Preliminary SAR of Novel Pleuromutilin–Polyamine Conjugates

While pleuromutilin (1) and its clinically available derivatives (2–6) are highly effective against Gram-positive bacteria, they remain inactive against many pathogenic Gram-negative bacteria due to the efflux pump AcrAB-TolC. In an effort to broaden the spectrum of activity of pleuromutilin (1), we developed a series of novel pleuromutilin–polyamine conjugates (9a–f) which exhibited promising intrinsic antimicrobial properties, targeting both Gram-positive and Gram-negative bacteria, including Staphylococcus aureus, methicillin-resistant S. aureus (MRSA), and Escherichia coli, along with the fungal strain Cryptococcus neoformans, and were devoid of cytotoxic and hemolytic properties with the exception of one conjugate. Furthermore, this series displayed moderate to low antibiotic potentiation of legacy antibiotics doxycycline and erythromycin, with three conjugates enhancing the activity four-fold in combination with doxycycline. In comparison to pleuromutilin (1) and tiamulin (2), one of the conjugates exhibited an expanded spectrum of activity, including Gram-negative bacteria and fungi, making it a promising option for combating microbial infections.


Introduction
The rise in antimicrobial resistance, caused by the excessive use of antibiotics and a lack of investment towards the discovery of new antibiotics, has resulted in a growing number of challenging microbial infections [1][2][3][4][5][6].The discovery of new antibiotics is one avenue for overcoming resistant microbes, while another is the discovery of compounds with little or no antimicrobial activity which can be used in combination with legacy antibiotics to restore their activities [6].
Upon its discovery in 1951, the diterpenoid antibiotic pleuromutilin (1), from the fungi Pleurotus mutilus (currently known as Clitopilus scyphoides) and P. passeckerianus (C.passeckerianus), was reported to exhibit bactericidal activity against Staphylococcus aureus and moderate activity against Mycobacterium smegmatis but showed no effect against Escherichia coli [7][8][9][10].However, the exact structure of pleuromutilin (1) (Figure 1) remained unknown until a decade after its initial discovery, after which researchers synthesized derivatives focusing on substituting the hydroxyl group at C-22 [8,10,11].This led to the development of tiamulin (2), an analogue achieved through substitution with sulphonate esters, which gained approval for veterinary use in 1979.While these derivatives, including the secondgeneration pleuromutilin antibiotic valnemulin (3), which boasted improved potency, were successful in veterinary applications, pharmaceutical companies aimed to develop pleuromutilin antibiotics for human use [8,10,11].In 1982, progress was made towards this goal with azamulin (4) entering phase I clinical trials, though it failed to pass due to poor bioavailability [8,12].Nevertheless, the breakthrough came with retapamulin (5), a potent pleuromutilin analogue approved by the FDA in 2007 for topical infections, sparking renewed interest in pleuromutilin and its derivatives [8,10,13].More recently, lefamulin (6), another pleuromutilin derivative, successfully passed phase III clinical trials for bacterial pneumonia and improved upon retapamulin (5) as it could be administered orally or intravenously [14,15].
The lack of activity of this structural class against Gram-negative bacteria was attributed to the expression of the efflux pump AcrAB-TolC, a tripartite efflux pump which exports small molecules from the bacterial cell, reducing drug efficacy [10,19,20].Encouragingly, a derivative of pleuromutilin showed promise against an E. coli strain when tested in the presence of the efflux pump inhibitor Phe-Arg-β-naphthylamide [10], indicating that inhibiting AcrAB-TolC could potentially enable pleuromutilin to be used in the treatment of Gram-negative bacterial infections.
The lack of activity of this structural class against Gram-negative bacteria was attributed to the expression of the efflux pump AcrAB-TolC, a tripartite efflux pump which exports small molecules from the bacterial cell, reducing drug efficacy [10,19,20].Encouragingly, a derivative of pleuromutilin showed promise against an E. coli strain when tested in the presence of the efflux pump inhibitor Phe-Arg-β-naphthylamide [10], indicating that inhibiting AcrAB-TolC could potentially enable pleuromutilin to be used in the treatment of Gram-negative bacterial infections.
SAR studies show that pleuromutilins with thioether or basic group linkers have enhanced antibacterial activity.Recent research suggests that increasing primary amines and positive charges in antibiotics can improve their uptake and accumulation in Gramnegative bacteria, potentially broadening their spectrum beyond Gram-positives.In our previous work on α,ω polyamines with disubstituted lipophilic head groups, we have successfully demonstrated examples that inhibit AcrAB-TolC [21].Thus, the present study aims to synthesize and evaluate a series of pleuromutilin-polyamine conjugates bearing varying polyamine chain lengths for their intrinsic antimicrobial activity and their ability to enhance the activity of doxycycline and erythromycin against Gram-negative bacteria.

