Selection of Amikacin Hyper-Resistant Pseudomonas aeruginosa After Stepwise Exposure to High Amikacin Concentrations

Abstract

Aerosolized amikacin reaches high concentrations in lung fluids, which are well above the minimum inhibitory concentrations (MICs) of resistant strains of Pseudomonas aeruginosa. However, P. aeruginosa can gain resistance to amikacin through different cumulative mechanisms; amikacin MICs are seldom reported beyond values of 1,000 μg/ml, as tested in clinical microbiology assays. To assess how high amikacin MICs can be reached by graded exposure, four amikacin-resistant P. aeruginosa isolates were grown in a 4-step increased exposure to amikacin; derivative strains were further characterized by measuring their comparative growth rate, biofilm-forming ability, and susceptibility to other antibiotics. In addition, the mechanism underlying the MIC increase was assessed phenotypically, using a set of 12 aminoglycoside disks, and measuring the effect of Phe-Arg-β-naphthylamide, an efflux pump inhibitor. Graded exposure to amikacin increased MICs of resistant strains up to 10,000–20,000 μg/ml, without apparent fitness cost, and having variable consequences on their biofilm-forming ability, and on their susceptibility to other antibiotics. Decreased permeability may have contributed to hyper-resistance, although evidence was inconclusive and variable between strains. Amikacin-resistant P. aeruginosa is able to gain in vitro hyper-resistance with minimal changes in the specific phenotypes that were tested; the ability to achieve high-level amikacin (AMK) resistance may confound the clinical utility of this aerosolized AMK, but clinical data would be required to assess this.

Introduction

Aminoglycosides, such as amikacin (AMK), are still widely used against infections by Gram-negative bacteria in the hospital setting. The use of these drugs is limited by their nephrotoxicity; also, although still active in vitro against most Gram-negative bacteria causative of hospital-acquired pneumonia (HAP, including ventilator-associated pneumonia, VAP), aminoglycosides achieve poor penetration into the lungs.¹ Therefore, a number of devices designed to aerosolize aminoglycosides exist; amikacin, alone or in combination with other antibiotics, has been recently tested in this manner.^2–4

By administering nebulized AMK, systemic exposure is minimized, while dramatically increasing the maximum concentration (C_max) in the lung and tracheal fluids, up to ≥10,000 μg/ml.⁵ Being a concentration-dependent antibiotic, such a high C_max is of significant therapeutic value: susceptible bacteria (minimum inhibitory concentration [MIC] ≤16 μg/ml for the Clinical and Laboratory Standards Institute [CLSI]; ≤8 μg/ml for the EUCAST) could be exposed to a concentration ≥625-fold MIC. Even resistant bacteria, with MICs of 128 μg/ml, for instance, would be exposed to a nearly 80-fold MIC, making these formulations potentially useful against HAPs caused even by AMK-resistant organisms. In this study, I tested the ability of AMK-resistant Pseudomonas aeruginosa clinical isolates to increase their AMK MIC up to 10,000 μg/ml.

Materials and Methods

Strains and amikacin treatment

Four AMK-resistant P. aeruginosa isolates from our strain collection were chosen, labeled A to D; they were recently obtained from blood cultures at a children's hospital in Mexico City. Pulsed-field gel electrophoresis (PFGE) as previously described,⁶ running SpeI-digested total DNA, was used to make sure no strains were identical to each other (not shown; Supplementary Fig. S1; Supplementary Data are available online at www.liebertpub.com/mdr). Cells growing at one dilution below the MIC of each strain (i.e., 64–128 μg/ml) were inoculated 1:50 into fresh cation-adjusted Mueller-Hinton (CAMH) broth (Fluka) containing AMK 1×MIC and incubated at 35°C/24 hr; these cultures were then inoculated 1:50 into fresh media containing AMK 1.5×MIC and incubated at 35°C/24 hr, then 2×MIC, and then 2,000 μg/ml. In addition, cells growing in CAMH broth, with or without AMK at one dilution below MIC, were plated on MH agar plates containing 1,000 and 2,000 μg/ml AMK, which were incubated at 35°C/24 hr before colony counting.

