Abstract
Bacterial resistance threatens the utility of currently available antibiotics. Rifampicin, a cornerstone in the treatment of persistent Gram-positive infections, is prone to the development of resistance resulting from single point mutations in the antibiotic’s target, RNA polymerase. One strategy to circumvent resistance is the use of ‘hybrid’ antibiotics consisting of two covalently linked antibiotic entities. These compounds generally have two distinct cellular targets, reducing the probability of resistance development and potentially providing simplified pharmacological properties compared to combination therapies using the individual antibiotics. Here we evaluate a series of semi-synthetic hybrid antibiotics formed by linking kanglemycin A (Kang A), a rifampicin analog, and a collection of fluoroquinolones. Kang A is a natural product antibiotic which contains a novel dimethyl succinic acid moiety that offers a new attachment point for the synthesis of hybrid antibiotics. We compare the activity of the Kang A hybrids generated via the acid attachment point to a series of hybrids linked at the compound’s naphthoquinone ring system. Several hybrids exhibit activity against bacteria resistant to Kang A via the action of the partnered antibiotic, suggesting that the Kang scaffold may provide new avenues for generating antibiotics effective against drug-resistant infections.
Keywords: Hybrid antibiotics, antibiotic resistance, rifamycin, kanglemycin, Staphylococcus aureus
Rifampicin, a member of the rifamycin class of antibiotics, is commonly used in the treatment of persistent Gram-positive bacterial infections, including those caused by Mycobacterium tuberculosis1. Rifampicin functions by inhibiting the bacterial RNA polymerase (RNAP)2. Resistance to rifampicin most frequently occurs as the result of point mutations in the polymerase, with mutations in residues H526 and S531 of the RNAP RpoB subunit representing the most common resistance mutations in clinical isolates of M. tuberculosis3. Due to the risk of developing resistance, rifampicin is generally used in combination with additional antibiotics, which may create complicated dosing and pharmacological issues4. One strategy for circumventing resistance and potentially simplifying the pharmacological considerations involves the use of ‘hybrid’ antibiotics composed of two covalently linked antibiotic entities5–6. Hybrid antibiotics should retain efficacy in the event that resistance develops to either one of the two component antibiotics. Several rifamycin hybrids containing quinolones, macrolides, oxazolidinones, and nitroimidazoles are described in the patent literature7–11. One rifamycin-quinolone hybrid called TNP-2092 entered clinical trials for the treatment of GI tract infections by Helicobacter pylori, as well as prosthetic joint infections and other infections associated with biofilm formation12–14.
The kanglemycins (Kangs) represent a unique class of natural product rifamycin analogs that contain a dimethylsuccinic acid (K-acid) and a deoxysugar (K-sugar) appended at C-20 and C-27, respectively (Figure 1)15. While the K-acid and K-sugar confer activity against a common RpoBS531L mutation, the Kangs have poor activity against some other rifampicin resistance mutations including a frequently encountered RpoBH526Y mutation16–17. In the present study, we sought to circumvent these refractory mutations through the generation of Kang hybrid antibiotics. The Kangs had not been previously assessed as partners in antibiotic hybrids.
Figure 1.
Potential attachment points for generating Kang A hybrids. The K-acid and K-sugar specific to the Kangs are highlighted in green. Potential entry points for hybridization are circled. The Kang-specific K-acid entry point is circled in blue.
There are several regions of the Kang structure that could be used as attachment points for a second antibiotic (Figure 1). We first evaluated whether the Kang acid could function in this capacity, as this structural feature is unique to the Kangs. As hybridization partners, we used fluoroquinolones, which have proved effective in a variety of other hybrid antibiotics6. Fluoroquinolones interfere with DNA replication by inhibiting the bacterial DNA gyrase and topoisomerase IV18. We took advantage of the primary or secondary amine found in a number of fluoroquinolones to couple these antibiotics to the K-acid of Kang A (1), the parent compound in the Kang family (Figure 2). We expected the K-acid to function as a flexible linker between the core structures of the two antibiotics. We tested a total of nine fluoroquinolones as partners in our Kang hybrids.
Figure 2.
Synthesis of K-acid hybrids 2a-i.
