Bacteriocin sublancin-168

Bacteriocin sublancin-168 is a 37-amino-acid, ribosomally synthesized and post-translationally modified antimicrobial peptide produced by Bacillus subtilis 168, encoded by the sunA gene within the SPβ prophage region. Originally classified as a lantibiotic, sublancin-168 was definitively reclassified in 2011 as an S-linked glycopeptide (glycocin) containing a β-S-linked glucose moiety attached to Cys22—a then-unprecedented post-translational modification in bacteria.

Bacteriocin Sublancin-168: A Structurally Distinct S-Linked Glycopeptide Antimicrobial for Gram-Positive Pathogen Research and Biodefense Procurement

Bacteriocin sublancin-168 is a 37-amino-acid, ribosomally synthesized and post-translationally modified antimicrobial peptide produced by Bacillus subtilis 168, encoded by the sunA gene within the SPβ prophage region [1]. Originally classified as a lantibiotic, sublancin-168 was definitively reclassified in 2011 as an S-linked glycopeptide (glycocin) containing a β-S-linked glucose moiety attached to Cys22—a then-unprecedented post-translational modification in bacteria [2]. The mature peptide (MW 3877.6 Da) possesses two disulfide bonds (Cys7–Cys36, Cys14–Cys29) and adopts a helix-loop-helix fold stabilized by hydrophobic and hydrogen-bonding interactions [3]. Sublancin-168 exhibits bactericidal activity against a range of Gram-positive bacteria including methicillin-resistant Staphylococcus aureus (MRSA), Bacillus cereus, Streptococcus pyogenes, and Listeria species [4].

• [1] Paik SH, Chakicherla A, Hansen JN. Identification and characterization of the structural and transporter genes for, and the chemical and biological properties of, sublancin 168, a novel lantibiotic produced by Bacillus subtilis 168. J Biol Chem. 1998;273(36):23134-23142. PMID: 9722542. View Source
• [2] Oman TJ, Boettcher JM, Wang H, Okalibe XN, van der Donk WA. Sublancin is not a lantibiotic but an S-linked glycopeptide. Nat Chem Biol. 2011;7(2):78-80. doi:10.1038/nchembio.509. View Source
• [3] Garcia De Gonzalo CV, Zhu L, Oman TJ, van der Donk WA. NMR structure of the S-linked glycopeptide sublancin 168. ACS Chem Biol. 2014;9(3):796-801. doi:10.1021/cb4008106. PDB: 2MIJ. View Source
• [4] Kouwen TRHM, Trip EN, Denham EL, Sibbald MJJB, Dubois JYF, van Dijl JM. The large mechanosensitive channel MscL determines bacterial susceptibility to the bacteriocin sublancin 168. Antimicrob Agents Chemother. 2009;53(11):4702-4711. doi:10.1128/AAC.00439-09. View Source

Why Bacteriocin Sublancin-168 Cannot Be Substituted by Nisin, Subtilin, or Glycocin F: Structural and Mechanistic Basis for Differentiated Procurement

Sublancin-168 occupies a structurally and mechanistically unique position among Gram-positive-targeting bacteriocins that precludes direct substitution. Unlike nisin and subtilin—prototypical lantibiotics that target lipid II to inhibit cell wall synthesis and form membrane pores—sublancin-168 is an S-glucosylated glycocin whose antimicrobial activity depends absolutely on glycosylation by the SunS glycosyltransferase and whose susceptibility in target cells is determined specifically by the mechanosensitive channel MscL, not by lipid II [1][2]. Compared to glycocin F (an O-linked, bacteriostatic glycocin from Lactobacillus plantarum carrying two GlcNAc moieties), sublancin-168 carries a single S-linked glucose, is bactericidal rather than bacteriostatic, and targets the glucose-specific PTS rather than the GlcNAc PTS [3]. Furthermore, sublancin-168's extraordinary aqueous stability at room temperature for 2 years exceeds that reported for subtilin, which is known to lose biological activity during storage [4]. These differences mean that experimental results, bioactivity profiles, resistance patterns, and formulation stability data obtained with any class analog cannot be extrapolated to sublancin-168.