Antibiotic susceptibility assays

MICs of AMK were assessed by serial twofold dilution in CAMH broth following CLSI guidelines and using reference P. aeruginosa strain ATCC 27853 as internal control in all experiments, as recommended;⁷ susceptibility to other antibiotics was assessed by disk diffusion on MH agar (using BBL disks). Underlying mechanisms of aminoglycoside resistance were phenotypically assessed using a set of disks containing 12 different aminoglycosides as reported previously.⁸ The activity of AMK, gentamicin, and ciprofloxacin was also measured in the presence of 50 μg/ml Phe-Arg-β-naphthylamide (PAβN; Sigma); PAβN is a known inhibitor of RND-type multidrug efflux pumps.⁹

Growth rate

Growth rate was measured by plating hourly dilutions of CAMH broth cultures (inoculated to ∼1 × 10⁵ colony forming units (cfu)/ml from overnight cultures and preincubated at 35°C/200 rpm for 2 hr to reach log phase), incubated for 8 hr at 35°C/200 rpm; selection coefficients, s_l ([No. of doublings of hyper-resistant strain]/[No. of doublings of original strain] −1),¹⁰ were calculated.

Virulence factors and phenotypes

Biofilm formation at 8 hr was measured using the crystal violet assay;¹¹ briefly, ∼1 × 10⁵ cells were inoculated into 100 μl CAMH broth in 96-well microplates, incubated for 8 hr at 35°C; media with suspended bacteria were removed, 0.1% crystal violet solution added, then removed, and wells washed with water; and remaining dye solubilized with ethanol and OD at 540 nm was determined. Production of rhamnolipids, and rhamnolipid-dependent swarming ability, was assessed using a rhamnolipid plate assay and a swarming plate assay, respectively, as previously described.¹² In addition, production of pyocyanin was measured spectrophotometrically in chloroform-acid extracts of overnight cultures in the PB medium (2% peptone, 1% K₂SO₄, and 0.14% MgCl₂) at 520 nm, as reported before.¹³

Results

Baseline characteristics of the four strains are in Table 1. Strains A, B, and C share AMK-resistance mechanism (and are likely related, according to their PFGE profile), while they differ in their susceptibility to other antibiotics and their biofilm-forming ability; strain D is the most distantly related of the group. After the four-step AMK passage, all strains were consistently growing at 2,000 μg/ml AMK; the procedure was repeated thrice with identical results. The new strains, named A^h, B^h, C^h, and D^h (h for hyper-resistant), were further characterized. Overall, AMK MICs for these new strains were 10,000–20,000 μg/ml (Table 2). This new phenotype was stable, as strains were passaged thrice on MH agar without antibiotics, and MICs persisted.

Table 1.

Baseline Characteristics of AMK-Resistant Pseudomonas aeruginosa Strains Used

	Strains
	A	B	C	D
AMK MIC^a (μg/ml)	128	256	128	256
Ceftazidime^b	R	R	R	R
Imipenem^b,c	I	R	R	R
Colistin^b	S	S	S	S
Ciprofloxacin^b	S	S	S	R
Piperacillin/tazobactam^b	S	R	R	R
Gentamicin^b	S	S	S	R
Aminoglycoside-resistance mechanism^d	AAC(6′)-I + APH(3′)-VI	AAC(6′)-I + APH(3′)-VI	AAC(6′)-I + APH(3′)-VI	AAC(6′)-II + APH(3′)-VI
Biofilm formation^e (SD)	0.318 (0.038)	0.534 (0.062)	1.237 (0.130)	1.269 (0.103)
“s₁”^f	0.121	0.073	0.081	0.025

AMK MIC for strain ATCC 27853 was 2 μg/ml.

By disk diffusion, interpreted per CLSI; same results for three independent assessments.

None of the strains produced carbapenemases, as determined by a Hodge test²⁰ using imipenem disks and Escherichia coli ATCC 25922 (an all-susceptible strain used for quality control in antibiotic susceptibility assays).

By disk diffusion of 12 aminoglycoside compounds (see text and Table 3).

OD₅₄₀, using the crystal violet technique, after 8-hr incubation (see text); SD, standard deviation of three independent determinations. Value for strain ATCC 27853 was 0.588 (0.073).

A “selection coefficient” (see text) was calculated for the 8-hr growth of each strain, compared to strain ATCC 27853, just for comparison purposes against a fully susceptible strain.

AMK, amikacin; CLSI, Clinical and Laboratory Standards Institute; MIC, minimum inhibitory concentration.

