Emerging Opportunities to Manipulate Metal Trafficking for Therapeutic Benefit

Abstract

The indispensable requirement for metals in life processes has led to the evolution of sophisticated mechanisms that allow organisms to maintain dynamic equilibria of these ions. This dynamic control of the level, speciation and availability of a variety of metal ions allows organisms to sustain biological processes while avoiding toxicity. When functioning properly, these mechanisms allow cells to return to their metal homeostatic setpoints following shifts in metal availability or other stressors. These periods of transition, when cells are in a state of flux in which they work to regain homeostasis, present windows of opportunity to pharmacologically manipulate targets associated with metal trafficking pathways in ways that could either facilitate return to homeostasis and recovery of cellular function, or further push cells outside of homeostasis and into cellular distress. The purpose of this Viewpoint is to highlight emerging opportunities for chemists and chemical biologists to develop compounds to manipulate metal trafficking processes for therapeutic benefit.

Graphical Abstract

Synopsis

Knowledge of metal trafficking processes can be leveraged to develop small molecules that help cells regain metal homeostasis for optimal cellular function, or conversely that push cells away from homeostasis and into cellular distress. This Viewpoint highlights lessons and opportunities for research in an area rich for therapeutic development at the interface of inorganic chemistry and cellular biology.

Introduction: Windows of Opportunity Beyond Metal Comfort Zones

It is estimated that approximately one-third of all proteins require a metal to function.^1–4 For processes carried out by these proteins to proceed appropriately, cells must orchestrate the acquisition, mobilization, and utilization of metal ions to ensure that the correct metal makes it to the correct place at the correct time, all while protecting against potential metal-induced damage. Maintaining a dynamic equilibrium of metal availability, i.e. metal homeostasis, is therefore critical for optimal cellular function, and indeed for life.

Adjustment of metal availability to achieve homeostasis occurs via the uptake, excretion, transport, or storage of metal ions, based on the needs of the cell. But how do cells “know” which of these metal trafficking functions to enact at any given time? Expression of proteins that carry out these functions are regulated in large part by metal-sensing transcription factors. Regardless of how much total metal is present, shifts in the “availability” of a particular metal ion that allow it to be sensed by its transcriptional regulator drive the cell to its homeostatic setpoint. If a metal is present in excess, cells return to homeostasis by excreting, using, or kinetically trapping that metal, and if the metal is limited, cells enact mechanisms to acquire it or labilize it from internal stores. These periods of transition are likely to occur during periods of rapid growth, in response to stress, or upon interaction with other cell types, for example at the host-pathogen interface in which the metallomes of each population of cells must adjust to the presence of the other.^5–12 The unique properties of metals mean that cells regulate their levels and speciation to different extents based on metal type, and likely by cell type as well. Recent studies in Salmonella rigorously defined the cellular metal buffering range set by a series of bacterial metal-sensing transcriptional regulators as a means to develop a quantitative model of metal homeostasis.¹³ Because these sensors are exquisitely specific, binding of their designated metal ion activates genes to express or repress proteins involved in that metal’s entry, expulsion, or sequestration. The cellular buffering capacity, or setpoint, varies greatly among metal ions, with Mn being most weakly buffered to 10⁻⁶ M, Zn more stringently to 10⁻¹² M, and Cu most tightly regulated to 10⁻¹⁸ M (Fig. 1).¹³ While these values were determined specifically in Salmonella, the results provide a useful framework for considering ramifications of stressors that push all cells, including eukaryotic cells, outside their metal comfort zones.¹³

Figure 1.

The cellular metal buffering capacity for Cu, Ni, Zn, Co, Fe, and Mn in Salmonella as set by the DNA occupancy of metal-sensing transcriptional regulators specific for each metal. Stressors that affect metal trafficking pathways and push cells outside of these buffering zones may have therapeutic potential. Figure adapted by permission from ref. 13 © 2019 Springer Nature.

These periods of transition in which cells are working to return to their homeostatic setpoints present windows of opportunity for manipulation of these systems. In some cases, it will be advantageous to prevent a return to homeostasis, while in other situations, it will be beneficial to facilitate return to homeostasis. Knowing how to do either, however, requires an understanding of metal trafficking pathways, how they intersect cellular function, and how they adapt to surrounding conditions.

The goal of this Viewpoint is to highlight emerging opportunities for chemists and chemical biologists to manipulate aspects of metal trafficking as strategies for potential therapeutic benefit.^{14, 15} Our discussions are framed in the context of identifying vulnerabilities created during transition periods in which cells work to return to metal homeostasis. Opportunities involve capitalizing on (or mitigating) vulnerabilities created when the cell is pushed outside of the homeostatic zone. When the system is in flux trying to achieve balance, what chemical perturbations could either insert a wedge or provide a boost to prevent or promote its recovery?

Developing medicinally-active inhibitors that target metalloenzymes directly is one powerful strategy to influence metal-dependent biological processes.^16–18 In addition to inhibiting the function of the target enzyme, the cell is likely to invoke metal trafficking pathways to repair the damage, potentially creating additional axes of vulnerability that could be exploited. For purposes of the current discussion, we focus our attention on alternative strategies to direct metalloprotein targeting that intervene during periods of transition before the metal reaches its final destination. While no means exhaustive, we highlight select examples of recent work in the areas of metabolism and the microbiome, pathogen and host metal adaptation processes, cancer proliferation, and other diseases where exciting recent advances in understanding metal trafficking offer new opportunities to apply principles of inorganic chemistry to affect biological outcomes.

Bioactive Small Molecules that Mobilize Metals

Metal ions can be toxic to cells through a variety of mechanisms. It is not surprising, then, that compounds that influence the cellular availability of metals have profound biological effects. As the curves in Fig. 1 show, Cu stands out as the most stringently regulated among the native biometals, being buffered in the Salmonella cytoplasm to 10⁻¹⁸ M. The fact that bacteria regulate detectable Cu to less than one “free” Cu atom per cell¹⁹ signals the acute danger that “free” Cu in particular presents to biological systems. The connotation “free Cu” here does not mean it is not coordinated by ligands; rather it means that whatever its speciation, it is kinetically-exchangeable and recognizable by its transcriptional regulatory machinery that directs acquisition, transport, and detoxification of this essential yet potentially toxic metal.^20–25 The cell may indeed have a significant amount of total Cu that it tolerates acceptably, but it clearly wants to keep the available levels exceedingly low. Given this sensitivity, compounds that mobilize Cu in particular are well-represented in the following subsections, although similar concepts hold for Zn, Fe, and other biometals.

Overriding Cu homeostatic mechanisms can have differentially desirable outcomes depending on the system: either Cu-potentiated toxicity or reversal of a Cu-related dysfunction (Fig. 2). For example, in harmful cell types, such as cancer and pathogenic microbes, it is advantageous to overwhelm the Cu buffering capacity of the cell to drive it toward cell death. On the other hand, in cells that are beneficial to humans, such as beneficial bacteria of the microbiome or host cells, the goal is to aid in the maintenance or restoration of Cu homeostasis. Direct metal supplementation and sequestration have been investigated as therapeutic strategies to accomplish these goals, but these approaches suffer from inefficient metal delivery and/or systemic toxicity.^26–28 Thus, there is a need to pharmacologically manipulate metals in these systems in a targeted way to avoid unwanted toxicity and increase efficacy.

Figure 2.

Cu-binding small molecules can bypass cellular Cu transport machinery. Possible impacts on Cu homeostasis include aiding recovery from Cu deficiency, restoring from Cu overload, or driving toward toxicity.

