Organohalogens Naturally Biosynthesized in Marine Environments and Produced as Disinfection Byproducts Alter Sarco/Endoplasmic Reticulum Ca²⁺ Dynamics

Abstract

Contemporary sources of organohalogens produced as disinfection byproducts (DBPs) are receiving considerable attention as emerging pollutants because of their abundance, persistence, and potential to structurally mimic natural organohalogens produced by bacteria that serve signaling or toxicological functions in marine environments. Here, we tested 34 organohalogens from anthropogenic and marine sources to identify compounds active toward ryanodine receptor (RyR1), known toxicological targets of non-dioxinlike polychlorinated biphenyls (PCBs) and polybrominated diphenyl ethers (PBDEs). [³H]Ryanodine ([³H]Ry) binding screening (≤2 μM) identified 10 highly active organohalogens. Further analysis indicated that 2,3-dibromoindole (14), tetrabromopyrrole (31), and 2,3,5-tribromopyrrole (34) at 10 μM were the most efficacious at enhancing [³H]Ry binding. Interestingly, these congeners also inhibited microsomal sarcoplasmic/endoplasmic reticulum (SR/ER) Ca²⁺ ATPase (SERCA1a). Dual SERCA1a inhibition and RyR1 activation triggered Ca²⁺ efflux from microsomal vesicles with initial rates rank ordered 31 > 34 > 14. Hexabromobipyrroles (25) enhanced [³H]Ry binding moderately with strong SERCA1a inhibition, whereas pyrrole (24), 2,3,4-tribromopyrrole (26), and ethyl-4-bromopyrrole-2-carboxylate (27) were inactive. Of three PBDE derivatives of marine origin active in the [³H]Ry assay, 4′-hydroxy-2,3′,4,5′,6-pentabromodiphenyl ether (18) was also a highly potent SERCA1a inhibitor. Molecular targets of marine organohalogens that are also DBPs of emerging environmental concern are likely to contribute to their toxicity.

GRAPHICAL ABSTRACT

■. INTRODUCTION

Halogenated organic compounds (HOCs) are pervasive in natural and built environments.^1–10 Despite restrictions and banning in most parts of the world through the Stockholm Convention on Persistent Organic Pollutants (POPs), levels of well-studied “legacy” compounds, including polychlorinated biphenyls (PCBs) and polybrominated diphenyl ethers (PBDEs), are routinely detected in human and wildlife tissues where they continue to pose significant health risks to highly exposed and susceptible populations.^11–17 In addition to anthropogenic sources, more than 4,500 HOCs are synthesized by microbes and algae,^1,18 and many of these natural products strongly resemble banned anthropogenic chemicals or their metabolites. For example, marine bacteria express enzymes that synthesize OH-PBDEs and their derivatives.^1,19–22 Like their anthropogenic counterparts, marine HOCs accumulate in marine mammals, Antarctic air, seabird eggs, and human breast milk, possibly adding to adverse health risks.^23–27 Recently, the biosynthetic pathways for a number of brominated pyrroles, bipyrroles and phenols have been also elucidated.^1,19,21,28 Tetrabromopyrrole (31) isolated from Pseudoalteromonas bacteria has been shown to elicit specific development cues to coral larvae necessary for metamorphosis and benthic attachment of Orbicella franksi and Acropora palmata onto crustose coralline algae.^29,30 Remarkably, high levels of several halopyrroles, including tetrabromopyrrole (31), are produced as disinfection byproducts (DBPs) during treatment of drinking water, agricultural irrigation water and saline wastewater, and are considered as emerging pollutants because of their cytotoxicity, genotoxicity, carcinogenicity and developmental toxicity.^31–35 For example, 31 is a prominent byproduct of chlorination of seawater for sewage-treatment in coastal areas and exhibits potent developmental toxicity to Platynereis dumerilii embryos.³⁴ Another DBP first identified in Israeli drinking water, 2,3,5-tribromopyrrole (34), is produced during sewage treatment protocols utilizing both chlorine and seawater that are common in coastal communities, and is genotoxic and cytotoxic to CHO (Chinese hamster ovary) cells.^31,32