• [1] Kouwen TRHM, Trip EN, Denham EL, Sibbald MJJB, Dubois JYF, van Dijl JM. The large mechanosensitive channel MscL determines bacterial susceptibility to the bacteriocin sublancin 168. Antimicrob Agents Chemother. 2009;53(11):4702-4711. doi:10.1128/AAC.00439-09. View Source
• [2] Garcia De Gonzalo CV, Denham EL, Mars RAT, et al. The phosphoenolpyruvate:sugar phosphotransferase system is involved in sensitivity to the glucosylated bacteriocin sublancin. Antimicrob Agents Chemother. 2015;59(11):6844-6854. doi:10.1128/AAC.01519-15. View Source
• [3] Wu C, Biswas S, Garcia De Gonzalo CV, van der Donk WA. Investigations into the mechanism of action of sublancin. ACS Infect Dis. 2019;5(3):454-459. doi:10.1021/acsinfecdis.8b00320. View Source
• [4] Paik SH, Chakicherla A, Hansen JN. Identification and characterization of the structural and transporter genes for, and the chemical and biological properties of, sublancin 168, a novel lantibiotic produced by Bacillus subtilis 168. J Biol Chem. 1998;273(36):23134-23142. PMID: 9722542. View Source

Quantitative Differential Evidence for Bacteriocin Sublancin-168: Head-to-Head and Cross-Study Comparisons Against Nisin, Subtilin, Pep5, and Glycocin F

MscL-Dependent Susceptibility: Sublancin-168 Requires the Mechanosensitive Channel MscL, Whereas Nisin and Pep5 Do Not

In a direct genetic study, deletion of the mscL gene in B. subtilis ΔSPβ and S. aureus completely abolished sublancin-168 susceptibility. Critically, the same ΔmscL mutants showed no alteration in sensitivity to 2.5 mg/mL nisin or 10 mg/mL Pep5, establishing that MscL is a selective determinant for sublancin-168 but not for these comparator bacteriocins [1]. This provides a unique, genetically defined susceptibility determinant not shared by the most widely used lantibiotic (nisin) or the pore-forming bacteriocin Pep5.

Evidence Dimension	MscL dependence of antimicrobial susceptibility
Target Compound Data	ΔmscL mutants fully resistant to sublancin-168; MscL required for susceptibility in both B. subtilis and S. aureus
Comparator Or Baseline	ΔmscL mutants: no altered sensitivity to nisin (2.5 mg/mL) or Pep5 (10 mg/mL) compared to parental ΔSPβ strain
Quantified Difference	Complete loss of sublancin-168 activity in ΔmscL vs. zero change in nisin/Pep5 susceptibility in the same genetic background
Conditions	B. subtilis ΔSPβ and S. aureus plate and liquid culture growth inhibition assays; NaCl concentration modulated susceptibility

Why This Matters

This MscL exclusivity means sublancin-168 can be deployed against MscL-expressing pathogens without cross-resistance to nisin-based preservatives or therapeutics, enabling orthogonal antimicrobial strategies.

• [1] Kouwen TRHM, Trip EN, Denham EL, Sibbald MJJB, Dubois JYF, van Dijl JM. The large mechanosensitive channel MscL determines bacterial susceptibility to the bacteriocin sublancin 168. Antimicrob Agents Chemother. 2009;53(11):4702-4711. doi:10.1128/AAC.00439-09. View Source

Glycosylation-Dependent Activity: S-Linked Glucose at Cys22 Is Essential for Sublancin-168 Antimicrobial Function, Unlike Nonglycosylated Lantibiotics

Sublancin-168 is the founding member of the glycocin family: its S-linked glucose at Cys22, installed by the SunS glycosyltransferase, is absolutely required for antimicrobial activity. In vitro reconstitution experiments demonstrated that the non-glycosylated SunA precursor peptide is devoid of antimicrobial activity, while glycosylation by SunS restores full bactericidal function [1]. This contrasts fundamentally with nisin, subtilin, and other lantibiotics, which depend on lanthionine/methyllanthionine thioether crosslinks rather than glycosylation for their bioactive conformation [2].