Using Small Molecules to Sabotage Cu Homeostasis.

To combat diseases such as cancer and those caused by infectious microbes, interfering with the ability of these cells to maintain Cu homeostasis is an attractive strategy for inducing cell death. For example, molecules that mobilize Cu specifically to target cell types such that they cannot recover homeostatic Cu balance can be potent antimicrobial or anticancer agents. A structure-function study of metal chelating agents with diverse chemical properties revealed that when ligands form neutral, lipophilic complexes of suitable affinity for Cu(II), these complexes induce Cu hyperaccumulation and cell death in Cryptococcus neoformans, the pathogenic fungal organism used in the study.²⁹ The metal-binding agents that were found to have the physical and chemical properties appropriate to induce cell death were also the same compounds that were capable of rescuing a growth defect of a C. neoformans strain lacking its Ctr1 and Ctr4 Cu import machinery.²⁹ This result is profound in its illustration that small molecules can provide chemical pathways to bypass biological metal trafficking processes.

The term ionophore is often used to distinguish small molecule metal-binding agents that mobilize and redistribute metal ions from those agents that sequester metals in high-affinity or kinetically trapped complexes that keep the metal biologically unavailable. It is important to keep in mind, however, that the mechanisms by which metal ionophores induce their biological effects are likely to be more diverse than a single term implies. Indeed, determining molecular mechanisms of how and why these compounds cause toxicity is an area of opportunity that could provide insight into how to tailor or direct them with specificity to desirable targets. For example, the small molecule pyrithione (PT, Chart 1), which binds divalent and trivalent metal ions, has been recognized as an antimicrobial Cu ionophore for some time.^{29, 32, 33} Yet, recent efforts to identify selective inhibitors of fungal Cu-only superoxide dismutase Sod5 uncovered PT as an inhibitor of this enzyme. The dual utility of this compound as a metalloenzyme inhibitor in addition to its delivery of toxic levels of intracellular Cu illustrates the notion that ionophores can have more than one mechanism contributing to their antimicrobial action.^{34, 35}

Chart 1.

Chemical structures of compounds discussed in the main text. PT, pyrithione; DTC, diethyldithiocarbamate; 8HQ, 8-hydroxyquinoline; TTM, tetrathiomolybdate; BCS, bathocuproine disulfonic acid; TPEN, N,N,N’,N’-tetrakis(2-pyridylmethyl)ethylenediamine; SnPPIX, tin protoporphyrin IX. GGTDTC and PcephPT are prochelators of DTC and PT, respectively.

†Many pyoverdines exist; the one produced by Pseudomonas aeruginosa discussed in the main text is shown.³⁰

‡Mycobacterium tuberculosis produces two mycobactins, hydrophobic mycobactin T (shown) and hydrophilic carboxymycobactin.³¹

Recognition that Cu-potentiated toxicity is an effective antimicrobial strategy has led to efforts to identify compounds that have biological activity specifically in the presence of Cu. For example, a high-throughput screen of existing small molecule chemical libraries recently identified 483 Cu-dependent inhibitors (CDIs) of Staphylococcus aureus.³⁶ Of note, there was a high representation of thioureas in the hits, suggesting this motif may be intrinsically prone to Cu-dependent antibacterial activity. Similar efforts led to the discovery of pyrazolopyrimidinones as CDIs, which are thought to increase intracellular Cu levels in S. aureus, overwhelming their intracellular buffering and efflux capacity.³⁷ Evaluation of existing drugs for Cu-potentiated activity is a promising way to revitalize and repurpose our current drug arsenal. Indeed, these CDIs exhibit antibiotic activity only in the presence of Cu, suggesting that identification of Cu-dependent activity will allow access to unrecognized value in existing drugs. Additionally, expanding the scope of considered properties in screens for new drug classes to include Cu-dependent activity (or metal-dependent more generally) could reveal an untapped reservoir of bioactive compounds.

Cu delivery by small molecules may also be useful in re-sensitizing resistant cells to existing drugs. Carbapenem antibiotics have suffered from resistance in the clinic due to bacterial expression of β-lactamase enzymes that cleave these drugs, including Zn-dependent metallo-β-lactamases such as NDM-1.^{38, 39} Cu ionophores that deliver intracellular Cu are being investigated as carbapenem adjuvants due to the ability of Cu to inhibit metallo-β-lactamases.⁴⁰ Thus, Cu mobilization by small molecules can be a viable strategy to increase efficacy of drugs that are not themselves Cu-dependent.

In another example of the role Cu potentiation can play in drug repurposing, the alcohol abuse drug disulfiram is being investigated for repurposing as an anticancer drug. Disulfiram is a disulfide dimer, reduction of which yields the diethyldithiocarbamate molecule (DTC, Chart 1) that contains negatively charged sulfur atoms and is capable of binding metal ions. The ability of DTC to form complexes with Cu is known to be important for its anticancer properties, and reactive oxygen species are thought to play a role in cell death.⁴¹ Recent work has enhanced the mechanistic understanding of Cu-potentiated toxicity, with identification of the NPL4 protein as a molecular target of the DTC–Cu complex.⁴² The DTC–Cu complex has been detected in mice treated with disulfiram and in plasma of patients receiving disulfiram treatment, further bolstering the idea that the DTC–Cu complex is the active anticancer metabolite.⁴² Intriguingly, tumor volume was suppressed by an additional 20% in mice fed high Cu diets during disulfiram treatment, relative to mice receiving a normal diet during treatment, suggesting that metal micronutrients in patient diets could benefit drug efficacy.⁴² Despite these positive results, significant obstacles exist for utilizing disulfiram in cancer therapy. For example, a phase 2 clinical trial of disulfiram in prostate cancer patients identified significant toxicity with no clear clinical benefits.⁴³ Thus, while disulfiram is an attractive candidate for Cu-dependent drug repurposing, additional work is needed to optimize efficacy while minimizing patient toxicity.

The examples described above exemplify the utility of using small molecules to redistribute metal ions to treat a variety of disease states. However, any non-specific delivery or chelation of these metals still carries the associated risk of toxicity. To avoid potential off-target effects of Cu chelators, we have developed molecules called prochelators as prodrugs of small molecule chelators in which the metal-binding site has been blocked with a masking group that is selectively removed only in the disease state.⁴⁴ For example, the prochelator GGTDTC (Chart 1) contains a masked diethyldithiocarbamate that is released from a self-immolative linker upon reaction with gamma-glutamyl transferase, an enzyme found to be overexpressed along the cell surface of several cancer types.⁴⁵ The approach relies on an activation mechanism unique to cancer cells in order to release a pharmacologically active agent that itself is conditionally active in cancer cells with amplified Cu metabolism, as has been shown in several prostate cancer cell lines.⁴⁶ Additional examples from our lab include a prochelator based on the hydroxyquinoline (8HQ, Chart 1) ionophore that requires oxidative activation for Cu delivery to fungal cells,⁴⁷ and the antibacterial PcephPT (Chart 1) that releases a PT ionophore upon activation by β-lactamase enzymes expressed in drug-resistant bacterial pathogens. In this way, the prochelator co-opts a mechanism of drug resistance in order to direct Cu toxicity preferentially against pathogenic bacteria.⁴⁸ Collectively, these strategies are designed to take advantage of the unique metallobiology and conditions at these disease states. Though the in vivo application of these approaches still needs to be established, these examples bolster the idea that knowledge of metal trafficking in disease presents opportunities for design and discovery of compounds that selectively manipulate metal distribution and availability.

