Volume 64, Issue 4, October 2020 Published by Johnson Matthey © Copyright 2020 Johnson Matthey

ISSN 2056-5135

Johnson Matthey’s international journal of research exploring science and technology in industrial applications

Johnson Matthey Technology Review is published by Johnson Matthey Plc.

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. You may share, copy and redistribute the material in any medium or format for any lawful purpose. You must give appropriate credit to the author and publisher. You may not use the material for commercial purposes without prior permission. You may not distribute modiﬁ ed material without prior permission.

The rights of users under exceptions and limitations, such as fair use and fair dealing, are not aﬀ ected by the CC licenses. www.technology.matthey.com www.technology.matthey.com

Johnson Matthey’s international journal of research exploring science and technology in industrial applications

Contents Volume 64, Issue 4, October 2020

394 Guest Editorial: Breaking Down Barriers and Borrowing from Biology By Tom Sturgeon 396 Preparation and Evaluation of a Composite Filler Micro-Embedded with Pseudomonas putida for the Biodegradation of Toluene By Yuxi Yan, Rencheng Zhu and Shunyi Li 407 Unlocking the Full Evolutionary Potential of Artificial Metalloenzymes Through Direct Metal-Protein Coordination By George S. Biggs, Oskar James Klein, Sally R. Boss and Paul D. Barker 419 Reduction of Biofilm Formation on Cooling Tower Heat Exchangers using Nano-silica Coating By Irfan Turetgen 425 A Mini-Review of Shape-Memory Polymer-Based Materials By Mathew J. Haskew and John G. Hardy 443 Application of Chitosan-Encapsulated Orange Oil onto Footwear Insock Leathers By Buket Yılmaz and Hüseyin Ata Karavana 452 Bacterial Community Composition in Produced Water of Diyarbakır Oil Fields in Turkey By Tuğçe Tüccar, Esra Ilhan-Sungur and Gerard Muyzer 466 The Biotechnological Potentials of Bacteria Isolated from Parsık Cave, Turkey By Begüm Çandiroğlu and Nihal Doğruöz Güngör 480 Antibacterial Potential of Six Lichen Species against Enterococcus durans from Leather Industry By Didem Berber, İpek Türkmenoğlu and Nüzhet Cenk Sesal 489 The Destructive Effects of Extremely Halophilic Archaeal Strains on Sheepskins, and Proposals for Remedial Curing Processes By Meral Birbir, Pinar Caglayan and Yasar Birbir 504 Johnson Matthey Highlights

507 Antibiotic and Heavy Metal Resistant Bacteria Isolated from Aegean Sea Water and Sediment in Güllük Bay, Turkey By Gülşen Altuğ, Mine Çardak, Pelin Saliha Çiftçi Türetken, Samet Kalkan and Sevan Gürün 526 “Nanomaterials and Environmental Biotechnology” A book review by Martin Hayes 529 Biocatalytic Reduction of Activated Cinnamic Acid Derivatives By Samantha Staniland, Tommaso Angelini, Ahir Pushpanath, Amin Bornadel, Elina Siirola, Serena Bisagni, Antonio Zanotti-Gerosa and Beatriz Domínguez https://doi.org/10.1595/205651320X15954136194837 Johnson Matthey Technol. Rev., 2020, 64, (4), 394–395

www.technology.matthey.com

Guest Editorial Breaking Down Barriers and Borrowing from Biology

Introduction others, such as hides and textiles, show that as boundaries within science are removed, previously As humans, we seem to desire structure, distant industries will have much to learn from one relationships and laws to understand the another. universe. Through increased understanding, we can solve the problems and challenges that we Themes on Interdisciplinary Science perceive. This method and the output are given the label of science. At its best, science provides I have reflected on three themes as this issue has exquisite understanding, life-changing solutions or come together. There are numerous examples on sometimes both. each theme and I would challenge the reader to The downside of the structures and rules we think “what next?” for each: impose is that they can create inertia. Because 1. Interdisciplinary understanding coming into the structure or rule served a purpose in the past, biology; for example, computational methods we can be more willing to stand by it blindly than and coding which go hand-in-hand with the openly seek the understanding or solutions we truly biological understanding required for directed desire; a dynamic seen in the natural and social evolution of proteins sciences alike and revealing more about human 2. Interdisciplinary understanding coming from nature than the universe. One such structure is biology; for example, improved understanding that of the disciplines within science. We should of biochemical pathways and the relevant challenge ourselves to be very clear on the purpose biological structures being coupled with of any structures we adhere to and be ready to synthetic chemistry understanding to remove barriers that get in the way of progress. allow much more targeted small molecule One such example is uncovering the fertile ground therapeutics to be designed of interdisciplinary research. In recent years 3. Platform technologies; for example, clustered interdisciplinary research has been of increasing regularly interspaced short palindromic importance across the sciences. Volume 64 of the repeats and CRISPR-associated protein 9 Johnson Matthey Technology Review started a (CRISPR‑Cas9) genome editing where you celebration of interdisciplinary science by looking at can custom design the edit while following when chemistry collaborates with physics (1) and standardised procedures. in this issue, we will celebrate the cross-disciplinary This third theme is perhaps the most important as contributions of biology with other fields. it turns niche expertise into something accessible This wide-ranging issue explores topics such as: to scientists across fields. Understanding the what we can continue to learn from organisms in technology may be beneficial but is not a prerequisite unusual environments; how we might leverage to accessing it. Biology follows favourably in the biology in artificial situations; and even how we footsteps of computing in producing such platform manage the interface between human-made, technologies and it is an attribute that perhaps we controlled systems and the outside world. In should value and prioritise more in other fields. To particular, the diversity of industrial applications is expand on this theme, it is exciting to look both striking. Some are familiar to Johnson Matthey and backwards and forwards to the contributions made this journal such as fine chemical synthesis, while possible by platform technologies from the field of

394 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15954136194837 Johnson Matthey Technol. Rev., 2020, 64, (4) biology. Often these point back to unlocking our Conclusions understanding of the structure and function of DNA at a molecular level and have resulted in some of As you read through this issue, I hope you enjoy the most impactful scientific contributions of the reading something outside of your current field. I last 50 years or so. Our health has been a significant would take you back to my earlier challenge and see beneficiary of these advances with cancer drugs if you can gain any greater insights by not seeing providing an illustrative case study. Looking back, the separation between your field and those of the we can see recent classes of therapeutics that were authors. Rather, question what you can leverage, significantly enabled by this flow of understanding what you can learn and what next? and platform technologies such as tyrosine kinase inhibitors and antibody-based therapies (2). Most TOM STURGEON importantly, patient outcomes have improved Immaterial Ltd, 25 Cambridge Science Park, substantially in part, thanks to these therapies (3). Milton Road, Cambridge, CB4 0FW, UK Looking to the future, gene and cell therapies Email: [email protected] appear to be following a similar pattern and will hopefully deliver similar patient benefits. Outside References of cancer treatments and healthcare, we can see many industries set to benefit from being able to 1. A. Smith, Johnson Matthey Technol. Rev., 2020, access biological understanding and technologies. 64, (2), 101 This is particularly as we seek to learn from biology 2. T. A. Baudino, Curr. Drug Discov. Technol., 2015, and reduce our impact on the planet by using 12, (1), 3 materials and energy in keeping with what Earth 3. M. Quaresma, M. P. Coleman and B. Rachet, The can sustain. Lancet, 2015, 385, (9974), 1206

www.technology.matthey.com

Preparation and Evaluation of a Composite Filler Micro-Embedded with Pseudomonas putida for the Biodegradation of Toluene Preparation of composite filler with high toluene removal efficiency

Yuxi Yan on the environment (1). Toluene is a common School of Ecology and Environment, Zhengzhou pollutant in VOCs and is produced in a large number University, Zhengzhou 450001, China; College of industrial activities, such as chemical refining of Chemical Engineering, Zhengzhou University, and dye processing. Toluene stimulates skin Zhengzhou 450001, China and mucosa, and when it reaches a certain high concentration, it also causes paralysis of the human Rencheng Zhu, Shunyi Li* nervous system. Compared with photocatalysis School of Ecology and Environment, Zhengzhou and chemical oxidation, using a biofilter to remove University, Zhengzhou 450001, China VOCs is more economical and environmentally friendly (2). More important is that it does not *Email: [email protected] produce secondary pollution. The key element to ensure the removal capacity of the biofilter is the preparation of the filler. As a carrier for the transfer The main objective of this study was to evaluate of pollutants, the filler can provide a suitable the performance of a self-developed filler micro- growth environment for microorganisms (3). embedded with Pseudomonas putida (P. putida) Micro-embedding technology is a method which for toluene removal in a biofilter under various uses physical or chemical methods to keep loading rates. The results show that the biofilter microorganisms in a defined space, ensuring could reach 85% removal efficiency (RE) on the microorganisms with high activity. The principle eighth day and remain above 90% RE when the of using micro-embedding technology to degrade empty bed residence time (EBRT) was 18 s and the VOCs is to use a hollow porous membrane to inlet loading was not higher than 41.4 g m–3 h–1. intercept microorganisms inside the filler. The pore Moreover, the biofilter could tolerate substantial size of the hollow porous membrane is smaller than transient shock loadings. After two shut-down that of microbial cells, so that microorganisms can experiments, the removal efficiency could be be embedded. The VOCs can enter the interior restored to above 80% after a recovery period of of the embedded carrier freely due to the small three days and six days, respectively. Sequence particle size, and the degradation products can analysis of the 16S rRNA gene of fillers in four flow out of the carrier through the pore size (4, 5). operating periods revealed that the highly efficient A large number of studies have been carried out bacterial colonies in fillers mainly included on different types of fillers. Chen et al. (6) used Firmicutes, Actinobacteria and Proteobacteria and a two-layer biofilter filled with new mixed packing that the abundance of Bacteroidetes increased materials to remove hydrogen sulfide gas. Dumont significantly during the re-start period. et al. (7) prepared a nutritional slow-release filler

(UP20) to biodegrade H2S. In the above studies, 1. Introduction the fillers were not embedded with microorganisms. The concentration of microorganisms in the filler The massive discharge of volatile organic was small, and the removal efficiency of the biofilter compounds (VOCs) has a great negative impact was low in the start-up phase, resulting in a longer

396 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15831468405344 Johnson Matthey Technol. Rev., 2020, 64, (4) start-up period. Zhu et al. (8) used a composite cross-linked in boric acid-calcium chloride solution packing material with functional microorganisms to and dried at room temperature for 24 h. Taking the remove H2S. However, toluene does not biodegrade mechanical strength as a single variable factor, the easily due to the presence of a benzene ring. proportions of polyvinyl alcohol, sodium alginate Zuo et al. (9) found that engineered P. putida and polypropylene fibre were adjusted to obtain could simultaneously degrade organophosphates, the optimal ratio. After many tests and modifying pyrethroids and carbamates. Muñoz et al. (10) the design, the optimum proportions of each studied the long-term performance and stability of component of the filler were determined as follows: P. putida in a toluene removal bioreactor. The above polyvinyl alcohol accounted for 30%~36%, sodium studies have found that P. putida is highly effective alginate accounted for 12%~18%, polypropylene in degrading organics containing benzene rings. fibre accounted for 4%~8%, decomposed plants However, there is a lack of studies on filler micro- accounted for 15%~25%, calcium carbonate embedded P. putida for toluene biodegradation. accounted for 15%~25%, activated carbon Existing problems with biofilters packed with fillers accounted for 4%~10% and P. putida accounted include bed clogging, low biomass concentration for 0.5%~1.5% (13). The schematic pictures of and pressure drops. These problems become more the size and the composition of the composite prominent when the biofilter is operated under high filler can be seen in Figure S1 and Figure S2 in the VOC loading rates or long-term operation (11). For Supplementary Information. example, Ryu et al. (12) found that the benzene removal efficiency of a well-designed biofilter 2.2 Experimental Setup decreased from greater than 90% to approximately 75% after 27 days of operation due to clogging The experimental system used in this experiment caused by the excess growth of biomass. is shown in Figure 1. Three biofilters were The main objective of this study was to evaluate constructed with transparent organic glass pipes. the performance of a self-developed filler micro- Each biofilter consisted of three modules (each embedded with P. putida for toluene removal under module is 105 mm in inner diameter and 500 mm various inlet loading rates. The variations in start‑up in height), and all of them were filled with 300 mm period, pressure drop, biomass concentration and composite fillers. A sampling port was set in the tolerance to transient shock loading were monitored top of each module. Toluene gas was prepared by throughout the experiments. Special attention was mixing fresh air with pure toluene in a mix chamber, paid to the analysis of the microbial community and then introduced into the bottom of each attached to these fillers and to monitoring the biofilter through the three models in sequence. evolution of the microbial community in various Three biofilters, namely biofilter 1 (BF1), biofilter periods. 2 (BF2) and biofilter 3 (BF3), were used in this experiment to evaluate the start-up performance. 2. Material and Methods BF1 was packed with the composite filler micro- embedded with P. putida, and both BF2 and BF3 2.1 Preparation of Filler were packed with the sterilised fillers without any microorganisms. However, the nutrient solution The composite filler was mainly composed of used for BF2 at the start-up period was mixed polyvinyl alcohol, sodium alginate, polypropylene with the P. putida suspension and the microbial fibre, decomposed plants, calcium carbonate and concentration of the suspension was the same activated carbon. First, polyvinyl alcohol and sodium as that of the P. putida suspension added in the alginate, as the embedding and protective agents, preparation of the composite filler in BF1. Specially, were heated, dissolved and cooled to 35°C. Then nutrient solution (0.11 K2HPO4, 0.04 KH2PO4, polypropylene fibre as the skeleton, decomposed 0.025 NH4Cl, 0.067 MgSO4, 0.036 CaCl2, 0.25 FeCl3, plants as nutrients and calcium carbonate as 0.03 MnSO4, 0.04 ZnSO4, 0.03 (NH4)2Mo7O4·4H2O; the pH buffer were added into the liquid agent, unit: g l–1; adjusted to pH = 7.0 with NaOH) for respectively. Additionally, activated carbon and microorganism growth was sprayed into the filler P. putida BRJC1032 (screening from the activated bed from the top of three biofilters throughout the sludge) were mixed with above agents to increase experiment. The nutrient solution was intermittently the physical adsorption capacity and biodegradation sprayed onto the top of the three biofilters with a capacity of toluene. After that, the mixtures were spray intensity of 1.5 l h–1 by a peristaltic pump for stirred in a container for 15 min and extruded to one hour out of every three hours and the nutrient spherical particles. Finally, these particles were solution was changed every seven days.

Gas outlet

Toluene

Sampling points

Flowmeter

Mixing chamber Gas inlet Drainage

Air pump Peristaltic pump

Valve Gas pump Circulation cistern

Fig. 1. Schematic diagram of the experimental set-up

2.3 Toluene Concentration Analysis volume of the whole biofilter (l) and Q is the gas flow rate (l min–1). The determination of toluene concentration was carried out by adsorption of activated carbon and 2.4 Physical and Chemical Property desorption of carbon disulfide, and then the toluene Analysis gas was injected into a gas chromatograph (GC- 2014, Shimadzu, Japan) equipped with a packed The specific surface area and the porosity of the column (free fatty acid phase (FFAP) capillary filler were measured by a surface area analyser column, 30 m × 0.25 mm × 0.25 μm) and a flame (Gemini® VII 2390, Micromeritics®, USA). Solid ionisation detector (FID). The gas chromatography samples were filtered and the pH value of the filtrate nitrogen was used as the carrier gas with a flow was detected using a Bioblock 90431 electrode rate of 1 ml min–1. Temperatures of the injection connected to a C-835 Bioblock multiparameter port, column and detection port were set to 150°C, analyser (Fisher Scientific, France). 65°C and 150°C, respectively. Gas samples were The mechanical strength of the composite filler was collected from the inlet and outlet of the biofilter measured by using a compressive strength-testing with a gas-tight syringe and injected into the GC instrument (YHKC-2A, Taizhou Yinhe Instrument daily (14). Data were obtained from the workstation Plant, China). The pressure drop of the packed bed by automatic comparison of the peak area of was measured using a digital pressure gauge (testo the inlet and outlet samples with the baseline 510, Testo SE & Co KGaA, Germany) connecting two of toluene. The performance of the biofilter was ends from the inlet and outlet. The pressure gauge evaluated in terms of (%) RE and the elimination had a measuring range of 0–100 kPa, a resolution of capacity (EC) as a function of toluene loading. The 1 Pa and an accuracy of ±0.3 Pa. RE and EC were calculated as in Equations (i)–(iii): The saturated moisture content: some packing fillers were chosen randomly and immersed into ()CC− distilled water for 2 h to adsorb as much water as Removalefficiency = in out × 100% (i) possible. Then the packing fillers were removed Cout and placed in a vacuum oven (DZF6050, Yiheng QC× Scientific Instrument Co Ltd, China) at 105°C for at Inletloading = in (ii) V least 12 h until its weight remained stable. QC×−()C The concentration of microorganisms in the filler Eliminationcapability = in out (iii) V was determined by plate counting. Approximately 10 g fillers were taken out homogeneously from where the Cin and Cout are the inlet and outlet the three modules of the running biofilter, and toluene concentration (mg m–3), the V is the then put into a conical flask with 90 ml distilled

398 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15831468405344 Johnson Matthey Technol. Rev., 2020, 64, (4) water. After that, the mixture was shaken in a of 30 s at 95°C, 30 s of annealing at 55°C, 45 s of thermostatic shaker bath for 2 h at 25°C to obtain elongation at 72°C and a final extension at 72°C the liquid containing microorganisms. Next, a series for 10 min. PCR was performed in triplicate in 20 μl of solutions were prepared by different dilution mixtures containing 4 μl of 5 × FastPfu Buffer, factors (1, 10, 102, 103, 104 and 105 times). Each 2 μl of 2.5 mM deoxyribonucleotide triphosphates 0.1 ml solution was taken and inoculated into three (dNTPs), 0.8 μl of each primer (5 μM), 0.4 μl of types of plate cultures (beef-protein, Rose Bengal FastPfu Polymerase and 10 ng of template DNA. medium and Gause’s No.1 medium) for bacteria, The resulting PCR products were extracted from a fungi and actinomycetes, respectively. The plates 2% agarose gel, further purified using the Axygen® were placed in a biochemical incubator (CLIN-250, AxyPrep DNA Gel Extraction Kit (Corning Inc, USA) Tianjin Huabei Experimental Instrument Co Ltd, and quantified using QuantiFluor®-ST fluorometer China) for 2–7 days at 28°C. Finally, the number (Promega, UK) according to the manufacturer’s of microorganism colonies in each plate was protocol (16). counted. Moreover, all the glass vessels used in Purified amplicons were pooled in equimolar this experiment were sterilised by using a seating fashion and paired-end sequenced on a MiSeq automatic electro-thermal pressure steam steriliser platform (Illumina Inc, USA) according to the (Model ZDX-35B, Shanghai Medical Instrument standard protocols established by Shanghai Majorbio Manufactory, China) (15, 16). Bio-Pharm Technology Co Ltd (Shanghai, China). The acquired sequences were compared with 16S 2.5 DNA Extraction and Sequencing rRNA gene sequences in the National Center for Biotechnology Information (NCBI) database. Approximately 10 g fillers were randomly sampled from the lowest module of BF1 system at the 25th 3. Results and Discussion day, 65th day, 95th day and 145th day. Then the samples were sealed with aluminium foil and frozen 3.1 Physicochemical Properties of at –4°C in a fridge. the Filler Microbial DNA was extracted from the above four samples using the E.Z.N.A.® soil DNA Kit (Omega Physicochemical properties of the experimental Bio-tek Inc, USA) according to the manufacturer’s filler used in this study and some other materials protocols. The final DNA concentration and from the references are listed in Table I (13). As purification were determined by a NanoDropTM shown in Table I, the experimental filler is spherical 2000 UV-vis spectrophotometer (Thermo with a diameter of approximately 10 mm. The bulk ScientificTM, USA), and DNA quality was checked by density of the experimental filler is approximately 1% agarose gel electrophoresis. Polymerase chain 271 kg m–3, similar to that of pine bark, and lighter reaction (PCR) was conducted according to the than most of the reference fillers. The mechanical following: 3 min of denaturation at 95°C, 27 cycles strength is greater than that of pine bark but

Table I Physicochemical Properties of the Fillers Bulk Saturated Specific Organic Size, Mechanical Porosity Filler density, pH moisture surface matter mm strength, N rate, % kg m–3 content, % area, m2 g–1 rate, % Experimental 10 ± 2 271 ± 17 153 ± 5 7.0 55.3 ± 3 13 ± 2 1.32 53 ± 4 filler Pine — 244 — 5.7 56.3 59.9 18.39 98.2 barka (17) Lava — 591 — 5.9 28.9 65.4 2.77 0.6 rocka (17) UP20 (7) 7 920 — 6.9 47 — — — Composite 12 471 427 10.5 49 38 3.91 — filler (8) Slow-release 50 164 — 7.9 46.7 88 — — filler (16) aDue to the irregular shape of pine bark and lava rock, there is no corresponding size data. Other “—” data show that the author did not determine its physicochemical property data in the corresponding literature

