGFP) control larvae that still possessed YFP‐ and EGFP‐marked pineal photoreceptor cells, respectively. In summary, here we describe a new transgenic line Tg(exorh:YFP‐NTR2.0) that facilitates ablation of photoreceptor cells within the pineal gland, with very low doses of MTZ. Metronidazole Treated Tg(exorh:YFP‐NTR2.0) Larvae Do Not Express aanat2 and Have Suppressed Melatonin Production The earliest circadian output signal in zebrafish is the rhythmic secretion of melatonin from the pineal gland [23, 42]. To confirm pineal photoreceptor cell ablation within MTZ‐treated Tg(exorh:YFP‐NTR2.0);Tg(lyz:DsRED2) larvae, we next explored expression of arylalkylamine‐N‐acetyltransferase (aanat2), that encodes a key pineal gland enzyme necessary for melatonin synthesis. Expression of aanat2 is known to demonstrate a robust circadian rhythm, with peak pineal gland expression coinciding with the dark phase [24, 42]. Examining aanat2 expression within age‐matched larvae exposed to the normal LD condition, or a reversed DL condition, revealed that DMSO‐treated Tg(exorh:YFP‐NTR2.0) larvae expressed high levels of aanat2 within the pineal gland under the DL condition, when assessed during the dark phase at 52 hpf/ZT16 (Figure 2A,B). This contrasted with DMSO‐treated Tg(exorh:YFP‐NTR2.0) larvae that lacked aanat2 expression during the light phase, LD condition at 52 hpf/ZT4 (Figure 2B). Confirming pineal photoreceptor cell ablation, aanat2 expression was not detected within Tg(exorh:YFP‐NTR2.0) larvae at ZT16 under the DL condition when treated with 100 μM of MTZ from 32 hpf (Figure 2B). Of note, similar MTZ treatment of WT embryos had no impact on aanat2 expression. Melatonin levels within DMSO‐treated Tg(exorh:YFP‐NTR2.0) larvae at 76 hpf/ZT16 under the DL condition were 17.0 ± 2.5 pg/mL (mean ± SD). Consistent with the absence of aanat2 expression, melatonin production was significantly reduced within MTZ‐treated Tg(exorh:YFP‐NTR2.0) larvae, with levels below the detection limit for the assay (Figure 2C,D). Tg(exorh:YFP‐NTR2.0) larvae lack pineal gland aanat2 expression and melatonin production following metronidazole‐mediated ablation of pineal gland photoreceptor cells. (A) Schematic illustrating conditions used for experiment in (B). (B) Expression of aanat2 in Tg(exorh:YFP‐NTR2.0) and WT larvae following 100 μM metronidazole (MTZ) or DMSO (control) treatments under standard LD or reversed DL conditions at 52 hpf (dorsal views, anterior top). Arrow marks aanat2‐expressing pineal gland. Numbers in parentheses represent proportion of larvae with displayed phenotype. (C) Schematic illustrating conditions used when measuring melatonin levels of DMSO‐ and 100 μM MTZ‐treated Tg(exorh:YFP‐NTR2.0) larvae at 76 hpf (ZT16 under DL condition). (D) Melatonin levels for larvae under conditions shown in (C), data collected from two independent experiments (n = 100 larvae/experiment). Error bars, mean ± SD; *p < 0.05; unpaired Student's t‐test. Scale bar 125 μm. ZT, zeitgeber time. These data show that treatment of Tg(exorh:YFP‐NTR2.0) larvae with 100 μM MTZ results in ablation of pineal gland photoreceptor cells, as evidenced by absence of aanat2 expression and melatonin production. bsx Crispants Do Not Develop a Functional Pineal Gland Consistent with its critical role in pineal gland development, genetic ablation of bsx results in the complete absence of a functional pineal gland [43, 44]. As an alternative approach to generate pineal gland‐deficient larvae, we generated crispants (first‐generation [F0] mutants [36]) for bsx. We first assessed two independent sgRNAs (bsx sgRNA 1 and 2) that target exon 1 of the bsx gene to investigate which edited their target sites with the greatest efficiency, when co‐injected with Cas9 protein into one