Evidence Dimension	Requirement of post-translational modification for antimicrobial activity
Target Compound Data	Non-glycosylated sublancin SunA precursor: no antimicrobial activity; S-glucosylated sublancin: fully active; sugar replacement tolerated but sugar removal abolishes activity
Comparator Or Baseline	Nisin and subtilin: activity depends on lanthionine ring formation and dehydroalanine residues, not glycosylation; glycocin F: activity depends on O-linked GlcNAc, not S-linked glucose
Quantified Difference	Glycosylation is binary essential (all-or-none) for sublancin-168 activity, whereas lantibiotics use a fundamentally different PTM chemistry
Conditions	In vitro glycosylation by His6-SunS followed by Factor Xa cleavage and oxidative folding; antimicrobial assays against B. subtilis

Why This Matters

The essential S-glycosylation creates a unique quality-control parameter for sublancin-168 procurement: any synthetic or recombinant preparation must be validated for correct S-glucosylation at Cys22, a specification irrelevant to nisin or subtilin purchasing.

• [1] Oman TJ, Boettcher JM, Wang H, Okalibe XN, van der Donk WA. Sublancin is not a lantibiotic but an S-linked glycopeptide. Nat Chem Biol. 2011;7(2):78-80. doi:10.1038/nchembio.509. View Source
• [2] Paik SH, Chakicherla A, Hansen JN. Identification and characterization of the structural and transporter genes for, and the chemical and biological properties of, sublancin 168, a novel lantibiotic produced by Bacillus subtilis 168. J Biol Chem. 1998;273(36):23134-23142. PMID: 9722542. View Source

Extraordinary Aqueous Stability: Sublancin-168 Retains Full Activity After 2 Years at Room Temperature, Contrasting with Subtilin Instability

Sublancin-168 exhibits extraordinary stability: no degradation or loss of biological activity was observed after storage in aqueous solution at room temperature for 2 years [1]. The NMR structure revealed that the two disulfide bonds (Cys7–Cys36, Cys14–Cys29), hydrophobic packing of the two α-helices, and the S-linked glucose in the interhelical loop collectively confer this exceptional stability [2]. In contrast, subtilin—a Bacillus-produced lantibiotic and structural analog of nisin—was reported to be highly unstable, with significant loss of biological activity observed during storage [3]. Nisin is considered relatively stable, but quantitative data on multi-year room-temperature aqueous stability comparable to that reported for sublancin-168 is not available in the published literature for nisin.

Evidence Dimension	Long-term aqueous stability at room temperature
Target Compound Data	No degradation or inactivation after 2 years in aqueous solution at room temperature
Comparator Or Baseline	Subtilin: highly unstable, biological activity lost during storage (exact duration not quantified in primary source); nisin: stable but no published report of 2-year RT aqueous stability
Quantified Difference	Sublancin-168: 2-year room-temperature aqueous stability documented; subtilin: known instability with activity loss during storage
Conditions	Aqueous solution, room temperature storage, biological activity assay following 2-year storage period

Why This Matters

For industrial or biodefense procurement, the 2-year room-temperature aqueous stability of sublancin-168 eliminates cold-chain requirements and lyophilization dependency, substantially reducing total cost of ownership compared to less stable lantibiotics such as subtilin.

• [1] Paik SH, Chakicherla A, Hansen JN. Identification and characterization of the structural and transporter genes for, and the chemical and biological properties of, sublancin 168, a novel lantibiotic produced by Bacillus subtilis 168. J Biol Chem. 1998;273(36):23134-23142. PMID: 9722542. View Source
• [2] Garcia De Gonzalo CV, Zhu L, Oman TJ, van der Donk WA. NMR structure of the S-linked glycopeptide sublancin 168. ACS Chem Biol. 2014;9(3):796-801. doi:10.1021/cb4008106. View Source
• [3] Gutowski-Eckel Z, Klein C, Siegers K, Hamm K, Entian KD. Patterns of reactivity of lantibiotics subtilin and nisin with molecular targets in Bacillus cereus and Bacillus subtilis 168. Appl Environ Microbiol. 1994;60(1):1-11. View Source

Spore Outgrowth Inhibition: Sublancin-168 Exhibits ~1000-Fold Greater Potency Against B. cereus Spore Outgrowth Than Vegetative Growth

Sublancin-168 inhibits B. cereus spore outgrowth, after the germination stage, approximately 1000-fold better than it inhibits exponential growth of the same cells [1][2]. This pronounced differential—inhibiting spore outgrowth at concentrations far below those needed to affect vegetative cells—is functionally analogous to the spore outgrowth inhibition reported for nisin and subtilin, placing sublancin-168 within the class of lantibiotics capable of targeting both spore and vegetative forms [3]. However, the specific MscL-dependent mechanism of sublancin-168 is distinct from the lipid II-mediated action of nisin, suggesting the spore outgrowth inhibition may operate through a different molecular pathway.