Legacy HOCs are known to disrupt a number of biological processes that mediate toxicological sequelae across life stages. Two well studied direct molecular targets include (1) the aryl hydrocarbon hydroxylase receptor (AhR) modulated by coplanar dioxin-like HOCs that alter expression of AhR responsive genes and (2) ryanodine receptors (RyRs), anchored within sarcoplasmic/endoplasmic (SR/ER) membranes of most invertebrate and vertebrate cell types, which bind noncoplanar (nondioxin-like; NDL) HOCs and alter Ca²⁺ homeostasis. RyRs, along with genetically related inositol-1,3,5trisphosphate receptors (IP₃R) affect the release of Ca²⁺ from SR/ER stores in response to physiological, pharmacological and environmental cues, and thus are critical components mediating a myriad of Ca²⁺-regulated cellular processes.^16,36,37 Ca²⁺ release into the cytoplasm is countered by highly active energy-dependent reaccumulation of Ca²⁺ into the microsomal lumen by the SR/ER Ca²⁺-ATPase (SERCA) pumps.

We observed that many of the natural HOCs identified to date are NDL structures that would be predicted to have activity toward RyRs and could disrupt Ca²⁺ dynamics and Ca²⁺-dependent signaling pathways.¹⁶ Experiments were performed to define the structure−activity relationship of 34 HOCs using [³H]Ry binding analysis. Ten natural/anthropogenic halogenated pyrroles, maleimides, bipyrroles, indoles, and diphenyl ether are identified as potent direct modulators of RyR1. Moreover, several of these RyR-active HOCs are also potent SERCA1a inhibitors. These two molecular actions result in rapid Ca²⁺ efflux from microsomal vesicles consistent with the opposing influences of RyR1 and SERCA1a on SR/ER Ca²⁺ balance. These results are the first to identify molecular targets of marine organohalogens that are also DBPs of emerging environmental concern and are likely to contribute to their cellular neurotoxicity reported (unpublished work).

■. MATERIALS AND METHODS

Chemicals and Reagents.

Organohalogens were custom synthesized as described in the Supporting Information (SI) or as described previously,^21,28 isolated from marine sponges³⁸ or purchased [pyrrole (24), ethyl-4-bromopyrrole-2-carboxylate (27) and 2,3-dibromo-N-methylmaleimide (28)] from SigmaAldrich (St. Louis, MO). Names and structures for the compounds tested are summarized in Table SI1. All the HOCs were dissolved in dry DMSO (10 mM) and stored at −80 °C. Phenylmethylsulfonyl fluoride (PMSF), leupeptin hydrochloride, Mg-ATP, phosphocreatine, creatine phosphokinase, phosphoenolpyruvate, pyruvate kinase/lactic dehydrogenase, Na₂ATP, NADH were purchased from Sigma-Aldrich. Arsenazo III was purchased from Santa Cruz Biotechnology (Dallas, TE). Ryanodine was purchased from Ascent Scientific (Cambridge, MA) and dissolved in 100% ethanol and stored at 4 °C until use. [³H]Ryanodine (specific activity: 56.6 Ci/ mmol) was obtained from PerkinElmer (Santa Clara, CA).

RyR1-Enriched Microsomal Preparations.

Microsomal membranes enriched in RyR1 and SERCA1a from the junctional region (JSR) were isolated from fast-twitch skeletal muscles of 1 year old male (2−2.5 kg) New Zealand White rabbits as detailed in SI and previously described.³⁹ Two independent microsomal preparations were used to conduct and replicate all assays. Protein concentrations were measured with the BCA protocol (Fisher, Waltham, MA).

Equilibrium [³H]Ry Binding Analysis.

[³H]Ry (56.6 Ci/ mmole) binding analysis was performed as previously described.⁴⁰ Briefly, 50 μg protein and 1 nM [³H]Ry were utilized in the presence of vehicle or 0.05−20 μM of each HOC, and incubated with 500 μL of binding buffer at 37 °C for 3 h as described in detail in SI. At least 2 independent experiments were conducted for each [³H]Ry binding analysis, each performed in triplicate or quintuplicate.

Measurements of SERCA1a Activity.