cell‐stage embryos. Both sgRNAs directed CRISPR/Cas9 editing to their target sites, as evidenced by the T7 endonuclease I assay (Supporting Information Figure S1A). Next‐generation sequencing of 2 dpf crispants revealed that bsx sgRNA 1 had the greatest editing efficiency, with 83.59% modified reads, compared to 72.93% for bsx sgRNA 2 (Figure 3A,B,D and Supporting Information Figure S1B,D). In contrast, bsx sgRNA 1‐ and 2‐only injected larvae had no editing at the target sites (Figure 3C and Supporting Information Figure S1C). Pineal glands were not detected within 2 dpf bsx sgRNA1 + Cas9 injected Tg(exorh:EGFP);Tg(lyz:DsRED2) larvae (hereafter referred to as bsx crispants), compared to tyrosinase (tyr) control crispants, a gene required for pigment development and previously used as controls for crispant experiments investigating immune responses (Figure 3E–G) [45, 46]. bsx crispants lack pineal gland development. (A) Schematic illustrating bsx‐targeting sgRNA1 and strategy to generate bsx crispants (sgRNA1 target in red, PAM site in blue). (B and C) Frequencies of unmodified and modified alleles, as detected by NGS of the bsx sgRNA1 target site, from 2 dpf bsx sgRNA1 + Cas9‐injected (B) and bsx sgRNA1 only‐injected (C) larvae. (D) Visualization of the 20 most common modified alleles as detected in (B) (insertions and deletions are shown by red rectangles and dashed lines, respectively). The predicted cleavage site is marked by the vertical dashed line. (E) Live imaging of Tg(exorh:EGFP);Tg(lyz:DsRED2) bsx and tyr (control) crispants at 2 dpf (dorsal views, anterior top). Arrow marks EGFP‐expressing pineal gland. Numbers in parentheses represent proportion of larvae with displayed phenotype. (F and G) Frequencies of unmodified and modified alleles, as detected by NGS of the tyr sgRNA target site, from 2 dpf tyr sgRNA + Cas9‐injected (F) and tyr sgRNA only‐injected (G) larvae. Scale bar 125 μm. Similar to MTZ‐treated Tg(exorh:YFP‐NTR2.0) larvae, examining aanat2 expression within bsx crispants under LD and DL conditions at 52 hpf revealed that expression of aanat2 observed in the pineal glands of non‐injected and tyr crispants at ZT16 (under the DL condition) was absent within bsx crispants (Figure 4A,B). bsx crispants lack pineal gland aanat2 expression. (A) Schematic illustrating conditions used for experiment in (B). (B) Expression of aanat2 in bsx and tyr crispants and non‐injected larvae under standard LD or reversed DL conditions at 52 hpf (dorsal views, anterior top). Arrow marks aanat2‐expressing pineal gland. Numbers in parentheses represent proportion of larvae with displayed phenotype. Scale bar 125 μm. ZT, zeitgeber time. Together, these data confirm that bsx crispants, like MTZ‐treated Tg(exorh:YFP‐NTR2.0) larvae, lack a functional pineal gland, as evidenced by aanat2 expression. The Pineal Gland Promotes Neutrophil Recruitment and Survival When Infections Are Delivered During the Dark Phase In our previous work, we have shown that light positively influences the larval zebrafish innate immune response to bacterial infection, with improved survival and neutrophil recruitment, compared to when bacterial challenges are delivered during the dark phase [12, 24]. Here, we used the same infection protocol (1000 CFU of Sal‐GFP injected into the hindbrain ventricle at 52 hpf) to examine whether the light‐detecting pineal gland contributes to post‐infection survival and neutrophil recruitment. We treated Tg(exorh:YFP‐NTR2.0);Tg(lyz:DsRED2) larvae and Tg(lyz:DsRED2) controls with 100 μM MTZ at 36 hpf/ZT12. Following confirmation of pineal photoreceptor cell ablation by fluorescence microscopy at 52 hpf/ZT4, larvae were infected and survival was followed thereafter for 4 days, under the LD condition (Figure 5A). This