Evidence Dimension	Selectivity ratio: spore outgrowth inhibition vs. vegetative growth inhibition
Target Compound Data	~1000-fold more potent against B. cereus spore outgrowth (post-germination) than against vegetative exponential growth
Comparator Or Baseline	Nisin and subtilin: also inhibit both spore outgrowth and vegetative growth, but potency ratio not consistently quantified as 1000-fold in published sources; nisin MIC against B. subtilis ATCC 6633 reported as 0.25 μM [supporting], whereas sublancin MIC against B. subtilis 168 ΔSPβ is 0.312 μM
Quantified Difference	~1000-fold spore vs. vegetative selectivity for sublancin-168 against B. cereus; class-level shared property with nisin/subtilin, but sublancin-168 mechanism is MscL-mediated rather than lipid II-mediated
Conditions	B. cereus spore germination and outgrowth assays; vegetative growth inhibition in parallel cultures

Why This Matters

The 1000-fold spore outgrowth selectivity positions sublancin-168 as a compelling candidate for biodefense applications targeting Bacillus anthracis spore contamination, where an MscL-dependent, non-lipid II mechanism provides an orthogonal mode of action to nisin-based spore control formulations.

• [1] UniProt Consortium. P68577 (SUNA_BACSU): SPbeta prophage-derived bacteriocin sublancin-168. UniProtKB/Swiss-Prot. Accessed 2026. View Source
• [2] NCBI Conserved Domain Database. pfam19151: Sublancin family. NCBI CDD. Accessed 2026. View Source
• [3] Paik SH, Chakicherla A, Hansen JN. Identification and characterization of the structural and transporter genes for, and the chemical and biological properties of, sublancin 168, a novel lantibiotic produced by Bacillus subtilis 168. J Biol Chem. 1998;273(36):23134-23142. PMID: 9722542. View Source

Dual Antibacterial and Immunomodulatory Activity: Sublancin-168 Protects Against MRSA Infection in Mice via Combined Direct Killing and Innate Immune Stimulation

Sublancin-168 demonstrates a dual pharmacological profile: direct bactericidal activity against MRSA (MIC = 15 μM) combined with host immunomodulatory effects including macrophage chemotaxis, enhanced phagocytosis, and selective chemokine induction (CXCL1, MCP-1) while reducing pro-inflammatory TNF-α [1][2]. In a mouse MRSA intraperitoneal infection model, sublancin-168 (0.5–4 mg/kg, single i.p. dose) conferred protection that was abrogated by depletion of monocytes/macrophages or neutrophils, formally demonstrating that in vivo efficacy depends on both direct antimicrobial and host-directed immunomodulatory mechanisms [2]. Nisin, the most widely commercialized lantibiotic, has reported immunomodulatory properties in some contexts, but quantitative in vivo protection data tied to specific immune cell populations at defined sublancin-168 doses have not been reported for nisin in a comparable MRSA infection model.

Evidence Dimension	In vivo MRSA protection and immunomodulatory activity
Target Compound Data	Sublancin-168 MIC = 15 μM against MRSA ATCC43300; in vivo protection at 0.5–4 mg/kg i.p.; protection blocked by macrophage/neutrophil depletion; induces CXCL1, MCP-1; reduces TNF-α
Comparator Or Baseline	Nisin: MIC values and food-preservation applications well-established, but comparable MRSA in vivo protection data with immune-cell-depletion controls not identified in the published literature; glycocin F: bacteriostatic, not bactericidal, and no reported in vivo MRSA protection
Quantified Difference	Sublancin-168 combines bactericidal activity (15 μM MIC) with validated in vivo immunomodulatory protection; nisin and glycocin F lack published equivalent dual-mechanism in vivo MRSA protection data
Conditions	MRSA ATCC43300; RAW264.7 macrophages and mouse peritoneal macrophages for in vitro immunomodulation; BALB/c mouse intraperitoneal MRSA infection model for in vivo protection

Why This Matters

For procurement in anti-MRSA therapeutic or prophylactic development, the dual direct-killing plus immunomodulatory mechanism of sublancin-168 may provide superior in vivo efficacy and reduced selective pressure for resistance compared to bacteriocins acting solely through direct antimicrobial mechanisms—a differentiating value proposition not met by nisin or glycocin F.