Activity of JSR Ca²⁺ ATPase was measured using a coupled enzyme assay monitoring the oxidation rate of NADH at 340 nm as described in SI and previous publication.⁴¹ Thapsigargin (TG), a specific blocker of SERCA, was used as a control to discriminate between SERCA-dependent ATPase and nonspecific ATPase (TG-insensitive) activities. At least two independent preparations were utilized to test each compound in triplicate.

Measurements of Microsomal Ca²⁺ Fluxes and Ca²⁺ -Induced Ca²⁺ Release.

The measurements of Ca²⁺ uptake and release from microsomal membrane vesicles were performed by monitoring absorbance using metallochromic Ca²⁺ dye Arsenazo III at 650−700 nm at 35 °C as detailed in SI, and similar to previous methods described with metallochromic Ca²⁺ dye Antipyrylazo III.⁴² Two different preparations were used for testing each compound in triplicate.

Data Analysis and Statistics.

Graphing and statistical analyses were performed using GraphPad Prism software (version 7.0; GraphPad Software Inc., San Diego, CA). Potency and efficacy values were determined by nonlinear regression with three-parameter equation and Bell-shaped equation. Statistical analysis of comparisons between the means of each group and means across all other groups, One-way ANOVA followed by post hoc Tukey’s test was used. For comparisons made between mean of each test compound and the mean of the control, Dunnett’s post hoc test was used to analyze the data at 95% confidence intervals.

RESULTS AND DISCUSSION

Structure−Activity Relationship for Marine HOCs toward RyR1.

The binding of Ry to multiple sites within the ion conduction pathway of RyRs results in complex influences on its channel gating behavior, stabilizing the open states at low (nM) and fully blocking the channel at higher (μM) concentrations.^39,43,44 Low nanomolar levels Ry binds preferentially to high-affinity sites only accessible when RyR channels are in the open state thus promoting SR/ER Ca²⁺ efflux (Ca²⁺ leak), which can have dramatic pathophysiological consequences.^45,46 Because of these properties, [³H]Ry binding analysis has been broadly used to identify ligands that interact with RyRs and understand their structure−function relationships, and serves as a rapid throughput screening tool. Of 34 HOCs initially tested (Table SI1), 10 halogenated pyrroles, bipyrroles, indoles, maleimides and diphenylethers (6, 13, 14, 17, 18, 19, 25, 28, 31, and 34) were identified to enhance [³H]Ry binding to RyR1 ≥ 1.5-fold at 2 μM when compared to baseline (binding measured in the presence of Veh alone, p < 0.05 vs Veh, Figure 1). These and an additional 4 HOCs (15, 24, 26, and 27) were selected for further study based on their structural similarities to establish an initial structure activity relationship (SAR) using [³H]Ry, Ca²⁺ ATPase, and Ca²⁺ flux measurements. On the basis of structural homologies, these compounds were classified into two major categories: (A) halogenated pyrroles, bipyrroles, maleimides, and indoles and (B) PBDE derivatives, as shown in Figure 2.

Figure 1.

Screening results of 34 organohalogens using [³H]Ry binding analysis. Compounds in red color represent significant effects toward RyR1 were found at 2 μM at a threshold of p < 0.05 compared with vehicle control using one-way ANOVA followed by Dunnett’s multiple comparisons test. Compounds in green color were those inactive toward RyR1, and they are selected because of their structure properties. Data shown in the figure were conducted from 2 to 3 different preparations with triplicates for each preparation and were normalized to 1% DMSO control (% Veh) and expressed as mean ± SD Table SI1 summarizes the chemical structures tested.

Figure 2.

Compounds selected for further evaluation for their activities toward RyR1, SERCA1a, and microsomal Ca²⁺ flux. (A) Structures of selected pyrroles, bipyrroles, maleimides and indoles. (B) Structures of PBDE derivatives of marine selected for further analysis in [³H]Ry binding, SERCA1 activity and Ca²⁺ assays. (C) Structure of ryanodine.

