aburaviensis strain Kut-8 [ 40 ]. The reason for differential sensitivity between Gram-positive and Gram-negative bacteria could be ascribed to the morphological differences between these microorganisms; Gram-negative bacteria have an outer polysaccharide membrane carrying the structural lipopolysaccharide components.

This makes the cell wall impermeable to lipophilic compounds; the Gram-positive bacteria, on the other hand, will be more susceptible as they have only an outer peptidoglycan layer which is not an effective permeability barrier [ 41 ].

The study on the production of antimicrobial agent usually involves a search of a suitable culture medium. Among the four media tried, GSB was found to be the optimum medium for antimicrobial agent production by S.

sannanensis SU It was evident from the finding that the antimicrobial agent production by S. sannanensis SU was positively affected by the nature and type of carbon and nitrogen sources in the medium. The results showed that antimicrobial agent production was higher in medium having glucose and soyabean meal as carbon and nitrogen source respectively.

This result is quite comparable with S. griseocarneus , for which glucose was found to be suitable carbon source for the antibiotic production [ 42 ]. Considering the carbon source, simple sugar such as glucose, fructose, sucrose as sole carbon source enhanced growth as well as bioactive metabolite production rather than more complex carbons [ 43 ].

Our result is also in conformity with the findings of Singh et al. tanashiensis A2D. The results indicated the dependence of the antimicrobial agent synthesis on the medium constituents. In fact, it has been shown that the nature of carbon and nitrogen sources strongly affect antibiotic production in different organisms [ 44 ].

sannanensis SU showed a narrow range of incubation temperature for relatively good growth and antimicrobial agent production. Highest growth and antimicrobial agent production was obtained at 28°C. The temperature range adequate for good production of secondary metabolites is narrow, for example, 5 ~ 10 degrees [ 45 ].

The maximum growth as well as highest antimicrobial activity by S. sannanensis SU was achieved at pH 7 while it does not exhibit any activity at pH 9. There are reports regarding the role of pH on the production of bioactive metabolite by microorganisms [ 46 ].

Incubation period upto seven days was found to be optimum for highest antimicrobial agent production by the strain SU Griffiths and Saker [ 47 ] reported that maximum secretion of bioactive metabolites by Cylindrospermopsis raciboskii was found to occur as cultures moved into the post-exponential phase of growth, while Egorov [ 48 ] postulated that maximum antimicrobial activity was attained after reaching the maximum value of the biomass.

The results indicated that environmental factors and cultural conditions like incubation temperature, pH and incubation period were found to have profound influence on antimicrobial agent production by S.

sannanensis SU as surveyed in streptomycetes by other investigator [ 49 ]. The result for the determination of specific rate of production of antimicrobial agent by the strain SU showed to be maximum on the sixth day 0.

The strain SU showed specific potential in antimicrobial agent production. Other observations on the specific rate of production of bioactive metabolite by different microorganism have also been reported [ 50 ],[ 51 ].

The solvent extraction of the culture broth of S. sannanensis SU using ethyl acetate and subsequent purification led to the recovery of a potent antibacterial agent active against Gram-positive bacteria.

Remya and Vijaykumar [ 52 ] have reported the antimicrobial potential of compound obtained from ethyl acetate extract of Streptomyces strain RM Similarly, the extraction of antibiotics has been carried out from streptomycetes by using various solvents including ethyl acetate and methanol [ 53 ],[ 54 ].

The resultant effect of incubating the bacteria at 2 × MICs was a significantly rapid reduction in the average log of the viable cells counts. This reduction is greater than the rate of kill observed in the test bacterial strains treated with the 1 × MICs.

The significant reduction in the cell counts between 4 and 8 h of incubation period acknowledged the fact that the antimicrobial agent was highly bactericidal seeing that the bacterial colonies were almost totally wiped out after incubating for 8 h.

The growth inhibition and efficacy of the antimicrobial agent were observed to be concentration and time dependent producing distinct time-kill profiles for the tested bacterial strains. Detection and identification of members of the genus Streptomyces are of great value because they provide a rich source of antibiotics.

The emergence and dissemination of antibacterial resistance is well documented as a serious problem worldwide [ 55 ]. Smith et al. express that "The emergence of bacterial resistance threatens to return us to the era before the development of antibiotics" [ 56 ].

The perspective of rapid emergence of drug resistance among bacterial pathogens shows that the potencies of prevalent antibiotics are decreasing steadily, leading to reduced useful-period of drugs.

This situation compounds the need for the investigation of new, safe and effective antimicrobials for replacement with invalidated antimicrobials or use in antibiotic rotation programs [ 57 ]. The MICs of the antimicrobial agent against the test pathogens indicated the potent activity which highlights its prospective and could be a candidate in the generation of new antimicrobial agents and also throws a light towards the fight against drug resistant pathogens especially of the methicillin resistant Staphylococcus aureus MRSA , since this particular isolate SU inhibits only the Gram positive bacteria.

The findings of the present study showed that naturally occurring actinomycetes have a great potential to produce metabolite active against bacteria enabling discovery of new antibiotics and hence merit future studies. Further ongoing detailed characterization and structure elucidation of the bioactive metabolite may be a new entity reported from this unique untapped ecological niche.

Takahashi Y: Exploitation of new microbial resources for bioactive compounds and discovery of new actinomycetes. Article Google Scholar. BMC Microbiol. Article PubMed Central PubMed Google Scholar.

Zin NM, Sarmin NIM, Ghadin N, Basri DF, Sidik NM, Hess WM, Strobel GA: Bioactive endophytic streptomycetes from the Malay Peninsula. FEMS Microbiol Lett. Article CAS PubMed Google Scholar. Berdy J: Bioactive microbial metabolites; a personal view. J Antibiot. Goshi K, Uchida T, Lezhava A, Yamasaki M, Hiratsu K, Shinkawa H, Kinashi H: Cloning and analysis of the telomere and terminal inverted repeat of the linear chromosome of Streptomyces griseus.

J Bacteriol. Article PubMed Central CAS PubMed Google Scholar. Bordoloi G, Kumari B, Guha A, Bordoloi MJ, Yadav RNS, Roy MK, Bora TC: Isolation and structure elucidation of a New antifungal and antibacterial antibiotic produced by Streptomyces sp.

Biosci Biotech Biochem. Article CAS Google Scholar. Singh LS, Mazumder S, Bora TC: Optimisation of process parameters for growth and bioactive metabolite produced by a salt-tolerant and alkaliphilic, Streptomyces tanashiensis strain A2D. J Myc Med.

Uyeda M: Metabolites produced by actinomycetes — antiviral antibiotics and enzymes inhibitors. Yakugaku Zasshi. Baltz RH: Antibiotic discovery from actinomycetes: will a renaissance follow the decline and fall?. SIM News. Google Scholar.

Saadoun I, Gharaibeh R: The Streptomyces flora of Badia region of Jordan and its potential as a source of antibiotic-resistant bacteria. J Arid Environ. Nedialkova D, Naidenova M: Screening the antimicrobial activity of actinomycetes strains isolated from Antarctica.

J Cult Collect. Singh LS, Baruah I, Bora TC: Actinomycetes of loktak habitat: isolation and screening for antimicrobial activities. Altschul SF, Thomas LM, Alejandro AS, Zhang J, Zhang Z, Miller W, Lipman DJ: Gapped BLAST and PSI-BLAST: a new generation of protein database search programs.

Nucleic Acids Res. Thompson JD, Higgin DG, Gibson TJ: CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice.

Nucl Acid Res. Saitou N, Nei M: The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol. CAS PubMed Google Scholar. Felsenstein J: Confidence limits on phylogenies: an approach using the bootstrap. Tamura K, Stecher G, Peterson D, Filipski A, Kumar S: MEGA6: molecular evolutionary genetics analysis version 6.