399 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15831468405344 Johnson Matthey Technol. Rev., 2020, 64, (4) smaller than that of volcanic stone (>500 N) (17). efficiency of BF1 increased from the initial 40% to The porosity rate is approximately 13%, which is 80%, and stabilised between 82% and 85% after significantly smaller than other fillers and helps the eighth day (21, 22). The removal efficiency toluene to better contact microorganisms in the of BF2 showed a downward trend in the first few filler when entering the biofilter (18, 19). The initial days and then rose to approximately 85% at the pH of the filler is 7.0 ± 0.2. The specific surface 14th day. The removal efficiency of BF3 gradually area is approximately 1.3 ± 0.1 m2 g–1, which is declined from the beginning, and it decreased similar to that of lava rock and composite filler. to almost zero on the 16th–18th days (14, 23). Compared with lava rock, UP20 and slow-release The results showed that fillers embedded with filler (7, 8, 16), the saturated moisture content and activated carbon and polypropylene fibres have a organic matter rate are higher, which can provide certain adsorption capacity. However, the removal water and nutrients for microorganisms in fillers. efficiency was gradually reduced when the filler In addition, the decomposed plant fibre contained reached adsorption saturation, as shown in the within the filler can provide nutrients for microbial BF3 trend line in Figure 2. For the same reason, growth during experimental operation (20). The the BF2 line also showed a downward trend at selected microbial source added to the filler was the beginning. Due to the substantial growth of P. putida, and the activated carbon was contained microorganisms, the subsequent removal efficiency in fillers, which can adsorb toluene quickly, gradually increased as shown in the BF2 trend line. promoting toluene to enter the biofilter. The filler Compared with BF2, the fillers in BF1 embedded in the biofilter did not appear to have deformation, with P. putida showed unique degradation of accumulation or other phenomena after operating toluene at the beginning. The filler‑embedded approximately 150 days. The results indicated that microorganisms entered the working state faster the fillers had favourable properties as biofilter than those cultured with the bacterial solution. media, and maintained characteristics under These results indicated that the biofilter packed long‑term operation. with the composite fillers prepared by micro- embedding could be quickly started up and the 3.2 Start-up Performance microorganisms in the biofilter could well utilise toluene as the carbon source (22). The removal efficiency of the three biofilters during the start-up period is presented in 3.3 Continuous Biodegradation Figure 2. Three biofilters, operated at low toluene Performance concentrations (100–120 mg m–3) and an EBRT of 35 s, demonstrated different removal performance Toluene continuous removal experiments were for toluene at the start-up period. The removal performed in three phases based on controlling the

100 Fig. 2. Removal performances of BF1 (packed with the fillers micro-embedded with P. putida), 80 BF2 and BF3 during the start-up period % , 60 l efficiency

va 40 BF3 BF1 emo R BF2 20

0 2 468 10 12 14 16 18 Time, d

EBRT of BF1 to 35 s (Phase 1, day 10 to day 49), 18 s biofilter. At a higher flow rate, the contact time (Phase 2, day 50 to day 80) and 12 s (Phase 3, day between the toluene and the microorganisms 81 to day 110). The results of these experimental in the fillers was shortened and that resulted in stages (Figure 3) are described below. Initially, deterioration of the biodegradation ability of the the biofilter was operated at a low loading rate of filter bed, leading to lower removal efficiency (24). toluene (10.5 g m–3 h–1) corresponding to a low Similarly, in Phase 3, the toluene inlet concentration inlet concentration (100–120 mg m–3) and high increased from 100 mg m–3 to 400 mg m–3, and EBRT (35 s) to facilitate proper microbial growth and the intake load increased to 123.3 g m–3 h–1 with a establish steady-state conditions (8, 23). Steady corresponding EBRT of 12 s. During this phase, the state was achieved on the 10th day of operation, removal efficiency of toluene gradually decreased which was evident from the constant value of the to 80%, and no significant improvement in removal removal efficiency (83%). On the 18th day, the inlet efficiency was observed (17, 22). concentration increased to 200 mg m–3, the removal Elimination capacity, another important indicator of efficiency was almost stable at 88% after a slight the biofilter, was also used to assess the ability of decrease. On the 28th day, the inlet concentration the biofilter in terms of toluene removal. Figure 4 increased to 400 mg m–3, and the removal efficiency demonstrates the relationship of elimination capacity dropped rapidly to 72% and finally stabilised at 90% upon the inlet loading. It could be seen from after five days of continuous operation. However, Figure 4 that the elimination capacity presented a when the inlet concentration was controlled at slow increase with the increase of inlet loading rates. 800 mg m–3, the removal efficiency did not reach The maximum elimination capacity of the biofilter a correspondingly high state (less than 80%). In was 101 g m–3 h–1, which is better than other typical Phase 1, the initial rapid increase within 90% of RE biofilters. For example, Zhuet al. (10) used composite may be due to some extent to competition among packing materials to remove H2S and observed a microorganisms in the filter unit (14, 21, 23). maximum elimination capacity of 65 g m–3 h–1. Liu Again, in Phase 2, the inlet loading rate was et al. (18) reported compost-based biofilter with a increased and maintained at 81.2 g m–3 h–1 with maximum elimination capacity of 50 g m–3 h–1 for a corresponding EBRT of 18 s, and the toluene toluene. inlet concentration varied between 100 mg m–3 The concentration of toluene in the nutrient and 400 mg m–3. The removal efficiency reached solution was 0.3 ± 0.1 g l–1 (the saturated solubility a maximum when the inlet loading rate was less of toluene in water was 0.5 ± 0.1 g l–1). This may than 41.4 g m–3 h–1 and was stable above 90%. be due to the short contact time between toluene However, the removal efficiency was only slightly and the nutrient solution. In addition, part of the decreased and then stabilised close to 86% at the toluene dissolved in the nutrient solution was end of this phase. This result might be attributed utilised by the filler with circulation of the nutrient to the decrease in residence time of toluene in the solution.

EBRT = 35s EBRT = 18s EBRT = 12s Fig. 3. Time course of the inlet 1000 100 and outlet concentration and the removal efficiency of BF1 –3

m 800 80 R emo

Inlet concentration va 60 600 Outlet concentration l efficiency ation, mg Removal efficiency

400 40 , %

luene concentr 200 20 To

0 0 10 20 30 40 50 60 70 80 90 100 110 Time, d

The above results showed that a sudden increase then biodegraded by microorganisms in the filler. in the inlet loading will cause the removal rate to A certain amount of toluene will be dissolved in decrease within a certain period of time. As the the nutrient solution, but with the circulation of experiment proceeds, the system will gradually the nutrient solution, part of the toluene will be return to a higher removal rate. When the degraded by the microorganisms in the filler again. microorganisms grew under suitable conditions, the recovery ability of the system also increased. 3.4 Tolerance for Transient Shock However, when the inlet loading rate was too high, Loading the degradation ability of the microorganisms was exceeded, resulting in a relatively low removal rate. To test the ability of the biofilter to resist sharp load After entering the biofilter, toluene is first adsorbed change, two interference-shutdown experiments by activated carbon and biofilms in the filler, and were operated after running for 114 days.

140 Fig. 4. Toluene elimination capacity of BF1 versus the 120 inlet loading –1 h

–3 100 m g

y, 80

40 y = 07989x + 2.9602 2

Elimination capabilit R = 0.995 20

0 20 40 60 80 100 120 140 Inlet loading, g m–3 h–1

3 days 7 days Fig. 5. Performance evaluation 1000 100 Inlet concentration during shutdown and restart Outlet concentration periods of BF1 under transient

–3 Removal efficiency shock loading

m 800 80 R emo va

600 60 l efficiency ation, mg

400 40 , %

luene concentr 200 20 To

0 0 110 115 120 125 130 135 140 145 Time, d

Figure 5 shows the performance evaluation forming units (CFU) g–1 (the filler was placed in during shutdown and restart periods of BF1 under the refrigerator for 1 month, and the biomass transient shock loading. When the inlet toluene concentration was reduced to 5 × 104 CFU g–1) concentration decreased from 400 mg m–3 to to 4 × 108 CFU g–1 on the 60th day, which was 200 mg m–3, the removal efficiency increased consistent with the trend in the pressure drop (24, to 90%. Then, the biofilter was subjected to 26). The above result indicates that the increase a three-day shutdown experiment and the in system pressure drop was mainly due to the removal efficiency was restored to 81.2% after rapid growth in microbial biofilm formation and running three days. Compared with the shutdown inlet loading rates. The efficient growth and experiments of Singh and Wang (22, 23), the reproduction of microbial biomass played an interrupt experiment in this study better reflects important role in the efficient operation of the the change of flow in actual operation. In the second system and the growth of the microorganisms experiment, when the inlet toluene concentration affected the pressure drop across the packed increased from 400 mg m–3 to 800 mg m–3, bed and the ease with which the packed bed the removal efficiency decreased drastically to was clogged. Low biomass reduces the removal 62%, and time for the RE to reach at 80.9% efficiency. In contrast, excess biomass reduces the was only six days after seven days of shutdown space required for gas and liquid to pass through operation. This result clearly indicates that a the biofilter, which leads to an increase in the certain amount of toluene absorbed in activated system pressure drop (27). Although the biomass carbon was supplied to the microorganisms during concentrations in the biofilter increased and the the shutdown operation of the system, and the porosity of the system was reduced, this process microbial activity was maintained; in addition, the did not cause blockage of the system and had no decomposed plant fibres also provided a carbon significant effect on the removal performance. source for the microorganisms, as found by Jorge and Livington (25). 4. Bacterial Community Analysis

3.5 Biomass Concentration and To explore the bacterial communities in the biomass attached to BF1, genetic sequencing Pressure Drop in the Biofilter analyses were carried out. Sequencing of 16S The attached growth biomass concentration and rRNA genes amplified from the active bacterial pressure inside the device were measured during communities during the operational stages revealed 1–60 days in the biofilter, as shown in Figure 6. 21 phyla, 41 classes, 96 orders, 184 families and The pressure drop increased more obviously from 347 genera (28, 29). The community analysis at 56 Pa to 373 Pa. The biomass concentration in the phylum level of the fillers is shown in Figure 7. biofilter gradually increased from 5 × 104 colony The four operational stages were sampled at the

500 1010 Fig. 6. Biomass concentration and pressure drop changes in Biomass concentr BF1 during the first 60 days 109 400

108 Pa 300 7

10 ation, CF

Pressure, 200 106 U

Pressure g Biomass concentration 105 –1 100

104 0102030 40 50 60 Time, d

Community barplot analysis

Firmicutes Actinobacteria Proteobacteria Day_25 Bacteroidetes Chloroflexi Gemmatimonadetes Others

Day_65 Samples Day_95

Day_145

0 0.2 0.4 0.6 0.8 1 Percent of community abundance on Phylum level

Fig. 7. Bacterial community analysis of the fillers sampled at the 25th day, the 65th day, the 95th day and the 145th day in BF1

25th day, the 65th day, the 95th day and the of Pseudomonas genus increased from 4.7 × 10–4 145th day, where the 25th day, the 65th day and to 1.9 × 10–3. After two intervention-shutdown the 95th day had a different EBRT and the same experiments, the abundance of Pseudomonas genus inlet toluene concentration, and the 145th day was decreased to 8.5 × 10–5, which indicates that the after two interference-shutdown experiments. The biofilter was not in a sterile environment and that dominant phyla were Firmicutes (63.4 ± 8.7%), there are other microorganisms competing with the followed by Actinobacteria (14.6 ± 3.9%) and P. putida added to the filler. When the environmental Proteobacteria (10.1 ± 4.2%). With decreased conditions and the nutrients in the biofilter became EBRT, the abundance of Firmicutes remained unsuitable for the added microorganisms and high, but the abundance of Actinobacteria were suitable for other microorganisms, the other decreased, and the abundance of Proteobacteria microorganisms were activated and enriched (32). increased. This is mainly due to a reduction However, in the start-up phase, the biofilter in residence time leading to the inability of embedded with P. putida started quickly, and the microorganisms to fully utilise toluene, and removal efficiency of toluene remained high, which a reduction in the carbon source leading to a indicated that the added P. putida contributed to change in the proportion of microorganisms (30). the efficient operation of the biofilter (33). These After two interference-shutdown experiments, results indicated that the biomass could maintain the abundance of Bacteroidetes increased and itself by microbial community changes, and the the normal microecological balance was broken, rapid re-adaptation of the biofilter could contribute which indicated that Bacteroidetes is a sensitive to the activity retention of its biomass during the biological indicator, similar to the results found by starvation period. Wolińska (31). Using this indicator (the increase in Bacteroidetes), it can be judged whether the 5. Conclusions biofilter is in an unstable state, which would provide some guidance for practical engineering A composite filler micro-embedded with P. putida applications. was prepared and evaluated for the biodegradation In the four operational periods, few Pseudomonas of toluene. The biofilter packed with the fillers could (abundance less than 1%, as shown in Figure S3) start up quickly with 85% RE on the eighth day, were found in the sampling of the above four periods. and tolerate substantial transient shock loadings. As the inlet loading rate increased, the abundance The RE of the biofilter remained above 90% when

404 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15831468405344 Johnson Matthey Technol. Rev., 2020, 64, (4) the EBRT was 18 s and the intake load was not 10. R. Muñoz, M. Hernández, A. Segura, J. Gouveia, higher than 41.4 g m–3 h–1. In the experimental A. Rojas, J. L. Ramos and S. Villaverde, Appl. period of 145 days, no filter plugging phenomenon Microbiol. Biotechnol., 2009, 83, (1), 189 was observed. Moreover, the high removal 11. J. V. Littlejohns, K. B. McAuley and A. J. Daugulis, efficiency and elimination capacity contributed J. Hazard. Mater., 2010, 175, (1–3), 872 to rich bacterial communities for the efficient 12. H. W. Ryu, K.-S. Cho and D. J. Chung, Bioresour. biodegradation of toluene. The communities Technol., 2010, 101, (6), 1745 mainly included Firmicutes, Actinobacteria and 13. Y. Nie, R. Zhu, S. Li, S. Li, M. Wang and Y. Yan, Proteobacteria, and the abundance of Bacteroidetes Chinese J. Environ. Eng., 2019, 13, (3), 678 increased significantly during the recovery 14. X. Chen, W. Qian, L. Kong, Y. Xiong and S. Tian, period. The composite filler exhibited favourable Biochem. Eng. J., 2015, 98, 56 physicochemical properties in this experiment and 15. W.-F. Yang, H.-J. Hsing, Y.-C. Yang and J.-Y. Shyng, its practicability in industrial engineering should be J. Hazard. Mater., 2007, 148, (3), 653 further investigated. 16. R. Zhu, S. Li, Z. Wu and É. Dumont, Environ. Technol., 2017, 38, (8), 945 Acknowledgments 17. Y. Luo, S. Li, H. Ma and Y. Wang, Trans. Chinese Soc. Agric. Eng., 2017, 33, (12), 218 (in Chinese) The authors would like to acknowledge the support 18. Y. Liu, X. Quan, Y. Sun, J. Chen, D. Xue and of the National Natural Science Foundation of China J. S. Chung, J. Hazard. Mater., 2002, 95, (1–2), (No. U1304216), the Science and Technology Plan 199 of He’nan Province, China (No. 122102310366), 19. R. Logares, S. Sunagawa, G. Salazar, the University Key Research Project of He’nan F. M. Cornejo-Castillo, I. Ferrera, H. Sarmento, Province, China (No. 19A610002 and 19A150010), P. Hingamp, H. Ogata, C. de Vargas, G. Lima-Mendez, and the China Postdoctoral Science Foundation J. Raes, J. Poulain, O. Jaillon, P. Wincker, (No. 2018M632794). S. Kandels-Lewis, E. Karsenti, P. Bork and S. G. Acinas, Environ. Microbiol., 2014, 16, (9), 2659 References 20. J. Zhang, L. Li and J. Liu, Biochem. Eng. J., 2017, 1. M.-C. Delhoménie and M. Heitz, Crit. Rev. 118, 105 Biotechnol., 2005, 25, (1–2), 53 21. Q. Hu, C. Wang and K. Huang, Chem. Eng. J., 2. R. Underhill, R. J. Lewis, S. J. Freakley, 2015, 279, 689 M. Douthwaite, P. J. Miedziak, O. Akdim, 22. K. Singh, B. S. Giri, A. Sahi, S. R. Geed, J. K. Edwards and G. J. Hutchings, Johnson M. K. Kureel, S. Singh, S. K. Dubey, B. N. Rai, Matthey Technol. Rev., 2018, 62, (4), 417 S. Kumar, S. N. Upadhyay and R. S. Singh, 3. Y. J. Tham, P. A. Latif, A. M. Abdullah, A. Shamala- Bioresour. Technol., 2017, 242, 351 Devi and Y. H. Taufiq-Yap, Bioresour. Technol., 23. M. Wang, S. Xu, S. Li and R. Zhu, J. Ind. Eng. 2011, 102, (2), 724 Chem., 2019, 75, 224 4. E. R. Rene, B. T. Mohammad, M. C. Veiga and 24. Y. Ding, W. Wu, Z. Han and Y. Chen, Biochem. Eng. C. Kennes, Bioresour. Technol., 2012, 116, 204 J., 2008, 38, (2), 248 5. Y. Deng, F. Yang, C. Deng, J. Yang, J. Jia and 25. F. Abbasian, R. Lockington, M. Megharaj and H. Yuan, Appl. Biochem. Biotechnol., 2017, R. Naidu, Appl. Biochem. Biotechnol., 2016, 183, (3), 893 178, (2), 224 6. Y. Chen, X. Wang, S. He, S. Zhu and S. Shen, 26. J. Song and K. A. Kinney, Biotechnol. Bioeng., J. Environ. Manage., 2016, 165, 11 2000, 68, (5), 508 7. E. Dumont and Y. Andrès, J. Chem. Technol. 27. Y. Hajizadeh, M.-M. Amin and I. Parseh, J. Ind. Biotechnol., 2010, 85, (3), 429 Eng. Chem., 2018, 62, 418 8. R. Zhu, S. Li, X. Bao and É. Dumont, Sci. Rep., 28. H. Li, S. Huang, Z. Wei, P. Chen and Y. Zhang, Sci. 2017, 7, 42241 Total Environ., 2016, 562, 533 9. Z. Zuo, T. Gong, Y. Che, R. Liu, P. Xu, H. Jian, 29. H. Liu, S.-J. Wang, J.-J. Zhang, H. Dai, H. Tang C. Qiao, C. Song and C. Yang, Biodegradation, and N.-Y. Zhou, Appl. Environ. Microbiol., 2011, 2015, 26, (3), 223 77, (13), 4547

30. L. Bergdoll, E. Point, F. Bayman and D. Picot, 32. S. R. Geed, M. K. Kureel, A. K. Shukla, R. S. Singh Biochim. Biophys. Acta., 2012, 1817, S138 and B. N. Rai, Resour. Eff. Technol., 2016, 2, (1), S3 31. A. Wolińska, A. Kuźniar, U. Zielenkiewicz, 33. M. Kumar, B. S. Giri, K.-H. Kim, R. P. Singh, D. Izak, A. Szafranek-Nakonieczna, A. Banach E. R. Rene, M. E. López, B. N. Rai, H. Singh, and M. Błaszczyk, Appl. Soil Ecol., 2017, 119, D. Prasad and R. S. Singh, Bioresour. Technol., 128 2019, 285, 121317

The Authors

Yuxi Yan received a bachelor’s degree from Zhengzhou University, China, in 2018 and is currently studying for a master’s degree at Zhengzhou University. His research interests include the biodegradation of VOCs.

Rencheng Zhu received his PhD from Nanjing University of Aeronautics and Astronautics, China, in 2017 and currently serves as an associate professor at Zhengzhou University. His research interests include the governance of VOCs and the characteristics of automobile exhaust emissions.