Chemical Synthesis General Methods
See Supplementary .

Determination of the MICs of Antibiotics in the Presence of Synergizing Compounds
Antibiotic restoring enhancer concentrations were determined using reported protocols (see Supplementary File) [26].

Real-Time Growth Curves
Real-time growth curves were determined using reported protocols (see Supplementary File) [26].

ATP Efflux Assay
ATP efflux assays were conducted using reported protocols (see Supplementary File) [28].Tiamulin (2) was selected to be used as a positive control in biological assays and was prepared by a one pot, two-step sequence, whereby 22-O-tosylate 8 was preincubated with 2.0 equivalents of KI to give the 22-iodo derivative which was then reacted with 1.1 equivalents of 2-(diethylamino)ethane-1-thiol HCl salt in the presence of excess DIPEA to afford (2) in 47% yield (Scheme 1) (Figure S1).

Synthesis of Pleuromutilin-Polyamine Conjugates
The target set of conjugates required for the synthesis of Boc-protected polyamine scaffolds 7a-f (Figure 2), which were prepared according to protocols reported in the literature [22][23][24].The polyamines (PA) chosen covered a range of overall lengths, from spermine (PA-3-4-3) through to the longer chain length PA-3-12-3.The set was chosen to allow the exploration of chain length, lipophilicity, and positioning of positive charges on antimicrobial and cytotoxicity/hemolytic properties.Tiamulin (2) was selected to be used as a positive control in biological assays and was prepared by a one pot, two-step sequence, whereby 22-O-tosylate 8 was preincubated with 2.0 equivalents of KI to give the 22-iodo derivative which was then reacted with 1.1 equivalents of 2-(diethylamino)ethane-1-thiol HCl salt in the presence of excess DIPEA to afford (2) in 47% yield (Scheme 1) (Figure S1).

Synthesis of Pleuromutilin-Polyamine Conjugates
The target set of conjugates required for the synthesis of Boc-protected polyamine scaffolds 7a-f (Figure 2), which were prepared according to protocols reported in the literature [22][23][24].The polyamines (PA) chosen covered a range of overall lengths, from spermine (PA-3-4-3) through to the longer chain length PA-3-12-3.The set was chosen to allow the exploration of chain length, lipophilicity, and positioning of positive charges on antimicrobial and cytotoxicity/hemolytic properties.

Antimicrobial Activities
The antimicrobial activities of pleuromutilin (1), tiamulin (2), and pleuromutilin-polyamine conjugates 9a-f were determined as the minimum inhibitory concentration (MIC) against a panel of Gram-positive (S. aureus and MRSA) and Gram-negative (E.coli, P. aeruginosa, K. pneumoniae, and A. baumannii) bacterial strains and two fungal strains (C.albicans and C. neoformans) (Table 1).Tiamulin (2) demonstrated potent activity against both S. aureus and MRSA with an MIC of 3.125 and ≤0.51 µM, respectively, while exhibiting no activity against any of the Gram-negative bacteria or fungi.Intriguingly, all of the conjugates exhibited moderate to good growth inhibition of S. aureus and the Gram-negative Scheme 2. General method for the synthesis of target pleuromutilin-polyamine conjugates (9a-f).

Antimicrobial Activities
The antimicrobial activities of pleuromutilin (1), tiamulin (2), and pleuromutilinpolyamine conjugates 9a-f were determined as the minimum inhibitory concentration (MIC) against a panel of Gram-positive (S. aureus and MRSA) and Gram-negative (E.coli, P. aeruginosa, K. pneumoniae, and A. baumannii) bacterial strains and two fungal strains (C.albicans and C. neoformans) (Table 1).Tiamulin (2) demonstrated potent activity against both S. aureus and MRSA with an MIC of 3.125 and ≤0.51 µM, respectively, while exhibiting no activity against any of the Gram-negative bacteria or fungi.Intriguingly, all of the conjugates exhibited moderate to good growth inhibition of S. aureus and the Gram-negative bacterium, E. coli, with the conjugates 9b-f also demonstrating strong growth inhibition of MRSA.Antifungal activity was observed for all conjugates against C. neoformans, with MIC's ranging from ≤0.17 to 0.72 µM, while against C. albicans, activity was only observed for the longer chained PA-3-10-3 9e (MIC 10.9 µM) and PA-3-12-3 9f (MIC 21.5 µM) conjugates.Meanwhile, the PA-3-12-3 conjugate 9f was the most active, exhibiting broad spectrum activity against all tested strains, notably against S. aureus, MRSA, E. coli, A. baumannii, and C. neoformans.