Williams ST, Cross T: Isolation, Purification, Cultivation and Preservation of Actinomycetes. Methods in Microbiology. Edited by: Norris JR, Ribbons DW. Locci R: Streptomycetes and Related Genera. Bergey's Manual of Systematic Bacteriology. Edited by: Williams ST. Pridham TG, Gottlieb D: The utilization of carbon compounds by some actinomycetales as an aid for species determination.

PubMed Central CAS PubMed Google Scholar. Gordon RE, Barnett DA, Handerhan JE, Pang CH: Nocardia coeliaca, Nocardia autotrophica and the Nocardia strain. Int J Syst Bacteriol. Shirling EB, Gottlieb D: Methods for characterization of Streptomyces species.

Becker B, Lechevalier MP, Gordon RE, Lechevalier HA: Rapid differentiation between Nocardia and Streptomyces by paper chromatography of whole-cell hydrolysates. J Appl Microbiol.

CAS Google Scholar. Lechevalier MP, Lechevalier H: Chemical composition as a criterion in the classification of aerobic actinomycetes. Int J Syst Evol Microbiol. Forbes BA, Sahm DF, Weissfeld AS, Trevino EA: Bailey and Scott's Diagnostic Microbiology.

Louis, Missouri. Grammer A: Antibiotic Sensitivity and Assay Test. Microbiological Methods. Edited by: Collins CH, Lyne PM. LeDuy A, Zajic JE: A geometrical approach for differentiation of an experimental function at a point: applied to growth and product formation.

Biotechnol Bioeng. Boruwa J, Kalita K, Barua NC, Borah JC, Mazumder S, Thakur D, Gogoi DK, Bora TC: Synthesis, absolute stereochemistry and molecular design of the new antifungal and antibacterial antibiotic produced by Streptomyces sp. Bioorg Med Chem Lett. Eliopoulos GM, Eliopoulos CT: Antibiotic combinations: should they be tested?.

Clin Microbiol Rev. Eliopoulos GM, Moellering RC: Antimicrobial Combinations. Antibiotics in Laboratory Medicine. Edited by: Lorain V. Cruishank R, Duguid JP, Marmion BP, Swain RHA: Medical Microbiology.

The Practice of Medical Microbiology. Pankuch GA, Jacobs MR, Appelbaum PC: Study of comparative antipneumococcal activities of penicillin G, RP , erythromycin, sparfloxacin, ciprofloxacin, and vancomycin by using time-kill methodology.

Antimicrob Agents Chemother. Baltch AL, Smith RP, Rite WJ, Bopp LH: Comparison of inhibitory and bactericidal activities and postantibiotic effects of LY and ampicillin used singly and in combination against vancomycin-resistant Enterococcus faecium.

Gomez KA, Gomez AA: A Statistical Procedure for Agricultural Research. Pridham TG, Hesseltine CW, Benedict RG: A guide for the classification of streptomycetes according to selected groups.

Placement of strain in morphological sections. Appl Microbiol. Pankey GA, Sabath LD: Clinical relevance of bacteriostatic versus bactericidal mechanisms of action in the treatment of gram-positive bacterial infections. Clin Infect Dis. Fiedler HP, Bruntner C, Bull AT, Ward AC, Goodfellow M, Potterat O, Puder C, Mihm G: Marine actinomycetes as a source of novel secondary metabolites.

Ant v Leeuw. Phoebe CH, Combie J, Albert FG, Van TK, Cabrera J, Correira HJ, Guo Y, Lindermuth J, Rauert N, Galbraith W, Selitrenikoff CP: Extremophilic organisms as an unexplored source of antifungal compounds. Glover JB: Applications of Fungal Ecology in the Search of new Bioactive Natural Products.

The Mycota IV: Environmental, Microbial Relationships. Edited by: Wicklow DT, Söderstr m BE. Thumar JT, Dhulia K, Singh SP: Isolation and partial purification of an antimicrobial agent from halotolerant alkaliphilic Streptomyces aburaviensis strain Kut World J Microbiol Biotechnol.

Scherrer R, Gerhardt P: Molecular sieving by the Bacillus megaterium cell wall and protoplast. Cruz R, Arias ME, Soliveri J: Nutritional requirements for the production of pyrazoloisoquinolinone antibiotics by Streptomyces griseocarneus NCIMB Appl Microbiol Biotechnol.

Calvo AM, Wilson RA, Bok JW, Keller NP: Relationship between secondary metabolism and fungal development. Microbiol Mol Biol Rev. Vilches C, Mendez C, Hardisson C, Salas JA: Biosynthesis of oleandomycin by Streptomyces antibioticus : influence of nutritional conditions and development of resistance.

J Gen Microbiol. Iwai Y, Omura S: Cultural conditions for screening of new antibiotics. Guimaraes LM, Furlan RL, Garrido LM, Ventura A, Padilla G, Facciotti MC: Effect of pH on the production of the antitumor antibiotic retamycin by Streptomyces olindensis.

Biotechnol Appl Biochem. Griffiths DJ, Saker ML: The Palm island mystery disease 20 years on: a review of research on cyanotoxin cylindrospermopsin.

Inc Environ Toxicol. Egorov NS: Antibiotics a Scientific Approach. Srinivasan MC, Laxman RS, Deshpande MV: Physiology and nutritional aspects of actinomycetes: an overview. Gogoi DK, Deka Boruah HP, Saikia R, Bora TC: Optimization of process parameters for improved production of bioactive metabolite by a novel endophytic fungus Fusarium sp.

DF2 isolated from Taxus wallichiana of North East India. Tabrizi SG, Hamedi J, Mohammadipanah F: Screening of soil actinomycetes against salmonella server typhi NCTC and characterization of the prominent active strains. Iran J Microbiol.

Remya M, Vijayakumar R: Isolation and characterization of marine antagonistic actinomycetes from west coast of India. Facta Univ Med Biol. Saha M, Ghosh D, Ghosh D, Garai D, Jaisankar P, Sarkar KK, Dutta PK, Das S, Jha T, Mukherjee J: Studies on the production and purification of an antimicrobial compound and taxonomy of the producer isolated from the marine environment of the Sundarbans.

Taechowisan T, Lu C, Shen Y, Lumyong S: Secondary metabolites from endophytic Streptomyces aureofaciens emuac and their antifungal activity. Gold HS, Moellering RC: Antimicrobial-drug resistance.

New Eng J Med. Smith TL, Pearson ML, Wilcox KR: Emergence of vancomycin resistance in Staphylococcus aureus. Niederman MS: Is "crop rotation" of antibiotics the solution to a "resistant" problem in the ICU?.

Am J Resp Crit Care Med. Download references. The authors thank the Department of Biotechnology DBT , Government of India, New Delhi for the financial assistance and deep appreciation is expressed to Dr.

Yogesh S. Shouche, Scientist, National Center for Cell Science, Pune, India for 16S rRNA gene sequencing of the strain. We are very grateful to anonymous reviewers and the editor of BMC Microbiology for their kind comments and suggestions to improve this manuscript.

Institute of Bioresources and Sustainable Development, Sikkim Centre, DBT, Tadong, Gangtok, , Sikkim, India. Institute of Bioresources and Sustainable Development, DBT, Takyelpat Institutional Area, Imphal, , Manipur, India. You can also search for this author in PubMed Google Scholar.

Correspondence to Laishram Shantikumar Singh. LSS: performed the entire experiment and prepared the manuscript. HS: assisted LSS in conducting the experiment and also in preparing the manuscript. NCT: Supervised and coordinated the study.

All authors have duly checked and approved the manuscript before submission. This article is published under license to BioMed Central Ltd. Reprints and permissions.