Using Small Molecules to Restore Metal Homeostasis.

The ability of small molecule chelators to override defects in metal trafficking makes them ideal candidates for application in diseases characterized by misregulation of metal homeostasis. Because maintenance of metal homeostasis is essential to cellular function, defects in these sensing and adaptation mechanisms are capable of giving rise to life-threatening diseases. In addition to forming complexes that can overwhelm biological systems by disturbing metal homeostasis, metal-specific ligands can conversely be used to rescue defects in metal trafficking. Indeed, the same ionophores that induce Cu-overload and cause cell death in some systems can be used to rescue Cu deficiency or misregulation in other systems, if the delivered Cu is recognized and properly incorporated into homeostasis networks.⁴⁹

Metal dyshomeostasis has been implicated in the progression of neurodegenerative diseases,⁵⁰ so compounds that are capable of correcting metal imbalances are attractive candidates for treatment of these diseases. Thiosemicarbazones are a class of molecules known for their rich metal coordination chemistry, their ability to transport metals across cell membranes, and their potential chemotherapy applications from tuberculosis to cancer.^{51, 52} Metal complexes of bis(thiosemicarbazonato) ligands have been studied for their suitability as therapies and diagnostic tools for neurological disorders due to the efficiency with which these ligands are able to bind and deliver metal ions. For example, the Cu bis(thiosemicarbazonato) complex CuATSM (Chart 1), has shown potential as a Cu-based therapeutic for the neurodegenerative disorder amyotrophic lateral sclerosis (ALS). CuATSM delivers Cu to mutated superoxide dismutase 1 (SOD1), resulting in increased activity and reduced aggregation of this protein.⁵³ Of note, antibacterial properties of CuATSM have also been demonstrated,⁵⁴ reinforcing the concept that the same metal-binding compounds can have more than one biological effect, depending on the context. CuATSM is also being investigated as a treatment for Parkinson’s disease. In animal models, CuATSM improved motor and cognitive functions, a result thought to arise, at least in part, from the complex’s ability to scavenge reactive nitrogen radicals.⁵⁵ In Alzheimer’s disease, aggregated amyloid-β peptide results in the formation of extracellular senile plaques in the brain, a key hallmark of the disease. Delivery of Cu via bis(thiosemicarbazonato) ligands has been shown to increase matrix metalloprotease activity, reducing levels of extracellular amyloid-β peptide.⁵⁶ Bis(thiosemicarbazonato) metal complexes have also been used as positron emission tomography (PET) tracers to visualize neurodegenerative disease states.^{57, 58}

The biological activity of Cu bis(thiosemicarbazone) complexes is related to their lipophilicity, ability to cross the blood-brain barrier, and Cu(II)/(I) reduction potential. Although delivery of Cu by these complexes is often described as a one-electron reduction that labilizes the metal ion as Cu(I), recent work raises the possibility that an oxidative pathway also exists in which the bis(thiosemicarbazone) ligand is oxidized and Cu(II) released.⁵⁹ This suggestion is interesting given the fact that the mechanism of Cu release will impact the oxidation state of the labilized metal and the stability of the ligand, which will in turn impact biological outcomes.

Copper is not the only target for compounds that mobilize metals. The small molecule natural product hinokitiol (Chart 1) forms a lipophilic complex with Fe ions and is capable of shuttling Fe across cell membranes to override defects in Fe transport machinery.⁶⁰ Hinokitiol binds both ferrous and ferric iron with high affinity (association constants of ~10¹⁵ and ~10²⁵, respectively), and, similar to the Cu ionophores discussed earlier, hinokitiol rescues growth of a yeast strain lacking its high affinity Fe import system. Hinokitiol-mediated Fe transport was also observed in mammalian cells lacking divalent metal transporter 1 (DMT1), mitoferrin, and ferroportin, illustrating that hinokitiol can substitute for a variety of Fe transport proteins. Notably, Fe supplied by hinokitiol was incorporated into heme, and hemoglobinization was restored in DMT1- and mitoferrin-deficient cells, indicating the delivered Fe is bioavailable.⁶⁰ The buildup of Fe gradients in the absence of Fe proteins is believed to bestow hinokitiol with site- and direction-specific transport of Fe.

No doubt, small molecules capable of bypassing or functionally restoring missing proteins hold therapeutic potential in diseases arising from impaired protein function. Friedreich’s ataxia is a neurodegenerative disease caused by mutations in the gene encoding frataxin, an essential mitochondrial protein involved in iron-sulfur (Fe-S) biosynthesis.^{61, 62} Loss of frataxin function leads to Fe accumulation in mitochondria, elevated oxidative stress, and loss of function of Fe-S enzymes, all of which have devastating multisystemic effects, including ataxia, cardiomyopathy, vision loss, and increased risk of diabetes. Although frataxin deficiency is known to perturb Fe homeostasis, the molecular mechanisms behind this observation are not fully understood. Fe chelators were proposed many years ago to deal with the mitochondrial Fe overload of Friedreich’s ataxia cells,^{63, 64} and clinical trials with deferiprone (Chart 1) have been performed in Friedreich’s ataxia patients. These trials suggested deferiprone at low doses could have some beneficial effects, but the compound became toxic at higher doses.⁶⁵ Interestingly, elevated levels of Zn and Cu have also been found in a model of Friedreich’s ataxia, and chelation of Cu via tetrathiomolybdate (TTM, Chart 1) or bathocuproine disulfonic acid (BCS, Chart 1), or Zn via N,N,N’,N’-tetrakis(2-pyridylmethyl)ethylenediamine (TPEN, Chart 1) was found to improve motor performance of Drosophila flies lacking frataxin.⁶⁶ The discovery that small molecule sequestration of metals other than Fe improved Friedreich’s ataxia disease phenotypes emphasizes the interconnected nature of metal homeostasis networks and highlights potential new areas for treatment development.

Additional work to fully characterize the pathways by which metal cofactors are synthesized may open new areas for therapeutic development with alternative mechanisms to those that directly chelate metals. Recent work from Barondeau and coworkers shows that Fe-S biogenesis function, and thus function of enzymes that depend on this cofactor, can be restored even in the absence of frataxin.⁶⁷ They found that a variant of the iron-sulfur cluster assembly scaffold protein (ISCU2) bypasses the need for frataxin by promoting the transfer of Fe-S clusters to acceptor protein glutaredoxin (GRX5).⁶⁸ They also discovered that hypoxic conditions enable Fe-S cluster synthesis to proceed in the absence of frataxin.⁶⁹

These examples suggest that fixing frataxin itself may not be necessary for ameliorating symptoms of Friedreich’s ataxia, provided that alternative interventions can be found to mend the Fe-S cluster pathway that is broken by mutant frataxin. More broadly, these examples underscore the importance of understanding metal and metal cofactor pathways to identify potential vulnerabilities and opportunities for therapeutic development.

Lessons and Opportunities

• Small molecules with appropriate properties can shift biological systems toward or away from metal homeostasis, with some molecules exhibiting dual utility.

• Including assays that test for Cu potentiation during drug discovery or drug repurposing campaigns could identify novel hits.

• Advanced bioanalytical and imaging techniques provide opportunities to enhance our understanding of how small molecules affect metal distribution and speciation.^{70, 71}

• Small molecules that correct metal imbalances and/or restore protein function have potential to restore flux through metal-dependent pathways and thereby ameliorate disease symptoms.