The K-acid-based hybrids were screened against wild-type Staphylococcus aureus as well as rifampicin and ciprofloxacin resistant strains. The compounds were also screened against Escherichia coli as a representative Gram-negative bacterium. The rifampicin resistant S. aureus strain contained an RpoBH481Y mutation (corresponding to M. tuberculosis RpoBH526Y), which confers a very high level of resistance to rifampicin. The ciprofloxacin resistant strain contained common S80F and S84L mutations in ParC and GyrA, respectively. The compounds were also tested against an additional S. aureus strain with a RpoBH481Y/ParCS80F/GyrAS84L triple mutation. While Kang A was completely inactive against the RpoBH481Y mutant (MIC > 64 μg/mL; Table 1), its activity compared to the wild-type S. aureus strain was unaffected by the ParCS80F/GyrAS84L double mutation. Kang A did not show Gram-negative activity against E. coli. The different fluoroquinolones showed distinct activities against the wild-type and ciprofloxacin resistant strains. Against wild-type S. aureus the MICs of the fluoroquinolones ranged from 0.25 μg/mL (ciprofloxacin) to 0.0078 μg/mL (trovafloxacin). The presence of the RpoBH481Y mutation did not affect their activity. Against the strains containing the ParCS80F/GyrAS84L double mutation, the fluoroquinolones exhibited reductions in activity between 16- and 128-fold. Against E. coli, all of the fluoroquinolones exhibited MICs less than 1 μg/mL.
Table 1.
Activity of K-acid hybrids against wild-type and drug-resistant S. aureus and E. coli.
| MIC (μg/mL) | ||||||
|---|---|---|---|---|---|---|
| S. aureus | ||||||
| Compound | Hybridization partner | WT | RpoBH481Y | ParCS80F GyrAS84L |
RpoBH481Y ParCS80F GyrAS84L |
E. coli |
| 1 | - | 0.25 | >64 | 0.25 | >64 | >64 |
| Ciprofloxacin | - | 0.25 | 0.25 | 32 | 16 | 0.016 |
| Pazufloxacin | - | 0.13 | 0.13 | 4 | 4 | 0.016 |
| Garenoxacin | - | 0.016 | 0.016 | 0.5 | 0.5 | 0.016 |
| Gatifloxacin | - | 0.063 | 0.063 | 2 | 2 | 0.016 |
| Trovafloxacin | - | 0.0078 | 0.016 | 1 | 1 | 0.0078 |
| Balofloxacin | - | 0.13 | 0.13 | 4 | 4 | 0.5 |
| Besifloxacin | - | 0.031 | 0.031 | 0.5 | 0.5 | 0.25 |
| Gemifloxacin | - | 0.016 | 0.016 | 1 | 1 | 0.0078 |
| Moxifloxacin | - | 0.016 | 0.016 | 1 | 1 | 0.031 |
| 2a | Ciprofloxacin | 16 | >64 | 32 | >64 | >64 |
| 2b | Pazufloxacin | 8 | >64 | 8 | >64 | >64 |
| 2c | Garenoxacin | 8 | 16 | 8 | >64 | >64 |
| 2d | Gatifloxacin | 64 | >64 | 32 | >64 | >64 |
| 2e | Trovafloxacin | 8 | 16 | 8 | >64 | >64 |
| 2f | Balofloxacin | 32 | >64 | 16 | >64 | >64 |
| 2g | Besifloxacin | 16 | 16 | 16 | >64 | >64 |
| 2h | Gemifloxacin | 16 | >64 | 8 | >64 | >64 |
| 2i | Moxifloxacin | 32 | >64 | 32 | >64 | >64 |
MIC assays were performed in duplicate with the same results obtained for each trial.