Shunyi Li received his PhD from Sun Yat-sen University, China, in 2005. He is currently an executive director of the Henan Environmental Protection Federation, China, and a professor at Zhengzhou University. His research interests include the management of VOCs and the management of odorous gases.

www.technology.matthey.com

Unlocking the Full Evolutionary Potential of Artificial Metalloenzymes Through Direct Metal-Protein Coordination A review of recent advances for catalyst development

George S. Biggs, Oskar James Klein, stabilise transition states. This highlights the Sally R. Boss*, Paul D. Barker** often neglected but crucial element of natural Department of Chemistry, University of systems that is the energetic contribution Cambridge, Lensfield Road, Cambridge, towards activating metal centres through protein CB2 1EW, UK fold energy. Finally, general principles needed for a different approach to the formation of ArMs are Email: *[email protected]; set out, utilising direct coordination inspired by **[email protected] the activation of an organometallic cofactor upon protein binding. This methodology, observed in nature, delivers true interdependence between Generation of artificial metalloenzymes (ArMs) metal and protein. When combined with the ability has gained much inspiration from the general to efficiently evolve enzymes, new problems in understanding of natural metalloenzymes. catalysis could be addressed in a faster and more Over the last decade, a multitude of methods specific manner than with simpler small molecule generating transition metal-protein hybrids have catalysts. been developed and many of these new-to- nature constructs catalyse reactions previously 1. Introduction reserved for the realm of synthetic chemistry. This perspective will focus on ArMs incorporating Metalloenzymes have been prominent in the 4d and 5d transition metals. It aims to summarise field of enzyme engineering since its emergence the significant advances made to date and asks some 40 years ago, at the birth of protein whether there are chemical strategies, used in and enzyme engineering (1, 2). Metal ions or nature to optimise metal catalysts, that have yet cofactors in solution have an intrinsic chemistry to be fully recognised in the synthetic enzyme that can be catalytic and these are accessible world, particularly whether artificial enzymes to detailed mechanistic study. These properties produced to date fully take advantage of the mean that co-localisation of substrate and metal structural and energetic context provided by the within a peptidic scaffold can be sufficient in protein. Further, the argument is put forward forming an ArM, without further influence from that, based on precedence, in the majority of the protein on the catalytic mechanism. With naturally evolved metalloenzymes the direct the advent of modern protein engineering and coordination bonding between the metal and design technologies, ArMs were developed by the protein scaffold is integral to catalysis. incorporating metal binding sites in or adjacent Therefore, the protein can attenuate metal to hydrophobic pockets. While the resulting ArMs activity by positioning ligand atoms in the form were active, they often displayed low efficiency of amino acids, as well as making non-covalent and specificity. Therefore, directed evolution (i.e. contributions to catalysis, through intermolecular iterative rounds of mutagenesis and selection interactions that pre-organise substrates and for activity, Figure 1) has become a key step in

407 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15928204097766 Johnson Matthey Technol. Rev., 2020, 64, (4) creating enzymes with new and useful properties. protein scaffold as well as the type of catalytic The choice of starting point for such a forced centre and reactions involved. Advances in ArMs evolution campaign, in this case the metal-protein have recently been reviewed and the reader complex formed initially, is of great importance. is referred to these for further details of the Since any particular enzyme follows a unique strategies used to find new systems (3–5). This evolutionary trajectory as new mutations move it article aims to provide an overview of the strengths along the fitness landscape towards (potentially and weakness of these different approaches and local) maxima, choice of the starting point may to provide a perspective of some challenges that directly predetermine the result. By nature of remain. the selection process, it is further possible, that trajectories leading to the global maximum fitness 2. Why Do We Want New Artificial fall beneath the cut-off limit for further evolution, Metalloenzymes? becoming inaccessible. For instance, a mutation introduced in the first round of mutagenesis One particular area that will greatly impact may lead to a destabilisation of the protein at chemical production on this planet is synthetic assay conditions, causing that initial variant biology. Replacing synthetic catalysts, acting on to be discarded through selection. However, petrochemical feedstocks in non-aqueous solvents, a compensating mutation to that variant in a with biocatalytic systems working in water with subsequent round of mutagenesis could result simple carbon neutral feedstocks (carbon dioxide in an enzyme which is stable, active and closer even?) is clearly highly desirable. But why engineer to a global fitness maximum. Finally, not every new enzymes, particularly using expensive and method of generating ArMs may be compatible relatively scarce transition metals, when the ability with current methods for directed evolution and to find new catalysts amongst gene products from therefore limit the extent of evolution that can all corners of the biological world has developed be achieved. at staggering pace (6–8)? As a consequence of In this perspective, different routes towards the latter, any target chemical can conceivably be ArMs are considered in the context of the starting obtained by recombining pre-existing metabolic

Gene library

Mutagenesis Expression

Selection and DNA Metal recovery modification

Activity assay

Fig. 1. The general overview of a directed evolution campaign for ArMs. The Darwinian algorithm can be reproduced in the laboratory, greatly increasing the speed of evolution. Mutagenesis methods introduce mutations with various levels of randomness, depending on the method used, to the starting point gene, forming a gene library. This library can then be expressed in a manner that couples expression products and genetic sequence information to yield the different proteins. Upon addition of the metal cofactor, the ArMs are formed and can be selected for improved variants in regard to desired parameters (reaction rates, yield, stereoselectivity, stability). The metal modification step must itself clearly be efficient and high yielding to avoid limiting the library size at that stage. The sequence information of the improved candidates is recovered and can be subjected to further rounds of directed evolution

408 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15928204097766 Johnson Matthey Technol. Rev., 2020, 64, (4) pathways (9). What will new and unnatural metal-substrate proximity may be enough to metalloenzymes provide? confer reactivity, directional metal-substrate orbital One clear feature is orthogonality: the objective overlap also plays a crucial role in activating the of introducing functionality into a cell that has no substrate to react. Indeed, it is via the formation counterpart in the natural world could provide of metal ligand, including metal substrate, chemistry that biology cannot currently catalyse, molecular orbitals that the substrate chemistry is alkene metathesis for example. As there is a limit to attenuated by the presence of the metal and that the number of additional transformations a viable catalytic reactivity can be achieved. Significant cell will perform, these orthogonal reactions may computational advancements have been made allow access to much shorter, and therefore more in the in silico design of catalytic metal binding efficient, pathways. If not for a synthetic purpose, sites (10, 11) and the mechanistic understanding one could also imagine orthogonal catalytic of reported ArMs (12–15). However, given the lack chemistry providing a diagnostic or reporter output of reliable parameters for defining transition metal without interference from the host endogenous bonding, and the immense complexity of the many processes. For it to be truly orthogonal, it is low energy interactions that determine the coupling difficult to imagine evolving a new enzyme based of protein folding to the binding of small molecules, around metals already abundant in nature and it is beyond current computational capabilities to already used as catalysts in biology. The transition predict what primary sequence and cofactors are metals used by nature are very carefully controlled necessary to achieve the optimal arrangement for by acquisition and regulatory networks that ensure metal catalysis. It therefore becomes important to catalytic metal ions are not free to operate outside have a malleable, promiscuous starting system that the endogenous metabolism. Therefore, there is can be used to sample a large space of different significant advantage in trying to introduce metals structures (16). Hence, while choosing proteins that biology currently has no evolved means of with well-defined properties and unique structures metabolising. This work therefore focuses primarily has some advantages from a design point of view, on non-biological transition metal cofactors as a starting points that do not fold into one specific route to introducing novel orthogonal activity into structure may be desirable, since they are not as a biologically viable system. closely constrained by any one particular energy well. For similar reasons, in choosing a particular 3. Evolutionary Routes to Optimised chemical strategy for introducing a metal cofactor into the protein, it becomes essential to use a Artificial Metalloenzymes method that allows for high throughput selection Natural evolution has provided numerous or screening (17). examples of metal ions used by enzymes for a plethora of different catalytic purposes. Rigorous 3.2 Considerations on Metal mechanistic and structural biochemistry has Chemistry in Proteins advanced understanding of the mechanistic detail of metalloenzyme activity significantly, to the point In addition to sampling sequence space to optimise that a few underpinning principles can be identified, the geometrical factor, protein evolution offers the linking protein structure and thermodynamics to unique possibility of sampling transition metal catalytic activity of metal centres. Together with chemistry by poising the metal in energised states. the knowledge garnered from extensive research In small molecule transition metal catalysis, on transition metal catalysts, it is possible to ligands will arrange around the metal centre establish key properties desirable for novel ArMs. to maximise bonding interactions and reach a thermodynamic minimum. In order to maintain 3.1 Considerations on the ligand exchange necessary for catalysis, some ligands tend to be weakly bonding, with the Protein-Substrate Interactions presence of strongly bonding ligands (for instance As mentioned above, the ability of enzymes to water or hydroxide) being a major factor in catalyst organise reactants cooperatively can in itself give poisoning. In enzymes however, the intramolecular rise to enhanced activity over background rates in bonds generated within the whole protein scaffold solution and in highly evolved systems this may can be used to place and maintain coordinating even be the greatest factor driving increased atoms from amino acids. These interactions can be reaction rates. It is important to realise that while seen as the second coordination sphere, shaping

409 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15928204097766 Johnson Matthey Technol. Rev., 2020, 64, (4) the metal complex and potentially leaving the first decorated with a linker moiety. These cannot sphere ligand atoms in a suboptimal configuration make use of the protein fold energy to optimise around the metal centre so that the energy of the chemical process of catalysis, a potential the resulting complex is not at a minimum on the factor in why directed evolution campaigns of coordination energy landscape. The stabilisation of ArMs have been of limited success. Whereas this complex is made possible by the favourable improvements in enantioselectivity and turnover intramolecular peptidic interactions (i.e. protein number have been reported, which can be traced fold energy) offsetting the steric and electronic to substrate binding and the hydrophobic micro- distortion of the optimum geometry (18). These environment respectively (23–25), significant energised, or entatic, states have a reactivity that is increases in the chemical turnover rate (in many not easily realised in conventional, synthetic metal systems characterised by the initial kcat) from the catalysts, if it is possible at all (19). This effect is free cofactor to the formed ArM have so far been most easily visualised by considering the common limited. Small changes in kcat can be explained biological process of activation of inert cofactors by organisational effects and indirect interactions by alteration of coordination upon binding to their with the substrate orbitals, such as charge respective apoenzymes. For instance, on their own compensation. As demonstrated by Hilvert et al., the cobalt metallo-organic cofactor, vitamin B12 significant increases in kcat have been shown to be and methionine synthase are catalytically inert; possible by fine tuning the actual centre of reaction, upon protein-cofactor binding and coordination which is the first coordination sphere of the metal of the cobalt centre to a specific histidine, methyl complex (26). From the perspective of the protein transfer activity is unmasked with great control scaffold, the formation of an entatic state requires and substrate specificity (20–22). Applying this the peptide to be at least partially folded before principle, it can be envisioned that even with the binding the metal. The more defined the fold, limited donor atoms available to proteins, a vast the greater the ability of the fold to energise the number of different complexes with different metal complex. This is in contrast to the desirable chemistries can be accessed, because the exact dynamic system for the evolutionary process. positioning as well as characteristics of the ligands A potential compromise can be struck by using dictate metal properties such as electron density, a starting scaffold that is partially folded as the redox potential, Lewis acidity and ligand exchange apoprotein and upon cofactor binding rigidifies to rates. Further, the metal cofactor does not need to a completely folded form. The initial folding energy be a bare metal ion but could be incorporated with can be used to poise the metal in an activated other ligands already attached. Interaction between state, while the folding process occurring during these ligands (for instance π–π stacking with an cofactor binding allows for the system to adapt arene ligand) and the protein can be relayed to during directed evolution. Once the ArM becomes the metal centre and allow for an even finer tuning more specialised after rounds of evolution, the of the metal centre. Again, current possibilities apoprotein will probably approach a more fully for design are insufficient to predict these effects folded form, yielding an ArM after cofactor addition which can be very subtle, highlighting the need for that is less promiscuous but contains a more biochemical high throughput screening methods. energised and active metal centre. To summarise, the number of different complex 3.3 The Optimal Method of ArM chemical factors required of ArMs demand the use of directed evolution in order to form enzymes Formation with industrially and medically relevant properties. The above considerations define a range of In order to ensure a high level of engineerability, requirements for potential methods of forming an optimal methodology for combining 4d and ArMs. Primarily, there needs to be a direct 5d metals starts with a highly promiscuous and connection between the protein scaffold and the malleable holoprotein that further has dative metal ion in the form of at least one coordination bonds between the metal ion and the peptidic bond, not only for localisation but also for poising moieties. A further point considering the cofactor the metal reactivity. As will be detailed below, attachment point is that the cofactor should be most of the successful methods of generating in a deep cleft within the protein topology rather ArMs published to date are efficient but rely on than at the surface. This is to allow the protein fully saturated, catalytically active cofactors to maximise substrate binding and secondary such as commercial transition metal catalysts transition state stabilising effects, as well as

410 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15928204097766 Johnson Matthey Technol. Rev., 2020, 64, (4) second sphere interactions influencing the metal been explored, with catalytic hydrogenation (31) complex. and hydroformylation (32) demonstrated. However, these rhodium metalloenzymes have 4. Strategies for Generating Artificial a much slower activity than commercial small molecule rhodium catalysts alone. Although in Metalloenzymes these examples it is demonstrated that unnatural ArMs are generated either from the combination metal complexes can coordinate to the natural of an unnatural transition metal cofactor being Zn(II) binding site, relatively low catalytic activity introduced into a protein scaffold or a natural is observed. The highly evolved zinc binding site metalloprotein being evolved in a laboratory to contains a complex secondary sphere architecture, enhance or alter its natural catalytic reactivity. A in order to modulate the Lewis acidity of zinc. detailed review of the field of the directed evolution The chemically different demands for rhodium of natural metalloproteins is out of the scope of catalysed hydrogenation and hydroformylation this perspective. However, the engineering and reactions will therefore not be met in this system. evolutionary approaches developed by Frances Further, evolution of such a specialised system Arnold and applied to haem metalloproteins (for may be difficult. example, cytochrome P450) are particularly Hartwig et al. reported taking the metal- noteworthy and applicable when evolving organic cofactor haem and substituting iron for unnatural metal-protein hybrid catalysts (27–29). a range of different 4d and 5d metals (including Four successful strategies have been employed rhodium, ruthenium, iridium and silver) (33). to localise an unnatural metal to a well-defined In one particularly comprehensive example, location within a protein matrix. an Ir(Me) porphyrin was incorporated into the cytochrome P450 enzyme CYP119 and catalytic 4.1 Metal Ion Substitution in Natural functionalisation of C–H bonds to C–C bonds by carbene insertion was demonstrated, capable of Enzymes high stereospecificity (25). Evolutionary campaigns Natural metal cofactors can be found in proteins on this artificial iridium metalloenzyme generated encapsulated by ligands supplied by the protein variants with an impressive 4000-fold increase in or with non-protein ligands also coordinated. This catalytic efficiency (defined by the kcat/KM), with enables two different methods of metal substitution: kinetic parameters and selectivities matching those (a) substituting the metal ion in a protein defined of native enzymes. These parameters highlight coordination site; or (b) substituting the metal ion the potential of this attachment method, and in in a natural metal-organic cofactor (such as haem) particular the advantages of introducing exogenous (Figure 2). metal cofactors with non-protein ligands remaining Many ArMs have been generated by substituting coordinated upon ArM formation. the catalytic Zn(II) ion located in a His3 binding site In this case, the mutations made to this iridium of carbonic anhydrase with different metals, for CYP119 metalloenzyme have greatly optimised the example, Coleman et al. reported esterase activity binding and pre-organisation of the substrate for of a Co(II) substituted carbonic anhydrase (30). catalysis, lowering the value for KM, (Figure 3). Replacement with different Rh(I) species has also In this system there is no direct iridium-protein coordination; the iridium metal is coordinatively saturated by four haem nitrogens, one methyl Remove ligand and coordination to the substrate. Therefore,

M1 M1 M2 the moderate increase in kcat cannot have come Add through an electronic (through bond) contribution M2 to catalysis from amino acid side chain ligands and protein fold energy but must arise from other minor contributions as discussed in the previous section. Fig. 2. Schematic representation of metal ion Another limitation of such a system is that it does substitution in natural enzymes. The natural not allow for the metal to interact with more than cofactor (red) can be substituted with a suitable one substrate at a time, an essential feature of unnatural cofactor (blue). This may include the many interesting organometallic transformations bare metal ion or larger cofactors such as haem such as metathesis.

20-fold increase in kcat M Bare metal cofactor

M M

WT ArM system Evolved ArM

–1 kcat = n/a kcat = 45.8 min KM ≥ 5 mM KM = 0.17 mM

4000-fold increase in k /K Protein cat M

Fig. 3. Comparisons between the activities of a bare cofactor and ArM before and after directed evolution. The data in this figure are taken from the work of Hartwiget al. (25). This elegant study is a good example of the issues encountered when using fully substituted artificial cofactors, even in highly optimised systems. Whereas directed evolution was able to achieve an impressive 4000-fold increase in kcat/KM, the actual chemical kcat was only moderately enhanced when compared to the cofactor in solution. This can be explained by the enzyme evolving to more strongly bind the substrate and optimise the orientation of the substrate-metal complex. However, as there is no direct metal-protein interface, directed evolution cannot influence the metal chemistry, capping the chemical potential at that observed for the free cofactor in solution

4.2 Supramolecular, Non-Covalent Binding of Tagged Complexes

There are many specific complexes between M proteins and small molecules which are well M understood and have very high affinity. ArMs have therefore been generated where a catalytic metal complex has been attached to a small molecule Fig. 4. Schematic representation of supramolecular, with high affinity for a protein target (Figure 4). non-covalent binding of tagged complexes. The This means of localising the new cofactor into a metal cofactor (red) is localised by non-covalent protein scaffold has been widely explored. Building interaction between a ligand bound recognition on the work of Wilson and Whitesides in the group (blue) and the protein 1970s (34), Ward and coworkers have assembled ArMs based on the high supramolecular affinity of small molecule biotinylated metal catalysts for the metathesis catalyst was localised to a hydrophobic protein streptavidin. As many as 12 different binding site in human serum albumin. The catalytic transformations have been performed metalloenzyme was directed to cancerous tissue by these metal-streptavidin hybrids, including (through specific glycosylation) and a pro-drug ruthenium-catalysed olefin metathesis (17), was administered which upon metathesis induced ruthenium-catalysed deallylation (35), iridium- cellular death (37). catalysed transfer hydrogenation (24) and One key benefit of supramolecular assembly dirhodium-catalysed cyclopropanation (36), all is apparent in the examples described above, in vivo. and that is that the conjugation between metal This strategy has also been employed in ArMs and protein is robust enough to be performed that were reported by Tanaka et al. for potential in complex cellular environments. Furthermore, therapeutic application. In this example, a unlike covalent attachment, supramolecular coumarin derivative tagged with a ruthenium assembly can be a reversible process, which allows

412 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15928204097766 Johnson Matthey Technol. Rev., 2020, 64, (4) for component recycling. In a recent report of scaffolds (44). Since then, the most successful Duhme-Klair et al. catalytic transfer hydrogenation generation of ArMs involving a covalent linkage is demonstrated from a siderophore-protein to an UAA were reported by Lewis et al. and combination that enables strong but redox- involve a reaction between an alkyne-substituted reversible catalyst anchoring (38). All current dirhodium catalyst and a genetically encoded examples of ArMs generated by supramolecular L-4‑azidophenylalanine residue through strain- assembly do, however, rely on the assembly of promoted azide-alkyne cycloaddition (SPAAC) proteins with known, highly catalytically active (45–47). Hypothetically, UAAs could be encoded metal complexes. As discussed previously, using into a specific residue of most proteins; here, the complexes which maintain their ligand set during protein scaffold selected was a β-barrel prolyl ArM formation does not allow the metal complex to oligopeptidase and the resulting metalloenzymes be subjected to evolutionary pressures limiting the generated catalysed olefin cyclopropanation. evolutionary potential. The effectiveness of introducing UAA via stop codon methodology is that theoretically the 4.3 Covalent Anchoring Through same conjugation technology is applicable to many different proteins to generate diverse ArMs Metal Ligands through a specific, fast and irreversible covalent Covalent anchoring relies on using a chemical conjugation. Beside commonly relying on pre- reaction to covalently link a protein side chain to formed metal complexes, an overarching issue of a strong ligand for a metal (Figure 5). Covalent covalent attachment and supramolecular assembly anchoring methods can be split into two broad is that the protein scaffold is used predominantly categories: (a) modification of a natural amino as an auxiliary providing a chiral and hydrophobic acid side chain (for example cysteine, lysine or micro-environment. Further, many reported tyrosine), via a nucleophilic–electrophilic reaction methods utilise a long flexible linker between the and (b) coupling through a genetically encoded point of attachment and the metal complex which unnatural amino acid (UAA). could remove the catalytic centre from the very There is a resurgence in research for developing interactions needed for the protein to exert an novel bioconjugation and protein modification influence on transition states. techniques of natural amino acids (such as cysteine, lysine or tyrosine) (39, 40). Generating ArMs 4.4 Direct Activation by Metal through cysteine modification is attractive due to Coordination to Protein Side Chains the high nucleophilicity and rarity of free cysteines allowing for greater control of reactivity. Salmain Dative ArMs have one or more coordination bonds and coworkers have modified the free Cys25 in directly from the metal to a Lewis basic amino acid the cysteine protease papain, using a variety residue (His, Cys, Ser, Glu, Asp) on the protein of ruthenium, rhenium and rhodium complexes scaffold (Figure 6). The protein therefore has all functionalised with either a maleimide or a direct electronic influence on the reactivity at chloroacetamide group (41–43). the metal centre. The active hybrid molecule is The pioneering work of the Schultz laboratory formed by substitution reactions from a precursor enabled incorporation of UAAs into protein metal species and the apoprotein. This allows

M M M E M Nu LB LB LB LB Nu E

Fig. 6. Schematic representation of cofactor Fig. 5. Schematic representation of covalent attachment via direct activation by metal anchoring through metal ligands. The metal coordination to protein side chains. The free metal cofactor (red) is attached to the protein by a cofactor (red) attaches to Lewis basic residues on reaction forming a covalent bond, for instance the protein (LB) via ligand substitution reactions, nucleophile (Nu) attacking an electrophile (E) forming a new metal-protein complex (blue)

413 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15928204097766 Johnson Matthey Technol. Rev., 2020, 64, (4) for potentially very clean reaction conditions for cyclopentadiene with medium selectivity, however, assembly of the metal-protein complex. Although to our knowledge no subsequent directed evolution advances have been made, the complexity of experiments have been reported. these metal-protein binding processes remain In contrast to these examples of forming the a major challenge for the design of competently complete coordination sphere by binding a bare folded and catalytically active metalloproteins from metal to the apoprotein state of the ArM, to the best scratch. It is important to distinguish between of our knowledge there are only very few examples metalloenzymes where coordination to the of adding exogenous metal complexes (particularly metal is provided only by amino acid sidechains, 4d and 5d metal complexes) as precursor cofactors substrates and solvents, and those in which the which then show catalytic activity upon direct metal brings its own specific ligands with it. The coordination to a protein (57). This is a particularly latter, metal cofactors would be artificial versions attractive methodology as the challenges of taking of commonly encountered natural examples such unnatural ligands such as arenes, carbenes and as haem, vitamin B12 and molybdopterin which phosphanes into biology become opportunities for are (bio)synthesised separately and bind to the expanding the repertoire of chemistries available protein through both non-covalent interactions and for catalysis. Controlling the ligand exchange coordination. As pointed out above, their activity is behaviour of 4d and 5d metal complexes with defined by the other ligands they carry to an active protein side chain ligands is challenging, not least site as well as the coordination by the protein. because coordination bonds between ligands and Degrado and coworkers have pioneered the heavier metals are often stronger than their 3d design of a number of synthetic proteins which counterparts and hence exchange rates are slower. directly coordinate bare metal atoms or metal This, however, remains an exciting area of research cofactors (10, 48). For example, in some of the due to the catalytic diversity demonstrated by earliest work, the His3-Zn(II) binding motif found many 4d and 5d metal complexes. In this specific in carbonic anhydrase was introduced into a area our own work has focused upon ruthenium designed four helical bundle protein, and hydrolytic complexes and their ligand exchange behaviour activity was observed (49). More recently, with biological systems, laying the foundation for de novo design has been coupled with directed future work into ArMs with direct metal-protein evolutionary approaches to generate an artificial coordination (58, 59). zinc metalloenzyme capable of accelerating ester cleavage with un-paralleled catalytic efficiency 6 –1 –1 5. Summary and Outlook (kcat/KM of 10 M s ) (26). In a range of studies (13, 50–52), Roelfes and Significant advances in the incorporation of coworkers use amber stop codon technology to organometallic complexes into proteins in order introduce the UAA (2,2′-bipyridin-5yl)alanine into to generate ArMs have been made. The studies a range of protein scaffolds. Upon addition of highlighted above reliably create hybrid molecules different bare metal ions, they were able to obtain where the stability and turnover number of the ArMs catalysing the Friedel-Crafts alkylation of metal centre is higher than the comparable small indoles, enantioselective metallohydration and the molecule organometallic complex in aqueous stabilisation of a semiquinone radical. By the use solution. Maybe unsurprisingly, the propensity of sophisticated computational design, the group for side reactions and catalyst decomposition was able to introduce beneficial point mutations in is lowered once the complex is in a hydrophobic many of the novel hybrid molecules, improving both protein environment, already showcasing the enantioselectivity and yield. The advances in stop usefulness of these hybrid systems. However, codon technology to introduce UAAs, especially in the question remains, as to whether or not these the context of directed evolution, make their use a strategies make full use of the protein component. promising option and provides an enticing method The unique and numerous demands of ArMs call for for expanding the ligand set available to the a highly integrated approach. To date, most of the protein scaffold (53–55). In another study, Reetz work described in the literature attempts to exploit and coworkers computationally designed a Cu(II) the chemistry of metal ions and their complexes ion binding site into the thermostable protein in a protein scaffold but with limited influence imidazole glycerol phosphate synthase (56). from the protein on any catalytic activity because The resulting ArM was able to catalyse the metal-protein coordination is largely indirect and Diels-Alder cycloaddition of an azachalcone and so cooperativity is limited.