analysis revealed no significant difference in the survival of pineal gland‐deficient larvae when infected during the light phase and maintained under standard LD lighting conditions, when compared to controls (Figure 5B). Of note, MTZ treatment had no impact on bacterial growth, when grown in culture (Supporting Information Figure S2). Next, we quantified neutrophil numbers at the hindbrain infection site at 3 h post‐infection (hpi), when numbers of recruited neutrophils typically peak in this infection model [12, 24]. Neutrophil abundance was not significantly different between MTZ‐treated Tg(exorh:YFP‐NTR2.0);Tg(lyz:DsRED2) and Tg(lyz:DsRED2) larvae (Figure 5C and Supporting Information Figure S3A). Similar survival analysis and neutrophil recruitment quantification within bsx crispants also revealed no impact of pineal gland ablation under the LD condition, when compared to tyr crispants and non‐injected control larvae (Figure 5D–F and Supporting Information Figure S3B). Ablation of the pineal gland does not impact larval survival and neutrophil recruitment when infected during the light phase. (A) Schematic illustrating conditions used for infection experiments in (B) and (C). (B) Kaplan–Meier graph of metronidazole (MTZ)‐treated Tg(exorh:YFP‐NTR2.0);Tg(lyz:DsRED2) and Tg(lyz:DsRED2) larvae following infection, up to 4 dpi under LD condition (infected at 52 hpf/ZT4). (C) Quantification of neutrophils in the hindbrain of infected MTZ‐treated Tg(lyz:DsRED2) and Tg(exorh:YFP‐NTR2.0);Tg(lyz:DsRED2) larvae at 3 hpi. (D) Schematic illustrating conditions used for infection experiments in (E) and (F). (E) Kaplan–Meier graph of bsx and tyr crispants and non‐injected Tg(exorh:EGFP);Tg(lyz:DsRED2) larvae following infection, up to 4 dpi under LD condition (infected at 52 hpf/ZT4). (F) Quantification of neutrophils in the hindbrain of infected non‐injected, bsx crispant and tyr crispant Tg(exorh:EGFP);Tg(lyz:DsRED2) larvae at 3 hpi. Data in (B) and (E) collected from three independent experiments/sample. Error bars, mean ± SD; ns, not significant; Gehan–Breslow–Wilcoxon test (B and E), unpaired Student's t‐test (C) and one‐way ANOVA with Dunnett's multiple comparisons test (F). ZT, zeitgeber time. Given that melatonin levels peak during the dark phase, we reasoned that any impact of pineal gland ablation and/or melatonin production may best be observed when infections are delivered during the dark phase. To address this, we repeated the infections on MTZ‐treated Tg(exorh:YFP‐NTR2.0);Tg(lyz:DsRED2) and Tg(lyz:DsRED2) larvae under the DL condition, with infections delivered at ZT16 (Figure 6A). These experiments revealed that larvae lacking pineal photoreceptor cells were more sensitive to infections, with a 16.5% reduction in survival (Figure 6B), and a 56.4% decrease in mean neutrophil numbers at the infection site (Figure 6C and Supporting Information Figure S4A), when compared to controls. Repeating infections within similarly treated bsx crispants, also revealed that pineal gland‐deficient larvae were more susceptible to infections (Figure 6D,E), when compared to non‐injected (18.3% decrease in survival) and tyr crispants (19.0% decrease in survival), and had reduced neutrophil recruitment (Figure 6F and Supporting Information Figure S4B), with 60.0% and 63.5% reductions in mean neutrophil numbers, when compared to non‐injected and tyr cispants, respectively. Ablation of the pineal gland impacts larval survival and neutrophil recruitment when infected during the dark phase. (A) Schematic illustrating conditions used for infection experiments in (B) and (C). (B) Kaplan–Meier graph of metronidazole (MTZ)‐treated Tg(exorh:YFP‐NTR2.0);Tg(lyz:DsRED2) and Tg(lyz:DsRED2) larvae following infection, up to 4 dpi under DL condition (infected at 52 hpf/ZT16). (C) Quantification of neutrophils in the hindbrain of infected MTZ‐treated Tg(lyz:DsRED2) and Tg(exorh:YFP‐NTR2.0);Tg(lyz:DsRED2) larvae at 3 hpi. (D) Schematic illustrating conditions used for infection experiments in (E) and (F). (E) Kaplan–Meier graph of bsx and tyr crispants and non‐injected Tg(exorh:EGFP);Tg(lyz:DsRED2) larvae following infection, up to 4 dpi under DL condition (infected at 52 hpf/ZT16). (F) Quantification of neutrophils in the hindbrain of infected non‐injected, bsx crispant and tyr crispant Tg(exorh:EGFP);Tg(lyz:DsRED2) larvae at 3 hpi. Data in (B) and (E) collected from three independent experiments/sample. Error bars, mean ± SD; ns, not significant, *p < 0.05, ****p < 0.0001; Gehan–Breslow–Wilcoxon test (B and E), unpaired Student's t‐test (C) and one‐way ANOVA with Dunnett's multiple comparisons test (F). ZT, zeitgeber time. These data show that while the pineal gland does not influence neutrophil recruitment or survival outcomes when infections are delivered during the light phase, it is required for antibacterial immunity to infections delivered in the dark. Collectively, these findings describe a previously unappreciated role for the pineal gland in supporting nighttime antibacterial immunity in larval zebrafish. Discussion Here we explored whether the larval zebrafish pineal gland contributes toward the host response to infection. Using two alternative approaches, we generated pineal gland‐deficient larvae, as evidenced by absent expression of pineal gland marking genes and melatonin production, to assess neutrophil migration toward, and survival following, bacterial infection. Our findings uncovered a role for the pineal gland in promoting nighttime recruitment of neutrophils to infection sites, thereby supporting nighttime antibacterial immunity. Although we did not specifically investigate a role for melatonin in this study, we speculate that this function of the pineal gland is, at least in part, dependent upon elevated melatonin production during the night. Melatonin influences target cells through two highly conserved G protein‐coupled receptors, MT1 and MT2. Through both receptor‐dependent and ‐independent mechanisms, melatonin has been shown to modulate immune responses against viral, parasitic and bacterial infections [6]. Furthermore, melatonin has been used to treat experimental and clinical sepsis and protects against septic organ injury [7]. Melatonin supports both pro‐inflammatory and anti‐inflammatory immune responses, with pineal gland‐derived melatonin promoting innate immune responses through its receptors on immune cells, while extra‐pineal gland melatonin dampens immune responses [6]. Being the most abundant circulating leukocyte, and providing the bulk of cellular antibacterial defense, neutrophils play a critical role in protecting the host from bacterial infections. As first responders to bacterial challenge, neutrophils extravasate from local vasculature and navigate to infection foci using chemotactic signals [47, 48]. Upon encountering bacteria, neutrophils employ their antibacterial effector functions (including phagocytosis, ROS production, neutrophil extracellular trap [NET] release, and degranulation) to kill invading microbes [49]. Melatonin has been shown to promote neutrophil chemotaxis in rodents and humans [8]. Intraperitoneal injection of melatonin into rats increases the number of peritoneal leukocytes, while neutrophils isolated from humans treated with melatonin have enhanced chemotactic responses and elevated expression of the leukocyte‐attracting chemokines MIP‐1, MCP‐1, and IL‐8 [8]. A conserved role for melatonin in promoting neutrophil recruitment has been demonstrated in zebrafish [26, 27]. Exploiting a larval tail fin