Table 2.

Phenotypic Changes Between Original and Amikacin Hyper-Resistant Strains

	A to A ^h	B to B ^h	C to C ^h	D to D ^h
AMK MIC^a (μg/ml)	15,000	10,000	20,000	15,000
s_l (SD)	−0.0752 (0.013)	+0.0198 (0.002)	−0.0071 (0.006)	−0.1214 (0.016)
Permeability^b	No change	Decreased	Decreased	Decreased
Ciprofloxacin MICs^c	×2	×0.5	×0.5	=
Biofilm formation^d (SD)	1.23 (0.224)	1.87 (0.319)	0.41 (0.260)	0.25 (0.102)
Pyocyanin production^e	nd	nd	nd	1.9^f

MIC for hyper-resistant strains only.

By disk diffusion of 12 aminoglycoside compounds (see text and Table 3).

Change from basal strain; results of three independent experiments, all with identical results.

Quotient of average OD₅₄₀, hyper-resistant/basal, using crystal violet method; not corrected for growth rate differences.

Quotient of average pyocyanin production, hyper-resistant/basal; nd, not detected; for strain D^h, 7.55 μg/ml (SD 0.318), for strain D, 3.82 μg/ml (SD 0.134). Strain ATCC 27853 produced 5.84 μg/ml (SD 0.112) of pyocyanin under same conditions.

p < 0.01 by χ² applied to pyocyanin amounts.

Plating ∼5 × 10⁹ previously untreated cfu of all four isolates directly on MH with 1,000 μg/ml AMK yielded no detectable growth. However, when cells growing on AMK at one dilution below MIC (64 μg/ml for strains A and C; 128 μg/ml for strains B and D) were plated directly on MH with 1,000 μg/ml AMK, 1.2–2.5% of cfu was able to grow, but none on plates with 2,000 μg/ml AMK after plating ∼5 × 10⁹ cfu.

The activity of other antibiotics changed slightly, measured as inhibitory halos; ciprofloxacin was the most affected (up or down for more than 5 mm). Ciprofloxacin MICs were assessed by serial dilution in CAMH broth: decreased one dilution in two strains (from 4 to 2 μg/ml in strains B^h and C^h), increased one dilution in one strain (from 1 to 2 μg/ml in A^h), and did not change in the one originally resistant to ciprofloxacin (32 μg/ml in D^h). Inhibitory halos of piperacillin/tazobactam increased, from 16 mm (indicative of resistance) in strain C to 19 mm (indicative of susceptibility) in strain C^h.

Growth rate, as s₁, showed slight and disparate changes (Table 2). The largest effect (−0.12) was detected in strain D, which was originally the most multiresistant, biofilm-forming strain; however, none of the growth differences was significant (p > 0.05 by χ² applied to number of doublings after 8-hr incubation). The ability to form biofilms decreased by ∼0.25–0.4× in the two originally most able biofilm formers (strains C and D) and increased by ∼1.2–2× in the less able biofilm formers (strains A and B; Table 2); none of these differences was significant (p > 0.05 by χ² applied to OD₅₄₀ values). None of the original strains produced rhamnolipids, as assessed in the plate assay and this did not change in hyper-resistant derivatives (not shown). Swarming was not detected after an 18-hr incubation and, after a 72-hr incubation, some motility was detected only in strain A, with a slight increase in derivative strain A^h (Fig. 1). Only strain D produced pyocyanin, and hyper-resistant derivative showed a twofold increase in pyocyanin production (Table 2). No detectable change in PFGE patterns was found between original and hyper-resistant strains (not shown).

FIG. 1.

Motility on swarming plates after 72-hr incubation. Strains were grown overnight in cation-adjusted Mueller-Hinton broth and inoculated on swarming plates using a toothpick; then incubated at 35°C. After 18-hr incubation, there was no growth outside of the boundaries of each colony. Squares are 13 mm by size.