Of the K-acid hybrids, compounds 2a-i, the most potent compounds against wild-type S. aureus were fusions of Kang A with pazufloxacin (2b), garenoxacin (2c), or trovafloxacin (2e), with MICs of 8 μg/mL (Table 1; since the hybrids differ considerably in size from the parent antibiotics, MIC values are also presented in Supplementary Table S1 in units of μM in for ease of comparison). Several of the hybrids (2c, 2e, and 2g) acquired low-level activity (MIC = 16 μg/mL) against the strain containing the RpoBH481Y mutation. This activity presumably results from the fluoroquinolone component of the hybrids, as the RpoBH481Y strain shows a high level of resistance to Kang A (MIC > 16 μg/mL). The K-acid hybrids also exhibited some activity against the ParCS80F/GyrAS84L fluoroquinolone resistant strains. Since this activity was lost against the RpoBH481Y/ParCS80F/GyrAS84L triple mutant, it likely stems from the Kang A portion of the compounds. These results show that while some of the K-acid hybrids achieved a low level of activity against strains carrying mutations conferring either Kang A or ciprofloxacin resistance, utilization of the K-acid as a point of hybridization was accompanied by a significant decrease in antibiotic potency. It is possible that this finding reflects a limit to the size of the modifications to the K-acid that are tolerated without impairing activity. Consistent with this idea, in a previous study we found that small hydrocarbon substituents appended to the acid were well-tolerated while bulkier modifications led to reduced activity19.
We next compared the activity of the K-acid hybrids to a series of hybrids generated at the C-3/C-4 position of the Kang A structure (Figure 1). This site has historically been the most common location of modifications in previously generated semi-synthetic rifamycins. We selected a subset of the fluoroquinolones used in the synthesis of the K-acid hybrids and fused them to Kang A via a benzoxazino modification introduced at C-3/C-420, generating hybrids 5a-e (Figure 3). Since we were uncertain whether the rigidity of the benzoxazino moiety would limit the activity of the hybrids, in addition to directly fusing the fluoroquinolones to the ring system, we also generated a series of hybrids with a short flexible linker between the two antibiotic components (Figure 3; compounds 6a-e). Details of the synthesis of these compounds is described in the Supplementary Information associated with this manuscript.
Figure 3.
Synthesis of a) C-3/C-4 hybrids 5a-e and b) linker-containing hybrids 6a-e.
The C-3/C-4 hybrids were screened against the same collection of strains as the K-acid hybrids. Against the wild-type S. aureus strain, the most potent C-3/C-4 fusions contained ciprofloxacin (5a) or moxifloxacin (5e), with MICs of 0.5 μg/mL, followed by garenoxacin (5c) or moxifloxacin with a flexible linker (6e), with MICs of 1 μg/mL (Table 2). Most of the hybrids had identical activities against the wild-type and ciprofloxacin resistant ParCS80F/GyrAS84L strains, suggesting that the observed activity against these strains was due to the Kang A portion of the molecules. The activity of the ciprofloxacin-containing hybrid 5a against the ParCS80F/GyrAS84L strain provided further evidence that the Kang A component was active. This hybrid showed a 64-fold improvement in its activity compared to the 32 μg/mL MIC of ciprofloxacin against the same strain. This represents a ~277-fold improvement in its MIC in units of μM (Supplementary Table S1). As expected, this activity was lost when the RpoBH481Y mutation was added in the RpoBH481Y/ParCS80F/GyrAS84L background. Several of the other hybrids (5b, 5c, 5e, and 6e) also surpassed the potency of their fluoroquinolone parents against the ParCS80F/GyrAS84L strain when the MIC data were analyzed in units of μM, again highlighting the activity of the Kang A component of these molecules (Supplementary Table S1).
Table 2.
Activity of C-3/C-4 hybrids against wild-type and drug-resistant S. aureus and E. coli.
| MIC (μg/mL)a | |||||||
|---|---|---|---|---|---|---|---|
| S. aureus | |||||||
| Compound | Hybridization partner | Flexible linker | WT | RpoBH481Y | ParCS80F GyrAS84L |
RpoBH481Y ParCS80F GyrAS84L |
E. coli |
| 5a | Ciprofloxacin | no | 0.5 | 32 | 0.5 | >64 | 8 |
| 5b | Pazufloxacin | no | 8 | 32 | 8 | >64 | 4 |
| 5c | Garenoxacin | no | 1 | 2 | 1 | >64 | 16 |
| 5d | Balofloxacin | no | 8 | 16 | 16 | >64 | 64 |
| 5e | Moxifloxacin | no | 0.5 | 8 | 0.5 | >64 | 32 |
| 6a | Ciprofloxacin | yes | 4 | 8 | 4 | >64 | 16 |
| 6b | Pazufloxacin | yes | 4 | >64 | 4 | >64 | 64 |
| 6c | Garenoxacin | yes | 4 | 8 | 16 | >64 | >64 |
| 6d | Balofloxacin | yes | 8 | 16 | 8 | >64 | >64 |
| 6e | Moxifloxacin | yes | 1 | 16 | 1 | >64 | >64 |
MIC assays were performed in duplicate with the same results obtained for each trial.