injury model of sterile inflammation, neutrophils were shown to have circadian rhythmicity in their recruitment, with peak numbers coinciding with the dark phase. Furthermore, exogenously supplied melatonin enhanced neutrophil abundance at the wound site, but not in the presence of luzindole, a melatonin receptor antagonist [26]. Examining neutrophil recruitment in aanat2 −/− larvae using the same injury model confirmed that endogenous melatonin was necessary for this rhythmic neutrophil recruitment, most likely through aanat2‐dependent expression of il1b and cxcl8 (that encode the zebrafish orthologs of IL‐1β and CXCL8/IL‐8) [27]. Our work here is the first to examine a role for the zebrafish pineal gland during the host response to infection. We have previously shown that neutrophil recruitment to bacterial infection peaks during the light phase, a phenotype that coincides with maximum protection to bacterial infections [24]. Our more recent work has shown that the positive impact of light on neutrophil recruitment (and survival) is, at least in part, dependent upon the clock gene per2 [12]. Collectively, these data suggest that the circadian rhythmicity in neutrophil recruitment toward sterile and infectious triggers of inflammation in larval zebrafish is antiphasic, with peak daytime antibacterial recruitment coinciding with the nadir for sterile wounds. It also suggests that while fewer numbers of neutrophils are recruited to bacterial challenges at night, those that do contribute to nighttime immune responses require an intact pineal gland. Here we also introduce a new tool to efficiently ablate pineal gland photoreceptor cells with very low doses of MTZ. The Tg(exorh:YFP‐NTR2.0) line generated for this study exploits the recently described rationally engineered NTR variant NfsB_Vv F70A/F108Y (NTR2.0). NTR2.0 circumvents the need for near‐toxic MTZ treatment doses with at least 100‐fold improvement in MTZ‐mediated cell ablation efficacy [34]. We and others have exploited the NTR2.0 genetic ablation system to ablate specific cell compartments, including macrophages, neutrophils, retinal rod photoreceptor cells and pancreatic β cells [34, 50, 51]. Previous versions of NTR necessitated high, near‐toxic, concentrations of MTZ (10 mM), precluding prolonged exposure for sustained cell loss [51, 52]. The improved efficacy of NTR2.0 enables sustained cell ablation through continuous exposure of larvae to low doses of MTZ [34, 50]. Here, we used a 100 μM dose of MTZ that enabled sustained ablation of pineal photoreceptor cells over several days. This continuous ablation of exorh‐expressing photoreceptive cells enabled us to examine the survival of pineal gland‐deficient larvae over several days, following infection. Of note, targeted cell ablation of the pineal gland has been previously performed by directing expression of conventional NTR using the aanat2 promoter (Tg(aanat2:CFP‐NTR)) [53]. In this study, two 20‐h treatments of 15 mM MTZ from 2 to 5 dpf were needed, with a recovery period between doses, for optimal ablation of pineal gland cells [53]. Two further NTR (NfsB) lines have been generated to ablate pineal gland photoreceptive cells, Tg(tph2:NfsB‐mCherry) and Tg(aanat2:Gal4);Tg(UAS:NfsB‐mCherry), using three separate 12‐h 10 mM MTZ treatments from 4 to 6 dpf [54]. Much of our understanding about how the pineal gland and/or melatonin influences the immune system has come from studies using nocturnal rodent models of inflammation. Here, using a well‐established diurnal zebrafish model of bacterial infection, we show that daytime antibacterial defenses can be maintained independently of the pineal gland. In contrast, at night, the pineal gland plays an important role in promoting the neutrophil response to bacterial