Three strains (B^h, C^h, and D^h) had apparently diminished outer membrane permeability, as activity of apramycin, fortimicin, and the netilmicin derivative Sch 21562 decreased; however, there was no change in the activity of these compounds upon strain A^h (Table 3). Amikacin MICs for all four hyper-resistant strains increased twofold or more (MIC for strain C^h exceeded the maximum tested of 20,000 μg/ml) when treated with PAβN, an efflux inhibitor. Gentamicin MICs, which were also increased in hyper-resistant strains, compared to original ones (from 4 μg/ml in strains A–C to 16–128 μg/ml in strains A^h–C^h; strain D was already resistant to gentamicin) increased 2- to 4-fold by PAβN treatment. In contrast, PAβN presence decreased MICs of ciprofloxacin 2- to 4-fold (Table 4).

Table 3.

Activity of 12 Aminoglycosides in Disk-Diffusion Assays (Inhibitory Halos, in mm)

	Strains
	A	A ^h	B	B ^h	C	C ^h	D	D ^h	ATCC
Apramycin	24	23	25	16	25	17	22	12	24
Fortimicin	23	23	25	14	25	16	18	7	25
Sch 21562	23	21	23	13	23	17	19	6	26
Sch 21561	6	6	6	6	6	6	6	6	25
Gentamicin	15	14	16	9	15	9	6	6	20
Tobramycin	6	6	6	6	6	6	6	6	24
Amikacin	6	6	6	6	6	6	7	6	23
Isepamicin	6	6	6	6	6	6	6	6	24
Netilmicin	6	6	6	6	6	6	6	6	22
Sch 22591	9	6	8	6	7	6	6	6	23
Kanamycin	6	6	6	6	6	6	6	6	19
Neomycin	10	9	8	6	7	6	6	6	20

Results of a representative experiment out of three; halos of other experiments were always within ±1 mm of the values reported in this study. ATCC, results for strain ATCC 27853, used as quality control.

Table 4.

Effect of 50 μg/ml PAβN Upon the Minimum Inhibitory Concentrations (in μg/ml) of Aminoglycosides and Ciprofloxacin

	Amikacin		Gentamicin		Ciprofloxacin
Strain	−	+PAβN	−	+PAβN	−	+PAβN
A	128	128	4	2	1	0.5
A^h	15,000	>20,000	64	128	2	1
B	256	256	4	4	4	1
B^h	10,000	20,000	16	64	2	0.5
C	128	128	4	2	4	1
C^h	20,000	>20,000	128	256	2	0.5
D	256	256	256	256	32	16
D^h	15,000	>20,000	>512	>512	32	8

Identical results for three independent experiments. MICs for strain ATCC 27853 were: 2 μg/ml of amikacin, 2 μg/ml of gentamicin, and 0.5 μg/ml of ciprofloxacin, with or without PAβN.

Discussion

Results reported in this study indicate that AMK hyper-resistant P. aeruginosa (i.e., MICs ≥10,000 μg/ml) can be selected in vitro, which could potentially be capable of withstanding the very high lung concentrations achieved by aerosolization of AMK. The acquisition of the hyper-resistance phenotype seems to require a stepwise increased exposure to AMK, as direct plating of resistant isolates on 1,000-μg/ml did not produce detectable growth. Although the ability of P. aeruginosa to gain increased resistance after graded exposure has been reported before, it has seldom been explored beyond MICs typically measured in clinical settings. Resistance to AMK is reportedly low in P. aeruginosa worldwide: data from 31,504 isolates indicate that 90.6% is susceptible, although this varies widely among different regions, from 72% in Latin America and South Asia to 96.8% in the United States; most resistant strains had an AMK MIC of 128 μg/ml, and none had a MIC of 256 μg/ml or higher (www.testsurveillance.com). This is not to say that high AMK MICs in P. aeruginosa had not been reported before: from a 7% of AMK-resistant isolates with MIC >512 μg/ml¹⁴ to a report of >1,024 μg/ml MICs (reduced to 256 μg/ml by adding fosfomycin¹⁵); and of 2 out of 53 multiresistant isolates with an MIC of 2,048 μg/ml,¹⁶ for instance. However, most surveys test only maximum concentrations in the 128–256 μg/ml range; and MICs as high as 10,000 μg/ml have apparently not been reported before.