Importantly, several of the hybrids acquired activity against the Kang A resistant RpoBH481Y strain, implying that the fluoroquinolone components function in cells resistant to Kang A (Table 2). Consistent with this hypothesis, these hybrids lose activity when the ParCS80F/GyrAS84L mutations were added to the RpoBH481Y mutation in the RpoBH481Y/ParCS80F/GyrAS84L background. The most potent hybrid against the RpoBH481Y strain was the garenoxacin fusion (5c), with an MIC of 2 μg/mL. Interestingly, some of the C-3/C-4 hybrids also acquired modest activity against E. coli (Table 2), with the most potent compounds containing pazufloxacin (5b; MIC = 4 μg/mL) or ciprofloxacin (5a; MIC = 8 μg/mL). This activity may be the result of improved Gram-negative cell penetration conferred by the fluoroquinolone component of the hybrids.
The addition of a short flexible linker led to notable variations in the activities of the C-3/C-4 hybrids (Table 2). To further explore the effects of a linker, we generated six different garenoxacin-linker conjugates with flexible or semi-rigid linkers of variable length (7a-f) and used these to produce an additional set of Kang A-garenoxacin hybrids (compounds 8a-f; Figure 4). We used garenoxacin in these hybrids as the garenoxacin fusion, 5c, exhibited the best activity against the RpoBH481Y strain in our previous round of synthesis. We found that the activity of the hybrids containing the different linkers varied significantly (Table 3). While some of the hybrids, in particular those with bulkier semi-rigid linkers had poor activity against the different strains tested, compound 8d, which contained a flexible 6-carbon linker, was active against both the Kang A resistant strain (MIC = 4 μg/mL) and ciprofloxacin resistant strain (MIC = 2 μg/mL).
Figure 4.
Synthesis of a) garenoxacin-linker conjugates 7a-f and b) hybrid antibiotics 8a-f.
Table 3.
Activity of garenoxacin-containing hybrids with different linkers against wild-type and drug-resistant S. aureus and E. coli.
| MIC (μg/mL)a | |||||
|---|---|---|---|---|---|
| S. aureus | |||||
| Compound | WT | RpoBH481Y | ParCS80F GyrAS84L |
RpoBH481Y ParCS80F GyrAS84L |
E. coli |
| 8a | 16 | 64 | 16 | >64 | >64 |
| 8b | 64 | >64 | 64 | >64 | >64 |
| 8c | 4 | 8 | 4 | >64 | >64 |
| 8d | 1 | 4 | 2 | >64 | >64 |
| 8e | 8 | 16 | 16 | >64 | >64 |
| 8f | 16 | >64 | 16 | >64 | >64 |
MIC assays were performed in duplicate with the same results obtained for each trial.
Finally, all hybrid antibiotics generated in this study were assessed for cytotoxicity against HEK293 cells (Table 4). None of the hybrids exhibited a high level of cytotoxicity. Only mild inhibition of cell growth was detected for K-acid hybrids 2a, 2c, 2d, 2f, and 2i (with IC50 values between 19 and 28 μg/mL) and C-3/C-4 hybrids 5d, 6a, and 6e (with IC50 values between 36 and 44 μg/mL). The remaining compounds did not exhibit cytotoxicity at the highest concentration tested (64 μg/mL).
Table 4.