Singh, L. Production of potent antimicrobial agent by actinomycete, Streptomyces sannanensis strain SU isolated from phoomdi in Loktak Lake of Manipur, India. BMC Microbiol 14 , Download citation. Received : 23 October Accepted : 28 October Published : 19 November Anyone you share the following link with will be able to read this content:.

Sorry, a shareable link is not currently available for this article. Provided by the Springer Nature SharedIt content-sharing initiative.

Skip to main content. Search all BMC articles Search. Download PDF. Abstract Background Actinomycetes have provided a wealth of bioactive secondary metabolites with interesting activities such as antimicrobial, antiviral and anticancer. Results An actinomycete strain isolated from Phoomdi soil of Loktak Lake of Manipur, India was identified as Streptomyces sannanensis SU Conclusion This study has therefore uncovered the potential of exploring virgin untapped habitats in the Indo-Burma biodiversity hot spot region as reservoir of promising antimicrobial metabolite producer.

Background Microbial natural products are the origin of most of the antibiotics on the market today. Culture media Starch casein agar SCA : Soluble starch, Test microorganisms The test bacterial strains used for screening of antimicrobial activity were procured from microbial type culture collection MTCC and Gene Bank of the Institute of Microbial Technology IMTECH , Chandigarh, India and are: Bacillus subtilis MTCC , Bacillus circulans MTCC , Bacillus megaterium MTCC , Staphylococcus aureus MTCC 96 , Micrococcus luteus MTCC , Mycobacterium phlei MTCC , Mycobacterium smegmatis MTCC 6, Escherichia coli MTCC , Klebsiela pneumoniae MTCC , Pseudomonas aeruginosa MTCC Isolation of the actinomycete strain During the course of screening for antibiotics from actinomycetes of Loktak habitats, the strain SU was isolated by serial dilution technique using SCA medium [ 12 ] by incubating at 28°C for 10 days from the soil sample collected from phoomdi in Loktak lake of Manipur, India.

Identification of the actinomycete strain The identification of the strain SU was carried out on the basis of the cultural characteristics, morphological, biochemical characteristics, through the courtesy of MTCC and Gene Bank of IMTECH, Chandigarh, India and also by the 16S rRNA gene sequencing carried out through the courtesy of National Centre for Cell Science, Pune, India.

Construction of phylogenetic tree The sequenced 16S rRNA gene of the strain SU was aligned with the nucleotide sequences of Streptomyces genera in GenBank database using BLAST [ 13 ]. Morphological, cultural, physiological and biochemical characteristics The morphological characteristics were studied by using cover slip method in which the culture was transferred to the base of cover slips buried in SCA medium [ 18 ] and incubated at 28°C for seven days.

Determination of antimicrobial potential of strain SU Spot inoculation on agar medium The antimicrobial activity was studied primarily by spot inoculation technique on agar medium as reported by Singh et al.

Antimicrobial agent production in liquid medium After preliminary testing for its antimicrobial potentiality, the isolate SU was further studied for the production of antimicrobial agent in liquid medium in shake flask condition.

Effect of temperature, pH and incubation period on growth and production of antimicrobial agent The effect of temperature on growth and production of antimicrobial agent was studied on GSB at different temperatures 25, 28, 31, 34 and 37°C at pH 7. Estimation of growth The biomass from the culture filtrate separated by means of centrifugation was transferred to a pre-weighed dry filter paper using a clean spatula and then placed in an oven at 50°C overnight to reach a fixed weight.

Results Characterization and taxonomy of the strain The actinomycetes strain SU isolated from soil samples of phoomdi in Loktak Lake of Manipur, India is a Gram-positive filamentous bacterium. Table 1 Morphological, physiological and biochemical characteristics of Streptomyces sannanensis SU Full size table.

Figure 1. Full size image. Table 2 Inhibition zone diameter and minimum inhibitory concentration of the active compound R f value 0. Figure 2. Figure 3. Figure 4. Figure 5. Figure 6. Figure 7. Table 3 In vitro time-kill assessment of the antimicrobial agent against the test microorganisms Full size table.

Discussion Actinomycetes have been recognized as the potential producers of metabolites such as antibiotics, growth promoting substances for plants and animals, immunomodifiers, enzyme inhibitors and many other compounds of use to man.

References Takahashi Y: Exploitation of new microbial resources for bioactive compounds and discovery of new actinomycetes. Article PubMed Central PubMed Google Scholar Zin NM, Sarmin NIM, Ghadin N, Basri DF, Sidik NM, Hess WM, Strobel GA: Bioactive endophytic streptomycetes from the Malay Peninsula.

Article CAS PubMed Google Scholar Berdy J: Bioactive microbial metabolites; a personal view. Article CAS PubMed Google Scholar Goshi K, Uchida T, Lezhava A, Yamasaki M, Hiratsu K, Shinkawa H, Kinashi H: Cloning and analysis of the telomere and terminal inverted repeat of the linear chromosome of Streptomyces griseus.

Article PubMed Central CAS PubMed Google Scholar Bordoloi G, Kumari B, Guha A, Bordoloi MJ, Yadav RNS, Roy MK, Bora TC: Isolation and structure elucidation of a New antifungal and antibacterial antibiotic produced by Streptomyces sp.

Article CAS Google Scholar Singh LS, Mazumder S, Bora TC: Optimisation of process parameters for growth and bioactive metabolite produced by a salt-tolerant and alkaliphilic, Streptomyces tanashiensis strain A2D.

Article Google Scholar Uyeda M: Metabolites produced by actinomycetes — antiviral antibiotics and enzymes inhibitors. Article Google Scholar Baltz RH: Antibiotic discovery from actinomycetes: will a renaissance follow the decline and fall?. Google Scholar Saadoun I, Gharaibeh R: The Streptomyces flora of Badia region of Jordan and its potential as a source of antibiotic-resistant bacteria.

Article Google Scholar Nedialkova D, Naidenova M: Screening the antimicrobial activity of actinomycetes strains isolated from Antarctica. Google Scholar Singh LS, Baruah I, Bora TC: Actinomycetes of loktak habitat: isolation and screening for antimicrobial activities.

Article Google Scholar Altschul SF, Thomas LM, Alejandro AS, Zhang J, Zhang Z, Miller W, Lipman DJ: Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Article PubMed Central CAS PubMed Google Scholar Thompson JD, Higgin DG, Gibson TJ: CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice.

In October , the FDA cleared superoxidized water as a high-level disinfectant FDA, personal communication, September 18, The exact mechanism by which free chlorine destroys microorganisms has not been elucidated.

Inactivation by chlorine can result from a number of factors: oxidation of sulfhydryl enzymes and amino acids; ring chlorination of amino acids; loss of intracellular contents; decreased uptake of nutrients; inhibition of protein synthesis; decreased oxygen uptake; oxidation of respiratory components; decreased adenosine triphosphate production; breaks in DNA; and depressed DNA synthesis , The actual microbicidal mechanism of chlorine might involve a combination of these factors or the effect of chlorine on critical sites Low concentrations of free available chlorine e.

Higher concentrations 1, ppm of chlorine are required to kill M. tuberculosis using the Association of Official Analytical Chemists AOAC tuberculocidal test One study reported that 25 different viruses were inactivated in 10 minutes with ppm available chlorine Several studies have demonstrated the effectiveness of diluted sodium hypochlorite and other disinfectants to inactivate HIV Chlorine ppm showed inhibition of Candida after 30 seconds of exposure In experiments using the AOAC Use-Dilution Method, ppm of free chlorine killed 10 6 —10 7 S.

aureus , Salmonella choleraesuis , and P. Because household bleach contains 5. A chlorine dioxide generator has been shown effective for decontaminating flexible endoscopes but it is not currently FDA-cleared for use as a high-level disinfectant Chlorine dioxide can be produced by mixing solutions, such as a solution of chlorine with a solution of sodium chlorite In , a chlorine dioxide product was voluntarily removed from the market when its use caused leakage of cellulose-based dialyzer membranes, which allowed bacteria to migrate from the dialysis fluid side of the dialyzer to the blood side tuberculosis , M.

chelonae , poliovirus, HIV, multidrug-resistant S. aureus , E. coli, Candida albicans , Enterococcus faecalis, P. aeruginosa in the absence of organic loading. However, the biocidal activity of this disinfectant decreased substantially in the presence of organic material e.