AMK hyper-resistance does not seem to have a significant fitness cost, measured as growth rate; it was related to a diminished ability to form biofilms in some strains, but an increased ability in others, although none of the differences between original and derivative hyper-resistant strains was significant. As to motility and pyocyanin production, only detected in one each of the original isolates (A and D, respectively), they actually increased in the hyper-resistant derivatives. Reduced motility and pyocyanin production have been reported in MexEF-OprN efflux overexpressing mutants.¹² Resistance to other antibiotics was also affected in a disparate way. Although the likely mechanism underlying this hyper-resistance seems to be a decreased permeability, this does not hold for all derivative strains reported in this study, suggesting that other mechanisms could be at play. Strain A^h did not display changes in the activity of other aminoglycosides, indicating a hyper-resistance mechanism different from decreased permeability. PAβN inhibits RND-type multidrug efflux pumps; MexXY-OprM is the efflux system responsible for decreased accumulation-mediated aminoglycoside resistance in P. aeruginosa.¹⁷ PAβN at 50 μg/ml reduced the MIC of gentamicin in MexXY-OprM mutants, from 4–8 to 0.5 μg/ml,¹⁸ but increased the MIC of AMK in a strain overexpressing MexXY from 0.25 to 2 μg/ml.¹⁹ Hyper-resistant strains reported in this study had all AMK MICs increased by PAβN treatment, supporting the notion that the pump inhibitor also antagonizes the action of AMK.¹⁹ However, gentamicin MICs were also increased by PAβN treatment, which either contradicts the inhibitory effect of PAβN upon MexXY-OprM efflux (should the overexpression of such efflux be the underlying hyper-resistance mechanism) or suggests an entirely different resistance mechanism. Aside, PAβN treatment reduced, as expected, the MIC of ciprofloxacin in all strains; PAβN is a nonspecific inhibitor of multidrug efflux pumps that diminishes the MIC of fluoroquinolones, even if efflux is mediated by MexXY-OprM.⁸

Overall, the nonuniformity of phenotypic changes (i.e., increases and decreases in biofilm-forming ability, growth rate, permeability, motility, pyocyanin production) accompanying the gained AMK hyper-resistance in P. aeruginosa suggests a potentially multifactorial phenomenon. This experiment included only a very limited number of isolates, and possibly only two P. aeruginosa lineages; therefore, it would not be possible to know from these results if all AMK-resistant strains may be able to gain hyper-resistance. Exploring a larger collection of clinical isolates may help address his question. Nevertheless, these changes were gained by all-tested resistant strains during a stepwise exposure to increased AMK concentrations, suggesting that the exposure to very high AMK concentrations by aerosolized administration might not ensure clinical success if used against already resistant strains. This, however, must be assessed by analyzing the results of clinical trials.

Footnotes

Disclosure Statement

The author has acted as speaker/consultant/researcher for several pharmaceutical companies, including one that is developing an amikacin nebulizator. However, this did not pose any conflict of interest with this article.

References

Chambers

H.F.

2006. Aminoglycosides. In Brunton

L.L.

, Lazo

J.S.

, and Parker

K.L.

(ed.), The pharmacological basis of therapeutics, 11th ed. McGraw-Hill, New York, pp. 1155–1171.

, Yang

, Liu

, Gutierrez

, Aymard

, Rouby

J.J.

, and Nebulized Antibiotics Study Group. 2011. Nebulized ceftazidime and amikacin in ventilator-associated pneumonia caused by Pseudomonas aeruginosa. Am. J. Respir. Crit. Care Med., 184:106–115.

Luyt

C.E.

, Clavel

, Guntupalli

, Johannigman

, Kennedy

J.I.

, Wood

, Corkery

, Gribben

, and Chastre

. 2009. Pharmacokinetics and lung delivery of PDDS-aerosolized amikacin (NKTR-061) in intubated and mechanically ventilated patients with nosocomial pneumonia. Crit. Care, 13:R200.

Montgomery

A.B.

, Vallance

, Abuan

, Tservistas

, and Davies

. 2014. A randomized double-blind placebo-controlled dose-escalation phase 1 study of aerosolized amikacin and fosfomycin delivered via the PARI investigational eFlow® inline nebulizer system in mechanically ventilated patients. J. Aerosol Med. Pulm. Drug Deliv., 27:441–448.

Niederman

M.S.

, Chastre

, Corkery

, Fink

J.B.

, Luyt

C.E.

, and Sánchez García

. 2012. BAY41-6551 achieves bactericidal tracheal aspirate amikacin concentrations in mechanically ventilated patients with Gram-negative pneumonia. Intensive Care Med., 38:263–271.

Maslow

J.N.

, Slutsky

A.M.

