Antimicrobial Activity of Isoniazid in Conjugation with Surface-Modified Magnetic Nanoparticles against Mycobacterium tuberculosis and Nonmycobacterial Microorganisms

Isoniazid, the choice antitubercular agent, has only been employed against Mycobacterium tuberculosis. This study evaluated if the enzyme-mimetic activities of magnetic nanoparticles could accelerate the activation process of isoniazid against mycobacterial and, more importantly, non-mycobacterial microorganisms. First, magnetic nanoparticles were synthesized and coated by lipoamino acid; then, isoniazid was conjugated to synthesized nanoparticles. Antibacterial activities of nanoconjugated isoniazid were evaluated against Mycobacterium tuberculosis and four Gram-positive and Gram-negative nonmycobacterial strains through a microdilution broth process. Results showed that the required amount of isoniazid against Mycobacterium tuberculosis would decrease to 44.8% and 16.7% in conjugation with naked and surface-modi ﬁ ed magnetic nanoparticles, respectively. Also, 32 μ g/mL and 38 μ g/mL of isoniazid in conjugation with naked and surface-modi ﬁ ed nanoparticles, respectively, could prevent the growth of Enterococcus faecalis , Escherichia coli , Pseudomonas aeruginosa , and Staphylococcus aureus . Hence, the vicinity of magnetic nanoparticles with isoniazid could declare promising aspects of isoniazid antibacterial capabilities.


Introduction
Despite years of investigations, tuberculosis (TB) caused by Mycobacterium tuberculosis (MTB) is still responsible for more than 4300 deaths every day, worldwide. The approved treatment consists of a two-phased multidrug regimen, which has faced failure in a considerable number of cases. The latest statistics of anti-TB-drug efficiency in 2017 showed that 3.5% of all new and 18% of previously treated TB patients are reported to be resistant to one or more than one of the TB drugs (MDR-TB). Furthermore, about 8.5% of these cases are categorized as extensively drug-resistant (XDR-TB) (1).
Isoniazid (INH), as an irreplaceable first-line anti-TB drug, is involved in both intensive and continuation phases of treatment (2,3). Nevertheless, in significant cases, MTB has managed to develop resistance against it. Generally, INH is known as an antitubercular agent with no known antibacterial activity against other microorganisms. However, previous studies have linked the activated INH bactericidal activities to mechanisms (4) that could be considered quite effective against other microorganisms, as well. Theoretically, INH-derived reactive species might interfere with nonmycobacterial metabolism and growth. Hence, the question is-granting that the proper enzymatic situation is provided-could activated INH reveal antibacterial activity against nonmycobacterial microorganisms?
Recently, the combination of antibiotics with magnetic nanoparticles (MNPs) reported promising perspectives for controlling drug-resistant infections (5,6). Therefore, the combination of INH with these nontoxic, biocompatible materials (7,8), could alleviate the developing resistance. MNPs also could display antibacterial activity, yet, due to unwanted oxidative effects of iron oxide on cellular structure, it is preferable to modify the particle surface by proper substances (9,10).
In this study, a combination of INH in electrostatic interaction with lipoamino acid-(LAA-) coated MNPs (INH@smMNPs) was prepared and characterized. Then, the antibacterial activities of this nanocomplex against four gram-positive and negative pathogenic bacterial strains, including Enterococcus faecalis, Staphylococcus aureus, Escherichia coli, and Pseudomonas aeruginosa were evaluated. Also, their activity against MTB was compared with free INH.

Synthesis of MNPs.
MNPs were synthesized through the chemical coprecipitation method, using FeCl 3 ·6H 2 O and FeSO 4 ·4H 2 O (2 : 1 molar ratio). An aqueous solution of iron salts was stirred at 70°C under a nitrogen atmosphere. After 1 hour, 5 mL of ammonium hydroxide 32% was added until pH 11 was obtained. After another 45 minutes of stirring, black sediments were separated magnetically. Finally, precipitates were washed 4-5 times with distilled water, dried in an oven at 80°C overnight, and then kept under N 2 atmosphere at 4°C.