Cytotoxicity of hybrid antibiotics against HEK293 cells.
| Compound | IC50 (μg/mL)a |
|---|---|
| 2a | 23 ± 5 |
| 2b | >64 |
| 2c | 19 ± 3 |
| 2d | 21 ± 7 |
| 2e | >64 |
| 2f | 23 ± 5 |
| 2g | >64 |
| 2h | >64 |
| 2i | 28 ± 5 |
| 5a | >64 |
| 5b | >64 |
| 5c | >64 |
| 5d | 36 ± 8 |
| 5e | >64 |
| 6a | 44 ± 7 |
| 6b | >64 |
| 6c | >64 |
| 6d | >64 |
| 6e | 38 ± 7 |
| 8a | >64 |
| 8b | >64 |
| 8c | >64 |
| 8d | >64 |
| 8e | >64 |
| 8f | >64 |
Cytotoxicity assays were performed in triplicate. Results shown represent the mean ± standard deviation.
In summary, we generated a series of Kang A-fluoroquinolone hybrids with the principal aim of overcoming common resistance mutations. We used Kang A as a hybridization partner as the compound had never been tested in this capacity and because it possesses a carboxylic acid functionality that could be used as a new entry point for a attaching a second antibiotic entity. While some of the K-acid hybrids exhibited a low level of activity against both Kang A and fluoroquinolone resistant bacteria, it appears that fusion of the two antibiotics at the K-acid may be suboptimal for retention of potency. A subset of the C-3/C-4 hybrids we generated showed better activity in both resistance backgrounds. While the TNP-2092 C-3/C-4 hybrid that entered clinical trials shows even more potent activity against resistant bacteria, it has undergone an extensive optimization program involving the screening of a much larger collection of quinolones, linkers, and modes of attachment12–13. Indeed, we found that alteration of any one of these features had dramatic effects on the activity of our Kang A hybrids. Our second round of synthesis of a collection of additional garenoxacin-containing Kang A hybrids demonstrates the importance of the linker alone in determining activity and suggests that broader optimization of our compounds could lead to Kang hybrids with improved potency.
Supplementary Material
Acknowledgements
This work was supported by NIH grant 5R35GM122559 to S.F.B. The initial synthetic strategy for generating C-3/C-4 hybrids was developed with the assistance of Ascendex Scientific.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Competing interests
The authors declare no competing financial interests.
Declaration of interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
REFERENCES
- 1.Chakraborty S; Rhee KY, Tuberculosis Drug Development: History and Evolution of the Mechanism-Based Paradigm. Cold Spring Harb Perspect Med 2015, 5 (8), a021147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Campbell EA; Korzheva N; Mustaev A; Murakami K; Nair S; Goldfarb A; Darst SA, Structural mechanism for rifampicin inhibition of bacterial rna polymerase. Cell 2001, 104 (6), 901–12. [DOI] [PubMed] [Google Scholar]
- 3.Ramaswamy S; Musser JM, Molecular genetic basis of antimicrobial agent resistance in Mycobacterium tuberculosis: 1998 update. Tuber Lung Dis 1998, 79 (1), 3–29. [DOI] [PubMed] [Google Scholar]
- 4.Tiberi S; du Plessis N; Walzl G; Vjecha MJ; Rao M; Ntoumi F; Mfinanga S; Kapata N; Mwaba P; McHugh TD; Ippolito G; Migliori GB; Maeurer MJ; Zumla A, Tuberculosis: progress and advances in development of new drugs, treatment regimens, and host-directed therapies. Lancet Infect Dis 2018, 18 (7), e183–e198. [DOI] [PubMed] [Google Scholar]
- 5.Domalaon R; Idowu T; Zhanel GG; Schweizer F, Antibiotic Hybrids: the Next Generation of Agents and Adjuvants against Gram-Negative Pathogens? Clin Microbiol Rev 2018, 31 (2). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Klahn P; Bronstrup M, Bifunctional antimicrobial conjugates and hybrid antimicrobials. Nat Prod Rep 2017, 34 (7), 832–885. [DOI] [PubMed] [Google Scholar]
- 7.Ding CZ; Ma Z; Li J; Harran S; He Y; Minor KP; Kim IH; Longgood JC; Jin Y; Combrink KD (R/S) rifamycin derivatives, their preparations and pharmaceutical compositions. U.S. Patent 7,226,931, 2007.