Active HOCs within group A enhanced the binding of [³H]Ry to RyR1 in a concentration-dependent manner and their relative efficacies at 10 μM ranked 14 > 31 > 13 = 34, whereas 24, 26, and 27 were inactive (Figure 3A, B).Interestingly 6 (2,3-dibromomaleimide), a naturally produced marine HOC, showed highly potent biphasic influences on RyR1, stabilizing the open state of RyR1 with an EC₅₀ of 0.36 μM, and inhibiting with an IC₅₀ of 3.0 μM (Figure 3A, C; table insert). Compound 6 was shown to form Michael adducts with protein Cys⁴⁷ and, therefore, raised the question whether its biphasic actions results from forming covalent adducts to hyper-reactive RyR1 Cys residues previously identified to confer redox-sensing properties and regulation to the RyR1 complex.^48–52 To better define the possible role of Cys-adducts in the biphasic response observed with 6, the concentrationeffect relationship was extended to 31 and 34, which are unable to form Cys-adducts, and commercially available 2, 3-dibromo-N-methylmaleimide (28) was also tested. Indeed, all three compounds produced biphasic modulation of RyR1 (Figure 3C), indicating that Michael addition to one or more reactive Cys was not required for the biphasic response. This conclusion was further supported by previous data indicating that 4′hydroxy-2,2′,4,5′-tetrabromodiphenyl ether (4-OH-BDE-49), also incapable of forming Michael adducts also exhibits biphasic actions toward RyR1, whereas BDE-49 activated in a monotonic manner.⁵³ Thus, biphasic activity was unlikely to be mediated by Cys modifications on RyR1, but more likely mediated by interaction with multiple binding sites on the homotetramer. 6-OH N-methylation of 6 decreased the apparent activation potency by 1.7-fold (p < 0.05) without altering inhibitory potency or maximum efficacy (Figure 3C, table insert). By comparison, 31 and 34 were 68- and 107-fold (p < 0.01) less potent activators, respectively, although they attained 2.5-fold higher maximum efficacy at 50 μM compared to that reached by 6 at 1 μM. Likewise 31 and 34 were weaker inhibitors than 6 (Figure 3C, table insert). Thus, not only is polybromination of the pyrrole essential for RyR1 activity, but also the 5-Br appeared to be especially important for optimizing activity. Hexabromobipyrrole is less active suggesting steric factors were important. Finally, we tested 2,3-dibromoindole (14) and its synthetic chlorinated analog 2,3-dichloroindole (13), and found that 14 is more active, indicating steric bulk and/or electron withdrawal from the indole ring were important for conferring activity toward RyR1.

Figure 3.

Halogenated pyrroles, bipyrroles, maleimides and indoles exhibit stringent structure−activity relationship toward RyR1. Panel A shows concentration−response curves expressed as specifically bound [³H]Ry to RyR1-enriched microsomal membranes (JSR) isolated as described in Materials and Methods. Data were fitted with three-parameter or Bell-shaped (compound 6) nonlinear regression with Graph Pad 7.03. Panel B shows stimulation relative to vehicle (1% DMSO) control (% of Veh; dashed line). Panel C shows results from an expanded concentration range to quantify biphasic parameters fitted as described in panel A for compounds 6, 28, and 31. The activation phase for 34 was fitted with a threeparameter equation but lacked sufficient data to fit the inhibition phase (dashed line). Parameters obtained from curve-fitting are summarized in inserted table. Data shown are from 2 to 3 independent preparations conducted in triplicates or quintuplicates and expressed as Mean ± SD. With Graph Pad 7.03, one-way ANOVA followed by Tukey’s multiple comparisons test was used to determine the significance among specific compounds. * p < 0.05 and ** p < 0.01.

A previous structure−activity study of anthropogenic PBDEs and OH-derivatives indicated that two ortho-Br substitutions were necessary for activating RyRs and that meta-Br further enhanced, whereas para-Br reduced, activity.⁵³ Of 13 naturally occurring PBDE derivatives identified in the marine environment and tested in the current study (Table SI1), only three were found to meet our criteria for efficacy toward RyR1 (i.e., significant enhancement of specific [³H]Ry binding at 2 μM, p < 0.05 vs 1% vehicle control (Figures 1 and 2, group B). Consistent with previous findings, compounds 18 and 19 have at least two ortho-Br substitutions with ranked efficacies at 10 μM of 18 > 19 (Figure 4). The importance of para-substituents was further underscored by the observation that para-OH is more active than para-methoxy (18 vs 19). With an additional ortho-hydroxy or ortho-methoxy, however, the 10 PBDE derivatives with a single ortho-Br that were tested, all failed to meet criteria for activity toward RyR1 (Table SI1 and Figure 1). Consistent with ortho-chlorination of NDL-PCBs,^40,53,54 monoortho-Br was generally less active than di-Br, although meta and para substitutions dictated the level of RyR1 activity. For example, para-methoxy, (17) was more active toward RyR1 than 15 with para-OH, further extending the structure−activity relationship for biosynthetic PBDEs.