• [1] Wang S, Wang Q, Zeng X, Ye Q, Huang S, Yu H, Yang T, Qiao S. Use of the antimicrobial peptide sublancin with combined antibacterial and immunomodulatory activities to protect against methicillin-resistant Staphylococcus aureus infection in mice. J Agric Food Chem. 2017;65(39):8595-8605. doi:10.1021/acs.jafc.7b02592. View Source
• [2] Li J, Chen J, Yang G, Tao L, Wang S, Cao B, Wang M. Sublancin protects against methicillin-resistant Staphylococcus aureus infection by the combined modulation of innate immune response and microbiota. Peptides. 2021;141:170533. doi:10.1016/j.peptides.2021.170533. View Source

Glucose-PTS Mediated Mechanism Distinct from Lipid II Targeting: Sublancin-168 Does Not Cause Membrane Pore Formation Even at 32× MIC, Unlike Nisin

Sublancin-168 inhibits DNA replication, transcription, and translation without compromising membrane integrity, even at concentrations as high as 32× the MIC [1]. This fundamentally differs from nisin, which binds lipid II, inhibits cell wall synthesis, and forms stable pores in the target membrane leading to leakage of essential metabolites [2]. Sublancin-168 suppresses glucose uptake via competition with the glucose-specific phosphotransferase system (PTS), and the histidine-containing phosphocarrier protein HPr and domain A of enzyme II (PtsG) are critical determinants for sublancin-168 sensitivity [3]. Moreover, unlike other bacteriocins for which PTS involvement leads to increased sensitivity in the presence of cognate sugars, addition of glucose to the growth medium increases resistance to sublancin-168, indicating a mechanistically distinct PTS interaction [3].

Evidence Dimension	Membrane integrity and mode of action
Target Compound Data	No membrane permeabilization at concentrations up to 32× MIC; inhibits DNA replication, transcription, translation; glucose-PTS (HPr, PtsG) involved; glucose in medium antagonizes activity
Comparator Or Baseline	Nisin: binds lipid II, inhibits cell wall synthesis, forms stable membrane pores, causes metabolite leakage; glucose-PTS not involved; glycocin F: bacteriostatic, GlcNAc-PTS involved, GlcNAc antagonizes activity
Quantified Difference	Sublancin-168: intact membrane at 32× MIC; nisin: membrane pore formation is a primary mechanism; glycocin F: bacteriostatic vs. sublancin-168 bactericidal
Conditions	Macromolecular synthesis assays (³H-thymidine, ³H-uridine, ³H-leucine incorporation); membrane integrity by SYTOX Green uptake; B. subtilis and S. aureus cultures

Why This Matters

The absence of membrane pore formation at 32× MIC suggests sublancin-168 may exhibit lower mammalian cytotoxicity than pore-forming lantibiotics, a relevant differentiator for procurement in therapeutic development programs where membrane-active peptides raise hemolytic and cytotoxic safety concerns.

• [1] Wu C, Biswas S, Garcia De Gonzalo CV, van der Donk WA. Investigations into the mechanism of action of sublancin. ACS Infect Dis. 2019;5(3):454-459. doi:10.1021/acsinfecdis.8b00320. View Source
• [2] Breukink E, de Kruijff B. Lipid II as a target for antibiotics. Nat Rev Drug Discov. 2006;5(4):321-332. doi:10.1038/nrd2004. View Source
• [3] Garcia De Gonzalo CV, Denham EL, Mars RAT, et al. The phosphoenolpyruvate:sugar phosphotransferase system is involved in sensitivity to the glucosylated bacteriocin sublancin. Antimicrob Agents Chemother. 2015;59(11):6844-6854. doi:10.1128/AAC.01519-15. View Source

Bacteriocin Sublancin-168: Evidence-Backed Application Scenarios for Research and Industrial Procurement

Biodefense Countermeasure Development Against Bacillus anthracis Spore Contamination

Sublancin-168's ~1000-fold selectivity for inhibiting B. cereus spore outgrowth over vegetative growth [1], combined with its 2-year room-temperature aqueous stability [2] and documented high stability against proteases and chemical degradation as noted in patent disclosures [3], makes it a uniquely qualified candidate for environmental decontamination formulations targeting Bacillus anthracis spores. The MscL-dependent mechanism [4] provides an orthogonal mode of action to nisin-based spore control, potentially enabling combination formulations with reduced resistance risk.

