2.1. Identification of active hits from the clinical compound library that had good activity against MRSA strain USA300
We screened the clincal compound library for activity against log phase culture of MRSA USA300 using the scheme as described in Methods and Figure 1
. Of the 1524 compounds in the clinical compound library, we identified 34 agents that inhibited the bacterial growth completely (Table 1
, Table 2
and Table 3
). Based on the results of the primary screen, we selected these 34 drug candidates that showed antibacterial activity for rescreens, and the results were found to be reproducible. Among them, 9 drug candidates (thonzonium, cetylpyridinium, trilocarban, benzododecinium, bithionol, brilliant green, chlorquinaldol, methylbenzethonium and gentian violet) are known antiseptics that are used as topical agents (Table 2
). Since these antiseptics cannot be used internally for humans, we excluded them from further analysis. However, these antiseptics may be of interest from a mechanistic point of view and may be tested for possible use in disinfecting MRSA contaminated objects or as topical agents for preventing or treating external or surgical wounds.
Among the remaining 25 drug hits, 11 candidates (vancomycin, daptomycin, doxycycline, clindamycin, linezolid, tetracycline, minocycline, meclocycline, chlortetracycline, rifampin, formylrifamycin) are known antibiotics (Table 3
), which are currently recommended by IDSA or approved by FDA for the treatment of MRSA [5
]. These findings confirm the validity of our screens.
In addition to the known antiseptics and currently recommended antibiotics for MRSA, we identified nine drug candidates (clofazimine, thiostrepton, carbomycin, spiramycin, chloroxine, quinaldine blue, closantel, dithiazanine iodide, pyrvinium pamoate) that had good activity against MRSA USA300 (Table 1
). Among the nine drug candidates, five drug candidates (clofazimine, thiostrepton, carbomycin, spiramycin, chloroxine) are antibiotics whose primary indications are for treating other infections than S. aureus
infections, and four drug candidates (quinaldine blue, closantel, dithiazanine iodide and pyrvinium pamoate) are antimalarial or anthelminthic drugs [12
]. The results suggest that MRSA treatment might utilize repurposed drugs previously used for treating other infections or disease conditions.
2.2. Ranking the Activity of the Active Hits by MIC Testing
The most active drug candidates against the MRSA USA300 were rifamycin antibiotics, tetracycline antibiotics, and clindamycin. Doxycycline, tetracycline, minocycline, meclocycline and chlortetracycline are all tetracyclines (TCs), which had relative low MICs and showed high activities against the MRSA strain. In particular, it was worth noting that minocycline and meclocycline seemed to be more active than doxycycline, chlortetracycline and tetracycline against the strain USA300 (Table 3
). This finding may be important for choosing the more active tetracyclines such as minocycline and meclocycline over less active doxycycline or tetracycline or chlortetracycline for treating MRSA. Further studies are needed to evaluate if the differences in activity in vitro
is correlated with clinical outcome.
It is worth noting that several cell wall inhibitors including nafcillin, cefotiam, cefmenoxime, cefdinir, and moxalactam had good activity against the MRSA strain USA300 (Table 1
). It is unclear why these agents had reasonable activity against the MRSA strain. The side chains in these agents may confer some additional antibacterial property as possibly occurs in tosufloxacin [11
To ensure that the hits we identified are also active against other S. aureus strains, we tested a few selected drug candidates including clofazimine, closantel, and cefotiam on the S. aureus Newman strain in comparison with the MRSA strain USA300 in an MIC test as described in Methods. We found the same MIC for both strains for clofazamine MIC = 5 μM or 2.36 μg/mL. For closantel, a two-fold lower MIC (2.5 μM or 1.66 μg/mL) was found for the Newman strain as compared with USA300 (MIC = 5 μM or 3.31 μg/mL), which is within the allowable variation of MICs for different strains. The only significant difference is with cefotiam where the Newman strain was quite susceptible (MIC= 1.25 μM or 0.65 μg/mL) but the USA300 strain had a four-fold higher MIC (5 μM or 2.62 μg/mL), which may be related to the methicillin resistance in the USA300 strain.
Clofazimine, which was developed as an anti-tuberculosis agent in the 1950s, is currently used in the treatment of leprosy and also used to treat MDR-TB [13
]. It is of interest that we found clofazimine had good activity against MRSA USA300 in vitro. This is consistent with the previous observation that clofazimine has anti-staphylococcal activity and appears to work by disrupting the membrane [16
]. On the other hand, as an immunomodulator [17
], clofazimine might have the advantage of regulating several aspects of immunity by Wnt signaling pathway for the treatment of chronic or drug-resistant infections in vivo [18
Thiostrepton is a natural macrocyclic thiopeptide antibiotic that inhibits protein synthesis by interfering with the function of elongation factor G (EF-G) [19
]. In our previous study, thiostrepton was also found to have activity against S. aureus
Closantel was previously shown to have activity against MRSA in vitro [20
] and also in C. elegans
model in vivo [9
]. Our result is consistent with these previous findings, indicating the reliability of our assay and results. More importantly, in addition to closantal, we identified other drug candidates (Table 1
) that have not been reported as being active against MRSA in previous studies.
Quinaldine blue is an antineoplastic and antimalarial drug, and we found it a high activity against USA300 (MIC = 1.2 μM). Quinaldine (2-methylquinoline) is a heterocyclic quinoline compound that is also used in dye manufacturing, food colorants, pH indicators, and pharmaceuticals [12
]. Pyrvinium pamoate is a cyanine dye, a substituted quinoline that has been used to treat pinworm (Enterobius vermicularis) infections. It has antimalarial and anti-cryptosporidium activities [21
]. Dithiazanine iodide has broad-spectrum activity against several intestinal parasites and also brain cancer stem cells [22
]. Because both of these anthelmintic agents have poor bioavailability and are not well absorbed through the gastrointestinal tract, they may only be used as topical agents for MRSA wound infections.
While our manuscript was in preparation, we noted a similar study was published by Lau et al. [23
]. However, there are several important differences between the two studies. First, we used MRSA strain USA300, while the other study used a different MRSA strain USA100, which may have different susceptibility to antibiotics. Second, we used a larger clinical compound library consisting of 1524 compounds than the other study which used a compound library with 1163 compounds. Third, while the identified known antibiotics share most similarity, the most important difference is the new compounds identified in the two studies are quite different. We identified nine drug candidates (quinaldine blue, thiostrepton, carbomycin, closantel, dithiazanine iodide, clofazimine, pyrvinium pamoate, chloroxine, spiramycin), which do not overlap with the other study that identified only six compounds: ivacaftor, 5-fluorouracil, penfluridol, niclosamide, gemcitabine hydrochloride and floxuridine, as having good activity for MRSA strain USA100 [23
]. This is quite surprising. The above differences in bacterial strains, drug libraries, and assay conditions used may account for the very different drug candidates identified by the two different studies.
Since clinical drugs have relatively clear safety and pharmacokinetic profiles in humans and their manufacturing processes have been established, studies examining whether existing clinical drugs that show high activities against MRSA represent a rapid and efficient approach to handle drug-resistant MRSA infections. Future studies are required to evaluate the newly identified drug candidates in drug combinations, especially those drug candidates with low blood concentrations in vitro and in animal models for treatment of MRSA infections before they can be evaluated in patients.