The potential for synthetic organometallic in that reactive promiscuity is reduced. As pointed chemistry to deliver cofactors which utilise ligand out above, such complexes would be designed to chemistry not available to naturally evolved have a latent catalytic activity which emerges once systems can vastly expand the orthogonal catalysis the metal complex is bound to a protein. The design available in synthetic biological applications. Using challenges raised by this approach are not just as a such molecules to embed novel metal-peptide result of a need to control the electronic and three- hybrid complexes in protein scaffolds allows for dimensional steric coordination sphere of the metal three-dimensional and electronic control around ion, but also to limit ligand exchange processes, the metal centre that reduces the need for intricate restricting lability of a precursor complex (in the synthetic catalyst generation. Instead, control of cellular milieu) until it reaches a specific protein the steric and electronic environment around target. Since the metal-ligand exchange processes the metal ion can be delivered via the protein for 4d and 5d metal complexes are typically slow, coordination sphere, particularly where a direct they are particularly attractive from this point of coordination bond is used to anchor the metal ion view but are hard to predict ab initio. to the protein. When combined with the ability to efficiently evolve enzymes, a sophisticated 6. Conclusion organometallic precursor complex together with a suitable apoprotein could potentially give rise to In conclusion, in order to optimise the chemistry a number of diverse reactivities. Therefore, new and biochemistry of ArMs, directed evolutionary problems in catalysis could be addressed in a faster campaigns coupled with high throughput screening and more specific manner than with small molecule methods rather than individually-designed synthetic catalysts. Together with non-covalent contributions strategies are much more likely to generate to catalysis and the intermolecular interactions that optimised orthogonal catalysts for new and efficient pre-organise substrates and stabilise transition metabolic processes. Direct coordination between states, such a system contains many readily metal ions and enzymes is essential in order to evolvable components. deliver truly interdependent systems, ideally where The majority of protein scaffolds selected for entatic states deliver enhanced reactivity, efficiency ArM construction have been chosen because of and selectivity that cannot easily be replicated their apparent engineerability. However, in most in conventional, synthetic metal catalysis. Going cases the focus seems to lie solely on the peptidic forward, methods of generating ArMs should be component with little consideration for evolution of evaluated and developed for both their ability to the metal complex. Although methods of selection be used in directed evolution procedures and the and directed evolution have been applied, these extent to which the protein scaffold participates in are often operating on already well-defined protein the activity of the metal complex. scaffolds that carry an abiotic cofactor but nota direct protein-metal complex, which inevitably Acknowledgements limits the scope for evolution. Arguably it is desirable, therefore, to select for a promiscuous George Biggs is supported by the Engineering and versatile protein starting point which is not and Physical Sciences Research Council (EPSRC) constrained by one energy minima but instead can (EP/N509620/1) and Peterhouse, University of potentially offer numerous distinct metal-binding Cambridge, UK. Oskar James Klein is supported environments, both in terms of direct coordination by the EPSRC (EP/R513180/1). Sally Boss and and through secondary, intramolecular spheres Paul Barker thank the Department of Chemistry, of influence, ultimately generating differential University of Cambridge. We thank Florian catalytic ArM activity. Hollfelder for deep discussions. Performing catalysis with exogenous metal complexes within cellular environments has References enormous potential applications in medicinal chemistry and synthetic biology. Given the potential 1. G. Winter, A. R. Fersht, A. J. Wilkinson, M. Zoller difficulties associated with cell-uptake, minimising and M. Smith, Nature, 1982, 299, (5885), 756 deactivation, overcoming toxicity of exogenous 2. A. R. Fersht, J-P. Shi, J. Knill-Jones, D. M. Lowe, metal ions and precise localisation of metal A. J. Wilkinson, D. M. Blow, P. Brick, P. Carter, cofactors in cells, the idea of using traditionally inert M. M. Y. Waye and G. Winter, Nature, 1985, organometallic complexes has obvious advantages 314, (6008), 235

3. F. Schwizer, Y. Okamoto, T. Heinisch, Y. Gu, Sci., 1968, 59, (2), 498 M. M. Pellizzoni, V. Lebrun, R. Reuter, V. Köhler, 20. C. L. Drennan, S. Huang, J. T. Drummond, J. C. Lewis and T. R. Ward, Chem. Rev., 2018, R. G. Matthews and M. L. Lidwig, Science, 1994, 118, (1), 142 266, (5191), 1669 4. M. Jeschek, S. Panke and T. R. Ward, Trends 21. M. L. Ludwig, C. L. Drennan and R. G. Matthews, Biotechnol., 2018, 36, (1), 60 Structure, 1996, 4, (5), 505 5. J. G. Rebelein and T. R. Ward, Curr. Opin. 22. R. Banerjee and S. W. Ragsdale, Annu. Rev. Biotechnol., 2018, 53, 106 Biochem., 2003, 72, 209 6. J. D. Tyzack, A. J. M. Ribeiro, N. Borkakoti and 23. M. T. Reetz, Acc. Chem. Res., 2019, 52, (2), J. M. Thornton, ACS Synth. Biol., 2019, 8, (11), 336 2494 24. J. Zhao, J. G. Rebelein, H. Mallin, C. Trindler, 7. A. Bhushan, P. J. Egli, E. E. Peters, M. F. Freeman M. M. Pellizzoni and T. R. Ward, J. Am. Chem. Soc., and J. Piel, Nat. Chem., 2019, 11, (10), 931 2018, 140, (41), 13171 8. M. C. Wilson, T. Mori, C. Rückert, A. R. Uria, 25. P. Dydio, H. M. Key, A. Nazarenko, J. Y.-E. Rha, M. J. Helf, K. Takada, C. Gernert, U. A. E. Steffens, V. Seyedkazemi, D. S. Clark and J. F. Hartwig, N. Heycke, S. Schmitt, C. Rinke, E. J. N. Helfrich, Science, 2016, 354, (6308), 102 A. O. Brachmann, C. Gurgui, T. Wakimoto, M. Kracht, M. Crüsemann, U. Hentschel, I. Abe, 26. S. Studer, D. A. Hansen, Z. L. Pianowski, S. Matsunaga, J. Kalinowski, H. Takeyama and P. R. E. Mittl, A. Debon, S. L. Guffy, B. S. Der, J. Piel, Nature, 2014, 506, (7486), 58 B. Kuhlman and D. Hilvert, Science, 2018, 362, (6420), 1285 9. A. Casini, F. Y. Chang, R. Eluere, A. M. King, E. M. Young, Q. M. Dudley, A. Karim, K. Pratt, 27. P. S. Coelho, E. M. Brustad, A. Kannan and C. Bristol, A. Forget, A. Ghodasara, R. Warden- F. H. Arnold, Science, 2013, 339, (6117), 307 Rothman, R. Gan, A. Cristofaro, A. E. Borujeni, 28. F. H. Arnold, Acc. Chem. Res., 1998, 31, (3), M.-H. Ryu, J. Li, Y.-C. Kwon, H. Wang, E. Tatsis, 125 C. Rodriguez-Lopez, S. O’Connor, M. H. Medema, 29. F. H. Arnold, Angew. Chem., Int. Ed., 2018, M. A. Fischbach, M. C. Jewett, C. Voigt and 57, (16), 4143 D. B. Gordon, J. Am. Chem. Soc., 2018, 140, (12), 30. J. E. Coleman, Nature, 1967, 214, (5084), 193 4302 31. Q. Jing, K. Okrasa and R. J. Kazlauskas, Chem. 10. W. F. DeGrado, C. M. Summa, V. Pavone, F. Nastri Eur. J., 2009, 15, (6), 1370 and A. Lombardi, Annu. Rev. Biochem., 1999, 68, 779 32. Q. Jing and R. J. Kazlauskas, ChemCatChem, 11. L. Falivene, Z. Cao, A. Petta, L. Serra, A. Poater, 2010, 2, (8), 953 R. Oliva, V. Scarano and L. Cavallo, Nat. Chem., 33. H. M. Key, P. Dydio, D. S. Clark and J. F. Hartwig, 2019, 11, (10), 872 Nature, 2016, 534, (7608), 534 12. L. Alonso-Cotchico, J. Rodrı́guez-Guerra, A. Lledós 34. M. E. Wilson and G. M. Whitesides, J. Am. Chem. and J.-D. Maréchal, Acc. Chem. Res., 2020, Soc., 1978, 100, (1) 306 53, (4), 896 35. T. Heinisch, F. Schwizer, B. Garabedian, E. Csibra, 13. I. Drienovská, L. Alonso-Cotchico, P. Vidossich, M. Jeschek, J. Vallapurackal, V. B. Pinheiro, A. Lledós, J.-D. Maréchal and G. Roelfes, Chem. P. Marlière, S. Panke and T. R. Ward, Chem. Sci., Sci., 2017, 8, (10), 7228 2018, 9, (24), 5383 14. L. Alonso-Cotchico, G. Sciortino, P. Vidossich, 36. A. D. Liang, J. Serrano-Plana, R. L. Peterson J. Rodríguez-Guerra Pedregal, I. Drienovská, and T. R. Ward, Acc. Chem. Res., 2019, 52, (3), G. Roelfes, A. Lledós and J.-D. Maréchal, ACS 585 Catal., 2019, 9, (5), 4616 37. S. Eda, I. Nasibullin, K. Vong, N. Kudo, M. Yoshida, 15. H. Feng, X. Guo, H. Zhang, L. Chen, P. Yin, A. Kurbangalieva and K. Tanaka, Nat. Catal., 2019, C. Chen, X. Duan, X. Zhang and M. Wei, Phys. 2, (9), 780 Chem. Chem. Phys., 2019, 21, (42), 23408 38. D. J. Raines, J. E. Clarke, E. V. Blagova, 16. R. B. Leveson-Gower, C. Mayer and G. Roelfes, E. J. Dodson, K. S. Wilson and A.-K. Duhme-Klair, Nat. Rev. Chem., 2019, 3, (12) 687 Nat. Catal., 2018, 1, (9), 680 17. M. Jeschek, R. Reuter, T. Heinisch, C. Trindler, 39. J. N. deGruyter, L. R. Malins and P. S. Baran, J. Klehr, S. Panke and T. R. Ward, Nature, 2016, Biochemistry, 2017, 56, (30), 3863 537, (7622), 661 40. E. A. Hoyt, P. M. S. D. Cal, B. L. Oliveira and 18. W. R. Hagen, Metallomics, 2019, 11, (11), 1768 G. J. L. Bernardes, Nat. Rev. Chem., 2019, 3, (3) 19. B. L. Vallee and R. J. Williams, Proc. Natl. Acad. 147

41. P. Haquette, B. Talbi, S. Canaguier, S. Dagorne, 51. N. Ségaud, I. Drienovská, J. Chen, W. R. Browne C. Fosse, A. Martel, G. Jaouen and M. Salmain, and G. Roelfes, Inorg. Chem., 2017, 56, (21), Tetrahedron Lett., 2008, 49, (31), 4670 13293

42. P. Haquette, M. Salmain, K. Svedlung, A. Martel, 52. M. Bersellini and G. Roelfes, Org. Biomol. Chem., B. Rudolf, J. Zakrzewski, S. Cordier, T. Roisnel, 2017, 15, (14), 3069 C. Fosse and G. Jaouen, ChemBioChem, 2007, 53. Q. Wang, A. R. Parrish and L. Wang, Chem. Biol., 8, (2), 224 2009, 16, (3), 323 43. N. Madern, N. Queyriaux, A. Chevalley, M. Ghasemi, 54. C. Mayer, C. Dulson, E. Reddem, A.-M. W. H. O. Nicolotti, I. Ciofini, G. F. Mangiatordi and Thunnissen and G. Roelfes, Angew. Chem. Int. M. Salmain, J. Mol. Catal. B: Enzym., 2015, 122, Ed., 2019, 58, (7), 2083 314 44. C. J. Noren, S. J. Anthony-Cahill, M. C. Griffith and 55. A. Fallah-Araghi, J.-C. Baret, M. Ryckelynck and P. G. Schultz, Science, 1989, 244, (4901), 182 A. D. Griffiths,Lab Chip, 2012, 12, (5), 882 45. J. C. Lewis, Curr. Opin. Chem. Biol., 2015, 25, 27 56. J. Podtetenieff, A. Taglieber, E. Bill, E. J. Reijerse and M. T. Reetz, Angew. Chem. Int. Ed., 2010, 46. H. Yang, P. Srivastava, C. Zhang and J. C. Lewis, ChemBioChem, 2014, 15, (2), 223 49, (30), 5151 47. P. Srivastava, H. Yang, K. Ellis-Guardiola and 57. J. de Jesús Cázares-Marinero, C. Przybylski and J. C. Lewis, Nat. Commun., 2015, 6, (1), M. Salmain, Eur. J. Inorg. Chem., 2018, (12), 7789 1383 48. A. Lombardi, F. Pirro, O. Maglio, M. Chino and 58. T. G. Scrase, M. J. O’Neill, A. J. Peel, P. W. Senior, W. F. DeGrado, Acc. Chem. Res., 2019, 52, (5), P. D. Matthews, H. Shi, S. R. Boss and P. D. Barker, 1148 Inorg. Chem., 2015, 54, (7), 3118

49. T. Handel and W. F. DeGrado, J. Am. Chem. Soc., 59. G. S. Biggs, M. J. O’Neill, P. Carames Mendez, 1990, 112, (18), 6710 T. G. Scrase, Y. Lin, A. M. Bin-Maarof, A. D. Bond, 50. I. Drienovská, A. Rioz-Martínez, A. Draksharapu S. R. Boss and P. D. Barker, Dalton Trans., 2019, and G. Roelfes, Chem. Sci., 2015, 6, (1), 770 48, (20), 6910

The Authors

George Biggs completed an MChem in Chemistry at the University of Bath, UK, in 2016. He is now a PhD student in the Department of Chemistry at the University of Cambridge. Supervised by Paul Barker and Sally Boss, his project is focused on understanding the reactivity of Ru(II) arene complexes with proteins for the development of novel ArMs.

Oskar James Klein obtained an MSc in Chemistry from the University of Cambridge in 2019, where he remains as a PhD student in the Department of Chemistry. Supervised by Paul Barker and Sally Boss and in collaboration with Professor Florian Hollfelder his project tries to develop a high throughout methodology for the formation and evolution of novel ArMs.

Sally Boss studied for an MSci in Chemistry at the University of Bristol, UK, and Heidelberg University, Germany, before moving to the University of Cambridge to begin a PhD on the synthesis and reactivity of Lewis acidic, heterobimetallic main group compounds. Shortly after obtaining her PhD in 2005, she was appointed to a joint College Lectureship in the Department of Chemistry and at Churchill College, University of Cambridge. Her time is split between teaching and research and her specific interest is in improving the utility of heavy metals in biology by careful design of complexes, targeted direction of metal- cofactors to protein targets and using spectroscopy to understand how they behave in situ.

Paul Barker is Senior Lecturer at the University of Cambridge, Department of Chemistry, and a Fellow of Downing College, University of Cambridge. His research has always been at the interface between inorganic chemistry and biology. It started in the field of electron transfer proteins studied by biophysical methods and mutagenesis, in the early days of protein engineering. After two, independent Medical Research Council (MRC), UK, and Biotechnology and Biological Sciences Research Council (BBSRC), UK, fellowships in Cambridge he joined the Chemistry faculty and has been combining protein engineering with synthesis and self-assembly for the purposes of generating novel protein based electronic and catalytic systems. His current interests span protein design and evolution, self-assembling materials and synthesis of organometallic complexes.

www.technology.matthey.com

Reduction of Biofilm Formation on Cooling Tower Heat Exchangers using Nano-silica Coating Environmentally sustainable antifouling coating demonstrated on stainless steel heat exchanger tubes

Irfan Turetgen towers provide cooling by spraying the heated Basic and Industrial Microbiology Section, water coming from the system onto a fill material Department of Biology, Faculty of Science, and rejecting the heat to the open atmosphere (1). Istanbul University, 34134 Vezneciler, Istanbul, The cooled water returns to a basin to recirculate Turkey again through the system. Common uses of wet cooling towers include air conditioner systems, Email: [email protected] manufacturing facilities, telecommunication devices and power plants. Such man-made installations provide an ideal environment for Cooling towers are industrial cooling units bacterial growth similar to an incubator, supported operating to dissipate heat. As with any surface in by water temperatures ranging between 24°C contact with aqueous systems, biofilm formation and 38°C (2–4). The heated water comes from appears on the surface of heat exchangers. the source to the heat exchanger that allows the Although biofilm formation on plastic tower fill in exchange of heat between two liquids at different wet cooling towers has been studied widely, no temperatures by indirect contact inside water studies were found regarding biofilm formation jacketed tubes (5). on steel heat exchangers in closed-loop systems. Wet cooling towers providing cool water In this study, heat exchangers were coated with for heating, ventilating and air conditioning nano‑silica, which is known to reduce the formation (HVAC) systems are known to be subject to of biofilm. Natural biofilm formation was monitored contamination. Organic and inorganic substances for six months. Biofouling was examined monthly in bulk water are deposited on the water contact using epifluorescence microscopy by assessing the surfaces, reducing the heat transfer significantly numbers of live and dead bacteria. It was observed and threatening the operating stability of that the biofilm layer formed on the nano-silica the whole system. Established biofilms offer coated heat exchanger surfaces was significantly cleaning challenges because they are resistant lower than on the control surfaces. 3 log microbial to most chemical and physical cleaning reduction was recorded on coated surfaces in the protocols and they also reduce the heat transfer first month. After six months, total biomass on efficiency (6, 7). HVAC systems are responsible control surfaces reached 1.28 × 1012 cell cm–2, for about half of the energy consumed in modern while the biomass on nano-silica coated surfaces buildings and industrial facilities. Therefore, was 6.3 × 104 cell cm–2. biofouling is always a significant issue for heat exchangers and should be taken into account 1. Introduction during heat exchanger design and production. As a solution, altering the surface properties could A cooling tower is a heat dissipation unit which be an effective approach to reduce biofouling in cools bulk water in industrial systems. Cooling such hard-to-reach articulated systems (2, 8, 9).

A biofilm layer is a community formed by bacterial Materials and Method cells living in a polymeric matrix that they produce; a functional partnership adhered to a living or A brand new fabricated closed-loop cooling tower inanimate surface organised by microorganisms was monitored for six months. A real size, fully in a dense exopolymer matrix. The capability of working closed-loop cooling tower system was microbes to stick to a substratum and to produce kept in operation by the manufacturer during the a biofilm layer has great significance in a diversity experimental period at the factory test laboratory. of cooling towers, where fouling can act as a The system was filled with distributed network perpetual source of contamination. Biofilm layer water. Regular blowdown was implemented to must be kept to a minimum in order to prolong limit the concentration of dissolved solids. In a the operating life of man-made water systems and circulation rig, hot process water was kept separate facilitate control of pathogens. Disinfectants may from the cooling water in a closed-loop system be used for this purpose (4, 9). (Figure 1). For the experiments, a half portion of Industrial cooling towers can be manufactured the stainless steel (316 SS) heat exchanger tubes from different materials. Generally, towers were coated with nano-silica and the other part was are made of reinforced concrete or fibreglass, left without coating. Coating was done by coaxial stainless steel, wood or reinforced plastic electrospraying before assembly and left to cure in sheets. The fill material is generally made of air for 24 hours. Coaxial electrospraying has several plastic sheets (polypropylene, polyethylene or implicit advantages such as high encapsulation polyvinylchloride) where heat dissipation occurs. efficiency and uniform particle distribution. The For corrosion resistance, towers are specially coating thickness was between 4–6 µm. Before treated, painted and covered with a protective coating, the heat exchanger tubes were sprayed film layer (7). In the case of corrosive water or with 96% ethanol to remove any dirt, oil or grime. atmospheric conditions, the use of plastic towers This application made the bonding of the coating is recommended. But heat exchanger units are stronger. made of stainless steel or copper for better Silica in powder form is hydrophilic. To produce thermal conduction (5). The critical issue that hydrophobic nano-silica, the silica particles were affects cooling is the aggregation of deposits transformed by fluorination to confer hydrophobicity. over the heat exchanger surfaces which includes The final particle size was about 40 nm. The biofouling. Conventional steel heat exchangers aqueous form of the nano-silica coating contains may have corrosion or deposits may have formed ethanol as solvent to keep it in liquid form before on the heat exchanger tubes. Both of these factors reduce the heat transfer rate (10). To solve this problem, novel anti-fouling coatings are considered. Nano-silica can be used in the form of liquid composites in many matrices as coating materials. Nano-silica is used in the textile and automotive industries because of its Fan self-cleaning, abrasion resistant, hydrophobic Heat and oleophobic features. It is known that exchanger nano-silica is able to create low-cost, hard and Hot tough coatings which are resistant to wear and water weathering (11). inlet Although biofilm formation on plastic fill surfaces in wet cooling towers has been studied widely, no studies were found on biofilm formation on steel Cooled heat exchangers in cooling towers. As coating of water heat exchangers is not common, the aim of the outlet current work was to limit tenacious biofouling on heat exchangers using a nano-silica coating, which will lead to longer material life, better Fig. 1. Schematic view of the cooling tower and cooling of water and less clogging in closed-loop heat exchanger systems.