Anti-MRSA Therapeutic Research Leveraging Dual Direct-Killing and Immunomodulatory Mechanisms

Sublancin-168's demonstrated in vivo protection against MRSA infection in mice at 0.5–4 mg/kg i.p., with formal proof that both macrophage and neutrophil compartments are required for efficacy [5], supports its procurement for anti-MRSA drug discovery programs seeking host-directed antimicrobial strategies. The absence of cytotoxicity in RAW264.7 macrophages, mouse peritoneal macrophages, and human Caco-2 epithelial cells at concentrations up to 500 μM [6] provides a favorable initial safety margin for preclinical development.

Orthogonal Antimicrobial Susceptibility Testing Panels Using MscL as a Diagnostic Marker

Because sublancin-168 susceptibility is strictly dependent on the presence of functional MscL—a property not shared by nisin, Pep5 [4], or any other clinically used lantibiotic—sublancin-168 can be incorporated into antimicrobial susceptibility testing panels to functionally probe MscL expression in clinical Gram-positive isolates. This provides a unique diagnostic dimension that cannot be replicated by lipid II-targeting bacteriocins such as nisin or daptomycin.

Glycocin Biosynthesis and Post-Translational Modification Research Using the SunS Glycosyltransferase System

The absolute requirement of S-glucosylation at Cys22 for sublancin-168 antimicrobial activity [7], combined with the demonstrated relaxed substrate specificity of the SunS glycosyltransferase enabling generation of sublancin analogs with alternative sugar moieties [7][8], makes the sublancin-168 biosynthetic system a unique research platform for studying and engineering S-linked glycopeptide natural products. This application cannot be pursued with nisin, subtilin, or other lantibiotic systems that lack glycosylation machinery.

• [1] UniProt Consortium. P68577 (SUNA_BACSU): SPbeta prophage-derived bacteriocin sublancin-168. UniProtKB/Swiss-Prot. Accessed 2026. View Source
• [2] Paik SH, Chakicherla A, Hansen JN. Identification and characterization of the structural and transporter genes for, and the chemical and biological properties of, sublancin 168, a novel lantibiotic produced by Bacillus subtilis 168. J Biol Chem. 1998;273(36):23134-23142. PMID: 9722542. View Source
• [3] Hansen JN. Construction of a structural variant of sublancin to facilitate its isolation and use in bioremediation of environmental contamination by gram-positive spore formers such as Bacillus anthracis. US Patent Application 20020165139. Filed 2001. View Source
• [4] Kouwen TRHM, Trip EN, Denham EL, Sibbald MJJB, Dubois JYF, van Dijl JM. The large mechanosensitive channel MscL determines bacterial susceptibility to the bacteriocin sublancin 168. Antimicrob Agents Chemother. 2009;53(11):4702-4711. doi:10.1128/AAC.00439-09. View Source
• [5] Li J, Chen J, Yang G, Tao L, Wang S, Cao B, Wang M. Sublancin protects against methicillin-resistant Staphylococcus aureus infection by the combined modulation of innate immune response and microbiota. Peptides. 2021;141:170533. doi:10.1016/j.peptides.2021.170533. View Source
• [6] Wang S, Wang Q, Zeng X, Ye Q, Huang S, Yu H, Yang T, Qiao S. Use of the antimicrobial peptide sublancin with combined antibacterial and immunomodulatory activities to protect against methicillin-resistant Staphylococcus aureus infection in mice. J Agric Food Chem. 2017;65(39):8595-8605. doi:10.1021/acs.jafc.7b02592. View Source
• [7] Oman TJ, Boettcher JM, Wang H, Okalibe XN, van der Donk WA. Sublancin is not a lantibiotic but an S-linked glycopeptide. Nat Chem Biol. 2011;7(2):78-80. doi:10.1038/nchembio.509. View Source
• [8] Wang H, Oman TJ, Zhang R, Garcia De Gonzalo CV, Zhang Q, van der Donk WA. Substrate selectivity of the sublancin S-glycosyltransferase. J Am Chem Soc. 2011;133(40):15910-15913. doi:10.1021/ja2067364. View Source