Copper in Human Metabolic Disease

Dietary metal deficiencies and misregulation of metal homeostasis have been associated with a number of metabolic disorders including diabetes, obesity, Wilson’s disease, Menkes disease, and cancer. Metabolism in this context refers to the sum of all the chemical reactions involved in the breakdown of molecules (catabolism) and the use of energy to synthesize new molecules (anabolism). Defects in Cu trafficking can influence metabolic processes by inhibiting the function of key cuproenzymes or interfering with Cu-dependent signaling, giving rise to cellular stress (Fig. 3).^{72, 73} This section highlights recent advances regarding the role of Cu in metabolic disease.

Figure 3.

Defects in Cu trafficking can cause localized Cu deficiency or overload, thereby interfering with Cu-dependent signaling and metalloenzyme activity. (a) Under homeostatic conditions, cells import, export, and transport Cu to regulate cellular processes, such as lipid metabolism. Cu-dependent enzymes (yellow) require cellular import of Cu via membrane ATPases (purple cylinder) and proper delivery via Cu chaperones (dark blue). Cu-regulated proteins (red) also require maintenance of Cu homeostasis for proper function. Export of Cu through Cu-exporting ATPases (blue cylinder) and sequestration via metallothioneins (MT, purple pentagon) ensure Cu does not accumulate to harmful levels. (b) Defects or blockages in Cu trafficking proteins prevent Cu from reaching its target and/or cause localized Cu accumulation or deficiency. Under these conditions of Cu misregulation, metalloenzymes can fail to receive their required Cu cofactor and remain in the nonfunctional apo form, while other proteins could be mismetalated with Cu. Cu deficiency or excess can also interfere with signaling that regulates metabolic processes. Therapeutic strategies involve using small molecules to either create blockages or bypass existing blockages, depending on the disease context.

Although metals like Cu have traditionally been studied as static cofactors in enzymes, there is an increasing appreciation for the role of these metal ions in regulation of metabolic processes. In fact, a new field termed “metalloendocrinology” explores the interface of inorganic chemistry and endocrinology to obtain a mechanistic understanding of how metal micronutrients, like Cu, impact hormone regulation.⁷⁴ Focus in this area has the potential to reshape nutritional recommendations and hormone-based therapies.

Interest in the impact of metals on hormone regulation likely stems in part from existing knowledge regarding the close link between lipid metabolism and Cu regulation. Though it is known that dysregulation of Cu homeostasis interferes with lipid metabolism, giving rise to a variety of pathologies, the underlying mechanisms of these relationships are not well understood.⁷⁵ However, investigation into the molecular explanations for these relationships has increased our understanding of the cross-talk between Cu status and important cellular processes. For example, a recent study established a link between Cu deficiency and increased fat accumulation and traced this observation to a loss in activity of Cu-dependent amino oxidase 3, a key regulator of adipocyte metabolism.⁷⁵ Loss of oxidase activity due to Cu deficiency or impaired Cu trafficking resulted in protein inhibition, defects in lipid metabolism, and increased fat accumulation.⁷⁵ Furthermore, Cu has been identified as a reversible inhibitor of the cyclic-AMP (cAMP)-degrading phosphodiesterase PDE3B through a single, conserved cysteine residue. In the absence of inhibition by Cu, PDE3B degrades cAMP and prevents activation of enzymes that break down fat. This study demonstrates that cellular Cu status directly impacts lipolysis.⁷⁶ It is notable to point out in this example the apparent role of Cu as a regulator of protein function, which is distinct from the better-recognized role of Cu in active sites of cuproenzymes. Yet, both examples illustrate the importance of sufficient Cu levels for breakdown of lipids, with Cu deficiency causing interference with lipid metabolism. Understanding the requirement for Cu and its interaction with Cu-dependent regulators in adipocyte metabolism will aid in the development of tools to treat obesity and related diseases such as diabetes and fatty liver.

In another example of consequences of Cu dysregulation, mutations in the Cu export protein ATP7B, a protein essential for mammalian Cu homeostasis, cause Cu accumulation in the liver, giving rise to Wilson’s disease. Although Cu buildup in hepatocytes is known to play a central role in the development of Wilson’s disease, it was unclear whether hepatocyte Cu overload is sufficient to cause Wilson’s disease pathology. Recently, it was determined that Cu overload in nonparenchymal cells, which make up 30—40% of total liver cells, impacts the inflammatory response, and Cu overload in both hepatocytes and nonparenchymal cells cause lipid imbalance in the liver.⁷⁷ Importantly, this study identified that cell-specific therapies targeting only hepatocytes may not work as well as they could since Cu imbalance in additional cell types contributes to this disease. In addition, the study expands known cell targets in which Cu homeostasis could be corrected by using small molecules.

Inherited mutations in genes responsible for delivering Cu to cytochrome c oxidase in the mitochondria result in dysfunctional mitochondrial energy metabolism that gives rise to severe pathologies in affected patients, including liver dysfunction, cardiomyopathy, metabolic acidosis, and neurological defects.⁷⁸ Following a targeted screen of known Cu ionophores, it was recently demonstrated that elesclomol (Chart 1) is capable of restoring function of the mitochondrial respiratory chain by delivering Cu across biological membranes in yeast models defective in normal mitochondrial Cu uptake.⁷⁸ Notably, direct Cu supplementation therapy has not been effective in human patients. The observation that elesclomol and other Cu ionophores, but not direct Cu supplementation, is able to rescue mitochondrial Cu defects illustrates the need for small molecules that efficiently shuttle Cu to where it is needed. Interestingly, elesclomol is also an investigative anticancer drug (vide infra).^{79, 80}

In the examples discussed in this section thus far, Cu dysregulation impairs important processes in host cell metabolism, and it is advantageous to restore Cu homeostasis to correct these imbalances (Fig. 2). Cancer, on the other hand, represents a different type of metabolic disease in which maintenance of Cu homeostasis drives cancer cell proliferation to the detriment of the host. Indeed, Cu is an essential factor in tumor proliferation, and as such, interference with Cu homeostasis in tumor cells via sequestration or delivery of this metal ion is a prospective anticancer strategy.^81–83

The ability of elesclomol to transport Cu has been linked to the anticancer activity of this drug, which is known to induce mitochondrial oxidative stress.^{79, 84} Strikingly, recently published work identified elesclomol as one of only three hits in a screen of 4,300 small molecules for their ability to kill proteasome inhibitor-resistant cells.⁸⁵ This work revealed that cancer cell adaptation to proteotoxic stress renders cells more sensitive to elesclomol in a manner that depends on Cu.⁸⁵ This adaptation involves a shift in cancer cell metabolism from glycolysis to high-mitochondrial dependency, characterized by increased mitochondrial energy metabolism. Intriguingly, ferredoxin, a mitochondrial enzyme critical to the Fe-S biosynthesis pathway, was identified as the single protein target of the elesclomol-Cu complex, which can serve as a substrate for reduced ferredoxin, leading to production of Cu(I) and a Cu-dependent form of cell death the authors refer to as “cuproptosis.” The adaptative shift from glycolysis to high-mitochondrial dependency increases ferredoxin enzymatic activity thereby potentiating toxicity of the elesclomol-Cu complex.⁸⁵ Thus, a shift in metabolic processes to adapt to one stressor creates a new vulnerability that can be leveraged for metal-dependent cell death. Of note, the adaptation presented in this example is not directly related to metal homeostasis, yet it facilitates the Cu-dependent activity of a small molecule. Interestingly, disulfiram was also identified in the screen and may promote Cu-dependent cell death via a similar mechanism to elesclomol. This example illustrates that opportunities to deploy metal-dependent small molecules may exist in unexpected places.