420 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15895565390677 Johnson Matthey Technol. Rev., 2020, 64, (4) use. The final nano-silica product was supplied by carried out using special software (NIS-Elements, a local company. After curing, the coating was solid Nikon Instruments Inc, Japan). Signals obtained on the surfaces, and no colour change, shedding from 20 randomly selected regions were recorded. or weight loss were observed on any of the coated Images were saved for later analysis. test surfaces after the experimental period. The The LIVE/DEAD® kit stains dead cells red and stability of the coating was tested in a different live cells green in colour. The LIVE/DEAD® test study by the present author (9) and the mean kit contains two DNA-binding dyes, propidium overall adhesion capability of the coating was iodide and SYTO® 9. These dyes differ in their recorded as 1.6 using a pull-off adhesion tester, spectral properties and their ability to enter the which matches very well with the general rating of living bacterial cell. The first dye in the kit is adhesion. Water was circulated over the stainless SYTO® 9, which can pass through the membrane steel (316 SS) heat exchanger tubes, where natural of all bacteria and stain the cells green. Propidium biofilm formation was allowed to occur. Sampling of iodide only enters into cells with a damaged cell the biofilm required dismantling the outer shell of membrane, allowing them to appear red under the heat exchanger unit every month. The system fluorescent light. The number of viable and dead temperature water was kept constant at 37°C using bacteria on surfaces can be determined in a single an electrical heating unit to eliminate temperature step using a dual emission filter cube (Chroma fluctuation which might influence biofilm formation Technology GmbH, Germany). over time. For both parameters over the six-month duration Pipe segments were cut monthly from the heat of the experiment, the difference between the exchanger using an angle grinder, kept in a average bacterial numbers were compared container filled with system water and brought by two-way analysis of variance. A follow-up quickly to the laboratory for analysis. LIVE/DEAD® post‑hoc analysis was done in order to determine BacLightTM Bacterial Viability Kit (InvitrogenTM, differences. The difference was considered Thermo Fisher Scientific, USA) dye was added significant when p < 0.05. SPSS® Version 18.0 immediately to cover the surfaces completely to software (IBM Corp, USA) was used for the stain the actively respiring and dead bacteria. statistical analyses. After 15 min, the surfaces were rinsed with sterile bi-distilled water to remove unattached cells, air Results and Discussion dried, covered with immersion oil and cover slip, then examined in the dark. This was repeated The bacterial numbers from the LIVE/DEAD® test every month until the study finished at the sixth kit were analysed in situ on the surfaces using the month. An epifluorescence microscope (Eclipse manufacturer’s software during the experimental 80i, Nikon Instruments Inc, Japan) was used period for six months. The results are given in to visualise the biofilm cells in situ. The camera Table I. The number of signals per cm2 were enables counting and taking images of bacteria on calculated using the magnification factor. Since solid surfaces, with the signals displayed on the the raw data were too scattered, the values are computer monitor. Counting and recording were given in the logarithmic (log10) base for better

Table I Numbers with Standard Deviation of Live-Dead Bacteria Counted on Heat Exchanger Surfaces Nano-silica coated test surfaces, cell cm–2 Uncoated control surfaces, cell cm–2

Months Dead (log10) Live (log10) Total (log10) Dead (log10) Live (log10) Total (log10)

1 3.6 ± 0.07 3.6 ± 0.05 4.6 ± 0.09 6.8 ± 0.11 6.0 ± 0.08 8.1 ± 0.14

2 3.3 ± 0.09 3.6 ± 0.10 4.3 ± 0.12 7.7 ± 0.13 7.5 ± 0.11 9.7 ± 0.16

3 3.3 ± 0.04 3.7 ± 0.07 4.2 ± 0.08 6.9 ± 0.11 8.0 ± 0.14 10.1 ± 0.09

4 3.0 ± 0.02 3.8 ± 0.04 4.5 ± 0.10 7.0 ± 0.13 8.2 ± 0.12 11.5 ± 0.17

5 3.1 ± 0.09 3.8 ± 0.10 4.7 ± 0.11 7.5 ± 0.15 8.1 ± 0.15 11.9 ± 0.17

6 3.2 ± 0.08 3.9 ± 0.06 4.8 ± 0.12 7.8 ± 0.17 9.1 ± 0.18 12.1 ± 0.21

421 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15895565390677 Johnson Matthey Technol. Rev., 2020, 64, (4) comparison. The logarithmic reduction was clearly properties of the material and supports less biofilm significant starting from the first sampling. formation (16–18). Hydrophobic coatings limit the The total bacterial numbers on coated tubes wettability of the surface, making it difficult for were recorded as 49,090 cell cm–2 after the initial organic and inorganic matter or microorganisms to month, and 13,016,957 cell cm–2 on uncoated adhere; and even if they do, they can easily be surfaces after the first month. The results distinctly detached from the surface by physical factors such showed that this type of coating reduces biofouling as laminar or turbulent water shear stress (19). formation on heat exchanger surfaces from the The issue of antimicrobial coatings has been start of the experimental set-up. The numbers extensively studied (20–24). The problem with these of surface associated bacteria on uncoated products is development of bacterial resistance control tubes gradually increased and reached against the agent (11, 25). Even antibiotic- 1.28 × 1012 cell cm–2 after the sixth month, at which containing coatings have been reported to promote time the biomass on nano-silica coated tubes was biofilm formation (26). Silver compounds combined 6.3 × 104 cell cm–2. No significant rise (p < 0.05) of with silica, silane and titanium coatings in particular bacterial numbers on nano-coated heat exchanger gave antimicrobial properties but the problem of tubes was recorded during the six‑month period toxicity in medical devices was mentioned (27). In in terms of total biofilm counts. This outcome industrial use, the resistance of microorganisms is demonstrates that a nano-silica coating can clearly at the top of the list as a disadvantage (28). In reduce the bacterial biofilm layers on coated heat addition, silver compounds in water systems will exchanger surfaces. reach the aquatic environment and appear as a As expected, nano-silica coating slowed down separate environmental problem. the adhesion and colonisation of bacteria on It is also emphasised that anti-biofilm coatings the substrata thanks to its strong hydrophobic are very important for preventing the formation of properties. The pH, dissolved oxygen, total a biofilm layer at an early stage (29, 30). However, dissolved matter and temperature values of the studies conducted to date are mostly aimed at water in the system during the six-month test solving clinical problems and have been done period were recorded and are given in Table II. in vitro with pure cultures (15, 17, 18, 31–33). The values in Table II were important to monitor Using monospecies biofilms is a sterile approach circulating water due to the blowdown regime. and cannot represent mixed cultures in the natural It is known that even with conventional cleaning environment and their interaction with each other. and disinfection regimens, there is a problem Sol-gel products and superhydrophobic coatings fighting against biofilm formation and development which are more strongly water repellent (31, 34) of microbial resistance (12). Based on previous have also been tried. It was observed that the life studies conducted in this field (13, 14), it is of these coatings was not as long as hydrophobic impossible to eliminate the formation of biofilm coatings. On the other hand, the high cost of layers on surfaces, but biofilm formation can be superhydrophobic coatings was a drawback. reduced (9, 15, 16). For this purpose, it is possible Contrary to hydrophobic coatings, some hydrophilic to modify surfaces with different coatings. The coatings were also found to be effective against nano-hydrophobic coating changes the surface biofouling. Holberg et al. (8) reported that

Table II pH, Dissolved Oxygen, Total Dissolved Solids and Temperature Values of Circulating Water in the Systema

Months pH Dissolved oxygen, mg l–1 Total dissolved solids, ppm Temperature, °C

1 7.33 7.40 110 37

2 7.48 7.54 113 37

3 7.28 7.34 109 37

4 7.24 7.24 108 37

5 7.30 7.55 107 37

6 7.36 7.39 110 37

aThe numerical data were the average of three consecutive measurements

422 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15895565390677 Johnson Matthey Technol. Rev., 2020, 64, (4) biocide-free silicone coatings showed promising 2. S. K. R. Namasivayam, A. L. Francis, real-life performance on fresh water-cooled heat R. S. A. Bharani and C. V. Nachiyar, J. Clean. exchangers and also performed well in laboratory Prod., 2019, 231, 872 tests. 3. H.-F. Tsao, U. Scheikl, C. Herbold, A. Indra, Ding et al. (35) tested an environmentally friendly J. Walochnik and M. Horn, Water Res., 2019, 159, antifouling coating product, butenolide, which was 464 designed for controlled release from biodegradable 4. I. Türetgen, Biofouling: J. Bioadhes. Biofilm Res., polyurethane. The anti-fouling effect was shown by 2004, 20, (2), 81 in situ tests. The main target was marine biofouling, 5. Z. Nourani, A. Naserbegi, Sh. Tayyebi and especially larval settlement on surfaces. Since the M. Aghaie, Therm. Sci. Eng. Prog., 2019, 14, adhesion of fouling organisms relies on a microbial 100406 biofilm layer, inhibition of primer settlement is 6. A. Zaza, N. E. Laadel, E. G. Bennouna, Y. El crucial. Hu et al. (36) sprayed bacterial-anti- Hammami and M. T. Janan, Energy Proc., 2019, adhesive modified polystyrene microspheres to 157, 1230 construct bacterially-anti-adhesive surfaces. It 7. T. V. Wagner, J. R. Parsons, H. H. M. Rijnaarts, P. can be used on any surface thanks to the lotus de Voogt and A. A. M. Langenhoff, J. Haz. Mater., effect. It was reported as robust and durable on 2020, 384, 121314 surfaces. Similar surface engineering strategies 8. S. Holberg, R. Losada, F. H. Blaikie, H. H. W. B. focus on altering the physicochemical properties Hansen, S. Soreau and R. C. A. Onderwater, of the material surface. In general, reduced Mater. Today Comm., 2020, 22, 100750 efficacy of regular disinfectants leads to progress 9. I. Türetgen, Water SA, 2015, 41, (3), 295 in development of antimicrobial surfaces and 10. M. Lemouari, M. Boumaza and A. Kaabi, Energy, coatings (37, 38). 2011, 36, (10), 5815 11. M. Malaki, Y. Hashemzadeh and M. Karevan, Prog. Org. Coat., 2016, 101, 477 Conclusion 12. M. Simões, L. C. Simões and M. J. Vieira, Food Sci. This is the first report of a nano-silica coating on Technol., 2010, 43, (4), 573 a stainless steel cooling tower heat exchanger. 13. I. W. Sutherland, Microbiology, 2001, 147, (1), 3 The study showed that the nano-silica coating 14. C. Gómez-Suárez, J. Pasma, A. J. van der Borden, significantly reduced bacterial fouling on surfaces. J. Wingender, H.-C. Flemming, H. J. Busscher and There are many similar surfaces with biofouling H. C. van der Mei, Microbiology, 2002, 148, (4), problems which have contact with water and 1161 require a solution. Nano-silica has proven to be 15. N. M. Dat, L. D. Manh, D. Hamanaka, D. V. Hung, effective at reducing the formation of biofilms on F. Tanaka and T. Uchino, Food Control, 2014, 42, surfaces and can be applied as a cost-effective, 94 effortless, non-toxic, readily available material. 16. M. Pasmore, P. Todd, B. Pfeifer, M. Rhodes and Due to growing restraints on environmental release C. N. Bowman, Biofouling: J. Bioadhes. Biofilm of biocidal agents and the growing restrictions Res., 2002, 18, (1), 65 on the use of disinfectants in man-made water 17. K. Naik and M. Kowshik, Mater. Sci. Eng., 2014, systems, as well as demand to decrease the cost 34, 62 of system maintenance, different ways to limit 18. J. Azeredo and R. Oliveira, ‘Biofilm Characteristics: biofilms in man-made water systems hold much Biofilm Formation: The Role of Hydrophobicity expectation. and Exopolymers in Initial Adhesion and Biofilm Formation’, in “Biofilms in Medicine, Industry and Environmental Biotechnology”, eds. Acknowledgments P. Lens, A. P. Moran, T. Mahony, P. Stoodley and This study was supported by ‘Research Fund of the V. O’Flaherty, Pt. 1, Section 1, IWA Publishing, Istanbul University’. Project number: 29220. London, UK, 2003, pp 16–32 19. B. Arkles, ‘Hydrophobicity, Hydrophilicity and Silanes’, Paint and Coatings Industry Magazine, References 2006, 22, (10), 114 20. V. Antoci, C. S. Adams, J. Parvizi, H. M. Davidson 1. F. Di Pippo, L. Di Gregorio, R. Congestri, V. Tandoi R. J. Composto, T. A. Freeman, E, Wickstrom, and S. Rossetti, FEMS Microbiol Ecol., 2018, P. Ducheyne, D. Jungkind, I. M. Shapiro and 94, (5), fiy044 N. J. Hickok, Biomaterials, 2008, 29, (35), 4684

21. A. L. Casey, L. A. Mermel, P. Nightingale and 30. M. Chen, Q. Yu and H. Sun, Int. J. Mol. Sci., 2013, T. S. J. Elliott, Lancet: Infect. Dis., 2008, 8, (12), 14, (9), 18488 763 31. L. J. Zhong, L. Q. Pang, L. M. Che, X. E. Wu and 22. S. A. McConnell, P. O. Gubbins and E. J. Anaissie, X. D. Chen, Coll. Surf. B: Biointerfaces, 2013, Clin. Infect. Dis., 2003, 37, (1), 65 111, 252 23. J. C. Hockenhull, K. M. Dwan, G. W. Smith, 32. A. Pagedar, J. Singh and V. K. Batish, J. Basic C. L. Gamble, A. Boland, T. J. Walley and Microbiol., 2010, 50, (S1), S98 R. C. Dickson, Crit. Care Med., 2009, 37, (2), 702 33. A. Okada, T. Nikaido, M. Ikeda, K. Okada, 24. B. S. Niël-Weise, T. Stijnen and P. J. van den Broek, J. Yamauchi, R. M. Foxton, H. Sawada, J. Tagami Intens. Care Med., 2007, 33, (12), 2058 and K. Matin, Dental Mater. J., 2008, 27, (4), 565 25. A. K. Epstein, B. Pokroya, A. Seminara and 34. B. J. Privett, J. Youn, S. A. Hong, J. Lee, J. Han, J. Aizenberg, Proc. Natl. Acad. Sci., 2011, J. H. Shin and M. H. Schoenfisch,Langmuir , 2011, 108, (3), 995 27, (15), 9597 26. L. R. Hoffman, D. A. D’Argenio, M. J. MacCoss, 35. W. Ding, C. Ma, W. Zhang, H. Chiang, C. Tam, Y. Xu, Z. Zhang, R. A. Jones and S. I. Miller, Nature, G. Zhang and P.-Y. Qian, Biofouling: J. Bioadhes. 2005, 436, (7054), 1171 Biofilm Res., 2018, 34, (1), 111 27. P. Jena, S. Mohanty, R. Mallick, B. Jacob and 36. J. Hu, J. Lin, Y. Zhang, Z. Lin, Z. Qiao, Z. Liu, A. Sonawane, Int. J. Nanomed., 2012, 7, 1805 W. Yang, X. Liu, M. Dong and Z. Guo, J. Mater. 28. H. Du, T.-M. Lo, J. Sitompul and M. W. Chang, Chem. A, 2019, 7, (45), 26039 Biochem. Biophys. Res. Comm., 2012, 424, (4), 37. P. S. V. V. S. Narayana and P. S. V. V. Sirihari, 657 Regen. Eng. Transl. Med., 2019 29. L. D. Renner and D. B. Weibel, MRS Bull., 2011, 38. S. Achinas, N. Charalampogiannis and 36, (5), 347 G. W. Euverink, Appl. Sci., 2019, 9, (4), 2801

The Author

Irfan Turetgen is a Full Professor at the Department of Biology, Faculty of Science, Istanbul University, Turkey. He received his PhD degree in Environmental Microbiology from Istanbul University. His major area of research is disinfection of heterotrophic biofilms in man‑made water systems, Legionella ecology in cooling towers and anti-biofilm coatings. He has authored research papers published in reputed journals and conference proceedings. Recently he is supervising thesis projects as a mentor. His current interest is modelling man-made water systems to mimic their function at laboratory scale and test disinfectants under conditions as close as possible to real life.

www.technology.matthey.com

A Mini-Review of Shape-Memory Polymer-Based Materials Stimuli-responsive shape-memory polymers

Mathew J. Haskew, John G. Hardy* The reversible transformation of SMPs functions Department of Chemistry and Materials by primary crosslinking net points (hard segments) Science Institute, Faraday Building, Lancaster memorising and determining the permanent shape, University, Lancaster, LA1 4YB, UK and secondary switching segments (soft segments)

with a transition (Ttrans) to reduce strain stress and

*Email: [email protected] hold the temporary shape. Below the Ttrans, the material will be in its permanent shape and be

stiffer than when Ttrans is reached and the SMPs are Shape-memory polymers (SMPs) enable the more malleable and can be deformed into a desired production of stimuli-responsive polymer-based shape (usually through application of an external materials with the ability to undergo a large force). The deformed state is maintained once the recoverable deformation upon the application of an external force has been removed and the system external stimulus. Academic and industrial research is no longer at or above Ttrans. SMPs revert to their interest in the shape-memory effects (SMEs) of these original state once the Ttrans conditions are met. SMP-based materials is growing for task-specific This process describes the SME pathway of SMP- applications. This mini-review covers interesting based materials that are thermally-induced (albeit aspects of SMP-based materials, their properties, how not for some light or chemical-induced systems). they may be investigated and highlights examples of While most SMP-based materials hold a single the potential applications of these materials. permanent shape and a single temporary shape, recent advances in SMP technology have allowed Introduction the generation of multiple-shaped-memory materials with different stimuli responses (light or SMEs refers to the ability of the material to chemical) (16, 20, 21). An interesting example of memorise a shape and materials that possess these this is a triple shaped-memory material generated properties have a multitude of exciting technical by combining two dual SMPs with different glass and medical applications (1–14). For materials such transition temperatures (Tg) (22, 23), where the as alloys this is commonly in a one-way SME (7, SMPs switch from one temporary shape to another at

15), however, there are a variety of materials the first transT , and then back to the permanent shape that are capable of reverting to their permanent at another, higher activation temperature (22). shape or original state upon exposure to a stimulus SMPs have a large range of properties from stable (such as a temperature change) or indeed multiple to biodegradable and transient, elastic to rigid or stimuli (16). SMP-based materials have been soft to hard, depending on the structural units that widely investigated since the 1980s because of the constitute the SMP. Consequently, SMPs not only abundance of potential applications imparted by respond to temperature (24) and magnetism (25) their interesting properties (for instance, stimuli- like shape-memory alloys (SMAs) (26), but also responsiveness and ability to change shape), to moisture (27), electricity (28), light (29) and which can lead to technological innovation and the chemical stimuli (such as a pH change) (30). generation of new high value products for technical Moreover, there are other principles of SME; for and medical applications (1, 17–19). instance, a thermal-responsive SMP can proceed

425 © 2020 Johnson Matthey https://doi.org/10.1595/205651319X15754757916993 Johnson Matthey Technol. Rev., 2020, 64, (4) via a Diels-Alder reaction (chemical crosslinking/ hard segments form the net points that link the reversible covalent bonds) (31). SMPs tend to have soft segments (acting as a fixed phase), whereas much milder processing conditions than SMAs the soft segments work as the molecular switches (<200°C, low pressure), have a greater extent (acting as a reversible phase). The fixed phase of deformation (strain more than 200% for most prevents free flow of the surrounding polymer materials) and tend to be based on cheap starting chains upon the application of stress. The reversible materials with simple synthetic procedures (12, phase, on the other hand, undergoes deformation 32). After the term ‘shape-memory’ was first in a shape-memory cycle and is responsible for proposed by Vernon in 1941 (32), the significance elasticity. For example, if the Ttrans is Tg, the micro- of SMPs was not fully realised until the 1960s, Brownian motion of the network chains is fixed at when crosslinked polyethylene (PE) was used low temperature (below Tg) and will be switched to make heat-shrinkable tubes and films (33). back on at high temperature (above Tg), recovering

Significant investment in the development of its original state. When Ttrans is the crystal melting

SMPs began in the 1980s (34) with rapid progress temperature (Tm), the switching segments realised in the last decade, particularly with a view crystallise at low temperature (below Tm), and then to the generation of shape-memory materials with recover their original state at high temperature exciting and versatile features. (above Tm). In addition, Tg normally extends over

a broader temperature range compared to Tm, Shape-Memory Polymer Function which tends to have relatively sharper transitions in most cases (26). Moreover, after the exposure

Two important quantities used to describe SMEs to a specific stimulus and the transT is achieved, the are the strain recovery rate (Rr) and the strain strain energy in the deformed state is released, fixity rate (Rf). Rr describes the ability of a material resulting in the shape recovery phenomenon. The to memorise its permanent shape, while Rf general process of this SME for SMPs is depicted in describes the ability of switching segments to fix Figure 1, wherein the polymer network structure is the mechanical deformation. Rr is calculated using either chemically or physically crosslinked and the Equation (i): switching units are made from a semi-crystalline or amorphous phase. εm(N) – εp(N) Rr(N) = × 100% (i) Shape-memory behaviour can be demonstrated ε (N) – ε (N–1) m p in various polymer systems that are significantly different in molecular structure and morphology. where N is the cycle number, εm is the maximum SME mechanisms differ according to the specific strain imposed on the material and εp is the strain SMP(s); for instance, the SME mechanism of the of the sample after recovery. Rf is calculated using chemically crosslinked semi-crystalline PE SMP.