Figure 4.

Activity of PBDE derivatives of marine origin toward [³H]Ry binding to RyR1. Panel A shows concentration−response curves for [³H]Rybinding to RyR1-enriched microsomal membranes (JSR) expressed as % of vehicle (Veh) control (1% DMSO). Panel B shows relative binding levels obtained at 0.5, 5, and 10 μM of test compound. Each color corresponds to the specific compound identified in the key shown in panel A. The dashed line indicates baseline binding in the presence of Veh (i.e., 100%). Data shown are from 2 to 3 independent preparations conducted in triplicates or quintuplicates and expressed as mean ± SD. With Graph Pad 7.03, one-way ANOVA, followed by Tukey’s multiple comparisons test was used to determine the significance among specific compounds. ** p < 0.01.

Collectively, these data identified for the first time that halogenated pyrroles, bipyrroles and indoles of marine origin were modulators of RyR1 and have a stringent structural requirement that dictated their potency and efficacy. It is interesting to note that the pyrrole moiety at 2-position of ryanodine was essential for not only conferring high-affinity binding to mammalian RyR1, but was also shown essential for conferring toxicity to mice,^44,55 suggesting that the HOCs identified here may represent reduced structures sufficient to occupy the essential pyrrole docking site(s) and confer the biphasic actions of ryanodine on multiple high- and low-affinity binding sites on the RyR1 homotetrameric complex.^45,56–58

Structure−Activity Relationship for Marine HOCs toward SERCA1a.

The efflux rate and amount of SR/ER Ca²⁺ released not only depends on the density and activities of RyR1 and IP₃R channels, but also is determined by the activity of ATP-dependent driven SERCA pumps necessary for reaccumulation of luminal Ca²⁺ and setting the ionic driving force that shapes Ca²⁺ transient properties. We therefore determined whether HOCs active toward RyR1 also influenced the activity of SERCA1a found in the same SR/ER membrane preparation tested for RyR1 activity.

Over 90% of the Ca²⁺-dependent ATPase activity measured in the microsomal preparations could be attributed to SERCA1a in that it was inhibited by TG (Figure 5). These studies were focused principally on the HOCs in Group A but also compared 18 in Group B. Compounds 18 and 25 were identified as the most potent inhibitors of SERCA1a activity with >80% inhibition at 2.0 μM (Figure 5D and F), and ~90% inhibition at 10 μM (Figure 7). At 10 μM compounds 31, 14, 6, and 34 showed intermediate SERCA1a inhibition (80, 60, 50, and 48%, respectively) (Figures 5B, 5E, SI1, and 5C), whereas 24, 26 and 27, which were inactive toward RyR1, were also inactive toward SERCA1a (Figure 5G).

Figure 5.

Inhibition of TG-sensitive SERCA1a Ca²⁺ ATPase activity by selected HOCs of marine origin. (A) Representative traces showing the oxidation of NADH in the coupled enzyme assay was monitored by measuring absorbance at 340−400 nm. Veh (0.1% DMSO), TG (thapsigargin, 20 μM) or a HOC was added into separate reactions cuvettes before addition of NADH and Na₂ATP to initiate the reactions. The oxidation rate of NADH in the presence of TG, Veh or test compounds were summarized in bar graph (B−G). Two different microsomal (JSR) preparations were used to calculate the summary data for each group, and each performed in triplicates or quintuplicates. Data shown as Mean ± SD and statistical differences assessed by one-way ANOVA, followed by Dunnett’s multiple comparisons test was performed using Graph Pad 7.03. * p < 0.05, **p < 0.01 vs Veh, ## p < 0.01 vs TG, and n.s. indicates no significant compared with Veh.

Figure 7.