The multiple hand-offs required to transport Cu from cell entry to metalloenzyme provide several steps that could in principle be targeted pharmacologically. For example, an innovative recent approach identified small molecules that bind and inhibit Cu-trafficking metallochaperone proteins Atox1 and CCS directly.⁸⁶ Blocking the Cu transport relay in this manner caused an accumulation of intracellular Cu, a decrease in lipogenesis, an increase in oxidative stress, and ultimately inhibited cancer cell proliferation in leukemia and lung cancer cell lines.⁸⁶ One of these compounds, DCAC50 (Chart 1), was further investigated for activity against triple negative breast cancer cells, the most aggressive breast cancer phenotype.⁸⁷ The efficacy of DCAC50 varied by cell line, but increased oxidative stress was observed for all cells treated with DCAC50 and is believed to be the cause of cytotoxicity.⁸⁷ This work highlights the opportunities for pharmacological inhibition of Cu trafficking in a way that is distinct from direct mobilization of metal ions.¹⁵

In addition to the metallochaperones, the Cu transporter ATP7A has also emerged as a potential target for cancer therapy. It has recently been shown that ATP7A is required for delivery of Cu to lysyl oxidase enzymes.⁸⁸ Lysyl oxidases are secreted Cu-dependent metalloenzymes that have been the subject of inhibitor development because of their role in tumor proliferation and metastasis.⁸⁹ Deletion of ATP7A significantly reduced tumor growth, suggesting that blocking Cu delivery to oncogenic metalloenzymes could open alternative strategies to inhibiting the enzyme directly.

Opportunities to leverage the need of cancer cells for Cu continue to be discovered. In this context, the emerging understanding of regulatory and signaling functions for Cu is creating new ways to approach therapies by focusing on metal-responsive signaling pathways.⁹⁰ The mitogen-activated protein kinase (MAPK) pathway is responsible for transmitting extracellular signals via propagation of phosphorylation events to the nucleus to activate cell growth, division, and differentiation.⁹¹ Mutations in components of this pathway underlie many cancers, including melanomas driven by mutant BRAF kinase, which phosphorylates the MEK kinases, which phosphorylate the ERK kinases, which then activate downstream signaling. The discovery that MEK1 binds Cu to promote its kinase activity is remarkable.^{92, 93} Not only was this the first example showing that Cu can directly regulate kinase activity, but it also demonstrated the prospect of using Cu chelators to inhibit the growth of specific mutant-positive tumors.⁹⁴ This work highlights two important notions: 1) selective manipulation of Cu-MEK interaction is a promising direction for chemotherapy development, and 2) it seems likely that additional signaling pathways are modulated by Cu or other transition metals in ways that await discovery.

In another example, a recent pharmacogenomic study of thiosemicarbazone NSC319726 (Chart 1) revealed Cu-dependent oxidative stress as a major contributor to cell cycle arrest induced by this compound in patient-derived glioblastoma cells.⁹⁵ The finding echoed similar Cu-dependent activity observed for other thiosemicarbazone derivatives.⁹⁶

Although this section focuses primarily on Cu, we would be remiss not to point out the therapeutic potential presented by manipulation of other metal ions. Interestingly, this same NSC319726 compound that showed Cu-dependent cytotoxicity in glioblastoma cells had previously been identified for its selective lethality against cancer cell lines carrying mutant tumor suppressor protein p53. Wild-type (WT) p53 requires a single Zn ion for proper folding and manipulating Zn concentrations can impact the structure and function of this protein. Specifically, p53 loses DNA binding specificity with Zn deficiency but can misfold and aggregate with surplus Zn. By serving as a small molecule “sink and source” of Zn, NSC319726 was found to reactivate mutant p53 with Zn in these cell lines by buffering Zn concentrations at levels optimal for restoring WT conformation to mutant p53.⁹⁷ While seemingly contradictory, it is not surprising that the same molecule can affect different mechanisms with different metal dependency in different cell types. Small molecules that induce lethality as a function of their metal binding are likely to have complex and situationally-nuanced mechanisms of action. The mechanisms may be complementary, or at least not mutually exclusive, as exemplified by a recent study that was able to visualize thiosemicarbazones complexed with Zn within lysozomes, yet found evidence of their transmetallation with Cu as the source of their cytotoxicity.⁵² The intersectionality of these examples harkens back to concepts that are ensconced within the cellular metal buffering data of Fig 1: the biological effect of metal-binding compounds will be dictated by the most favorable free energy of metalation given the availability of metal and metal buffers within a particular system.

The anticancer properties of another small molecule, PAC-1 (Chart 1), also arise from its ability to chelate Zn. Apoptosis relies on activation of the zymogen procaspase-3 to caspase-3, a process that is inhibited by Zn.^{98, 99} Procaspase-3 is elevated in cancer cells and compounds capable of activating it present attractive chemotherapeutics.¹⁰⁰ PAC-1 is capable of sequestering Zn ions with high affinity (42 nM), allowing procaspase-3 to autoactivate to caspase-3 and induce apoptosis.^{101, 102}

Lessons and Opportunities

• Cu plays an integral role in the regulation of a wide array of metabolic processes. It is not only a static cofactor, but also has roles in signaling and metabolic regulation.

• Understanding how Cu regulates these processes will inform new strategies for pharmacologically altering hormone regulation, lipid metabolism, cancer cell proliferation, and likely other processes through manipulation of Cu trafficking.

• Although this section focused primarily on Cu, similar principles apply for strategically manipulating other metal ions involved in metabolic processes.

Microbial Adaptation Strategies During Changes in Metal Availability

Like host cells, microbes adapt their metabolic processes in response to changes in metal availability within their microenvironments. There are a multitude of conditions that can lead to changes in metal availability at the host-microbe interface, from host-imposed metal withholding or overload, to dietary choices, drug treatment, or genetic variation, all of which require microbes to adapt in ways that promote their survival (Fig. 4b). This section examines adaptation strategies and compensatory mechanisms employed by microbes when they sense changes in metal availability.

Figure 4.

Changes in metal availability require host and microbe to adapt. (a) Metals at the center of interactions between host, commensal or pathogenic microbes, and xenobiotics. (b) Changes to cellular metal availability can be caused by the host immune response, dietary choices, drug treatment, or genetic mutations. Genetic mutations can cause localized metal deficiency or surplus, requiring cells to adapt metal utilization and potentially impacting metalloenzyme activity. Host-imposed metal withholding (e.g. by calprotectin) or overload causes microbes to adjust metal transport and utilization as well as alter their metabolite production (e.g. yersiniabactin and phenazines). Xenobiotics, including dietary nutrients and antibiotics, influence the makeup of the microbiota and the equilibrium between commensal and pathogenic microbes. In turn, the makeup of the microbiota influences how dietary metals are processed by the host. Changes in metal availability can impact metalloenzyme function in host or microbe.

Impact of Metal Availability on Metabolite Production and Function.

In response to infection, hosts restrict pathogen access to essential metals, a process termed nutritional immunity.^103–105 While best understood for Fe, elucidation of mechanisms for restricting Mn, Zn and even Cu are expanding the overall appreciation of these processes.^{6, 7, 106, 107} For example, the host defense protein calprotectin is capable of sequestering Zn, Mn, Fe, and Cu to induce metal starvation in invading bacterial and fungal pathogens.^108–110 A key strategy microbial pathogens have evolved for tolerating conditions of metal starvation is production of small molecule metabolites, which can interact with metals and impact metal balance between host and pathogen (Fig. 4a). Ironically, Pseudomonas aeruginosa secretes redox-cycling secondary metabolites called phenazines (e.g. pyocyanin, Chart 1), which make Fe(II) available to calprotectin and promote Fe withholding.¹² Sensing reduced Fe availability, P. aeruginosa secretes the siderophore pyoverdine (Chart 1) to scavenge Fe and decreases production of phenazines. This example highlights interactions between microbial metabolites and the immune response, and the adaptation strategies employed by microbes during nutritional immunity.