Cytotoxic and Hemolytic Activities
Tiamulin (2) and pleuromutilin-polyamine conjugates 9a-f were evaluated for cytotoxicity towards HEK293 cells (human kidney epithelial cell line), reported as the concentration of compound at 50% cytotoxicity (IC 50 ), and for hemolytic activity against human red blood cells, reported as the concentration of compound at 10% hemolytic activity (HC 10 ) (Table 2).The only conjugate to exhibit either of these properties was the longest PA-3-12-3 variant (9f), which was both cytotoxic (IC 50 8.3 µM) and hemolytic (HC 10 13.2 µM). a Concentration of compound at 50% cytotoxicity on HEK293 human embryonic kidney cells with tamoxifen as the positive control (IC 50 24 µM) and values presented as the mean (n = 2); b Concentration of compound at 10% hemolytic activity on human red blood cells with melittin as the positive control (HC 10 0.95 µM) and values presented as the mean (n = 2).
When taken together, the combination of intrinsic antimicrobial activities and cytotoxicity/hemolytic activities identified the PA-3-10-3 conjugate 9e as being of particular interest.Of note, in addition to the strong activity towards Gram-positive bacteria, the conjugate also exhibited activity towards the Gram-negative bacterium E. coli.

Real-Time Growth Inhibition Assay
To investigate the kinetics of the antibacterial activity exhibited by 9e and tiamulin (2), real-time growth inhibition curves were determined against S. aureus ATCC 25922 and P. aeruginosa PAO1 via measurement of optical density at 490 nm during an 18 h culturing period.Although appearing to be qualitatively different, both compounds inhibited the growth of S. aureus (Figure 3), with the 18 h time point values in close agreement with the MIC values obtained for the two compounds (Table 1) using classical microdilution techniques.Further investigation is required to determine the factors leading to the differences observed in the sub-MIC growth curves for the two compounds.No such inhibition was observed against P. aeruginosa (Figure S8).

Antibiotic Enhancement Activities
Tiamulin (2) and polyamine conjugates 9a-f were also evaluated for their ability to enhance the antibiotic action of doxycycline towards P. aeruginosa ATCC 27853 and of erythromycin towards E. coli ATCC 25922.For the doxycycline assay, the antibiotic was present at a concentration of 4.5 µM (2 µg/mL), 20 times below the MIC value of 90 µM (40 µg/mL).Amongst the test compounds, only modest levels of enhancement were detected, with 4-fold increases in activity observed for 9a, 9b, and 9f, while the remaining compounds, including tiamulin (2), were essentially unable to enhance the antibiotic action of doxycycline (Table 3).For the erythromycin assay, the antibiotic was present at a concentration of 2.7 µM (2 µg/mL), well below the MIC value of 174 µM (128 µg/mL).Insignificant levels of enhancement of the action of the lipophilic antibiotic erythromycin was detected towards E. coli, though of note is the observance of improved activity for three of the conjugates (9a, 9c, and 9d) in comparison to tiamulin (2).

Antibiotic Enhancement Activities
Tiamulin (2) and polyamine conjugates 9a-f were also evaluated for their ability to enhance the antibiotic action of doxycycline towards P. aeruginosa ATCC 27853 and of erythromycin towards E. coli ATCC 25922.For the doxycycline assay, the antibiotic was present at a concentration of 4.5 µM (2 µg/mL), 20 times below the MIC value of 90 µM (40 µg/mL).Amongst the test compounds, only modest levels of enhancement were detected, with 4-fold increases in activity observed for 9a, 9b, and 9f, while the remaining compounds, including tiamulin (2), were essentially unable to enhance the antibiotic action of doxycycline (Table 3).For the erythromycin assay, the antibiotic was present at a concentration of 2.7 µM (2 µg/mL), well below the MIC value of 174 µM (128 µg/mL).Insignificant levels of enhancement of the action of the lipophilic antibiotic erythromycin was detected towards E. coli, though of note is the observance of improved activity for three of the conjugates (9a, 9c, and 9d) in comparison to tiamulin (2).