No bacteria or viruses were detected on artificially contaminated endoscopes after a 5-minute exposure to superoxidized water and HBV-DNA was not detected from any endoscope experimentally contaminated with HBV-positive mixed sera after a disinfectant exposure time of 7 minutes Hypochlorites are widely used in healthcare facilities in a variety of settings.

A — dilution of 5. For small spills of blood i. Because hypochlorites and other germicides are substantially inactivated in the presence of blood 63, , , , large spills of blood require that the surface be cleaned before an EPA-registered disinfectant or a final concentration solution of household bleach is applied If a sharps injury is possible, the surface initially should be decontaminated 69, , then cleaned and disinfected final concentration Extreme care always should be taken to prevent percutaneous injury.

At least ppm available chlorine for 10 minutes is recommended for decontaminating CPR training manikins Full-strength bleach has been recommended for self-disinfection of needles and syringes used for illicit-drug injection when needle-exchange programs are not available.

The difference in the recommended concentrations of bleach reflects the difficulty of cleaning the interior of needles and syringes and the use of needles and syringes for parenteral injection Clinicians should not alter their use of chlorine on environmental surfaces on the basis of testing methodologies that do not simulate actual disinfection practices , Other uses in healthcare include as an irrigating agent in endodontic treatment and as a disinfectant for manikins, laundry, dental appliances, hydrotherapy tanks 23, 41 , regulated medical waste before disposal , and the water distribution system in hemodialysis centers and hemodialysis machines Chlorine long has been used as the disinfectant in water treatment.

Water disinfection with monochloramine by municipal water-treatment plants substantially reduced the risk for healthcare—associated Legionnaires disease , Chlorine dioxide also has been used to control Legionella in a hospital water supply.

Thus, if a user wished to have a solution containing ppm of available chlorine at day 30, he or she should prepare a solution containing 1, ppm of chlorine at time 0.

Sodium hypochlorite solution does not decompose after 30 days when stored in a closed brown bottle The use of powders, composed of a mixture of a chlorine-releasing agent with highly absorbent resin, for disinfecting spills of body fluids has been evaluated by laboratory tests and hospital ward trials.

The inclusion of acrylic resin particles in formulations markedly increases the volume of fluid that can be soaked up because the resin can absorb — times its own weight of fluid, depending on the fluid consistency. One problem with chlorine-releasing granules is that they can generate chlorine fumes when applied to urine Formaldehyde is used as a disinfectant and sterilant in both its liquid and gaseous states.

Liquid formaldehyde will be considered briefly in this section, and the gaseous form is reviewed elsewhere The aqueous solution is a bactericide, tuberculocide, fungicide, virucide and sporicide 72, 82, OSHA indicated that formaldehyde should be handled in the workplace as a potential carcinogen and set an employee exposure standard for formaldehyde that limits an 8-hour time-weighted average exposure concentration of 0.

The standard includes a second permissible exposure limit in the form of a short-term exposure limit STEL of 2 ppm that is the maximum exposure allowed during a minute period Ingestion of formaldehyde can be fatal, and long-term exposure to low levels in the air or on the skin can cause asthma-like respiratory problems and skin irritation, such as dermatitis and itching.

For these reasons, employees should have limited direct contact with formaldehyde, and these considerations limit its role in sterilization and disinfection processes. Key provisions of the OSHA standard that protects workers from exposure to formaldehyde appear in Title 29 of the Code of Federal Regulations CFR Part Formaldehyde inactivates microorganisms by alkylating the amino and sulfhydral groups of proteins and ring nitrogen atoms of purine bases Varying concentrations of aqueous formaldehyde solutions destroy a wide range of microorganisms.

Four percent formaldehyde is a tuberculocidal agent, inactivating 10 4 M. tuberculosis in 2 minutes 82 , and 2. anthracis The formaldehyde solution required 2 hours of contact to achieve an inactivation factor of 10 4 , whereas glutaraldehyde required only 15 minutes.

For these reasons and others—such as its role as a suspected human carcinogen linked to nasal cancer and lung cancer , this germicide is excluded from Table 1. When it is used, , direct exposure to employees generally is limited; however, excessive exposures to formaldehyde have been documented for employees of renal transplant units , , and students in a gross anatomy laboratory Formaldehyde is used in the health-care setting to prepare viral vaccines e.

To minimize a potential health hazard to dialysis patients, the dialysis equipment must be thoroughly rinsed and tested for residual formaldehyde before use.

Paraformaldehyde, a solid polymer of formaldehyde, can be vaporized by heat for the gaseous decontamination of laminar flow biologic safety cabinets when maintenance work or filter changes require access to the sealed portion of the cabinet. Glutaraldehyde is a saturated dialdehyde that has gained wide acceptance as a high-level disinfectant and chemical sterilant Aqueous solutions of glutaraldehyde are acidic and generally in this state are not sporicidal.

Once activated, these solutions have a shelf-life of minimally 14 days because of the polymerization of the glutaraldehyde molecules at alkaline pH levels.

This polymerization blocks the active sites aldehyde groups of the glutaraldehyde molecules that are responsible for its biocidal activity. Novel glutaraldehyde formulations e. However, antimicrobial activity depends not only on age but also on use conditions, such as dilution and organic stress.

However, two studies found no difference in the microbicidal activity of alkaline and acid glutaraldehydes 73, The biocidal activity of glutaraldehyde results from its alkylation of sulfhydryl, hydroxyl, carboxyl, and amino groups of microorganisms, which alters RNA, DNA, and protein synthesis.

The mechanism of action of glutaraldehydes are reviewed extensively elsewhere , The in vitro inactivation of microorganisms by glutaraldehydes has been extensively investigated and reviewed , Spores of C. Microorganisms with substantial resistance to glutaraldehyde have been reported, including some mycobacteria M.

chelonae , Mycobacterium avium-intracellulare, M. xenopi , Methylobacterium mesophilicum , Trichosporon , fungal ascospores e. chelonae persisted in a 0. Two percent alkaline glutaraldehyde solution inactivated 10 5 M.

tuberculosis cells on the surface of penicylinders within 5 minutes at 18°C However, subsequent studies 82 questioned the mycobactericidal prowess of glutaraldehydes.

tuberculosis and compares unfavorably with alcohols, formaldehydes, iodine, and phenol Suspensions of M. avium, M. intracellulare, and M. tuberculosis estimated time to complete inactivation ~25 minutes The rate of kill was directly proportional to the temperature, and a standardized suspension of M.

tuberculosis could not be sterilized within 10 minutes An FDA-cleared chemical sterilant containing 2. tuberculosis per membrane Several investigators 55, 57, 73, 76, 80, 81, 84, have demonstrated that glutaraldehyde solutions inactivate 2. tuberculosis in 10 minutes including multidrug-resistant M.

tuberculosis and 4. tuberculosis in 20 minutes. Glutaraldehyde is commonly diluted during use, and studies showed a glutaraldehyde concentration decline after a few days of use in an automatic endoscope washer , This emphasizes the need to ensure that semicritical equipment is disinfected with an acceptable concentration of glutaraldehyde.