Equation (ii): The crystalline phase, with a Ttrans being Tm, is used as the molecular switching unit providing εu(N) Rf(N) = × 100% (ii) shape fixity. The chemically crosslinked PE network ε (N) m memorises the permanent shape after deformation upon heating (12, 36, 37), and the mechanism of where εu is the strain in the fixed temporary shape. the thermally-induced shape-memory PE (SMPE) is SMPs respond to specific stimuli through changes depicted in Figure 2. in their macroscopic properties (for example, The associated modulus of elasticity is dictated shape) (26). The polymer network underlying by configurational entropy reduction that occurs active movement involves a dual system, one that with deformation of the constituent chains and is highly elastic and another that can reduce the is therefore often termed entropy elasticity. For stiffness upon application of a certain stimulus. T>Ttrans (Tg, Tm or other), polymer networks The latter system incorporates either molecular exhibit super-elasticity wherein the polymer chain switches or stimulus sensitive domains (35). segments between crosslink points can deform Their shape-memory feature is a result of the quite freely and are prone to being twisted combination of the polymer’s architecture, and a randomly via rotations about backbone bonds, programming procedure that enables the formation maintaining a maximum entropy and minimum of a temporary shape. Net points consist of covalent internal energy as macroscopic deformation bonds or intermolecular interactions and the SMP’s occurs (12). The classic prediction from rubber

(a)

+ ∆conditions SMP SMP SMP

Permanent Deformation stage, may Deformed shape require an external force temporary shape TTtrans tensile stress T>Ttrans)

– ∆conditions

+ ∆conditions SMP SMP

Permanent Deformed temporary shape, shape shape, if force applied recovery then the external force T>Ttrans is removed T

(b) External applied force, Strain energy Strain energy released, bending of polymer stored polymers shape recovery

Heating, a hot Cooling, taken off Heating, a hot plate hot plate plate

Original state T>Tg/Tm Deformed state T>Tg/Tm Recovered shape, T

Fig. 1. (a) The general SME mechanism of SMPs; (b) thermally-responsive SMP

elastic theory is that the resulting elastic shear low temperature, and the ability to trigger shape modulus (G) is proportional to both crosslink recovery at high temperature; super-elasticity density and temperature (Equation (iii)): above Ttrans that leads to the eventual shape pRT recovery and avoids residual strain (permanent G = vKBT = (iii) deformation); and complete and rapid fixing of the MC temporary shape by immobilising the polymeric chains without creep thereafter (12, 37). Thus where ν is the number density of network chains, p far, the SME models describing how SMPs recover the mass density, R the universal gas constant and their original state prominently involve thermo-

MC the molecular weight between crosslinks. From responsive SMPs. However, careful design of the a macroscopic viewpoint, the SME in SMPs can be polymers allows the opportunity for SMPs to possess graphically represented in three-dimensions (3D). different stimuli responses and applications. Tensile strain vs. temperature and tensile stress (for example, elongation) is depicted in Figure 3. Shape-Memory Polymer Triggers Using the shape-memory strain-temperature- stress relationship description in Figure 3, the A multitude of different triggers for SMEs and SMPs features of SMPs that allow for good shape- exist. However, an in-depth review is outside the memory behaviour include: a sharp transition that scope of this mini-review, and therefore a few can be used to quickly fix the temporary shape at examples are highlighted below.

Force of T>Tm stretching

PE cross-linked T>Tm Heating at or by radiation above Tm

Cooled at stretched shape

Fig. 2. Molecular model of the thermally-induced SME mechanism of crosslinked SMPE

SMPs, with a schematic of the SME mechanism for

Shape fixity thermally-induced SMPs with Tg (amorphous cases) ε Strain removal m and Tm (crystalline cases). Figure 2 presents a 200 specific example of the SME mechanism for SMPE ε u with the Ttrans being Tm. In addition, advanced 150 thermomechanical constitutive models have been used to study the materials’ behaviour (for 100 example strain-temperature-stress development

Recovery with time) in a very accurate way (41). By applying

% ), ε ( in ra St Deformation 3.0 these models to SME mechanistic studies and the 50 2.5 detailed characterisation of the SMPs (crosslinks, 2.0 a 0 intermolecular and intramolecular interactions 20 1.5 40 ε 1.0 (σ), MP involving the SMPs) (12), a deeper understanding Temper 60 p 80 * 0.5 of the SME of SMPs can be achieved, which has ature (T), 100˚C 0 Stress 120 proven beneficial for the development of new SMPs and their proposed applications (31). For Fig. 3. A general 3D plot of an SMP during a example, poly(ε-caprolactone) (PCL), typically thermomechanical shape-memory cycle a biodegradable polymer, has been reported to possess high shape fixity and recovery. This was achieved by integrating reversible bonds within the Thermally-Induced Shape-Memory PCL polymer network via the Diels-Alder addition of Polymer 1,2,4-triazoline-3,5-dione (TAD)-anthracene and Alder-ene addition of TAD-indole (42). These PCL It is possible to generate thermally-induced SMEs SMPs were reported to attain recovery ratios greater in a variety of materials (18–20, 38–40), however than 99% (43). Furthermore, a dual-functional a comprehensive overview is outside the scope (self-healing and shape-memory) polymer network of this mini-review. As previously discussed, the was achieved by crosslinking a polydimethylsiloxane SME of SMPs can be thermally-induced, and these (PDMS) polymer containing dense carboxylate SMPs are the most common (26). Figure 1 depicts groups (100% mol) (PDMS-COOH) with small a general overview of the SME mechanism of amount of poly(ethylene glycol) diglycidyl ether

(PEGDGE) (44). This SMP (PDMS‑COO-E) actuates Light-Induced Shape-Memory at body temperature (37°C) with possible strain Polymer ca. 200% and shape recovery ratios at 98.06%. In addition, a 25 mm × 4 mm × 1 mm sample It is possible to generate light-induced SMEs in a cut into two separate pieces healed (the variety of materials (18–20, 38, 40, 45), however a two pieces become one whole piece with no comprehensive overview is outside the scope of this evidence of a cut) when the two cut surfaces mini-review. Light-activated SMPs (LASMPs) (46) were brought into contact after 6 h at 25°C. typically use photothermal or photochemical Thus, the unique material, PDMS‑COO-E, (photocrosslinking or photocleavage) triggers for may have a wide range of applications in SMEs. For instance, photothermal LASMPs typically many fields, including wearable electronics, employ photo-absorber molecules and particles biomedical devices and four‑dimensional (4D) that convert light to heat, thereby increasing printing (1, 19). Interestingly, the material was the temperature at the desired region within also reported to possess a greater than 85% the LASMP. Photochemical LASMPs incorporate light transmittance (425 nm to 700 nm) (44), photosensitive molecules to create or cleave bonds therefore PDMS-COO-E has potential applications during irradiation with light, imparting potentially in transparent electronic devices. Figure 4 very swift SMEs (47, 48). It is possible to improve illustrates the possible SME mechanism of the response time of SMPs by increasing the PDMS-COO-E. The short PDMS linear chains are thermal conductivity with various conductive crosslinked by chemical covalent interactions and additives (49). However, the heating and cooling abundant hydrogen bonds into a 3D network. The of materials with substantial thickness takes time, covalent crosslinked networks of PDMS-COO-E which can be minimised by using light to trigger maintain the permanent shape and resilience, transitions in LASMPs (46). It is also possible to whereas, at ca. 37°C the weak hydrogen bonds generate multistimuli-responsive materials using are broken, and the dynamics of polymer chains components of the materials that respond to increase, resulting in recovering the permanent different wavelengths of light (for example, one shape. Meanwhile, a large number of hydrogen wavelength of light to induce photocrosslinking, bonds enable the samples to heal at temperature while a second wavelength of light cleaves bonds). without external stimulus (44). It is possible to produce materials that can be reversibly switched between an elastomer and a rigid polymer employing polymers containing cinnamic groups (48) that can be fixed into pre‑determined shapes utilising ultraviolet (UV) light illumination (>260 nm), and then recovered their original state when exposed to UV light at a different wavelength (<260 nm) (49). Figure 5 depicts one example of the process of LASMPs shape recoverability.

Cooling Heating Electrically-Induced Shape-Memory reshape recovery Polymer

It is possible to generate electrically-induced SMEs in a variety of materials (18, 20, 50–55), however a comprehensive overview is outside the scope of this mini-review. A variety of electrically conductive

Stretched materials including organic electronic materials (including conductive polymers such as polypyrrole = PDMS = PEGGE

O H O (PPy) (28, 56–58) and carbon nanotubes = –COOH = O H O (CNTs) (59, 60)) and inorganic electronic materials (such as alloys, metals (61) and silver nanowires Fig. 4. The possible mechanism about shape (NWs)), have been incorporated in materials memory effect of PDMS-COO-E polymer. Reprinted displaying SMEs to impart swift triggers to the with permission from MDPI (44) SMEs, enabling a variety of interesting applications.

A hv2 Shape External force applied recovery (e.g. elongation under tensile stress) Permanent shape, D elastomer B

hv1

Deformed temporary shape with external = elastomer network cross-link force removed, = photo-cross-link site rigid polymer = polymer chain backbone

Fig. 5. Schematic of an example of the SME function of LASMPs

Highlighting some of the potential of electrically- of SMP designs driving technological innovation. A induced SMEs, electrically-induced SMP composites schematic of the composite is shown in Figure 6. incorporating shape-memory polyurethane Polymeric blend SMPs can be constructed from (SMPU) and Ag NWs in a bilayer structure exhibits two immiscible polymeric matrices. The shape- flexibility and electrical conductivity (62–64), recovery of these systems can be controlled with which may find applications as capacitive sensors, relative ease by varying the ratio of the polymer healable transparent conductors and wearable blends (68). However, this process may have electronics (65). In such materials the Ag NWs adverse effects on shape-memory characteristics are randomly distributed on the surface layer of and diminish the material’s performance, thereby the composite to form a conductive percolating limiting potential applications. On the other hand, network that retains conductivity (200 Ω sq–1) after SMP functionality may also be enhanced with other a 12% elongation. However, continual increase capabilities. For instance, it was recently reported in elongation causes a dramatic increase to the that a new hybrid SMP was developed by combining composites’ resistance value and the eventual loss single-walled CNTs (SWCNT) into a poly(lactic acid) of electrical conductivity (66). When the material (PLA) and thermoplastic polyurethane (TPU) SMP (deformed or in its original state), is connected to system, containing poly(ethylene glycol) (PEG) a typical circuit, a low voltage of 1.5 V was enough plasticiser (68). By incorporating PEG, the hybrid to activate a light-emitting diode (LED) (65). The SMP composite achieved a lower temperature Tg (for composites possessing a higher Ag NW content example, 10 wt% of PEG lowered Tg of the PLA/TPU exhibited a higher recovery ratio and reached the sample from 60°C to 40°C), meanwhile enhancing maximum recovery speed quicker (66). It was the dispersion of SWCNT (for instance, even at assumed that all the heat from electrical (Joule) 4 wt% of SWCNT loading, 100% SMP tensile strain heating was absorbed by the sample, i.e. no was possible, much greater than previously reported convective loss (67). Therefore, the composites with electrically-induced SMP studies, i.e. 12% discussed higher Ag NW content had a lower resistance value previously). In addition, the presence of the SWCNT and the heating effectiveness was promoted. Heat can stabilise the SMP system and enhance its shape- initiates the thermal Ttrans of the SMPU leading to an fixity after deformations at room temperature improved shape recovery, and voltages as low as conditions (68). Furthermore, the material was 5 V reverted bent composites to their original state capable of a conductivity above 10–7 S cm–1, which within 3 s (66). This represents a good example of a can be considered conductive, as documented (68). multifunctional SMP and demonstrates the potential The PLA/TPU SMP composite (2 wt% SWCNT and

(a) (b) 20 um 412.7 nm 500 nm 15 6.7 10 20 5 15

10 20 um 15 5 10 5

Power Spin coating

AgNWs LED Composite electrode

Release substrate Release substrate LED

Coating SMP Conductive layer composite

SMP layer Peel off AgNW layer AgNW layer SMP layer Release substrate

Fig. 6. (a) transmission electron microscopy (TEM) image of Ag NWs; (b) atomic force microscopy (AFM) image of Ag NWs; (c) schematic illustration of composites fabrication process; (d) the LED turned on as the composite was applied with voltages (the inset shows the circuit connecting with the composites). Reprinted with permission. Copyright 2014 Elsevier (66)

10 wt% PEG) also achieved shape-recovery, via however a comprehensive overview is outside the Joule heating derived from electricity, in 80 s when scope of this mini-review. Water is an important currents of 125 mA were applied. The high stiffness stimulus due to the fact it is abundant in a multitude of SWCNT filler results in decreasing shape-recovery of different environments, non-toxic and safe for a performance because of the hindrance on the variety of applications. polymer chain movements (68). As a result, under An interesting example highlighting the room temperature stretching, the Rf and Rr values potential of such materials is based on strong obtained were ca. 80% and 65%, respectively. and flexible composite films (73) utilising the Therefore, when its shape-recoverability is compared combination of a flexible interpenetrating polyol- to other SMPs (shape-recovery ratios being upwards borate network (74) and electroactive PPy (75, of 98%), the material is lacking. However, the hybrid 76) that exchange water with the environment SMP composite does possess electroactive ability, resulting in film expansion or contraction. The thus a trade-off relationship between shape-memory/ free-standing multi-functional SMP films were recovery and electroactive ability needs to be carefully prepared by electropolymerisation of pyrrole considered when designing similar materials. in the presence of the polyol-borate complex (composed of pentaerythritol ethoxylate (PEE) coordinated to boron(III)) (74), wherein the Water-Induced Shape-Memory interpenetrating network enables water-gradient- Polymer induced displacement, converting chemical It is possible to generate water-induced SMEs in potential energy in water gradients to mechanical a variety of materials (18, 20, 38, 39, 69–72), work (73), and results in adaptation of the

431 © 2020 Johnson Matthey https://doi.org/10.1595/205651319X15754757916993 Johnson Matthey Technol. Rev., 2020, 64, (4) architecture in response to an environmental pH-Induced Shape-Memory Polymer condition change (i.e. sorption and desorption of water which drives the SME process, as depicted It is possible to generate pH-induced SMEs in a in Figure 7). The design of the water-responsive variety of materials (18, 20, 38, 77–80), however PPy-PEE composites was creatively applied to a comprehensive overview is outside the scope of prepare actuators and generators driven by this mini-review. An example of the interesting water gradients. The film actuator can generate properties of such pH-responsive SMPs and their contractile stress up to 27 MPa, lift objects composites is produced by blending poly(ethylene 380 times heavier than itself and transport cargo glycol)-poly(ε-caprolactone)-based polyurethane 10 times heavier than itself (73). An assembled (PECU) with functionalised cellulose nanocrystals generator associating the actuator with a (CNCs) displaying pH responsive pyridine moieties piezoelectric element driven by water gradients, (CNC-C6H4NO2) (81, 82). At high pH values the outputs alternating electricity at ca. 0.3 Hz, with pyridine is deprotonated, facilitating hydrogen a peak voltage of ca. 1 V (73). The electrical bonding interactions between the pyridine energy can be stored in capacitors that could groups and hydroxyl moieties on the cellulose, power micro and nanoelectronic devices (73). whereas at low pH values, the protonation of the The SME mechanism for this SMP differs to that pyridine moieties diminishes these interactions. of Figure 1 and Figure 2, utilising water as the By comparison, carboxylic acid functionalised shape-memory trigger for Ttrans, and the original cellulose nanocrystals (CNC-CO2H) responded to and deformed state interchange automatically via pH variation in the opposite manner (83–85). When water sorption and desorption states. However, the functionalised CNCs were combined with PECU the shape-memory phenomenon remains the polymer matrix to form a nanocomposite network, same, further demonstrating the potential of SMP the mechanical properties of PECU were improved designs driving technological innovation. along with the pH-responsiveness of CNCs (85).

(a) (b) (c) Metal electrode Insulating

layer Vcap

Voutput RL G PVDF

Wire Vrec G VG PEE-PPy

(d) (e) 1.5 0.7

1.0 0.6 0.5 V V 0.5 0.4 0.0 ltage, ltage, 0.3 0.55 V Vo Vo –0.5 0.2 ltage,

–1.0 Vo 0.1 0.50 290 310 Time, s –1.5 0 20 40 60 80 100 120 0 100 200 300 400 Time, s Time, s

Fig. 7. Design and performance of a water-gradient–driven generator: (a) the assembly of a piezoelectric polyvinylidene fluoride (PVDF) element with a PEE-PPy actuator to form the generator; (b) the connection of the generator with a 10 MW resistor as load; (c) the configuration of the rectifying circuit and charge storage capacitor; (d) the generator’s output voltage onto the 10 MW resistor; (e) voltage across a capacitor when being charged by the generator. The inset shows a stepwise increase in the capacitor voltage accompanying each cycle of the energy conversion process. Reprinted with permission. Copyright 2013 The American Association for the Advancement of Science (73)

The percolated network of pH-sensitive CNC in the triggered by transition along the digestive polymer matrix served as the switching units for tract) (83). the shape-memory composite, the SME process of this material is depicted in Figure 8 (81, 82). Magnetically-Induced Shape-Memory The CNC serves as the switching unit of the SMP Polymer composite within the matrix of PECU which is physically crosslinked and microphase separated It is possible to generate magnetically-induced to yield the net points. Such pH-responsive shape- SMEs in a variety of materials (18, 20, 38, 86– 88), memory nanocomposites have promise in the however a comprehensive overview is outside design of biomaterials for biomedical applications the scope of this mini-review. The SMP devices (for example, SMP-based drug delivery systems discussed thus far are being researched with potential application into wearable electronics, nanoelectronics (such as actuators), biomaterials and biomedical devices (1, 18, 19). However, in some instances (such as medical devices) a key challenge is the design and implementation of a safe and effective method of actuating a variety of device geometries in vivo. As previously discussed, a pH‑triggered SMP design can be potentially effective when utilised as drug delivery devices, when the target environment has a substantial pH difference (for instance, the digestive system) (83). Force pH = 8 However, the development of electrically and Deformation pH = 4 thermally-triggered devices that safely operate in vivo is difficult due to the (generally) high temperatures these SMPs can reach (relative to

Fixity biological systems). For instance, the electroactive PLA-TPU SMP composite (2 wt% SWCNT and

pH = 4 pH = 8 10 wt% PEG) reaches temperatures greater than 70°C in 80 s as shape-recovery is achieved (68). Recovery An alternative method of achieving actuation is inductive heating by loading ferromagnetic particles into an SMP system and exposing the doped device to an alternating electromagnetic field (89), benefiting from the innate thermoregulation offered by a ferromagnetic material’s Curie temperature

(Tc, at which a ferromagnetic material becomes paramagnetic, losing its ability to generate heat via a hysteresis loss mechanism) (90). By using particle sizes and materials that will heat mainly via a magnetic hysteresis loss mechanism over an eddy current mechanism, it is possible to have PECU matrix in solution Hydrogen bonding an innate thermoregulation mechanism that limits O + the maximum achievable temperature to Tc (89). CNC–C6H4NO2 CNC + C NH Therefore, by selecting ferromagnetic particle materials with a T within safe medical limits, Fig. 8. Schematic representation of the pH- c responsive shape-memory materials, which rely Curie thermoregulation eliminates the danger of on hydrogen bonding switching mechanism in the overheating and the need for a feedback system interactions between cellulose nanocrystals (CNC– to monitor implanted device temperatures (89). C6H4NO2) within polymer matrix upon immersion However, this technology is not only useful when in hydrochloric acid solution (pH = 4) or sodium applied to medical devices. Other useful applications hydroxide solution (pH = 8). Reprinted with include remote activation in which wires or permission. Copyright 2015 American Chemical Society (81) connections to SMP devices could be eliminated, simplifying the design and reducing possible points

433 © 2020 Johnson Matthey https://doi.org/10.1595/205651319X15754757916993 Johnson Matthey Technol. Rev., 2020, 64, (4) of failure. An example of this method of actuation composites did not exceed temperatures above involves the incorporation of 10% by volume the respective Ni Zn particle Tc values, signifying nickel zinc ferrites (for example C2050 (Ceramic a thermoregulation characteristic. In addition, it Magnetics Inc, USA) and CMD5005 (Ceramic was stated that the 10% volume of Ni Zn particles Magnetics Inc), particle sizes ca. 50 µm with did not impact the SMPs shape-memory properties spherical shapes) with an ester-based thermoset significantly (89). The Tg increased from 55°C to polyurethane (PU) SMP, MP5510 (SMP Technologies 61.4°C and the shape-recovery of a flower and

Inc, Japan) (Tg of 55°C) (89). The magnetic field foam-based device was achieved within 15 s to utilised to achieve shape-recovery was a copper- 25 s, at a temperature range of 23°C to 78.6°C. wound solenoid coil with a 2.54 cm diameter, The potential applications for this device are 7.62 cm length and with a total of 7.5 turns. The illustrated in Figure 9. Optimisation of this device/ unit possessed an adjustable power setting capable design is still required before it can be considered of outputting 27 W to 1500 W at between 10 MHz clinically viable, however, this SMP composite and 15 MHz frequency (note: this high frequency highlights very interesting characteristics, remote may induce eddy currents in the tissue, causing activation (via magnetic fields inducing thermally- undesirable direct heating of the human body in triggered actuation) and thermoregulation (via Tc medical applications) (91). However, an alternating temperature of the material being employed). magnetic field of 12.2 MHz and approx. 400 A –1m (centre of the inductive coil) at room temperature Shape-Memory Polymer was used for actuation to demonstrate proof of Classification concept for the device. It was also reported that clinically useable frequencies (50 kHz to 100 kHz) As highlighted above, SMP materials are diverse (92) should still be effective (89), albeit this and respond to many different external stimuli could result at a different quantitative level (i.e. (including temperature, light, electricity, water, shape-recovery and memory performance may pH and electromagnetic fields) by a variety of be reduced). Furthermore, C2050 and CMD5005 mechanisms. Although SMPs can be classified possess a Tc of 340°C and 130°C, respectively. based on their composition and structure, stimulus These temperatures exceed physiological limits and shape-memory function, their classification and are therefore not practical for medical can be difficult, as organising these polymeric devices currently, however, these doped SMP smart materials into one or two simple categories