- 8.Ding CZ; Kim IH; Wang J; Ma Z; Jin Y; Combrink KD; Lu G; Lynch AS Nitroheteroaryl-containing rifamycin derivatives. US Patent 7,678,791, 2010.
- 9.Li J; Ding CZ; Ma Z Rifamycin C-11 oxime derivatives effective against drug-resistant microbes. U.S. Patent 7,265,107, 2007.
- 10.Ding CZ; Jin Y; Combrink KD; Kim IH Quinolone carboxylic acid-substituted rifamycin derivatives. U.S. Patent 7,884,099, 2011.
- 11.Ma Z; Jin Y; Li J; Ding CZ; Minor KP; Longgood JC; Kim IH; Harran S; Combrink KD; Morris TW Rifamycin derivatives effective against drug-resistant microbes. U.S. Patent 7,247,634, 2007.
- 12.Ma Z; Lynch AS, Development of a Dual-Acting Antibacterial Agent (TNP-2092) for the Treatment of Persistent Bacterial Infections. J Med Chem 2016, 59 (14), 6645–57. [DOI] [PubMed] [Google Scholar]
- 13.Robertson GT; Bonventre EJ; Doyle TB; Du Q; Duncan L; Morris TW; Roche ED; Yan D; Lynch AS, In vitro evaluation of CBR-2092, a novel rifamycin-quinolone hybrid antibiotic: studies of the mode of action in Staphylococcus aureus. Antimicrob Agents Chemother 2008, 52 (7), 2313–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Robertson GT; Bonventre EJ; Doyle TB; Du Q; Duncan L; Morris TW; Roche ED; Yan D; Lynch AS, In vitro evaluation of CBR-2092, a novel rifamycin-quinolone hybrid antibiotic: microbiology profiling studies with staphylococci and streptococci. Antimicrob Agents Chemother 2008, 52 (7), 2324–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Wang NJ; Fu Y; Yan GH; Bao GH; Xu CF; He CH, Isolation and structure of a new ansamycin antibiotic kanglemycin A from a Nocardia. J Antibiot (Tokyo) 1988, 41 (2), 264–7. [DOI] [PubMed] [Google Scholar]
- 16.Mosaei H; Molodtsov V; Kepplinger B; Harbottle J; Moon CW; Jeeves RE; Ceccaroni L; Shin Y; Morton-Laing S; Marrs ECL; Wills C; Clegg W; Yuzenkova Y; Perry JD; Bacon J; Errington J; Allenby NEE; Hall MJ; Murakami KS; Zenkin N, Mode of Action of Kanglemycin A, an Ansamycin Natural Product that Is Active against Rifampicin-Resistant Mycobacterium tuberculosis. Mol Cell 2018, 72 (2), 263–274 e5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Peek J; Lilic M; Montiel D; Milshteyn A; Woodworth I; Biggins JB; Ternei MA; Calle PY; Danziger M; Warrier T; Saito K; Braffman N; Fay A; Glickman MS; Darst SA; Campbell EA; Brady SF, Rifamycin congeners kanglemycins are active against rifampicin-resistant bacteria via a distinct mechanism. Nat Commun 2018, 9 (1), 4147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Blondeau JM, Fluoroquinolones: mechanism of action, classification, and development of resistance. Surv Ophthalmol 2004, 49 Suppl 2, S73–8. [DOI] [PubMed] [Google Scholar]
- 19.Peek J; Xu J; Wang H; Suryavanshi S; Zimmerman M; Russo R; Park S; Perlin DS; Brady SF, A Semisynthetic Kanglemycin Shows In Vivo Efficacy against High-Burden Rifampicin Resistant Pathogens. ACS Infect Dis 2020, 6 (9), 2431–2440. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Linghu X; Segraves NL; Abramovich I; Wong N; Muller B; Neubauer N; Fantasia S; Rieth S; Bachmann S; Jansen M; Sowell CG; Askin D; Koenig SG; Gosselin F, Highly Efficient Synthesis of a Staphylococcus aureus Targeting Payload to Enable the First Antibody-Antibiotic Conjugate. Chemistry 2018, 24 (12), 2837–2840. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.