Data suggest that 1. Chemical test strips or liquid chemical monitors , are available for determining whether an effective concentration of glutaraldehyde is present despite repeated use and dilution.

The frequency of testing should be based on how frequently the solutions are used e. The bottle of test strips should be dated when opened and used for the period of time indicated on the bottle e.

The results of test strip monitoring should be documented. The glutaraldehyde test kits have been preliminarily evaluated for accuracy and range but the reliability has been questioned To ensure the presence of minimum effective concentration of the high-level disinfectant, manufacturers of some chemical test strips recommend the use of quality-control procedures to ensure the strips perform properly.

In December , EPA issued an order to stop the sale of all batches of this product because of efficacy data showing the product is not effective against spores and possibly other microorganisms or inanimate objects as claimed on the label Other FDA cleared glutaraldehyde sterilants that contain 2.

Glutaraldehyde is used most commonly as a high-level disinfectant for medical equipment such as endoscopes 69, , , spirometry tubing, dialyzers , transducers, anesthesia and respiratory therapy equipment , hemodialysis proportioning and dialysate delivery systems , , and reuse of laparoscopic disposable plastic trocars Glutaraldehyde is noncorrosive to metal and does not damage lensed instruments, rubber.

or plastics. Glutaraldehyde should not be used for cleaning noncritical surfaces because it is too toxic and expensive. Colitis believed caused by glutaraldehyde exposure from residual disinfecting solution in endoscope solution channels has been reported and is preventable by careful endoscope rinsing , Healthcare personnel can be exposed to elevated levels of glutaraldehyde vapor when equipment is processed in poorly ventilated rooms, when spills occur, when glutaraldehyde solutions are activated or changed, , or when open immersion baths are used.

Acute or chronic exposure can result in skin irritation or dermatitis, mucous membrane irritation eye, nose, mouth , or pulmonary symptoms , Epistaxis, allergic contact dermatitis, asthma, and rhinitis also have been reported in healthcare workers exposed to glutaraldehyde , Glutaraldehyde exposure should be monitored to ensure a safe work environment.

The silica gel tube and the DNPH-impregnated cassette are suitable for monitoring the 0. The passive badge, with a 0. ACGIH does not require a specific monitoring schedule for glutaraldehyde; however, a monitoring schedule is needed to ensure the level is less than the ceiling limit.

For example, monitoring should be done initially to determine glutaraldehyde levels, after procedural or equipment changes, and in response to worker complaints In the absence of an OSHA permissible exposure limit, if the glutaraldehyde level is higher than the ACGIH ceiling limit of 0.

Engineering and work-practice controls that can be used to resolve these problems include ducted exhaust hoods, air systems that provide 7—15 air exchanges per hour, ductless fume hoods with absorbents for the glutaraldehyde vapor, tight-fitting lids on immersion baths, personal protection e.

If engineering controls fail to maintain levels below the ceiling limit, institutions can consider the use of respirators e. In general, engineering controls are preferred over work-practice and administrative controls because they do not require active participation by the health-care worker. Even though enforcement of the OSHA ceiling limit was suspended in by the U.

Court of Appeals , limiting employee exposure to 0. If glutaraldehyde disposal through the sanitary sewer system is restricted, sodium bisulfate can be used to neutralize the glutaraldehyde and make it safe for disposal.

The literature contains several accounts of the properties, germicidal effectiveness, and potential uses for stabilized hydrogen peroxide in the health-care setting. Published reports ascribe good germicidal activity to hydrogen peroxide and attest to its bactericidal, virucidal, sporicidal, and fungicidal properties Tables 4 and 5 The FDA website lists cleared liquid chemical sterilants and high-level disinfectants containing hydrogen peroxide and their cleared contact conditions.

Hydrogen peroxide works by producing destructive hydroxyl free radicals that can attack membrane lipids, DNA, and other essential cell components. Catalase, produced by aerobic organisms and facultative anaerobes that possess cytochrome systems, can protect cells from metabolically produced hydrogen peroxide by degrading hydrogen peroxide to water and oxygen.

This defense is overwhelmed by the concentrations used for disinfection , Hydrogen peroxide is active against a wide range of microorganisms, including bacteria, yeasts, fungi, viruses, and spores 78, Bactericidal effectiveness and stability of hydrogen peroxide in urine has been demonstrated against a variety of health-care—associated pathogens; organisms with high cellular catalase activity e.

aureus , S. marcescens , and Proteus mirabilis required 30—60 minutes of exposure to 0. Synergistic sporicidal effects were observed when spores were exposed to a combination of hydrogen peroxide 5. Other studies demonstrated the antiviral activity of hydrogen peroxide against rhinovirus The product marketed as a sterilant is a premixed, ready-to-use chemical that contains 7.

The mycobactericidal activity of 7. tuberculosis after a minute exposure When the effectiveness of 7. No complaints were received from the nursing or medical staff regarding odor or toxicity. A new, rapid-acting Manufacturer data demonstrate that this solution sterilizes in 30 minutes and provides high-level disinfection in 5 minutes This product has not been used long enough to evaluate material compatibility to endoscopes and other semicritical devices, and further assessment by instrument manufacturers is needed.

Under normal conditions, hydrogen peroxide is extremely stable when properly stored e. Corneal damage from a hydrogen peroxide-soaked tonometer tip that was not properly rinsed has been reported Hydrogen peroxide also has been instilled into urinary drainage bags in an attempt to eliminate the bag as a source of bladder bacteriuria and environmental contamination Although the instillation of hydrogen peroxide into the bag reduced microbial contamination of the bag, this procedure did not reduce the incidence of catheter-associated bacteriuria As with other chemical sterilants, dilution of the hydrogen peroxide must be monitored by regularly testing the minimum effective concentration i.

Compatibility testing by Olympus America of the 7. Iodine solutions or tinctures long have been used by health professionals primarily as antiseptics on skin or tissue. Iodophors, on the other hand, have been used both as antiseptics and disinfectants. FDA has not cleared any liquid chemical sterilant or high-level disinfectants with iodophors as the main active ingredient.

An iodophor is a combination of iodine and a solubilizing agent or carrier; the resulting complex provides a sustained-release reservoir of iodine and releases small amounts of free iodine in aqueous solution.

The best-known and most widely used iodophor is povidone-iodine, a compound of polyvinylpyrrolidone with iodine. This product and other iodophors retain the germicidal efficacy of iodine but unlike iodine generally are nonstaining and relatively free of toxicity and irritancy , Several reports that documented intrinsic microbial contamination of antiseptic formulations of povidone-iodine and poloxamer-iodine caused a reappraisal of the chemistry and use of iodophors The reason for the observation that dilution increases bactericidal activity is unclear, but dilution of povidone-iodine might weaken the iodine linkage to the carrier polymer with an accompanying increase of free iodine in solution Iodine can penetrate the cell wall of microorganisms quickly, and the lethal effects are believed to result from disruption of protein and nucleic acid structure and synthesis.

Published reports on the in vitro antimicrobial efficacy of iodophors demonstrate that iodophors are bactericidal, mycobactericidal, and virucidal but can require prolonged contact times to kill certain fungi and bacterial spores 14, , , Three brands of povidone-iodine solution have demonstrated more rapid kill seconds to minutes of S.

aureus and M. chelonae at a dilution than did the stock solution The virucidal activity of 75— ppm available iodine was demonstrated against seven viruses Other investigators have questioned the efficacy of iodophors against poliovirus in the presence of organic matter and rotavirus SA in distilled or tapwater Besides their use as an antiseptic, iodophors have been used for disinfecting blood culture bottles and medical equipment, such as hydrotherapy tanks, thermometers, and endoscopes.