, and Arbeit

R.D.

. 1993. Application of pulsed-field gel electrophoresis to molecular epidemiology. In Persing

D.H.

, Smith

T.F.

, Tenover

F.C.

, and White

T.J.

(ed.), Diagnostic molecular microbiology. American Society for Microbiology, Washington DC, pp. 435–444.

CLSI. 2007. Performance standards for antimicrobial susceptibility testing, seventeent informational supplement, document M100-S17. Clinical and Laboratory Standards Institute, Wayne, PA.

Över

, Gür

, Ünal

, Miller

G.H.

, and Aminoglycoside Resistance Study Group. 2001. The changing nature of aminoglycoside resistance mechanisms and prevalence of newly recognized resistance mechanisms in Turkey. Clin. Microbiol. Infect., 7:470–478.

Chevalier

, Brein

, Mahamoud

, Malléa

, Barbe

, and Pagès

J.-M.

. 2004. Inhibitors of antibiotic efflux in resistant Enterobacter aerogenes and Klebsiella pneumoniae strains. Antimicrob. Agents Chemother., 48:1042–1046.

10.

Melnyk

A.H.

, Wong

, and Kassen

. 2014. The fitness costs of antibiotic resistance mutations. Evol. Appl. [Epub ahead of print]; DOI: 10.1111/eva.12196:

11.

O'Toole

G.A.

, Pratt

L.A.

, Watnick

P.I.

, Newman

D.K.

, Weaver

V.B.

, and Kolter

. 1999. Genetic approaches to study of biofilms. Methods Enzymol., 310:91–109.

12.

Köhler

, van Delden

, Curty

L.K.

, Hamzehpour

M.M.

, and Pechere

J.C.

. 2001. Overexpression of the MexEF-OprN multidrug efflux system affects cell-to-cell signaling in Pseudomonas aeruginosa. J. Bacteriol., 183:5213–5222.

13.

Essar

D.W.

, Eberly

, Hadero

, and Crawford

I.P.

. 1990. Identification and characterization of genes for a second anthranilate synthase in Pseudomonas aeruginosa: interchangeability of the two anthranilate synthases and evolutionary implications. J. Bacteriol., 172:884–900.

14.

Kim

J.Y.

, Park

Y.J.

, Kwon

H.J.

, Han

, Kang

M.W.

, and Woo

G.J.

. 2008. Occurrence and mechanisms of amikacin resistance and its association with β-lactamases in Pseudomonas aeruginosa: a Korean nationwide study. J. Antimicrob. Chemother., 62:479–483.

15.

Montgomery

A.B.

, Rhomberg

P.R.

, Abuan

, Walters

K.A.

, and Flamm

R.K.

. Potentiation effects of amikacin and fosfomycin against selected amikacin-nonsusceptible gram-negative respiratory tract pathogens. Antimicrob. Agents Chemother., 58:3714–3719.

16.

Hayashi

, Fukushima

, Hayashi-Nishino

, and Nishino

. 2014. Effect of methylglyoxal on multidrug-resistant Pseudomonas aeruginosa. Front. Microbiol., 5:180.

17.

Morita

, Tomida

, and Kawamura

. 2012. MexXY multidrug efflux system of Pseudomonas aeruginosa. Front. Microbiol., 3:408.

18.

Mesaros

, Glupczynski

, Avrain

, Caceres

N.E.

, Tulkens

P.M.

, and Van Bambeke

. 2007. A combined phenotypic and genotypic method for the detection of Mex efflux pumps in Pseudomonas aeruginosa. J. Antimicrob. Chemother., 59:378–386.

19.

Mao

, Warren

M.S.

, Lee

, Mistry

, and Lomovskaya

. 2001. MexXY-OprM efflux pump is required for antagonism of aminoglycosides by divalent cations in Pseudomonas aeruginosa. Antimicrob. Agents Chemother., 45:2001–2007.

20.

Anderson

K.F.

, Lonsway

D.R.

, Rasheed

J.K.

, Biddle

, Jensen

, McDougal

L.K.

, Carey

R.B.

, Thompson

, Stocker

, Limbago

, and Patel

J.B.

. 2007. Evaluation of methods to identify the Klebsiella pneumoniae carbapenemase in Enterobacteriaceae. J. Clin. Microbiol., 45:2723–2725.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.04 MB