challenge and maintaining nighttime immunity. To our knowledge, this is the first report showing that the pineal gland helps to maintain protection against nighttime infections. Low melatonin levels, for example in the elderly and people with chronic insomnia or naturally disrupted circadian rhythms (e.g., night shift work), are associated with a higher risk of infections and poor prognosis for inflammatory conditions [55, 56]. Given that neutrophils provide the bulk of cellular antibacterial defense, a better understanding of how the pineal gland/melatonin regulates nighttime neutrophil activity may reveal new therapeutic possibilities to treat infectious disease. Author Contributions Concept and design: Christopher J. Hall, Guy R. Warman, James F. Cheeseman, and Jonathan W. Astin. Experiments, data processing, and data analysis: Gabriel Anasarias‐Dumo, Yevetta Y. Xiang, David Cumim, and Christopher J. Hall. Drafting the manuscript: Gabriel Anasarias‐Dumo, Yevetta Y. Xiang, and Christopher J. Hall. All authors read, revised, and approved the final manuscript. Conflicts of Interest The authors declare no conflicts of interest. Supporting information Figure S1: T7 endonuclease I assay for bsx sgRNAs 1 and 2 and editing efficiency for bsx sgRNA2. (A) Schematic illustrating bsx gene with target sites for bsx sgRNA1 and 2 (sgRNA target in red, PAM site in blue), location of primer pairs for T7 endonuclease I assay (Fwd1/Rev1) and T7 endonuclease I assay (with and without T7 endonuclease I digestion) from pools of 10 2 dpf larvae following injection with individual sgRNAs + Cas9 protein and non‐injected controls (arrows and arrowheads mark 731 bp WT amplicons and cleaved products indicative of CRISPR‐Cas9 editing, respectively). (B and C) Frequencies of unmodified and modified alleles, as detected by NGS of the bsx sgRNA2 target site, from 2 dpf bsx sgRNA2 + Cas9‐injected (B) and bsx sgRNA2 only‐injected (C) larvae. (D) Visualization of the twenty most common modified alleles as detected in B (insertions and deletions are shown by red rectangles and dashed lines, respectively). The predicted cleavage site is marked by the vertical dashed line. Figure S2: Metronidazole treatment has no impact on bacterial growth. CFU measurements of Sal‐GFP in LB broth grown under LD conditions over a 24‐hour period in the presence of 100 μM MTZ, compared to no treatment and DMSO controls. Error bars, mean ± SD; ns, not significant; one‐way ANOVA with Dunnett's multiple comparisons test. Figure S3: Immunofluorescence detection of neutrophils recruited to the hindbrain infection site in larvae lacking a functional pineal gland under LD condition. (A) Immunofluorescence detection of neutrophils within the hindbrain (dorsal views, anterior to left) of MTZ‐treated Tg(lyz:DsRED2) and Tg(exorh:YFP‐NTR2.0);Tg(lyz:DsRED2) larvae infected under LD condition (at 3 hpi). (B) Immunofluorescence detection of neutrophils within the hindbrain (dorsal views, anterior to left) of bsx and tyr crispant Tg(exorh:EGFP);Tg(lyz:DsRED2) larvae infected under the LD condition, compared to non‐injected larvae (at 3 hpi). Scale bar 100 μM. Figure S4: Immunofluorescence detection of neutrophils recruited to the hindbrain infection site in larvae lacking a functional pineal gland under DL condition. (A) Immunofluorescence detection of neutrophils within the hindbrain (dorsal views, anterior to left) of MTZ‐treated Tg(lyz:DsRED2) and Tg(exorh:YFP‐NTR2.0);Tg(lyz:DsRED2) larvae infected under DL condition (at 3 hpi). (B) Immunofluorescence detection of neutrophils within the hindbrain (dorsal views, anterior to left) of bsx and tyr crispant Tg(exorh:EGFP);Tg(lyz:DsRED2) larvae infected under the DL condition, compared to non‐injected larvae (at 3 hpi). Scale bar 100 μM.