Antiseptic iodophors are not suitable for use as hard-surface disinfectants because of concentration differences. Iodophors formulated as antiseptics contain less free iodine than do those formulated as disinfectants Iodine or iodine-based antiseptics should not be used on silicone catheters because they can adversely affect the silicone tubing Ortho-phthalaldehyde is a high-level disinfectant that received FDA clearance in October It contains 0.

OPA solution is a clear, pale-blue liquid with a pH of 7. Tables 4 and 5. Preliminary studies on the mode of action of OPA suggest that both OPA and glutaraldehyde interact with amino acids, proteins, and microorganisms. However, OPA is a less potent cross-linking agent.

This is compensated for by the lipophilic aromatic nature of OPA that is likely to assist its uptake through the outer layers of mycobacteria and gram-negative bacteria OPA appears to kill spores by blocking the spore germination process Studies have demonstrated excellent microbicidal activity in vitro 69, , , , For example, OPA has superior mycobactericidal activity 5-log 10 reduction in 5 minutes to glutaraldehyde.

Studies have shown that by , if the global problem of antimicrobial resistance AMR is not addressed, 10 million people will die from AMR annually Mulpuru et al. The development of new antibiotics is typically abandoned due to unresolved regulatory issues and lack of commercial appeal Årdal et al.

Furthermore, industry experts have analyzed unsuccessful antibacterial programs and found that high-throughput screening of synthetic chemicals often fails to find promising compounds with expected properties Brown and Wright, Therefore, identifying and developing novel antibiotic drugs remains very challenging Ismail et al.

In recent years, computational methods, especially deep learning, have emerged as a promising avenue for drug discovery by combining domain knowledge and data-driven learning Button et al.

Among them, deep generative methods have been shown to be ideal for drug candidate discovery Gómez-Bombarelli et al. In view of this, deep generative methods can be applied to generate ideal antibiotic molecules, thereby increasing the success rate of new antibiotic discovery and reducing the cost in the discovery process.

To date, the related works of deep generative methods that have been reported in industry and academia Dao et al. In generation-based work, many models represent a molecule with the corresponding simplified molecular input line entry specification SMILES notation so that the molecule generation task is transformed into a sequence-to-sequence generation task Born et al.

However, this approach often requires large-scale pretraining, and grammatical errors will often lead to the generation of invalid SMILES, which cannot be converted into realistic molecular structures.

There are also models that are represented as molecular graphs so that the task of generating molecules is transformed into a graph-to-graph generation task. Graph-based generation methods are further divided into autoregressive Li et al. Autoregressive generation generates one atom at a time, borrowing ideas from sequence generation, and similarly, this method of generation also leads to chemically invalid intermediate graphs.

Non-autoregressive generation generates the entire graph at once; this method has also been proven to be feasible, but its disadvantage is that it requires an additional graph matching process Kwon et al.

In guided optimization-based work, the optimization can be in the generated molecular map or SMILES space or in the encoder—decoder latent space. For the former, the combination of a generative model and a property prediction model or scoring function for reinforcement learning is a common paradigm Born et al.

For the latter, many models utilize different optimization strategies or sampling strategies for molecule generation Button et al. In this article, we develop a novel generative method that combines attribute prediction models with an efficient guided search strategy to optimize and generate potent antibiotic molecules with expected antibacterial activity.

Inspired by Graph2Seq models in the natural language domain Tu and Coley, ; Xu et al. We also decouple the encoder and decoder to reduce the complexity of the model. Furthermore, the encoder and attribute prediction model Yang et al.

In this latent space, a guided search based on attribute evaluation Yang et al. We applied this method to generate antibiotic molecules with enhanced antibacterial activity compared to existing antibiotics, and these generated molecules achieved the greatest similarity to the expected functional antibiotic, although we did not explicitly constrain the similarity.

Our approach combined a designed antibacterial activity space with guided searches to produce antibiotic molecules with enhanced inhibition for the first time. We believe our work will contribute to the development of new antibiotics. We obtained the antibiotic and natural product dataset from Stokes et al.

Another are natural products isolated from plant, animal and microbial sources. After deduplication, compounds remained. For convenience, we call these compounds the main dataset. coli BW The distribution of growth inhibition values in the main dataset is shown in Supplementary Figure S1.

The MOSES set is based on the ZINC Irwin and Shoichet, Clean Leads collection. The set excludes molecules containing charged atoms or atoms other than C, N, S, O, F, Cl, Br, H or cycles of more than eight atoms.

The molecules were filtered via two filters—medicinal chemistry filters and PAINS filters Polykovskiy et al. The dataset contains 1 molecular structures. The GuacaMol dataset is derived from the ChEMBL 24 database Mendez et al. The advantage of ChEMBL is that it contains only molecules that have been synthesized and tested against biological targets Brown et al.

The overall framework of the MDAGS model is shown in Figure 1. Overview of MDAGS. A The encoder and the predictor are jointly trained to obtain the latent vector representation of the input molecular graph and the predicted growth inhibition value, respectively, and the combination of the two constructs the potential growth inhibition space.

The generator generates SMILES corresponding to the latent space optimization result after being pretrained with a large number of unlabeled molecules.

B Latent space optimization consists of two parts: i sampling and ii updating. A given seed molecule is sampled and the sampled molecule is updated to find the molecule with the global minimum growth inhibition.

Molecules satisfying the growth inhibition threshold during sampling and updating are considered successful molecules and are generated. A latent property growth inhibition space of molecular graphs is learned by joint training of the encoder and predictor and optimized based on this space.

where Q is query, K is key and V is value vectors, respectively. d k is the dimension of query and key vectors, and T is transpose of the matrix.

σ is the activation function GeLU and b is bias. We feed the latent vector into the model, and given a start marker, the model predicts the next marker in the sequence until a stop symbol is generated, thus generating a molecular SMILES. Since the latent representation of the molecular graph and the input dimension of the decoding block are inconsistent, we add a fully connected layer between the two to map the latent space vector representation to the corresponding dimension vector of the decoding block.

Although the latent space is already dimensionally reduced relative to the broad drug space, it is still a high-dimensional space. With high-dimensional data, it is difficult to mitigate the impact of dimensional disaster and to continue to explore in the desired direction of growth inhibition in the process of optimization.

Inspired by the work of Yang et al. We define a particle as a point in latent space and the fitness as the growth inhibition corresponding to that point. For a seed molecule, the initial population is obtained by sampling around the potential representation, and the population is updated in the direction of minimizing growth inhibition under the guidance of fitness until a point with the minimum global growth inhibition is found.

During the update process, points that satisfy the threshold range for growth inhibition are saved and generated see Supplementary Algorithm S1 for the detailed process. The optimization mainly includes two parts: sampling and updating.

The experimental settings for MDAGS model could be found in Supplementary Material. In machine learning, especially in the field of deep learning, the choice of hyperparameters has a great impact on the performance of the model.

Therefore, in our work, the joint model is also optimized with hyperparameters before jointly training the encoder and predictor. We apply Bayesian optimization BO to perform this process Shahriari et al. The main idea of BO is that the posterior distribution of a given complex objective function of optimization is computed with a GP.

The objective function can have just the input and output specified with unknown internal structure and mathematical properties. The posterior distribution of the objective function is modified until the posterior distribution reasonably fits the real objective distribution so that the current parameters can be better adjusted.

In our experiments, the optimized hyperparameters are the number of message-passing steps, neural network hidden size, number of feed-forward layers and dropout probability. We performed 20 optimizations on the training set of the main dataset and used the validation set to select the best hyperparameter combination.

The result of hyperparameter optimization is as follows: the number of message-passing steps is 5, the neural network hidden size is , the number of feed-forward layers is 2 and the dropout probability is 0.

The generation and optimization of molecules is often concerned with maximizing or minimizing certain properties. In this article, we jointly train an encoder and a predictor to learn the growth-inhibitory property of molecules.