(a) Fig. 9. SMP devices used to evaluate feasibility of actuation by inductive heating: (a) flower shaped device shown in collapsed and actuated form; (b) SMP foam device shown in collapsed and actuated form. Reproduced with permission. 1.5 mm Copyright 2006 IEEE Transactions on Biomedical Engineering (89)

(b) 5 mm

434 © 2020 Johnson Matthey https://doi.org/10.1595/205651319X15754757916993 Johnson Matthey Technol. Rev., 2020, 64, (4) is an over-simplification of their abilities and polyesterurethanes (PEU), oligourethane segments characteristics (93). are the hard-elastic segments, while polyester SMPs are considered to consist of net points and serves as the switching segment (99). molecular switches or stimuli sensitive domains. These net points can be achieved by covalent Thermoset Shape-Memory Polymers bonds (chemically crosslinked) or intermolecular interactions (physically crosslinked). Chemically For chemically crosslinked SMPs, two methods crosslinked SMPs involve suitable crosslinking are commonly used to synthesise covalently chemistry and are referred to as thermosets (94, crosslinked networks (36, 41). The first method 95). Physically crosslinked SMPs involve a polymer relies on addition of a multi-functional crosslinker morphology consisting of at least two segregated during polymerisation (41), whereas the second domains and are referred to as thermoplastics (96). method relies on the subsequent crosslinking of The network chains of the SMP can be either a linear or branched polymer (36). The networks amorphous or crystalline and therefore, the Ttrans are formed based on many different polymer is either a Tg or Tm. The network architectures are backbones. Covalently crosslinked SMPs possess thought to be constructed through crosslinking net chemically interconnected structures determining points, with polymer segments connecting adjacent the original macroscopic shape. The switching net points. The strongly crosslinked architectures segments of these materials are generally the ensure the polymer can maintain a stable shape network chains between net points, and a Ttrans on the macroscopic level (93). Thermoplastic of the polymer segments is used as the shape- polymers exhibit a more reversible nature (97), memory switch. The chemical, thermal, mechanical meaning the physical crosslinked net points can and shape-memory properties are determined be disrupted and reformed with relative ease. The by the reaction conditions, curing times, the interconnection of the individual polymer chains type and length of the network chains and the in a physically crosslinked network is achieved by crosslinking density (35). Comparing physically the formation of crystalline or glassy phases. For crosslinked SMPs with chemically crosslinked thermoset polymers, the individual polymer chains SMPs, the chemically crosslinked SMPs often show are connected by covalent bonds and are therefore less creep, thus, any irreversible deformation of more stable than physically crosslinking networks the polymer during shape recovery is less. This is and show an irreversible nature (98–100). because covalent crosslinked networks are more Regarding thermo-responsive SMPs, they can stable than physical crosslinked networks. As a be classified according to the nature of their result, chemically crosslinked SMPs usually show permanent net points and the Ttrans related to the better chemical, thermal, mechanical and shape- switching domains into four different categories: (a) memory properties than physically crosslinked physically crosslinked thermoplastics, Ttrans = Tg; SMPs (96). For example, the shape recovery ratio (b) physically crosslinked thermoplastics, of thermoplastic SMPU is usually in the range of

Ttrans = Tm; (c) chemically crosslinked amorphous 90% to 95% within 20 shape recovery cycles, polymers, Ttrans = Tg; (d) chemically crosslinked and the elastic modulus is between 0.5 GPa and semi-crystalline polymer networks Ttrans = Tm (93). 2.5 GPa at room temperature (26). Additionally, when exposed to air, it is sensitive to moisture Thermoplastic Shape-Memory and therefore possesses unstable mechanical properties. In contrast, an epoxy SMP shows Polymers better overall performance as a shape-memory For the physically crosslinked SMPs, the formation material. The shape recovery ratio typically of a phase-segregated morphology is the reaches 98–100%, the elastic modulus between fundamental mechanism behind the thermally- 2 GPa and 4.5 GPa, and it is generally stable in induced SME of these materials (93, 99). One the presence of moisture (26). Thermoplastic SMPs phase provides the physical crosslinks while the (such as SMPU) are mostly researched and used other acts as a molecular switch. They can be as functional materials at a small scale, such as further classified into linear polymers, branched for biomaterials (30, 97). However, thermosetting polymers or a polymer complex. Linear SMPs may SMPs (for example styrene-based SMP (SSMP) consist of block copolymers and high molecular and epoxy SMPs) are generally used for structural weight polymers, the typical physically crosslinked materials, such as space deployable structures and SMP is linear block copolymers, such as PU. In automobile actuators (97, 98).

Shape-Memory Functionality Figure 10, and an integrated insight into the classification of SMPs is shown inFigure 11. The approaches to designing different shape- An example of a selective triple shape memory functions become more abundant as multicomposite SMP was documented to incorporate a scientists and engineers better understand the SME neat SSMP (112) and two SSMP composites (113). mechanism of SMPs. For instance, discussed thus One incorporated iron(II, III) oxide nanoparticles far are examples of SMPs with polymeric blends, while the other CNT nanoparticles. This unique addition of crosslinking species, incorporation of SMP composite successfully possessed three electroactive and ferromagnetic substances. All different regions within the sample: neat SSMP, of which enhances an SMP device functionality, SSMP-Fe3O4 and SSMP-CNT. Because of this, the enabling unique and interesting characteristics material also possessed distinct shape-memory which can be tailored to a plethora of applications capabilities with different triggers. For instance, (for example, self-healing and wearable the material was documented undergoing a three- electronics, drug delivery and implantable medical step shape-memory recovery process, subjected devices) (101–110). Further still, one-way SMEs, to an alternating magnetic field of 30 kHz, a radio two-way SMEs (such as dual shape PPy-PEE, frequency (RF) field of 13.56 MHz and direct oven discussed previously), triple SMEs, multiple SMEs heating at 130°C (113). Furthermore, the Rf and Rr and even temperature-memory effects (TMEs) for the original shape to the first temporary shape have been widely investigated in SMPs (34). As the (and back to the original shape) was reported at types of SMP materials increasingly diversify, two 93% and 93%, respectively. Meanwhile, the Rf and even three different types of shape-memory and Rr for the first temporary shape to the second functions can be achieved simultaneously in the temporary shape (and back to the first temporary same SMP material (34, 111). These types of shape) was at 95% and 99%, respectively (113). materials can usually be achieved when combining The SME mechanism for this multicomposite is different SMPs possessing different properties. represented in Figure 12 and it was concluded that A schematic of one-way, two-way, dual shape this unique material has promising characteristics and triple shape functionality SMPs is shown in to be used in biomimetic materials. Examples of

One-way SME Two-way SME

>Ttrans >Ttrans

Dual shape >Ttrans

Deformed Deformed shape shape maintained

1 1 2 >Ttrans >Ttrans2 >Ttrans >Ttrans

1 1 2 Triple shape >Ttrans >Ttrans2

Fig. 10. The varying shape-memory functionality of SMPs

Composition and structure Stimulus Shape-memory function type

Block copolymer TEMPERATURE ONE-WAY SME

ELECTRICITY Thermo- Supramolecular polymer sensitive

TWO-WAY SME

MAGNETIC

Polymer blend/composites

Water- TRIPLE SHAPE SME WATER SENSITIVE sensitive e.g. PPy-PEE discussed in Figure 7

Cross-linked homopolymer

Redox- OXIDATION- MULTI-SHAPE SME REDUCTION sensitive

Polymer IPN/semi-IPN

Light- LIGHT/RADIATION sensitive MULTI-FUNCTIONALITY

Fig. 11. The classification of SMPs based on composition and structure, stimulus triggers and the possible type of shape-memory functions

applications of SMP-based materials and their SMPs by imparting new functional characteristics, composites are highlighted in Table I. broadening the potential applications of these materials and enabling a multipurpose material. Conclusion SMPs and their composites are capable of industrially important applications (examples of As the understanding of SMPs continually develops which include: self-healing (101–104), generators among the academic and industrial communities, driven by water gradients (73), sensors (72), task- the generation of new and potentially innovative specific medical devices (18, 105) and wearable SMPs will be more rapid while we realise the full electronics (106–110), a few examples of which potential of these materials. SMPs are one of the are highlighted in Table I. The literature published most interesting of polymer classes within the field to date de-risks investment from governments and of functional polymers. In addition, SMP composites industry to raise the technology readiness levels can enhance the already impressive capabilities of towards products on the market.

a b 1 b1 c1 a2 b2 c2 Deformed a1 1 c1 a2 b2 c2 at 130 C

Fixed at 20 C Temporary shape A Original shape

Alternating magnetic field heating of 30 kHz a 1 c 1 a2 a1 c2 a2 b RF field heating 1 b 2 b1 c 1 b2 c 13.56 MHz 2 Temporary shape C Temporary shape B

Oven heating 130 C a b SSMP-Fe O Neat SSMP a1 b c 3 4 1 1 a2 b2 c2

c SSMP-CNT Recovered shape an, bn, cn

Fig. 12. Schematic of the selective shape recoveries of the multicomposite SSMP induced by alternating magnetic field heating, RF field heating and oven heating, respectively n(a , bn and cn stand for the n sections of SSMP–Fe3O4, neat SSMP and SSMP–CNT, respectively). Reproduced by permission of The Royal Society of Chemistry. Copyright 2015 The Royal Society of Chemistry (113)

Table I Examples of Applications of SMP-Based Materials and Their Composites Application References Actuators (for example, for generators) (73) Biomedical devices (such as drug delivery systems, expanding foam and endovascular (44, 83, 89) thrombectomy device) Multipurpose/multifunctionality (for example, self-healing, biocompatible, body (44, 113) temperature actuation and selective triple shape-memory) Thermoregulators (89, 90) Wearable electronics (65, 68)

Acknowledgements research aligned with some of the topics discussed in this review, specifically the Glycoscience Tools We acknowledge the Faculty of Science and for Biotechnology and Bioenergy (IBCarb) Network Technology, Lancaster University, UK, for an in Industrial Biotechnology and Bioenergy (NIBB, Early Career Internal Grant and The Royal BB/L013762/1); the BBSRC FoodWasteNet NIBB Society, UK, for a Research Grant (RG160449). (BB/L0137971/1), the BBSRC From Plants to We acknowledge the Engineering and Physical Products (P2P) NIBB (BB/L013819/1) and the Sciences Research Council (EPSRC), UK, for Lignocellulosic Biorefinery Network (LBNet) NIBB an EPSRC First Grant (EP/R003823/1) and (BB/L013738/1). a Pathfinder Grant from the EPSRC Centre for Innovative Manufacturing in Large Area Electronics (EP/K03099X/1). We also thank the References Biotechnology and Biological Sciences Research 1. A. Lendlein and O. E. C. Gould, Nat. Rev. Mater., Council (BBSRC), UK, for financial support of 2019, 4, (2), 116

2. T. Biggs, M. B. Cortie, M. J. Witcomb and 22. I. Bellin, S. Kelch, R. Langer and A. Lendlein, L. A. Cornish, Platinum Metals Rev., 2003, Proc. Natl. Acad. Sci., 2006, 103, (48), 47, (4), 142 18043 3. D. Kapoor, Johnson Matthey Technol. Rev., 23. F. Pilate, A. Toncheva, P. Dubois and J.-M. Raquez, 2017, 61, (1), 66 Eur. Polym. J., 2016, 80, 268 4. Y. V. Kudriavtsev and E. L. Semenova, Platinum 24. F. Ji, Y. Zhu, J. Hu, Y. Liu, L.-Y. Yeung and G. Ye, Metals Rev., 2014, 58, (1), 20 Smart Mater. Struct., 2006, 15, (6), 1547 5. R. Oshima, S. Muto and T. Hamada, Platinum 25. R. Mohr, K. Kratz, T. Weigel, M. Lucka-Gabor, Metals Rev., 1988, 32, (3), 110 M. Moneke and A. Lendlein, Proc. Natl. Acad. 6. J. M. Jani, M. Leary, A. Subic and M. A. Gibson, Sci., 2006, 103, (10), 3540 Mater. Des., 2014, 56, 1078 26. J. Leng, X. Lan, Y. Liu and S. Du, Prog. Mater. 7. C. Naresh, P. S. C. Bose and C. S. P. Rao, Sci., 2011, 56, (7), 1077 ‘Shape Memory Alloys: A State of Art Review’, 27. B. Yang, W. M. Huang, C. Li and L. Li, Polymer, International Conference on Advances in 2006, 47, (4), 1348 Materials and Manufacturing Applications 28. N. G. Sahoo, Y. C. Jung and J. W. Cho, Mater. (IConAMMA-2016), Bangalore, India, 14th–16th Manuf. Processes, 2007, 22, (4), 419 July, 2016, IOP Conference Series: Materials 29. A. Lendlein, H. Jiang, O. Jünger and R. Langer, Science and Engineering, Vol. 149, IOP Publishing Nature, 2005, 434, 879 Ltd, Bristol, UK, 2016 30. A. Lendlein and R. Langer, Science, 2002, 8. D. Patil and G. Song, Smart Mater. Struct., 2017, 296, (5573), 1673 26, (9), 093002 31. K. Inoue, M. Yamashiro and M. Iji, J. Appl. Polym. 9. N. Ma, Y. Lu, J. He and H. Dai, J. Text. Inst., Sci., 2009, 112, (2), 876 2019, 110, (6), 950 32. L. B. Vernon and H. M. Vernon, The Vernon 10. C. Wen, X. Yu, W. Zeng, S. Zhao, L. Wang, Benshoff Company, ‘Process of Manufacturing G. Wan, S. Huang, H. Grover and Z. Chen, AIMS Articles of Thermoplastic Synthetic Resins’, US Mater. Sci., 2018, 5, (4), 559 Patent 2,234,993; 1941 11. W. M. Huang, Z. Ding, C. C. Wang, J. Wei, 33. W. C. Rainer, E. M. Redding, J. J. Hitov, Y. Zhao and H. Purnawali, Mater. Today, 2010, A. W. Sloan and W. D. Stewart, W. R. Grace & Co, 13, (7–8), 54 ‘Polyethylene Product and Process’, US Patent 12. C. Liu, H. Qin and P. T. Mather, J. Mater. Chem., 3,144,398; 1964 2007, 17, (16), 1543 34. J. Hu, Y. Zhu, H. Huang and J. Lu, Prog. Polym. 13. Y. Liu, H. Du, L. Liu and J. Leng, Smart Mater. Sci., 2012, 37, (12), 1720 Struct., 2014, 23, (2), 023001 35. A. Lendlein and S. Kelch, Angew. Chem. Int. Ed., 14. W. Sokolowski, A. Metcalfe, S. Hayashi, L. Yahia 2002, 41, (12), 2034 and J. Raymond, Biomed. Mater., 2007, 2, (1), 36. S. Ota, Radiat. Phys. Chem., 1981, 18, (1–2), S23 81 15. P. K. Kumar and D. C. Lagoudas, ‘Introduction 37. Q. Zhao, H. J. Qi and T. Xie, Prog. Polym. Sci., to Shape Memory Alloys’, in “Shape Memory 2015, 49–50, 79 Alloys”, ed. D. C. Lagoudas, Springer, Boston, USA, 2008, pp. 1–51 38. X. Wu, W. M. Huang, Y. Zhao, Z. Ding, C. Tang and J. Zhang, Polymers, 2013, 5, (4), 1169 16. K. Yu, T. Xie, J. Leng, Y. Ding and H. J. Qi, Soft Matter, 2012, 8, (20), 5687 39. Y. Bai, J. Zhang and X. Chen, ACS Appl. Mater. Interfaces, 2018, 10, (16), 14017 17. J. G. Hardy, M. Palma, S. J. Wind and M. J. Biggs, Adv. Mater., 2016, 28, (27), 5717 40. K. Yu, Y. Liu and J. Leng, RSC Adv., 2014, 4, (6), 2961 18. K. Wang, S. Strandman and X. X. Zhu, Front. Chem. Sci. Eng., 2017, 11, 143 41. I. A. Rousseau and P. T. Mather, J. Am. Chem. Soc., 2003, 125, (50), 15300 19. M. Behl and A. Lendlein, Mater. Today, 2007, 10, (4), 20 42. J. Caprasse, T. Defize, R. Riva and C. Jérôme, ‘Comparative Study of PCL Shape-Memory 20. H. Meng and G. Li, Polymer, 2013, 54, (9), 2199 Networks with Diels-Alder or Alder-ene Adducts’, 21. L. Sun, W. M. Huang, Z. Ding, Y. Zhao, Advanced Functional Polymers for Medicine C. C. Wang, H. Purnawali and C. Tang, Mater. (AFPM), Montpellier, France, 16th–18th May, Des., 2012, 33, 577 2018

43. T. Defize, R. Riva, J.-M. Raquez, P. Dubois, 59. G. Zhou, H. Zhang, S. Xu, X. Gui, H. Wei, J. Leng, C. Jérôme and M. Alexandre, Macromol. Rapid N. Koratkar and J. Zhong, Sci. Rep., 2016, 6, Commun., 2011, 32, (16), 1264 24148 44. H.-Y. Lai, H.-Q. Wang, J.-C. Lai and C.-H. Li, 60. M. Dahmardeh, M. S. M. Ali, T. Saleh, T. M. Hian, Molecules, 2019, 24, (18), 3224 M. V. Moghaddam, A. Nojeh and K. Takahata, 45. D. Iqbal and M. H. Samiullah, Materials, 2013, Phys. Status Solidi A, 2013, 210, (4), 631 6, (1), 116 61. H. B. Gilbert and R. J. Webster, IEEE Robot. 46. D. J. Maitland, M. F. Metzger, D. Schumann, Autom. Lett., 2016, 1, (1), 98 A. Lee and T. S. Wilson, Lasers Surg. Med., 62. M. Xie, L. Wang, J. Ge, B. Guo and P. X. Ma, ACS 2002, 30, (1), 1 Appl. Mater. Interfaces, 2015, 7, (12), 6772 47. H. Xie, K.-K. Yang and Y.-Z. Wang, Prog. Polym. 63. H. Tanaka and K. Honda, J. Polym. Sci. Pol. Sci., 2019, 95, 32 Chem., 1977, 15, (11), 2685 48. Z. Yuan, A. Muliana and K. R. Rajagopal, Math. 64. H. Luo, J. Hu and Y. Zhu, Mater. Letters, 2012, Mech. Solids, 2016, 22, (5), 1116 89, 172 49. E. Havens, E. A. Snyder and T. H. Tong, 65. C. Gong, J. Liang, W. Hu, X. Niu, S. Ma, H. T. Hahn ‘Light-Activated Shape Memory Polymers and and Q. Pei, Adv. Mater., 2013, 25, (30), 4186 Associated Applications’, SPIE Smart Structures 66. H. Luo, Z. Li, G. Yi, X. Zu, H. Wang, Y. Wang, and Materials + Nondestructive Evaluation and H. Huang, J. Hu, Z. Liang and B. Zhong, Mater. Health Monitoring, San Diego, California, USA, Letters, 2014, 134, 172 5th May, 2005, “Smart Structures and Materials 67. T. Akter and W. S. Kim, ACS Appl. Mater. 2005: Industrial and Commercial Applications Interfaces, 2012, 4, (4), 1855 of Smart Structures Technologies”, Vol. 5762, 68. Y.-C. Sun, M. Chu, M. Huang, O. Hegazi and Society of Photo-Optical Instrumentation H. E. Naguib, Macromol. Mater. Eng., 2019, Engineers (SPIE), Bellingham, USA, 2005, 8 pp 304, (10), 1900196 50. Y. Liu, H. Lv, X. Lan, J. Leng and S. Du, Compos. 69. Y. Guo, Z. Lv, Y. Huo, L. Sun, S. Chen, Z. Liu, Sci. Technol., 2009, 69, (13), 2064 C. He, X. Bi, X. Fan and Z. You, J. Mater. Chem. 51. H. Lu, Y. Yao and L. Lin, Pigm. Resin Technol., B, 2019, 7, (1), 123 2014, 34, (1), 26 70. I. T. Garces, S. Aslanzadeh, Y. Boluk and 52. J. Alam, A. Khan, M. Alam and R. Mohan, C. Ayranci, Materials, 2019, 12, (2), 244 Materials, 2015, 8, (9), 6391 71. K. Fan, W. M. Huang, C. C. Wang, Z. Ding, 53. J. Zhang, X. Ke, G. Gou, J. Seidel, B. Xiang, P. Yu, Y. Zhao, H. Purnawali, K.C. Liew and L. X. Zheng, W.-I. Liang, A. M. Minor, Y. Chu, G. Van Tendeloo, eXPRESS Polym. Lett., 2011, 5, (5), 409 X. Ren and R. Ramesh, Nat. Commun., 2013, 4, 72. L. Sun, T. X. Wang, H. M. Chen, A. V. Salvekar, 2768 B. S. Naveen, Q. Xu, Y. Weng, X. Guo, Y. Chen 54. X. Gong, L. Liu, Y. Liu and J. Leng, Smart Mater. and W. M. Huang, Polymers, 2019, 11, (6), 1049 Struct., 2016, 25, (3), 035036 73. M. Ma, L. Guo, D. G. Anderson and R. Langer, 55. J. Zhou, H. Li, R. Tian, R. Dugnani, H. Lu, Science, 2013, 339, (6116), 186 Y. Chen, Y. Guo, H. Duan and H. Liu, Sci. Rep., 74. M. Shibayama, M. Sato, Y. Kimura, H. Fujiwara 2017, 7, 5535 and S. Nomura, Polymer, 1988, 29, (2), 336 56. S.-K. Lee, S.-J. Lee, H.-J. An, S.-E. Cha, J.- 75. E. Smela, Adv. Mater., 2003, 15, (6), 481 K. Chang, B. Kim and J. J. Pak, ‘Biomedical 76. R. H. Baughman, Science, 2005, 308, (5718), Applications of Electroactive Polymers and 63 Shape-Memory Alloys’, SPIE’s 9th Annual 77. Q. Song, H. Chen, S. Zhou, K. Zhao, B. Wang International Symposium on Smart Structures and P. Hu, Polym. Chem., 2016, 7, (9), 1739 and Materials, San Diego, USA, 11th July, 2002, “Smart Structures and Materials 2002: 78. H. Xiao, C. Ma, X. Le, L. Wang, W. Lu, P. Theato, Electroactive Polymer Actuators and Devices T. Hu, J. Zhang and T. Chen, Polymers, 2017, (EAPAD)”, Vol. 4695, Society of Photo-Optical 9, (4), 138 Instrumentation Engineers (SPIE), Bellingham, 79. X.-J. Han, Z.-Q. Dong, M.-M. Fan, Y. Liu, J.-H. Li, USA, 2002, 15 pp Y.-F. Wang, Q.-J. Yuan, B.-J. Li and S. Zhang, 57. N. G. Sahoo, Y. C. Jung, H. J. Yoo and J. W. Cho, Macromol. Rapid Commun., 2012, 33, (12), Compos. Sci. Technol., 2007, 67, (9), 1920 1055 58. N. G. Sahoo, Y. C. Jung, N. S. Goo and J. W. Cho, 80. J. Li, Q. Duan, E. Zhang and J. Wang, Adv. Mater. Macromol. Mater. Eng., 2005, 290, (11), 1049 Sci. Eng., 2018, 7453698