Membrane Perturbation Activities
Perturbation or disruption of the bacterial membrane is a validated mechanism of action of particular classes of antibiotics, including the polymyxin family of lipopeptides.In addition to being lethal to bacteria, this mechanism of membrane perturbation can also be used to enhance the action of other antibiotics, whereby a sub-MIC dose of the can facilitate the entry of the antibiotics into the microorganism [30].Membrane perturbation appears to be one mechanism of both antibiotic activity and antibiotic enhancement activity exhibited by α,ω-disubstituted polyamines, being well documented in a number of studies [21].
The mechanism of antibiotic action of 9e towards the Gram-positive bacterium Staphylococcus aureus was attributed to its ability to disrupt the bacterial membrane and cause the release of intracellular ATP.The test organisms, being S. aureus ATCC 25923 and MRSA, were briefly exposed to 9e at a single dose of 100 µg/mL, with the leakage of ATP determined by the use of a bioluminescence assay (Figure 4).While considered active in the assay, the level of disruption induced by exposure to 9e was significantly less than that induced by the positive control squalamine.Nevertheless, a 40% level of ATP release is detrimental to the survival of the bacteria.After 3 min, it will be clear that the bacteria will be killed.disruptor can facilitate the entry of the antibiotics into the microorganism [30].Membrane perturbation appears to be one mechanism of both antibiotic activity and antibiotic enhancement activity exhibited by α,ω-disubstituted polyamines, being well documented in a number of studies [21].
The mechanism of antibiotic action of 9e towards the Gram-positive bacterium Staphylococcus aureus was attributed to its ability to disrupt the bacterial membrane and cause the release of intracellular ATP.The test organisms, being S. aureus ATCC 25923 and MRSA, were briefly exposed to 9e at a single dose of 100 µg/mL, with the leakage of ATP determined by the use of a bioluminescence assay (Figure 4).While considered active in the assay, the level of disruption induced by exposure to 9e was significantly less than that induced by the positive control squalamine.Nevertheless, a 40% level of ATP release is detrimental to the survival of the bacteria.After 3 min, it will be clear that the bacteria will be killed.We also examined the ability of 9e to act as membrane disruptor of the Gram-negative bacteria P. aeruginosa PAO1 (Figure 5), using a nitrocefin colorimetric assay.This assay makes use of a chromogenic cephalosporin derivative, which in the presence of an outer membrane disruptor gains entry to the periplasm, where upon the action of β-lactamases leads to substrate hydrolysis with a detectable color change from yellow to red.While the positive control, polymyxin B, demonstrated potent ability to perturb the outer membrane of P. aeruginosa PAO1, polyamine conjugate 9e was inactive at all test doses.Tiamulin (2) was also evaluated in the same assay, also failing to exhibit any detectable membrane perturbation.We also examined the ability of 9e to act as membrane disruptor of the Gram-negative bacteria P. aeruginosa PAO1 (Figure 5), using a nitrocefin colorimetric assay.This assay makes use of a chromogenic cephalosporin derivative, which in the presence of an outer membrane disruptor gains entry to the periplasm, where upon the action of β-lactamases leads to substrate hydrolysis with a detectable color change from yellow to red.While the positive control, polymyxin B, demonstrated potent ability to perturb the outer membrane of P. aeruginosa PAO1, polyamine conjugate 9e was inactive at all test doses.Tiamulin (2) was also evaluated in the same assay, also failing to exhibit any detectable membrane perturbation.

Conclusions
Six pleuromutilin-polyamine conjugates 9a-f were successfully synthesized from pleuromutilin-22-OTs (8) and Boc-protected polyamines 7a-f, in addition to the veterinary drug tiamulin (2).These compounds primarily inhibited the growth of Gram-positive bacteria, with limited effects on Gram-negative bacteria and two fungal strains.Notably, all polyamine conjugates exhibited anti-E.coli and antifungal activity against C. neoformans, unlike pleuromutilin (1) or tiamulin (2).Cytotoxicity and hemolysis assays showed that all compounds, except the longest polyamine conjugate, 9f, were non-toxic.Moreover, when combined with doxycycline against P. aeruginosa, three conjugates (9a, 9b, and 9f) exhibited better antibiotic potentiation than tiamulin (2).Preliminary investigations suggest that while 9e exhibits intrinsic Gram-positive antibacterial by a mechanism related to membrane perturbation, the ability of the compound class to enhance the action of doxycycline towards Gram-negative bacteria does not appear to be linked to outer membrane disruption.Further studies will be required to determine their precise mechanism.