In this way, the distribution of molecules in the latent space is organized by growth inhibition. We apply the results of hyperparameter optimization to set up the model and train the model to generate the latent space with the main dataset. Figure 2 shows the mapping of attribute values to latent space attribute values, visualized using t-SNE.

From Figure 2A , it can be seen that the molecular representations learned by joint training of the encoder—predictor have a more concentrated distribution for the molecules with obvious growth inhibition, and the molecules with low values are almost all clustered directly below the latent space.

To compare with this distribution, we chose the Morgan fingerprint, which is one of the classic fingerprints for feature vectorization of molecules, to compute the fingerprint of the main dataset.

We performed the same visualization and found that the calculated molecular latent characterization does not possess the ability to aggregate molecular inhibition Fig. Comparison of the latent space learned by the encoder—predictor and the latent space corresponding to the Morgan fingerprint.

A Learned by the encoder—predictor. B Calculated using Morgan fingerprints. The value of the color bar represents the strength of antibacterial activity, the stronger the antibacterial activity, the smaller the value.

We analyzed the predictive performance of the two designed fitness functions GP and FFN to ensure that the two fitness functions supported the prediction of growth inhibition Table 1. The experimental growth inhibition R 2 correlation values of the two fitness functions were 0.

It can be seen from this table that the performance of the two fitness functions is similar in these indicators, and both have good predictive performance.

Therefore, it can be determined that the two fitness functions support the prediction of growth inhibition and can be used for the subsequent potential space-guided search optimization.

The first three of these four compounds are cephalosporin antibiotics and the fourth is a tetracycline antibiotic. Among them, Compounds 1 and 3 have broad-spectrum antibacterial ability and Compounds 2 and 4 are active against gram-positive and gram-negative bacteria.

For the four selected seed compounds, predicted growth inhibition was calculated with GP and FFN and compared with the actual growth inhibition Supplementary Table S2.

The prediction results of the two functions are slightly different. In addition, we counted the number of potential threshold range representations corresponding to the four seeds generated by the GP and FFN fitness function optimization Supplementary Table S3.

The FFN fitness function has more latent space representations than the GP optimization. We believe that this is related not only to setting different thresholds but also to the prediction accuracy of the two fitness functions.

Exploring more successful latent space representations may mean exploring a wider space. Supplementary Figure S2 verifies our conjecture. The latent representations optimized with the GP fitness function are more evenly spread around the seeds, but the latent representations optimized with the FFN fitness function have been further explored in a certain direction.

Our goal is to find new molecules with stronger growth inhibition. We plotted the inhibition values corresponding to the main dataset and the latent representations optimized by the GP and FFN fitness functions Supplementary Fig.

It can be seen from the figure that the predicted inhibition value of the latent representation obtained by the optimization of the two fitness functions is significantly lower than that of the main dataset. It is important for the generative model to generate chemically valuable, novel and unique molecules.

This ensures that the generator has the ability to broadly explore new spaces beyond the existing chemical space without overfitting existing data.

Therefore, we evaluate the generative ability of the model on multiple metrics and compare it with two benchmarks—MOSES and GuacaMol. It is worth noting that although we use the MolGPT framework in this article, our generation idea and method differ from the original.

We believe these have an impact on the generative power of the model. Therefore, we also compared with the MolGPT model. On the MOSES benchmark, we compare our model with five benchmarks, CharRNN, AAE, VAE, JTN-VAE and LatentGAN, in addition to the MolGPT model Table 2.

In addition to validity, novelty and uniqueness, evaluation metrics include filters and internal diversity.

Filters are mainly used to evaluate the proportion of filtered generated molecules applied during dataset construction. The resulting molecules, while chemically valid, may contain unnecessary fragments.

Therefore, when building the MOSES training dataset, molecules with these fragments were removed, and the model is expected to avoid producing them Polykovskiy et al. Internal diversity assesses the chemical diversity within the generated set of molecules to avoid model collapse models tend to produce several fixed types of data Benhenda, The results showed that our model significantly outperformed the other models on the novelty metric, suggesting that the model can generate new molecules that are completely different from the training set.

In addition, it also has comparable performance to the baseline model on the other four metrics. The bold value in each column corresponds to the optimal value in all comparison models under this indicator.

On the GuacaMol benchmark, we compare our model with four benchmarks, Graph MCTS, AAE, ORGAN and VAE, in addition to the MolGPT model Table 3.

In addition to validity, novelty and uniqueness, the evaluation metrics also include KL divergence and FCD metrics. KL divergence measures how well one probability distribution approximates another Kullback and Leibler, This metric reflects the ability of the model to fit the distribution of the data.

FCD is a measure of how close the distribution of the generated data is to the distribution of molecules in the training set Preuer et al.

The results show that our model has the best performance on uniqueness, KL divergence and FCD. This shows that the model can generate novel new molecules while fitting the distribution of the dataset.

The bold value in each row corresponds to the optimal value in all comparison models under this indicator. For language generation models, it is necessary to train on a large dataset to learn grammar rules. Considering the small number of main datasets, direct training cannot learn SMILES grammar rules well, which affects the generation effect.

We chose to learn grammar rules on two large datasets, MOSES and GuacaMol, and then use the main dataset for fine-tuning to generate antibiotic-specific compounds. We investigated whether generators benefit from transfer learning. In addition, the test set of the main dataset was generated with three trained generators after removing the duplicate data in the MOSES and GuacaMol datasets, and the ability of the generators to regenerate was evaluated Supplementary Table S4.

The results show that the model trained directly on the master dataset performs less well than the pretrained and fine-tuned models.

Additionally, we found that the model pretrained and fine-tuned on the GuacaMol dataset outperformed its MOSES dataset counterpart. We speculate that this may because the GuacaMol dataset has a more similar distribution to the main dataset. Supplementary Figure S5 verifies our estimate.

Considering the excellent performance of the model on the GuacaMol-pretrained dataset, we use this model in subsequent generation tasks. We generated and 11 chemically usable molecules with the trained generators for GP and FFN, respectively, as successful latent representations of fitness optimization.

We explored the spatial intersection generated by the two methods Supplementary Fig. S6a ; there were repeated molecules explored in the two methods, and and 11 different molecules were generated by each exploration, respectively.

The two spanning spaces are more intuitively displayed using TMAP Supplementary Fig. S6b , which calculates the similarity based on ECFP fingerprints to construct a minimum spanning tree.

The tree is drawn and displayed by Fearun, and a point in the figure represents a molecule Probst and Reymond, Since the two modalities explore different chemical spaces, further analysis of both chemical spaces is warranted. We further explored the property distribution of the molecules in the two generative spaces with the main dataset Fig.

The distribution of the two generated datasets is almost the same as that of the main datasets of several important attributes [MW, LogP, number of hydrogen bond donors HBD , number of hydrogen bond acceptors HBA , topological polar surface area TPSA and synthetic accessibility SAS ] on the whole, but the individual properties are slightly different: for MW, LogP and HBD, the generated molecules are consistent with the main datasets.

Provided by the Springer Nature SharedIt content-sharing initiative. Skip to main content Thank you for visiting nature. nature british dental journal product news article. Download PDF. Subjects Antimicrobials Health care. Reader response number Rights and permissions Reprints and permissions.

About this article Cite this article Potent antimicrobial formula. Copy to clipboard. About the journal Journal Information Open access publishing About the Editors About the BDA CPD BDJ Books For Advertisers Jobs Contact.

However, in combination with other classes of drug indicates no interaction. Different classes of antibiotics show antibacterial effects through different model of actions.