Access to Fe is essential for microbial growth, and many pathogenic organisms secrete siderophores to acquire extracellular Fe. Intriguingly, the siderophore yersiniabactin (Chart 1) can also chelate Cu, a property that has been shown to protect Escherichia coli from Cu toxicity imposed in host phagocytic cells while preserving nutritional access to that metal, a strategy termed “nutritional passivation.”¹¹¹ More recently, yersiniabactin has also been shown to participate in nickel acquisition and sequestration.¹¹² This dual function of yersiniabactin presents a new paradigm in which bacteria not only secrete these metabolites to scavenge metals for nutritional use but also to buffer this metal extracellularly and protect from toxicity. With yersiniabactin now implicated in Fe, Cu, and Ni binding, there are questions about how it specifies which metal to bind and whether other siderophores also functionally bind metals other than Fe. Indeed, small molecule metabolites are capable of interacting with metals other than Fe, with fungal natural products revealed to be transcriptionally responsive to external Cu levels.^{113, 114} Key insights into these questions regarding metal specificity are likely to come from comprehensive analyses that relate the free energy of metalation of various compounds in the context of metal availability under different conditions.¹³

Metal availability is not the only cue that signals microbes to secrete siderophores. The marine organism Vibrio harveyi has been shown to tune the quantity and type of siderophores it produces based on cell density, a process regulated by quorum sensing.¹¹⁵ Siderophores repressed by quorum sensing are thought to be used in Fe uptake for growth, enabling just enough siderophore production for Fe uptake. Siderophores stimulated by quorum sensing likely serve other purposes, such as signaling molecules or virulence factors. Thus, environmental cues on cell density signal the cell to alter production of metabolites involved in Fe acquisition, allowing this organism to avoid unnecessary siderophore production.

Maintaining Metalloprotein Function During Stress.

In addition to altering metabolite production, microbial cells directly shift their metallobiology in order to deal with changes in metal availability or other stressors. They have evolved sophisticated adaptation mechanisms to propagate in hostile and varying host environments, which means they have back-up plans if a metal becomes limiting (Fig. 5). One of these adaptations includes the ability to switch between cofactors based on metal availability (Fig. 5a). For instance, the opportunistic fungal pathogen Candida albicans fine-tunes its superoxide dismutase expression as a function of Cu availability, utilizing Cu-requiring Sod1 under Cu-replete conditions and Mn-requiring Sod3 when intracellular Cu levels are low.¹¹⁶ This is not an on/off response, but rather a readjustment of expression to conserve Cu for use in other enzymes, particularly cytochrome c oxidase. Similarly, S. aureus resists Mn starvation by populating Mn-utilizing SodM with Fe.¹¹⁷ Surprisingly, this enzyme has equal activity whether loaded with Mn or Fe. In contrast, SodA in bacterial pathogen Streptococcus pyogenes has lower activity when it contains Fe than when it is populated with Mn. Thus, this pathogen coordinates Mn import with Fe efflux to ensure that Mn-requiring SodA is properly metallated.¹¹⁸ This coordination of metal transport is another effective strategy microbes employ to maintain cellular function during stress (Fig. 5b).

Figure 5.

Compensatory mechanisms allow microbial pathogens to adapt to a variety of hostile host environments. (a) The ability to switch metal cofactors allows pathogens to maintain key cellular functions, including superoxide dismutase (Sod) activity and heme biosynthesis, even under conditions in which the default metal cofactors cannot be used. (b) Cells also adapt to metal limitation and oxidative stress by coordinating metal transport.

During oxidative stress, E. coli adapts to sustain heme synthesis and catalase activity, with the primary responses involving compensatory adjustments in mechanisms of Fe-cofactor synthesis. The activity of coproporphyrinogen III oxidase, an enzyme of the heme biosynthesis pathway, can be supplied by either of two isozymes, HemF or HemN.^{119, 120} In the presence of hydrogen peroxide, Mn-utilizing HemF displaces HemN, which contains an Fe-S cluster and is more vulnerable to damage by reactive oxygen species.¹²¹ S. aureus, on the other hand, responds to oxidative stress by exporting Mn through a Mn efflux pump recently identified in this organism.¹²² Although Mn can serve as an antioxidant,¹²³ it is toxic if present in excess, and export of Mn was shown to be essential for survival during oxidative stress and for promoting infection in a mouse model.¹²²

A proteomics-based investigation into the response of P. aeruginosa to Fe limitation identified an upregulation of Fe-sparing metabolic pathways to compensate for downregulation of Fe-rich metabolic pathways. Notably, though, Fe-S cluster biosynthesis proteins remain upregulated in low Fe conditions, indicating a basic requirement for these cofactors for P. aeruginosa metabolism, a vulnerability that could be leveraged in antibacterial therapy. The study also revealed a surprising Fe-dependent regulation of proteins encoded by genes that are induced by Zn starvation. Thus, during Fe starvation in P. aeruginosa there is an increased reliance on Zn as a metal cofactor (Fig. 5b).¹²⁴

Lessons and Opportunities

• Metal availability impacts production and function of metabolites. Several of these metabolites interact with metals and influence metal availability to host and microbe.

• Cross-talk exists between microbial metabolites, metals, and the immune system.

• Production of metal-binding metabolites can be regulated by factors other than metal availability. Cues like cell density can also impact the quantity and identity of these metabolites.

• Open questions and areas ripe for investigation include discovery of other small molecule metabolites that are likely to interact with metals in addition to Fe for nutritional, signaling, and protective functions.

• Identifying the compensatory mechanisms employed by pathogens during stress provides the knowledge required to strategically design therapies that anticipate and prevent adaptation.

Metallobiology of the Microbiome

Pathogenic microbes are not the only bugs that are influenced by metals. There has been a recent surge in appreciation for the role the microbiome plays in human health, and the metallobiology of these systems is an exciting emerging area of investigation. This section provides a glimpse of recent investigations concerning the metallobiology of the microbiome. It includes a discussion of the interesting questions being uncovered regarding metabolites synthesized by metalloenzymes and competition between beneficial and pathogenic microbes for metal micronutrients in the gut.

Copper and zinc are required nutrients for animal species and are common additives in livestock feed for nutritional purposes and growth promotion. For decades, metals have been used in conjunction with antibiotics, but increasing bans on the use of antibiotics in livestock has correlated with a rise in the use of metal supplements. Though it has been assumed that Cu and Zn supplementation do not affect the microbiome of these animals, recent evidence suggests that heightened use of metals in feed has contributed to the evolution of bacterial strains that are pathogenic to humans.¹²⁵

The link between metals and antibiotics is perhaps more significant than has been conventionally appreciated. Antibiotics are known to perturb the makeup of the microbiome,¹²⁶ and recently, implications for these perturbations on how metal micronutrients are processed have been identified (Fig. 4b). Isotopic analysis revealed that Cu processing in the gastrointestinal tract is significantly impacted by antibiotic-mediated depletion of gut microbiota, with reduced expression of Cu transporters Ctr1 and ATP7A observed in the colons of mice treated with antibiotics.¹²⁷ The authors point out that the antibiotics used, notably ampicillin, neomycin, vancomycin, and metronidazole, are capable of interacting with Cu(II) in vitro, but Cu–antibiotic interactions leading to Cu sequestration would likely lead to widespread impacts to Cu processing in the GI tract rather than solely in the colon and do not explain the observed effects.¹²⁷ Though the molecular mechanisms giving rise to these changes are currently unknown, it is noteworthy that antibiotic treatment can impact processing of dietary metal micronutrients. This concept evokes a recurring theme in which the influence of drug treatment intersects with processes controlling metal availability, and vice versa, such as the example described earlier for disulfiram and Cu.