Conclusions
Six pleuromutilin-polyamine conjugates 9a-f were successfully synthesized from pleuromutilin-22-OTs (8) and Boc-protected polyamines 7a-f, in addition to the veterinary drug tiamulin (2).These compounds primarily inhibited the growth of Gram-positive bacteria, with limited effects on Gram-negative bacteria and two fungal strains.Notably, all polyamine conjugates exhibited anti-E.coli and antifungal activity against C. neoformans, unlike pleuromutilin (1) or tiamulin (2).Cytotoxicity and hemolysis assays showed that all compounds, except the longest polyamine conjugate, 9f, were non-toxic.Moreover, when combined with doxycycline against P. aeruginosa, three conjugates (9a, 9b, and 9f) exhibited better antibiotic potentiation than tiamulin (2).Preliminary investigations suggest that while 9e exhibits intrinsic Gram-positive antibacterial by a mechanism related to membrane perturbation, the ability of the compound class to enhance the action of doxycycline towards Gram-negative bacteria does not appear to be linked to outer membrane disruption.Further studies will be required to determine their precise mechanism.

a
S. aureus ATCC 25923, streptomycin (MIC 21.5 µM) and chloramphenicol (MIC 1.5-3 µM) used as positive controls and values presented as the mean (n = 3); b MRSA ATCC 43300, vancomycin (MIC 0.7 µM) used as a positive control and values presented as the mean (n = 2); c E. coli ATCC 25922, streptomycin (MIC 21.5 µM) and colistin (MIC 2 µM) used as positive controls and values presented as the mean (n = 3); d P. aeruginosa PAO1, streptomycin (MIC 21.5 µM) and colistin (MIC 1 µM) used as positive controls and values presented as the mean (n = 3); e K. pneumoniae ATCC 700603, colistin (MIC 0.2 µM) used as a positive control and values presented as the mean (n = 2); f A. baumannii ATCC 19606, colistin (MIC 0.2 µM) used as a positive control and values presented as the mean (n = 2); g C. albicans ATCC 90028, fluconazole (MIC 0.4 µM) used as a positive control and values presented as the mean (n = 2); h C. neoformans ATCC 208821, fluconazole (MIC 26 µM) used as a positive control and values presented as the mean (n = 2); i Not tested.

Microorganisms 2023 , 15 Figure 3 .
Figure 3. Bacterial growth inhibition exhibited by 2 (left) and 9e (right) against S. aureus ATCC 25923 at different concentrations.Positive control was bacteria only and negative control was media only.

Figure 3 .
Figure 3. Bacterial growth inhibition exhibited by 2 (left) and 9e (right) against S. aureus ATCC 25923 at different concentrations.Positive control was bacteria only and negative control was media only.

Figure 4 .
Figure 4. ATP release in S. aureus ATCC 25923 (left) and MRSA (right) exhibited by polyamine conjugate 9e as determined using ATP efflux assay.Squalamine (100 µg/mL) was the positive control and water was the negative control.Compounds were tested at a final concentration of 100 µg/mL, and results are reported as percentage (%) relative to positive control.

Figure 4 .
Figure 4. ATP release in S. aureus ATCC 25923 (left) and MRSA (right) exhibited by polyamine conjugate 9e as determined using ATP efflux assay.Squalamine (100 µg/mL) was the positive control and water was the negative control.Compounds were tested at a final concentration of 100 µg/mL, and results are reported as percentage (%) relative to positive control.
a Concentration (µM) required to restore doxycycline activity at 4.5 µM against P. aeruginosa ATCC 27853.Fold change shown in parentheses is the ratio between the intrinsic MIC of the test compound and the combination MIC; b Concentration (µM) required to restore erythromycin activity at 2.7 µM against E. coli ATCC 25922.Fold change shown in parentheses is the ratio between the intrinsic MIC

Table 3 .
Cont.Concentration (µM) required to restore doxycycline activity at 4.5 µM against P. aeruginosa ATCC 27853.Fold change shown in parentheses is the ratio between the intrinsic MIC of the test compound and the combination MIC; b Concentration (µM) required to restore erythromycin activity at 2.7 µM against E. coli ATCC 25922.Fold change shown in parentheses is the ratio between the intrinsic MIC of the test compound and the combination MIC. a