Synthesis of Lipoamino Acid.
A total of 110 mmol of diethyl acetamidomalonate and 150 mmol of hexadecanoic bromide were dissolved in sodium with 85 mL of ethanol solution (3% w/v). The solution was refluxed at 60°C for 24 hours. The obtained precipitates were refluxed with 180 mL of HCL 1 M and 20 mL of dimethylformamide for another 24 hours. Ammonium hydroxide 32% was added to the obtained precipitates until pH 7 was gained. Finally, the resulting 2-amino-hexadecanoic acid was filtered and washed three times by absolute ethanol. The final product was kept at room temperature overnight to dry (7).

Synthesis of Surface-Modified MNPs (smMNPs). Coated
MNPs were synthesized through a one-step coprecipitation method, as explained. The procedure was initiated with FeCl 3 , FeSO 4 , and synthetic LAA in a molar ratio of 2 : 1 : 4 in aqueous solution.

In Vitro Evaluation of the Antimycobacterial Activity.
The test compounds were initially dissolved in DMSO to produce a concentration of 1 μg/mL. All microplate wells received 100 μL of freshly prepared Middlebrook's 7H9 medium, except the first column. 200 μL of distilled water was added to the first column of the 96-well plates to minimize the medium evaporation in the test wells during incubation. Then, 100 μL quantities of the test compounds with the desired concentrations were added to the wells of the first row (each concentration was assayed in duplicate), and serial dilution was made from the first row to the last. A microbial suspension of 100 μL that was prepared with a standard concentration of 0.5 McFarland and diluted with a 1 : 10 proportion by distilled water was added to all the test wells. Plates were then sealed and incubated for four days at 37°C. After that, 12 μL of Tween 80 10% and 20 μL of Alamar blue 0.01% were added to each well. The plates were reincubated at 37°C. The results were assessed after 24 and 48 h.  15.625 μg/mL), and 10 μL of inoculated bacteria was applied for each microorganism. The first and last rows of the microplate were left empty to achieve a better optical contrast after plate reading. The prepared microplates were incubated for 24 hours at 37°C; then, the optical density was measured at 600 nm by a microplate reader (BioTek, PowerWave XS2). This procedure was repeated three times.

In Vitro Evaluation of Antibacterial
A blank 96-well microplate consisting of 45 μL of culture media and 45 μL of the sample (as explained) was prepared. At the end, 10 μL of culture media was added to each well. The turbidity of each well in a sample microplate (a microplate with the concerned microorganism) was compared to an equivalent well in a blank microplate. Microorganism viability was calculated as follows: 2.9. Statistical Analysis. All antimicrobial studies were evaluated using IBM SPSS software. To determine the differences between the means of the results, the one-way ANOVA procedure and the post hoc Tukey test were performed. P value ≤ 0.05 was considered to be statistically significant. The experiment was repeated three times.