And the cell wall or membrane disruptors, like SPR, prone to show synergistic antimicrobial activities with antibiotics targeting inner cellular components. Similarly, Zurawski et al. reported that SPR could be synergy with rifampin against XDR A.

baumannii Zurawski et al. Corbett et al. shown that SPR demonstrated synergistic activities with azithromycin to combat E. coli , K. pneumoniae , and A.

baumannii Corbett et al. Thus, based on SPR targeting the outer membrane of Gram-negative strains, we assume that is the reason why the synergistic antimicrobial effects are observed. In consistence, there are no synergistic effects observed between the combination of and other non-outer membrane targeting polymyxins, such as PB, PE, PMBN, and SPR According to the time-killing curve, in combination with SPR demonstrated synergistic bacteriostatic activities against E.

Other studies also reported similar regrowth during the time-killing curve Yau et al. Skarp et al. made hypotheses for th e bacteria regrowth phenomenon, such as drug degradation, inoculation effect, and selection for existing or emerging resistant subpopulation Skarp et al.

But triple combinations could be a resort for sustaining the bacteriostatic activities and theoretically preventing bacteria regrowth Li et al. Further studies to assess such combinations are needed. In the present study, our data show that the combination treatment of and SPR indicates antagonism effects against P.

aeruginosa Supplementary Figure S3. Antagonism is interpreted as drug interactions which produce an overall activity less than the sum of the two individual drugs, eventually leading to an attenuated activity Acar, Drug antagonism is emerging as a powerful tool to study underlying mechanisms of drug action Yeh et al.

Previously, several studies have shown that, when drug combinations indicate pharmacodynamically antagonistic effects, there are two possible combinational target relationships: one is the two drugs have the same target; the other one is the two drugs have different targets, which affect related pathways and regulate the same phenotype Jia et al.

Based on this theory, we can conduct further investigations to better understand the action mechanism of In vitro and in vivo toxicity assays all show that alone or in combination with SPR has good safety.

According to this, we evaluated the synergistic therapeutic effects between and SPR in vivo. Skin and soft tissue infections have been prevalent and increased significantly since the mids, leading to an increasing clinical and economic burden on health facilities and hospitalized patients Amin et al.

A Report from the SENTRY Antimicrobial Monitoring program claimed that the occurrence rate of E. coli isolated from skin and soft tissue infections is 6. Besides, to minimize the differences derived from the pharmacokinetics of different species, we constructed subcutaneous abscess infection murine model.

As we expected, synergistic antimicrobial activities were observed by the combinational treatment of and SPR in vivo. The water solubility enhancement is necessary and can be optimized by molecule-modified, nano-technology, and complexation with other compounds such as large ring cyclodextrins and protein hydrolysate Inada et al.

In conclusion, we identified a novel bioactive molecule, , which shows antibacterial activity against E. Besides, combined with SPR exhibited significant synergistic combinational effects against both standard strains and drug-resistant clinical strains.

In addition, the combination of and SPR significantly decreases the viable bacterial loads in the E. coli -induced subcutaneous abscess model in vivo. Our results presented here are encouraging due to the bright future of SPR in combination with , which could be a promising treatment in the fight against E.

coli and its drug-resistant strains-related infections, but the underlined mechanism still needs to be investigated. The action mechanism of SPR is solely destroying the outer membrane structure, often potentiating the penetration of co-administered antibiotics to exhibit synergistic antimicrobial effects against Gram-negative bacteria.

Thus, we hypothesize that could target inside components of the bacterial cells, like protein synthesis, DNA replication, ROS production, and so on. And further investigations to validate our hypothesis about the mechanisms will be studied in the future. The animal study was reviewed and approved by the Ethic Committee of the Third Xiangya Hospital, Central South University NO: sydw SP and WY designed and perfected the research.

LL conducted most of the experiments and wrote this manuscript. SP, LS, LY, and LZ analyzed research results.

LL, YY, and ZL conducted the experiments and data collection. SP and WY assisted with revision and editing the manuscript. All authors contributed to the article and approved the submitted version.

This study was supported by the National Natural Science Foundation of China, grant numbers and , the Natural Science Foundation of Hunan Province, grant numbers JJ and JJ, and the Key Research and Development Program of Hunan Province of China, grant number SK We are grateful to Li Min Shanghai Jiaotong University, Shanghai, China for providing bacterial strains.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers.

Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Abernethy, J. Thirty day all-cause mortality in patients with Escherichia coli bacteraemia in England. doi: PubMed Abstract CrossRef Full Text Google Scholar. Acar, J.

Antibiotic synergy and antagonism. North Am. CrossRef Full Text Google Scholar. Akhoundsadegh, N. Outer membrane interaction kinetics of new Polymyxin B analogs in gram-negative bacilli.

Agents Chemother. Alamneh, Y. Minocycline and the SPR adjuvant are an efficacious antibacterial combination for infections. Antibiotics Alfei, S. Bactericidal activity of a self-biodegradable lysine-containing Dendrimer against clinical isolates of genus.

Amin, A. Hospitalist perspective on the treatment of skin and soft tissue infections. Mayo Clin. Antimicrobial Resistance Collaborators Global burden of bacterial antimicrobial resistance in A systematic analysis.

Lancet , — Arias, C. Antibiotic-resistant bugs in the 21st century--a clinical super-challenge. Basmaci, R.

Escherichia Coli meningitis features in children from to in France. Bollenbach, T. Antimicrobial interactions: mechanisms and implications for drug discovery and resistance evolution. CLSI M Performance Standards for Antimicrobial Susceptibility Testing.

Wayne, PA: Clinical and Laboratory Standards Institute. Google Scholar. Corbett, D. Potentiation of antibiotic activity by a novel cationic peptide: potency and Spectrum of activity of SPR Da Silva, F. Nanonization techniques to overcome poor water-solubility with drugs. Expert Opin.

Drug Discovery 15, — Eckburg, P. Safety, tolerability, pharmacokinetics, and drug interaction potential of SPR, an intravenous potentiator, after single and multiple ascending doses and when combined with β-lactam antibiotics in healthy subjects.

Escherich, T. The intestinal bacteria of the neonate and breast-fed infant. Falagas, M. Colistin: the revival of polymyxins for the management of multidrug-resistant gram-negative bacterial infections.

French, S. Potentiation of antibiotics against gram-negative bacteria by Polymyxin B analogue SPR from unique perturbation of the outer membrane. ACS Infect. Harkins, C. Methicillin-resistant Staphylococcus aureus emerged long before the introduction of methicillin into clinical practice.

Genome Biol. Huang, D. Assessment of combined Polymyxin B and minocycline therapy against Klebsiella pneumoniae Carbapenemase KPC -Producing K. Imai, Y. A new antibiotic selectively kills gram-negative pathogens. Nature , — Inada, A.

Enhancement of water solubility of indomethacin by complexation with protein hydrolysate. Ismail, A. Solubility enhancement of poorly water soluble domperidone by complexation with the large ring cyclodextrin. Jia, J. Mechanisms of drug combinations: interaction and network perspectives.

Drug Discov. Kaper, J. Pathogenic Escherichia coli. Kerantzas, C. Origins of combination therapy for tuberculosis: lessons for future antimicrobial development and application. MBio 8, e— Khalil, W.

Can mineral trioxide aggregate and Nanoparticulate EndoSequence root repair material produce injurious effects to rat subcutaneous tissues? Lefort, A.

Host factors and portal of entry outweigh bacterial determinants to predict the severity of Escherichia coli bacteremia. Li, Y. Synergistic antimicrobial activity of Colistin in combination with rifampin and azithromycin against Escherichia coli producing MCR Liu, Y.

Antibiotic adjuvants: an alternative approach to overcome multi-drug resistant gram-negative bacteria. Llor, C. Antimicrobial resistance: risk associated with antibiotic overuse and initiatives to reduce the problem. Drug Saf. Mendes, R.