Indeed, in humans, dietary metal intake influences the makeup of the microbiota and susceptibility to pathogen colonization.¹²⁸ Under conditions of metal homeostasis, a diverse population of commensal bacteria thrive in the gut that protect against colonization by pathogenic bacteria. However, when a diet low or high in Zn or high in Fe interferes with gut metal homeostasis, colonization by commensal bacteria decreases, and colonization by pathogenic species such as Enterobacteriaceae increases.¹²⁸ Fascinatingly, intake of dietary metals tunes the levels of harmful versus commensal bacteria in the gut, with these species competing for colonization, and the availability of metal sways the outcomes of these competitions. Yet, there is still much to be learned about mechanisms of how this balance is regulated, and vast opportunities exist in understanding the molecular basis of interactions between xenobiotics and the microbiota.^{129, 130}

Another interesting avenue of investigation is the role of small molecule metabolites synthesized by bacteria of the microbiome in mediating microbe-microbe and microbe-host interactions. For example, though the siderophore enterobactin (Chart 1) has been assumed to have negative effects on the host, recent work has identified that the host cells can use bacterial enterobactin for their own Fe acquisition.¹³¹ This work raises questions about how the immune system distinguishes between pathogen- or commensal-produced enterobactin and how it interacts with Fe-bound enterobactin based on this distinction.

The study of metalloenzymes in gut bacteria is an equally compelling area of study. These enzymes of the gut microbiota are important for promoting commensal colonization, protecting against pathogenic colonization, synthesizing essential nutrients and metabolites, and breaking down xenobiotics.¹³² Recently, the N-nitrosating non-heme-iron-dependent enzyme SznF was discovered, expanding our understanding of the synthesis of N-nitroso groups in bioactive metabolites.¹³³ Homologues of this enzyme are thought to be present in an array of bacteria, including human pathogens. Continued exploration of metalloenzymes in gut bacteria and their metabolites is sure to shed light on cross-talk between metal homeostasis, metabolite production, and interactions of microbes with each other and their host (Fig. 4a).

Lessons and Opportunities

• Understanding how the microbiome responds to and metabolizes dietary nutrients, xenobiotics and metals will inform optimal nutritional requirements.

• Links have been established between diet and disease, but there is still much to learn about the molecular mechanisms associated with various metal-dependent processes behind these observations.

• Mechanistic understandings of interactions between microbiota communities and how these interactions are impacted by diet, xenobiotics and metal availability will aid in understanding how to promote colonization by beneficial bacteria while reducing colonization by pathogenic strains.

• Discovery of metalloenzymes and their metabolites will improve our understanding of the players in microbiota processes and could represent new targets.

Iron Transport as an Antibacterial Drug Target

Access to Fe is an essential component of the ability of pathogenic microbes to establish and promote infection, and they have developed a variety of mechanisms to acquire this Fe. At the same time, Fe can be toxic if it is not tightly controlled. Understanding these acquisition and detoxification mechanisms informs antimicrobial treatment strategies that center on interference with these processes. Indeed, strategies that help prevent pathogens from acquiring Fe would be powerful tools in fighting microbial infections. For these strategies to be effective, it is important to keep in mind that successful microbes have redundant and alternative pathways for acquiring required nutrients like Fe. These pathways are likely to be different depending on both the organism and its context within the host, meaning that developing inhibitors of Fe uptake requires deep understanding of how Fe is acquired under diverse scenarios. This section describes opportunities to exploit the microbial requirement for Fe by co-opting their Fe acquisition, utilization, and detoxification pathways (Fig. 6).

Figure 6.

Simplified depiction of microbial Fe assimilation and detoxification systems that are viable therapeutic targets. Microbial pathogens acquire Fe primarily via import of ferric siderophores or heme, which are then processed to release Fe for cellular use. Export of heme to prevent toxicity has also been identified as important for some species, like C. difficile. Blocking any of these pathways could be effective antimicrobial strategies.

As part of nutritional immunity, the host innate immune response starves microbes of Fe.^{6, 106} Interestingly, macrophages are capable of adjusting their Fe regulation based on whether the bacteria is localized intra- or extracellularly.¹³⁴ In the case of intracellular bacteria, Fe is excreted from the macrophage, while Fe is imported and sequestered when bacteria are extracellular. The macrophage effectively serves as a barrier between the bacteria and the Fe it seeks and is capable of using opposing regulatory mechanisms to starve the bacteria of this nutrient.

The bacterial pathogen P. aeruginosa, meanwhile, has been found to shift its Fe acquisition pathways during the course of infection by dialing down siderophore production in favor of heme uptake as an Fe source, a process that requires heme oxygenase HemO.¹³⁵ Indeed, HemO was found to be consistently expressed in P. aeruginosa strains isolated from infected lungs of patients with cystic fibrosis, suggesting that heme regulatory pathways are critical for these bacteria to adapt and thrive in their niche host environment.¹³⁵ A growing understanding of these heme uptake pathways is providing insight into therapeutic targets and paving the path to new inhibitors. For example, iminoguanidines are being investigated as allosteric inhibitors of HemO.¹³⁶

Targeting heme oxygenase is a promising approach in other host-pathogen systems, too. Infections caused by Mycobacterium tuberculosis (Mtb) are notoriously deadly and difficult to treat, and new therapies are urgently needed.^{137, 138} Interestingly, host heme oxygenase, HO-1, has emerged as a promising host-derived target.¹³⁹ By catalyzing heme degradation to generate carbon monoxide, biliverdin, and Fe, HO-1 has been shown to induce anti-inflammatory effects that can be paradoxically both protective and detrimental to the host, depending on the infection.¹³⁹ A recent study in Mtb showed that pharmacological blockade of HO-1 with unnatural metal porphyrin SnPPIX (Chart 1) decreased the bacterial load in mice with acute Mtb infection.¹³⁹ While promising, the mechanism by which HO-1 inhibition in the host aids the clearance of Mtb remains unclear.

Opportunities also exist to target Mtb directly by interfering with Fe acquisition, as Mtb uses both siderophore and heme-mediated Fe uptake.³¹ Potential strategies include inhibiting the biosynthesis or uptake of the Mtb siderophore mycobactin (Chart 1).¹⁴⁰ Heme uptake in this pathogen was only recently discovered and is still a new area of investigation.¹⁴¹ Inhibitors of heme uptake would need to selectively target Mtb heme proteins while having a low affinity for host heme-binding proteins. Once again, there is still much to learn about the intricacies of how Mtb adapts its metal acquisition strategies within the host.