Results
3.1. Characterization. The FTIR spectra of synthesized LAA were recorded in the region of 750-4000 cm -1 ( Figure 1). The vibrations around 1520 cm -1 and 3302 cm -1 might be related to NH band stretching and bending vibrations, respectively. The sharp peak that appeared at around 1749 cm -1 might be attributed to the carbonyl stretching band of carboxylic acid. Also, the stretching vibration of aliphatic C-C bands appeared at around 2940 cm -1 . The FTIR spectra of the nanoconjugates of INH recorded in the region of 500-4000 cm -1 are shown in Figure 2. The stretching vibration of the Fe-O band around 570 cm -1 was able to confirm the formation of Fe 3 O 4 . The peak that appeared at around 1660 cm -1 also confirmed the existence of an acrylic carbonyl group of the INH structure. Peaks at around 3010 cm -1 and 3100 cm -1 are attributed to asymmetric and symmetric aromatic CH stretching bands, respectively, and NH stretching vibrations appeared at around 3300 cm -1 . In the case of INH@smMNP, the sharp peak at around 1749 cm -1 might be attributed to the C=O stretching bands of the LAA structure. Also, the vibrations at around 2850 cm -1 and 2930 cm -1 might be related to aliphatic stretching CH bands that exist in the LAA carbonic side chain.
TEM micrographs of drug-loaded MNPs-as shown in Figure 3-represent all spherical particles with average sizes of 11 nm and 12.93 nm for INH@MNP and INH@smMNP, respectively. To estimate the average sizes of each sample, 30 isolated particles were selected, and the mean size was reported as explained.
Hysteresis loop behavior on 0O e point is shown in Figure 4 suggesting that samples caused no coercivity and remanence, which confirms their standard superparamagnetic properties. Saturation magnetization values were measured as 43.95 and 7.76 emu/g for INH@MNP and INH@smMNP, respectively.
The X-ray pattern of the prepared compound was performed at a 2θ range of 20°to 70°at room temperature ( Figure 5). Intensity peaks of INH@MNP and INH@smMNP crystallites were located at 2θ degrees of 30°, 35.5°, 43°, 57°, and 62°equivalent to (220), (311), (400), (511), and (440) Bragg reflection, respectively, which confirms the cubic spinel structure of synthesized magnetite. The peak positions were confirmed by X'Pert HighScore software. Crystallite sizes were measured by the Scherrer formula: where λ is the wavelength of X-ray radiation, which is considered 1.54 Å for all samples. β is considered as the full width at half maximum intensity at the diffracting angle. K is a constant close to unity, which is considered 0.9 when the half-width of the peak is taken. The crystallite sizes were measured as 20.80 nm and 34.66 nm for INH@MNP and INH@smMNP, respectively.

Calculation of INH Loaded on MNPs or smMNPs.
The UV spectra of all samples showed the most absorption at 212 nm and 263 nm ( Figure 6). Therefore, UV absorption of serially diluted concentrations of INH was read at both mentioned wavelengths. Due to Beer-Lambert's law, acquired values of optical density at 263 nm were in a more acceptable range. Figure 7 shows  INH@smMNP) have inhibited more than 90% of the growth of Enterococcus faecalis. On the other hand, antibacterial activity evaluation of pure INH up to 500 μg/mL showed that the free drug is not able to efficiently prevent nonmycobacterial growth. Table 1 represents the antibacterial activity based on INH concentration in each sample.