3D trajectory plot showing the relationships among parameters calculated from [³H]Ry-binding, SERCA1a activity and Ca²⁺ release rates from microsomal (JSR) membranes. Each individually colored sphere represents data extracted from one compound at 10 μM that are presented in the Figures 3–6. Veh indicates DMSO control. See text for further explanation.

The degree and distribution of pyrrole bromination was an important factor for both RyR1 and SERCA1a activities, with 31 (tetrabromination) being most active, 34 being intermediate (2,3,5-bromination), and 26 (2, 3, 4-bromination) being inactive in both assays. These data indicate di-ortho-bromine substitution is an important contributor for both RyR1 and SERCA1a activity. This also extended to compound 25 (hexabromobipyrroles) that has two ortho-bromine substituents and possesses similar RyR1 activity as compound 34 (Figure 1). Compound 25 also exhibited the strongest SERCA inhibition (Figure 5D). As shown in the Figure SI3, compound 25 diminished the loading capacity of calcium into the microsomal vesicles in a concentration-dependent manner, an effect compounded by targeting both RyR1 activation and SERCA1a inhibition.

Interestingly, 0.6−1.0 μM compound 6, a dibromomaleimide maximally activated RyR1 (Figure 3) without inhibition of SERCA1a (Figure SI 1). In fact, the RyR1 inhibitory phase of 6 was detected at <5 μM, whereas SERCA1a inhibition was detected only at the highest concentration tested (10 μM) (Figure SI1).

Impact of RyR1 and SERCA1a Activities on Microsomal Ca²⁺ Dynamics.

The net uptake and release of Ca²⁺ across the SR/ER (microsomal) membrane compartments has been shown to depend on three principal factors: (1) activity of ATP-dependent SERCA pumps sequestering free cytoplasmic Ca²⁺ to a resting level of ~100 nM; (2) active release of luminal Ca²⁺ through mechanisms regulating RyRs and IP³Rs channel activities; and (3) passive Ca²⁺ leakage out of the microsomal compartment through mechanisms that remain unclear, but are likely to involve distinct leaks states of RyRs and IP³R, although additional mechanisms have been proposed.^59,60 We therefore set out to further evaluate how 9 HOCs that showed distinct patterns of RyR1 and SERCA1a activities influenced net Ca²⁺ flux across microsomal membranes (JSR) (Figure 6 and Figure SI2). Figure 6A depicts the Ca²⁺ transport assay that utilized the membrane-impermeant indicator Arsenazo III to record extra-vesicular Ca²⁺ in real-time. Flux of Ca²⁺ into (net uptake) or out of (net release) the vesicles depends on all three aforementioned factors. We therefore proceeded to test nine HOCs for their ability to trigger the release of Ca²⁺ accumulated after bolus addition of 4 × 45 nmole Ca²⁺ was completed during the loading phase of the experiment (Figure 6A, B). Since the Ca²⁺ driving force (i.e., the amount of Ca²⁺ loaded into the lumen of the vesicles) was kept consistent across experiments, the rates of release induced by HOCs could be directly compared.

Figure 6.

Microsomal Ca²⁺ release triggered by selected HOCs of marine origin. (A) Schematic illustration showing the Ca²⁺ transport system monitored with Ca²⁺ dye Arsenozo III. (B) Representative traces of Ca²⁺ fluxes. Microsomal (JSR) vesicles were loaded with 4 sequential additions of 45 nmole CaCl₂, 60−90 s after the final bolus of Ca²⁺ was accumulated into vesicles, vehicle (0.2% DMSO, a) or test compounds (b or c) was introduced into one of the cuvettes. Ruthenium red (RR, 2 μM), a RyR blocker, was then added to each cuvettes to confirm the engagement of RyR1 followed by addition of SERCA inhibitor, cyclopiazonic acid (CPA, 50 μM), and CaCl₂ were added to calibrate absorbance unit into absolute Ca²⁺ in nmol. The Ca²⁺ release rate of initial 60 s upon the addition of Veh or test compounds (5 or 10 μM) are summarized in Panels C−H. Data shown are from 3 to 6 independent experiments using 2 different microsomal membrane preparations. Data expressed as Mean ± SD and one-way ANOVA followed by Dunnett’s multiple comparisons test was performed using Graph Pad 7.03. *p < 0.05, **p < 0.01 vs Veh. And n.s. indicates no significant difference compared with Veh.