An alternative to blocking bacterial Fe uptake systems for limiting pathogen growth is to use pathogens’ Fe uptake systems against them in a “Trojan-horse” strategy.^{142, 143} This approach involves conjugating antibiotics to siderophores that an unsuspecting pathogen will import via their siderophore uptake pathway. Siderophore–antibiotic conjugates (sideromycins) have been developed in different bacterial systems using varied combinations of siderophores and antibiotics.^144–146 An understanding of how siderophores are recognized and imported has aided in the design of these conjugates.^147–149 Hijacking Fe acquisition pathways in Mtb may be particularly powerful in overcoming the challenge of getting drugs through the impenetrable Mtb cell wall.³¹

Though bacteria need Fe and heme, too much of either can be toxic. For Clostridium difficile, the ability to sense and export excess intracellular heme is now known to play a key role in pathogenesis.¹⁵⁰ A recent study established for the first time the requirement for a heme detoxification system in this obligate anaerobic pathogen.¹⁵⁰ Pharmacological interference with heme sensing or efflux could therefore be worthy of investigation.

Lessons and Opportunities

• Pathogens require Fe, so interfering with their ability to acquire it is a viable treatment strategy.

• Effective ways to interfere with Fe trafficking will depend on the pathogen.

• Understanding how each pathogen acquires Fe will inform these strategies.

Iron-Dependent Cell Death: Ferroptosis

Ferroptosis is an Fe-dependent cell death pathway that is distinct from apoptosis. This section provides an overview of ferroptosis, including why we would want to accelerate it, why we would want to inhibit it, and a summary of what is known about how to do both.

The term ferroptosis was introduced in 2012 and refers to a type of regulated cell death that is characterized by Fe-dependent accumulation of lipid-based peroxide species when repair systems that depend on glutathione are compromised (Fig. 7).^151–153 Specifically, ferroptosis results from loss of activity of glutathione peroxidase 4 (GPX4), an enzyme that reduces lipid peroxides.¹⁵⁴ Despite the integral role of Fe in ferroptosis, its exact function is not fully understood.^{153, 155} Generation of lipid reactive oxygen species (ROS) is the most commonly cited function of Fe, but it is possible that the Fe requirement for ferroptosis has additional origins. For example, there are a number of metabolic enzymes involved in ROS generation that require Fe as a cofactor. Ferroptosis has been linked to neurodegenerative diseases and may also have a role in tumor-suppressor function. Thus, the benefits of inhibiting or initiating ferroptosis are dependent on the disease context. There are many mechanistic aspects of ferroptosis that still need to be elucidated, but pharmacological induction and inhibition of ferroptosis are being investigated.¹⁵²

Figure 7.

Ferroptosis is characterized by loss of activity of GPX4, which causes the accumulation of lipid peroxides. There are four classes of ferroptosis inducers (red text), which are described in the main text. Inhibitors of ferroptosis are denoted in green text. Abbreviations: glutathione peroxidase 4 (GPX4), glutathione (GSH), glutathione disulfide (GSSG), lipoxygenase (LOX). Figure adapted with permission from ref. 151 © 2018 Feng, Stockwell.

There are currently four classes of ferroptosis inducers, which are classified based on their mechanism of ferroptosis induction.¹⁵¹ Class 1 inducers starve cells of cysteine, a consequence of which is depletion of intracellular glutathione (GSH), an essential cofactor for GPX4 activity. Class 2 inducers, on the other hand, bind to and inhibit GPX4 directly. Class 3 inducers deplete both GPX4 protein and coenzyme Q₁₀ (CoQ₁₀), an endogenous lipophilic antioxidant. The 1,2-dioxolane FINO₂ (Chart 1) has been shown to initiate ferroptosis through indirect inhibition of GPX4 and direct oxidation of Fe, which combined cause significant lipid peroxidation.¹⁵⁶ Its mechanism is thus distinct from existing ferroptosis inducers, and its discovery led to a new class 4 of ferroptosis inducers, of which FINO₂ is currently the only known member.

The first discovered inhibitors of ferroptosis were Fe chelators (class 1), such as deferoxamine (Chart 1) and ciclopirox, and lipophilic antioxidants (class 2), like α-tocopherol, butylated hydroxytoluene (BHT), and ferrostatin-1 (Chart 1).^{151, 157} Fe chelators inhibit ferroptosis by preventing Fenton chemistry and making Fe unavailable to lipoxygenases (LOXs), non-heme Fe-dependent enzymes that catalyze the dioxygenation of polyunsaturated fatty acids.¹⁵¹ Lipophilic antioxidants are thought to prevent lipid peroxidation by a radical trapping mechanism.¹⁵⁸ A recent study aiming to more thoroughly characterize the mechanism by which ferrostatins inhibit apoptosis found that ferrostatins accumulate in lysosomes, mitochondria, and the endoplasmic reticulum (ER), but it is localization to the ER that enables ferrostatins to exert their effect.¹⁵⁹ Deuterated polyunsaturated fatty acids (D-PUFAs) make up the class 3 inhibitors and work by preventing initiation and propagation of lipid peroxidation.¹⁶⁰ Finally, class 4 ferroptosis inhibitors suppress lipid peroxidation by inhibiting LOX directly without Fe chelation and by serving as radical trapping antioxidants.^{151, 158} Of note, ferroptosis inhibitors have recently been implicated as potential therapeutics in Friedreich’s ataxia. Indeed, mitochondrial Fe accumulation, oxidative stress, and lipid peroxidation are the hallmarks of ferroptosis and are also present in Friedreich’s ataxia cells, suggesting ferroptosis may play a role in Friedreich’s ataxia. Intriguingly, ferroptosis inhibitors were capable of rescuing cell death that was induced by knockdown of frataxin.¹⁶¹

Lessons and Opportunities

• Pharmacological agents can be used to inhibit cellular processes that play a role in ferroptosis. Depending on the target, small molecules can be used to either promote or inhibit ferroptosis.

• Compounds that increase cellular Fe abundance present opportunities as initiators of ferroptosis, while Fe chelators inhibit ferroptosis.

Summary and Outlook

A requirement for all living systems is the ability to sense deviations in metal availability and enact responses to reclaim metal homeostasis. Understanding how cell systems interact with one another, either symbiotically or pathogenically, provides us with knowledge we can use to manipulate these systems to restore health and fight disease. Although there are common themes in the discussion of metal homeostasis of these systems, it is important to note that each system is different, and knowledge regarding the intricacies of cellular control over metals and an appreciation of how they differ across organisms and cell types will inform the most effective metal manipulation strategies. This Viewpoint has highlighted examples related to cancer, metabolic diseases and microbes, with a heavy focus on Cu and Fe, but these concepts are applicable to other diseases and the full array of biologically important metals.

Biographies

Katherine J. Franz is the Alexander F. Hehmeyer Professor and Chair of Chemistry at Duke University in Durham, North Carolina, USA. After formative undergraduate research experiences with Prof. James Loehlin at Wellesley College and Dr. Richard Fish at the Lawrence Berkeley National Lab, she earned a PhD in inorganic chemistry with Prof. Stephen J. Lippard at MIT and an NIH postdoctoral fellowship with Prof. Barbara Imperiali, also at MIT. Since 2003, Kathy and her research group at Duke use principles of inorganic chemistry to develop new chemical tools and bioactive compounds to manipulate the location, speciation, and reactivity of metal ions in biological systems for potential therapeutic benefit.

Elizabeth W. Hunsaker obtained a B.S. in Chemistry in 2014 from Catawba College, a small liberal arts school located in Salisbury, NC. She is currently a Ph.D. candidate in the Department of Chemistry at Duke University, working under the guidance of Professor Kathy Franz. Her research focuses on understanding how biologically relevant transition metals like copper influence the response of pathogenic fungi to antimicrobial drug stress.