Discussion
The incidence of drug-resistant TB cases is rising, as one of every three cases of TB in eastern Europe and also one of every seven cases in other areas has been reported to be resistant to INH, and the rate of this resistance is increasing every year (14). This study has focused on the resistance developed against isoniazid, one of the first-line drugs in TB treatment. Generally, INH enters the mycobacterium through the cytoplasm by passive infusion as a form of a prodrug. In the    Previous studies proved that conjugation of MNPs with antibiotics is helpful in combatting the developed resistance against antibiotics (5,18). Also, there is a chance that in therapeutic doses, the imposition of oxidative stress by naked iron oxide NPs causes irreparable effects on mammalian cells   9.5 38 * E. faecalis = Enterococcus faecalis. † S. aureus = Staphylococcus aureus. ‡ E. coli = Escherichia coli. § P. aeruginosa = Pseudomonas aeruginosa. * * MIC50 represents the minimum required concentration (μg/mL) of an antibacterial agent to inhibit at least 50% of bacterial growth. † † MIC90 represents the minimum required concentration (μg/mL) of an antibacterial agent to inhibit at least 90% of bacterial growth. 6 Journal of Nanomaterials (7,19). Naked MNPs, usually by adhering to the cell membrane, release reactive oxidative species (ROS) that could initiate a series of oxidation reactions leading to protein and lipid peroxidation (7,10,20). So far, different natural and synthetic structures have been applied to reform MNP surfaces. Nevertheless, the main feature of interest in the field of nanomaterial surface modification is not only the ability to cover the destructive effects but also the capability of drug efficiency enhancement. Previous studies on biomembrane models revealed that amphiphilic structures could cause stronger interactions with the cellular membranes. One of such materials is lipoamino acid which has been widely used in various aspects of pharmaceutical optimization (21)(22)(23). Also, the surfactant-like nature of LAAs would improve the water solubility of MNPs and prevent nanoparticle agglomeration (24,25). Hence, the antibacterial activity of INH was compared by conjugating with naked and surfacemodified MNPs. FTIR spectra of prepared samples confirmed the presence of expected functional groups. TEM micrographs showed that all the synthesized nanoparticles were spherical. Crystalline structures of MNPs were approved by the XRD test. Also, crystalline and particle sizes of MNPs were categorized as ultra-small superparamagnetic iron oxide nanoparticles. The obtained results from VSM showed that the magnetic property of MNP-conjugated INH was considerably decreased when particle surfaces were modified by LAA, which could be probably be resolved by more optimized synthetic methods.
The utilization of MNPs in conjugation with INH showed promising improvements against MTB. According to the reported results, the minimum inhibitory concentrations of INH, INH@MNP, and INH@smMNP are equal to 1.26 μg/mL, 65.52 μg/mL, and 32.21 μg/mL, respectively. If the calculated proportion is exerted, it could be concluded that the concentrations of pure INH in INH@MNP and INH@smMNP are equivalent to 0.87 μg/mL and 1.08 μg/mL, respectively.
As mentioned previously, INH is a prodrug; when exposed to the KatG enzyme or catalase and peroxidase, it converts to the active form of the drug. A mutation in the KatG gene or the absence of catalase and peroxidase in MTBs has been listed as one of the causes that could result in INH resistance (4,26). On the other hand, MNPs were reported to have intrinsic enzyme-mimetic properties (27,28). Conclusively, it is expected that enzyme-mimetic activities of iron oxide NPs, specifically their peroxidase-like and catalase-like activities, have resulted in activating the prodrug. Generally, no antibacterial activity has been reported for INH, except against Mycobacterial strains. Nevertheless, its activity in conjugation with MNP compounds showed that the active form of the drug could be considered effective against Enterococcus faecalis, Staphylococcus aureus, Escherichia coli, and Pseudomonas aeruginosa. Thus, utilizing MNPs in the vicinity of INH has resulted in drug activation leading to the benefit of the other antibacterial mechanisms of activated INH. The activated drugs could release oxygen-, carbon-, and nitrogen-centered free radicals, which interfere with lipid, protein, and carbohydrate synthesis, and cause nucleus leakage (4).
As expected, surface modification of nanoparticles has slightly reduced the oxidative activities of iron oxide NPs. Thus, the required amount of INH against bacteria would increase when LAA is coated on nanoparticles. Despite the weaker antibacterial activity of INH@smMNP, this paper prefers to reduce the potential unintended effects of MNPs by applying surface modification. Besides, the particular amphiphilic structure of lipoamino acid could partially mimic the natural structure of the cell membrane (29,30). Several studies have taken advantage of this characteristic to increase drug efficacy, especially antibiotics (11). Also, in the vicinity of the lung tissue (the most targeted tissue for TB), the influence of more lipophilic structures is more likely. As in the past, the use of liposomal structures was explicitly considered in the design of pharmaceutical molecules for the lung tissue (31,32).
Khoshneviszadeh et al. showed that generally, magnetic nanoparticles display a double-edged impact on microorganism growth. So that, contrary to their antibacterial effects at therapeutic concentrations, at low concentrations (more specifically lower than 62.5 μg/mL), magnetic nanoparticles have promotive impacts on microorganism growth (8). Hence, the antibacterial effects of MNPs have been partially eliminated at the studied concentrations.
Previous investigations indicated that almost half of the TB patients could be hospitalized at least once or more during the treatment, and the average length of stay was estimated at 7-13 days or more (29). This fact would strengthen the chance of other hospital-acquired infections (HAIs). The mentioned bacteria are four of the most common causes of HAIs, which during TB treatment procedure could be easily transmitted through respiratory ventilators, infectious wounds, or pneumonia (33). Hence, the declared new aspect of INH potential utility could promise that the use of its nanoconjugates may decrease the risk of these coinfections.

Conclusion
Due to the resistance developed against specific drugs such as isoniazid, the treatment of tuberculosis as a serious fatal infectious disease has faced failure in a considerable number of cases.