Results of the analyzed initial release rates from these flux experiments generally followed predictions from the relative activities of HOCs toward RyR1 and SERCA1a (Figure 6C-G). Specifically, compounds 24, 26, and 27 failed to trigger net release of accumulated Ca²⁺ from the microsomal vesicles (Figure 6H), consistent with their lack of activity toward both biochemical targets. By contrast, active HOCs triggered dose-dependent Ca²⁺ efflux with a rank order of 31 > 18 > 34 > 25 > 14. The initial rates of Ca²⁺ efflux triggered by these HOCs followed their respective activities toward RyR1 and SERCA1a with the exception of the dibrominated indole 14 whose activity was significantly under-predicted by the Ca²⁺ flux rates it produced at either 5 or 10 μM, concentrations anticipated to largely stabilize a RyR1 open channel state while inhibiting SERCA1a ~55% (Figure 7). The discrepancy with compound 14 could be explained if an extended amount of time is needed to fully manifest stabilization of the open channel state, which is afforded by the 3-h incubation in [³H]Ry binding assays but not the Ca²⁺ assay. Moreover, compound 14 may also inhibit RyR1 leak-states, which if blocked, can dramatically increase the Ca²⁺ loading capacity of the microsomal vesicles. Brominated macrocyclic bastadins isolated from the marine sponge Ianthella basta^61–66 and anthropogenic PCBs¹⁶ have been previously shown to influence the balance of channel and leak states of RyR1, although whether this mechanism explains the results obtained with compound 14 remains to be determined.

Considering the potent biphasic activity of compound 6 on RyR1 (Figure 3) and its modest inhibitory activity toward SERCA1a (Figure SI1), it was not unexpected that high concentrations (5 and 10 μM), concentrations expected to completely inhibit RyR1 channels, failed to trigger net Ca²⁺ efflux added to Ca²⁺-loaded microsomal vesicles (when the extravesicular free Ca²⁺ is very low; ~100 nM) (Figure SI2). However, an additional a large bolus (90 nmol) of Ca²⁺ to trigger Ca²⁺-induced Ca²⁺-release resulted in significantly higher rates of release than control (** p < 0.01, Figure SI2A and B). The sensitizing influences on calcium-induced calcium release (CICR) were observed in the presence of nanomolar 6 that should partially activate RyR1 without any SERCA1a inhibition, although their magnitude did not reach statistical significance (Figure SI2C and D).

The present results are significant because they identify several prominent naturally synthesized marine HOCs that directly target two major proteins whose functions work in opposition to tightly regulate net ER/SR Ca²⁺ dynamics, and thus, are essential for shaping localized and global Ca²⁺ signals that regulate a plethora of cellular functions.^16,67 From the perspective of chemical ecology, tetrabromopyrrole (31) has been recently isolated from Pseudoalteromonas bacteria and shown to serve specific habitat cues for the attachment and metamorphosis of benthic marine invertebrates, including coral Acropora millepora larvae and multiple Caribbean corals (Orbicella franksi and Acropora palmata) that settle onto crustose coralline algae.^29,30,68 The current findings may have widespread implications since motility, secretion and metamorphosis of planula are associated with the creation of crystalline substratum calcification mediated by precise Ca²⁺ signals.²⁹ HOCs, including tetrabromopyrrole and hexabromobipyrroles, extracted from Pseudoalteromonas bacteria that often associate with coralline algae-associated bacterium and Chromobacterium were reported to possess a spectrum of antibiotic, antifungal, antiprotozoal, antipathogen and antineoplastic activities.^69–73 Recently, natural sources of HOCs were shown to accumulate in marine organisms, especially apex predators such as seabirds and dolphins.^74–77

A second, equally significant implication is the recent detection of several halopyrroles in DBPs, including compounds 31 and 34, found in waste and drinking water treatment streams are cytotoxic, genotoxic and carcinogenic to mammalian cells.^31,32,34 The newly identified targets reported here and their neuroactive properties reported in the following paper (in preparation) provide important mechanistic clues about their potential disruptive influences on marine ecology and contributions to human health risks, especially given recent evidence that HOCs of biosynthetic and anthropogenic origins accumulate in marine mammals and human breast milk.24,27,74,75,77,78