81. Y. Li, H. Chen, D. Liu, W. Wang, Y. Liu and 97. S. Kelch, S. Steuer, A. M. Schmidt and A. Lendlein, S. Zhou, ACS Appl. Mater. Interfaces, 2015, Biomacromolecules, 2007, 8, (3), 1018 7, (23), 12988 98. J. Leng, H. Lu, Y. Liu, W. M. Huang and S. Du, 82. T. Wu, Y. Su and B. Chen, ChemPhysChem, MRS Bull., 2009, 34, (11), 848 2014, 15, (13), 2794 99. B. K. Kim, S. Y. Lee and M. Xu, Polymer, 1996, 83. H. Chen, Y. Li, Y. Liu, T. Gong, L. Wang and 37, (26), 5781 S. Zhou, Polym. Chem., 2014, 5, (17), 5168 100. S. Rimdusit, M. Lohwerathama, K. Hemvichian, 84. K. H. M. Kan, J. Li, K. Wijesekera and P. Kasemsiri and I. Dueramae, Smart Mater. E. D. Cranston, Biomacromolecules, 2013, Struct., 2013, 22, (7), 075033 14, (9), 3130 101. H. Luo, H. Wang, H. Zhou, X. Zhou, J. Hu, G. Yi, 85. A. E. Way, L. Hsu, K. Shanmuganathan, C. Weder Z. Hao and W. Lin, Appl. Sci., 2018, 8, (3), 392 and S. J. Rowan, ACS Macro. Lett., 2012, 1, (8), 102. J. Zhang, M. Huo, M. Li, T. Li, N. Li, J. Zhou and 1001 J. Jiang, Polymer, 2018, 134, 35 86. N. Gabdullin and S. H. Khan, ‘Review of Properties 103. G. Ji, P. Zhang, J. Nji, M. John and G. Li, ‘11 of Magnetic Shape Memory (MSM) Alloys and - Shape Memory Polymer-Based Self-Healing MSM Actuator Designs’, 2014 Joint IMEKO TC1- Composites’, in “Recent Advances in Smart Self- TC7-TC13 Symposium: Measurement Science healing Polymers and Composites”, eds. G. Li Behind Safety and Security, Madeira, Portugal, and H. Meng, Woodhead Publishing Series in 3rd–5th September, 2014, Journal of Physics: Composites Science and Engineering, ch. 11, Conference Series, Vol. 588, IOP Publishing Ltd, Woodhead Publishing, Cambridge, UK, 2015, Bristol, UK, 2015, 6 pp pp 293–363 87. E. Faran and D. Shilo, Exp. Tech., 2016, 40, (3) 104. A. V. Menon, G. Madras and S. Bose, Polym. 1005 Chem., 2019, 10, (32), 4370 88. J. Karger-Kocsis and S. Kéki, Polymers, 2018, 105. A. Lendlein, M. Behl, B. Hiebl and C. Wischke, 10, (1), 34 Expert Rev. Med. Devices, 2010, 7, (3), 357 89. P. R. Buckley, G. H. McKinley, T. S. Wilson, 106. M. P. Gaj, A. Wei, C. Fuentes-Hernandez, W. Small, W. J. Benett, J. P. Bearinger, Y. Zhang, R. Reit, W. Voit, S. R. Marder and M. W. McElfresh and D. J. Maitland, IEEE Trans. B. Kippelen, Org. Electron., 2015, 25, 151 Biomed. Eng., 2006, 53, (10), 2075 107. S. J. Park and C. H. Park, Sci. Rep., 2019, 9, 90. A. Goldman, “Modern Ferrite Technology”, 9157 Springer Verlag, New York, USA, 2006 108. S. Thakur ‘Shape Memory Polymers for Smart 91. P. R. Stauffer, T. C Cetas, A. M. Fletcher, Textile Applications’, in “Textiles for Advanced D. W. Deyoung, M. W. Dewhirst, J. R. Oleson and Applications”, eds. B. Kumar and S. Thakur, ch. R. B. Roemer, IEEE Trans. Biomed. Eng., 1984, 12, Intech, Rijeka, Croatia, 2017, 432 pp BME-31, (1), 76 109. L. Li, P. Shi, Li Hua, J. An, Y. Gong, R. Chen, 92. A. Jordan, R. Scholz, P. Wust, H. Fähling and C. Yu, W. Hua, F. Xiu, J. Zhou, G. Gao, Z. Jin, R. Felix, J. Magn. Magn. Mater., 1999, 201, (1– G. Sun and W. Huang, Nanoscale, 2018, 10, (1), 3), 413 118 93. X. Fu, Y. Yuan, Z. Liu, P. Yan, C. Zhou and J. Lei, 110. Y. Huang, M. Zhu, Z. Pei, Q. Xue, Y. Huang and Eur. Polym. J., 2017, 93, 307 C. Zhi, J. Mater. Chem. A, 2016, 4, (4), 1290 94. A. M. Kushner, J. D. Vossler, G. A. Williams and Z. Guan, J. Am. Chem. Soc., 2009, 131, (25), 111. S. Ahn, P. Deshmukh and R. M. Kasi, 8766 Macromolecules, 2010, 43, (17), 7330 95. A. Li, J. Fan and G. Li, J. Mater. Chem. A, 2018, 112. J. Leng, D. Zhang, Y. Liu, K. Yu and X. Lan, Appl. 6, (24), 11479 Phys. Lett., 2010, 96, 111905 96. F. Xie, L. Huang, J. Leng and Y. Liu, J. Intell. 113. W. Li, Y. Liu and J. Leng, J. Mater. Chem. A, Mater. Syst. Struct., 2016, 27, (18), 2433 2015, 3, (48), 24532

The Authors

Mathew John Haskew received his BSc in Chemistry from Lancaster University, UK, and subsequently undertook an MSc by research on SMP-based materials (with John Hardy at Lancaster University). He is currently undertaking a PhD in Engineering with Samuel Murphy and John Hardy at Lancaster University. His PhD involves the development of biodegradable biomaterials, and a combination of computational modelling and experimental validation of their efficacy.

John George Hardy received his MSci and PhD in Chemistry from the University of Bristol, UK, and the University of York, UK, respectively. Thereafter he undertook postdoctoral research in Biochemistry, Biomedical Engineering, Materials Science and Pharmacy (in France, Germany, Northern Ireland and the USA) before returning to the UK to lead a research group developing stimuli-responsive materials for technical and medical applications.

www.technology.matthey.com

Application of Chitosan-Encapsulated Orange Oil onto Footwear Insock Leathers Spray drying technique for an environmentally sustainable antibacterial formulation

Buket Yılmaz leathers acquiring antibacterial properties. The Department of Materials Science and products and process are biodegradable, nontoxic Engineering, Graduate School of Natural and biocompatible. and Applied Science, Ege University, 35100 Bornova-Izmir, Turkey 1. Introduction

Hüseyin Ata Karavana* Footwear is the most commonly worn apparel in Department of Leather Engineering, Faculty of daily life, and its design features must prioritise Engineering, Ege University, 35100 Bornova- anatomy, comfort and hygiene. For this reason, it Izmir, Turkey; Department of Materials Science is important to develop sustainable improvements and Engineering, Graduate School of Natural to footwear’s functional properties. and Applied Science, Ege University, 35100 Footwear that carries the body’s weight during Bornova-Izmir, Turkey the day can affect foot health physically, chemically and microbiologically. Continuous contact with *Email: [email protected]; the external environment exposes footwear to [email protected] microorganisms during normal use. All sorts of footwear play a role in the transport, spread and contamination of pathogenic or non-pathogenic The purpose of this study was to devise an antibacterial microorganisms (1). treatment for footwear insock leathers. Orange oil- There are different microorganisms in every part loaded chitosan microparticles were utilised for of the human body. Sweat is regularly secreted this purpose. Emulsion formulations with different from the body under normal conditions. It contains ratios were prepared, and from these formulations 98% water and urea, uric acid, fatty acid, lactic acid microparticles were manufactured using a spray and sulfates (2). The feet have more sweat glands drying technique. Microparticles obtained in this way than other parts of the body. Sweat secreted from were applied to the insock leathers using a finishing feet during the usage of footwear is decomposed process. Successful encapsulation was confirmed by means of foot microbiota; as a result, bad by ultraviolet-visible (UV-vis) spectrophotometry, odours emerge in footwear. Brevibacterium linens, Fourier transform infrared (FTIR) spectroscopy and Staphylococcus epidermidis, Staphylococcus aureus scanning electron microscopy (SEM) techniques. The and Escherichia coli are some microorganisms microparticles exhibited highly spheroid shape with that make footwear unhygienic. As a result of the a size range of 3–5 µm. Microparticle encapsulation breakdown of amino acids in sweat and skin by efficiencies ranged from 79.41% ± 3.36% to 86.60% these microorganisms, bad odour arises in feet, ± 1.13%. After performing microbiological tests socks and footwear (3, 4). and in vitro release studies on the insock leathers, Nowadays, in addition to individual foot care it was determined that the prepared microparticles and hygiene for odour prevention in shoes, are able to perform core material delivery. Also, commercial materials with various deodorising successful microparticle application resulted in these and antimicrobial effects are also employed (5).

Footwear insock is a thin layer of materials put into Analytical grade chemicals were used in the the shoe after manufacture to cover the insole. It analyses. Insock leathers, without dye and ready directly contacts the sole of the wearer’s foot and for experimental application, were donated from can provide a more sanitary environment when Ata Dilek Leather (Izmir, Turkey). specially treated for antimicrobial purposes (6). For the microparticle production, chitosan (shell Spray drying is an advantageous way to encapsulate material) was added to 1% w/w aqueous acetic active substances and essential oils. Spray drying acid for preparing the chitosan solution. This is a common and accepted encapsulation method solution was stirred at 45°C by using a magnetic for industrial applications. With this method, it is stirrer until wholly dissolved. During the pre- possible to mass produce capsules. The distribution emulsion preparation, orange oil (core material) of particles is uniform (7–10). was gradually mixed into the chitosan solution and Microencapsulation technology has been used for stirred for 1 min at 10,000 rpm. The surfactants the application of orange oil to textiles and leathers, as compound emulsifiers used for pre-emulsion being an economically viable, fast and efficient preparations were Tween 40 with Span 20 at method by combining core and shell materials, the ratio of 8:2 w/w. Then, the microparticles desirable perceptual and functional characteristics, were prepared by using an SD Basic spray dryer and also allowing functional substances to be (Labplant, UK) with nozzle diameter of 0.5 mm. released in a controlled manner. This technique The orange oil to chitosan ratios in the four has also been used to microencapsulate a wide encapsulating compounds came to 1:1, 1:1.33, range of active, functional, sensitive or volatile 1:1.67 and 1:2 w/w. The ingredients of the substances (11–14). Tea tree oil containing formulations in the spray-drying process are shown melamine formaldehyde microcapsules, essential in Table I. Homogeneous emulsions were fed to oils (eucalyptus, lavender or oregano), polyurethane the spray dryer under the following conditions: dispersions containing photoactive antimicrobial pump speed 12 ml min–1, outlet air temperature agents, zinc oxide and silver nanoparticles are 114°C and inlet air temperature 175°C. some substances that protect upper leathers from The microparticles’ morphology was examined by the harmful propagation of microorganisms (3, 15). a Quanta 250 FEG scanning electron microscope In addition, aromas confined to microcapsules are (FEI, USA) at 2 kV accelerating voltage. Before also used to prevent bad odours in footwear (16, coating in an argon atmosphere with gold-palladium 17). Application of antibacterial and aromatic by a K550X sputter coating machine (Quorum materials onto footwear insocks to control bad Emitech, UK), the samples were mounted onto an odours is good for foot hygiene and desired shoe aluminium stub. The grain side of leathers coated comfort. with microparticles was examined by a TM1000 The use of orange oil presents as an ecological tabletop scanning electron microscope (Hitachi, alternative to synthetic chemicals, attracting Japan) after coating with gold-palladium. the attention of the scientific community to the The FTIR spectra of the spray-dried microparticles development of eco-friendly antimicrobials. In this and leathers with microparticles were procured by study, microparticles were produced by a spray a Spectrum 100 FTIR attenuated total reflectance drying method after the emulsions with orange (ATR) spectrometer (PerkinElmer, USA). The oil and chitosan were prepared in different ratios. measurements were made using four scans with a Microparticles manufactured in this way were then resolution of 4 cm–1 between 4000 cm–1 to 650 cm–1 transferred to the surface of the footwear insock wavenumber ranges at room temperature. leathers using a finishing process. Afterward, some Encapsulation efficiencies of microparticles were tests and analyses were performed on microparticle calculated as the amount of orange oil (core coated footwear insock leather samples to material) encapsulated in the microparticles. The evaluate the effectiveness of the microparticles, their presence on the leather surface and their Table I Composition of the Formulations antimicrobial properties. Formulation code Orange oil:chitosan, w/w 2. Experiment T3 1:1 T4 1:1.33 Pharmaceutical grade cold pressed orange oil T5 1:1.67 was donated from Ephesus, Turkey. Chitosan T6 1:2 was purchased from Acros OrganicsTM (Belgium).

444 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15901340190139 Johnson Matthey Technol. Rev., 2020, 64, (4) encapsulation efficiency was calculated using Table II Basic Finishing Recipe Applied to Equation (i) (18). Insock Leathers Materials Amount, part Practice B (total amount of oil content – the amount of Water 100 3 × Spray surface oil content) Anionic wax 50 %EE = × 100 (i) A (total amount of oil Non-ionic aliphatic content) polyurethane 25 binder A solvent extraction method was used to determine Orange oil loaded 12 total oil content. A 0.1 g measurement of orange oil microparticles loaded microparticles was dissolved in 10 ml of 1% acetic acid solution at room temperature for 45 min. were placed into an incubator for incubation at Released orange oil which was obtained from the 37°C for 18 h in the Mueller Hinton broth (MHB) completely dissolved microparticles was placed in medium. Then, microorganisms were inoculated a beaker containing 50 ml n-hexane for extraction in petri dishes containing 105 colony forming unit 45 min. So as to determine the total amount of (CFU) ml–1 of Mueller Hinton agar (MHA) medium. orange oil in the microparticles, this extract was Next, microparticle coated insock leather samples filtered through a syringe filter (0.22 μm). Orange with 12.7 mm diameter were placed into the petri oil content in the filtrate was measured using a UV- dishes (20, 25). All petri dishes were placed into 1800 UV-vis spectrophotometer (Shimadzu, Japan) an incubator for incubation at 37°C for 24 h, and at 202 nm in triplicate. Surface oil content was also inhibition zones were measured to determine determined by the same solvent extraction method antibacterial activity. described above, except for a dissolving process in 1% acetic acid solution (11). 3. Results and Discussion In vitro release studies of microparticles and microparticle loaded leathers were carried out at In this study, orange oil microparticles were a speed of 100 rpm in phosphate-buffered saline successfully prepared by spray drying method. (PBS) and methanol at 37°C. 1 mg orange oil This method is a simple, viable method to obtain loaded microparticles was suspended in beakers microparticles, suitable to prevent active substance containing 4 ml methanol and 16 ml of PBS. Insock biological activity loss, avoiding exposure to leathers with 2.5 cm2 area were placed in beakers elevated heating and to organic solvents. containing 16 ml methanol and 16 ml of PBS for in vitro release studies of microparticle loaded 3.1 Surface Appearance of leathers. At suitable time intervals, the medium Microparticles and Microparticle in the beakers was filtered through a 0.22 μm Coated Insock Leathers syringe filter. Sink conditions were maintained in the receptor compartment during in vitro release A scanning electron microscope was used to studies. The released amount of orange oil was examine the morphology of the spray-dried analysed by UV method, as previously described, microparticles. SEM micrographs revealed that all for 5 h. Experiments were performed five times. microparticle formulations have a highly spheroid A spraying pistol with nozzle diameter of 0.5 mm shape with a morphology approximating an orange was used to apply microparticles to the insock peel effect. Microparticles of non-uniform size leathers during the finishing process. Spray-dried were observed with clear distinction between shell microparticles were added to the finishing recipe and core materials. These shape features indicate as 20 g m–2 (19). The basic finishing recipe for the that orange oil is spread on the surface of the insock leathers is given in Table II (20). microparticles. The morphology of spray-dried The efficacy of microparticle coated insock leathers microparticle formulations is shown in Figure 1. against test microorganisms Staphylococcus Particle morphology (surface, size and distribution) aureus ATCC® 6538TM, Escherichia coli ATCC® was not affected by the polymer ratio or core:shell 25922TM, Candida albicans ATCC® 10231TM, ratio. It is observed that there was formation of Klebsiella pneumoniae ATCC® 4352TM and Bacillus microcapsules, but they have stuck one to another subtilis ATCC® 6633TM was examined by agar disc and an agglomerate of microcapsules occurred. diffusion method (21–24). Test microorganisms Microparticles with similar morphology were also

3.2 Fourier Transform Infrared (a) (b) Spectroscopy Studies

Interactivity between the core material and shell

3.780 μm material usually leads to characteristic alterations in

2 μm 10 μm the FTIR spectra. FTIR spectra of chitosan, orange oil, microparticles and insock leather samples are shown in Figure 3 and Figure 4. Characteristic (c) (d) peaks at 1029 cm–1, 1149 cm–1, 1373 cm–1, 3.173 μm 1419 cm–1, 1585 cm–1, 2867 cm–1 and 3362 cm–1 were demonstrated in the FTIR spectrum of chitosan –1 (Figure 3). The peak at 3362 cm (OH and NH2 2 μm 10 μm stretching) was attributed to the amino group of chitosan. An intense absorption peak was seen at 2867–2922 cm–1 owing to C–H stretching in all (e) (f) spectra. The peak at 1585 cm–1 was attributed to 4.523 μm + N–H bending of the NH3 functional group present in the chitosan (28, 29). The peak at 1373 cm–1 confirmed the presence of an amide III band in the 2 μm 10 μm chitosan. The C–O–C stretching resulted from the spectra at 1149 cm–1 and 1029 cm–1. The spectrum –1 (g) (h) at 660 cm was attributed to stretching vibration of pyranoside ring (30–34). 5.330 μm The FTIR spectrum of the orange oil showed the distinctive bands of D-limonene, which is the primary constituent in orange oil (Figure 3). 2 μm 10 μm Especially, the bands between 2919–2834 cm–1 were attributed to the C–H stretching vibrations in Fig. 1. SEM micrographs of microparticle formulations: –1 (a) T3 formulation, 50,000 × magnification; –CH–, –CH2– and –CH3. The spectrum at 2965 cm (b) T3 formulation, 10,000 × magnification; (c) was attributed to the stretching vibrations of =C–H. T4 formulation, 50,000 × magnification; (d) The band 1644 cm–1 was attributed to the stretching T4 formulation, 10,000 × magnification; (e) vibrations of C=C. The band seen at 1435 cm–1 T5 formulation, 50,000 × magnification; (f) T5 was attributed to the C–H bending vibrations in formulation, 10,000 × magnification; (g) T6 –CH–, –CH – and –CH . The peaks at 885 cm–1 and formulation, 50,000 × magnification; (h) T6 2 3 –1 formulation, 10,000 × magnification 797 cm were attributed to the bending vibrations (out of plane) in =CH2 and =C(R)–H, respectively. The band at 1376 cm–1 was also attributed to the

C–H bending vibrations in –CH3 (mostly used to obtained in other spray drying experiments carried describe the existence of methyl) (35, 36). out using natural polymeric mixtures (11, 26, 27). As seen in Figure 3, most bands in the FTIR spectra We also examined the surface appearance of of the microparticles belonged to chitosan, which microparticle-free and microparticle-coated insock indicated that orange oil droplets were trapped in leather samples. The different microparticle chitosan (shell material) and that distinctive band formulations were clearly observed on insock of orange oil vanished or declined. Evidently, the leather surfaces after successful application of the free vibrations of orange oil molecules were blocked finishing process. Micrographs of insock leather by the chitosan because of physical interactions surfaces are shown in Figure 2. such as van der Waals or electrostatic interaction. After the finishing process, the presence of Furthermore, the intensity of microparticle peaks microparticles on the insock leather can be seen on the FTIR spectrum was lower than that of very clearly for all formulations. The images indicate chitosan because of the interaction between orange that the fixation was successfully achieved. Hence, oil and chitosan. The FTIR spectra of microparticles the leather samples preserve the capsule content demonstrated the C–H bending vibrations of –1 even after the finishing process. –CH3 at 1376 cm , except the =C(R)–H bending

(a) (b) (c)

30 μm 100 μm 30 μm

(d) (e) Fig. 2. SEM micrographs of the insock leather after finishing process: (a) microparticle free; (b) T3 formulation; (c) T4 formulation; (d) 100 μm 100 μm T5 formulation; (e) T6 formulation

Chitosan 1029.40 Fig. 3. The FTIR 2867.87 1585.84 1419.87 3362.83 1373.85 1149.73 spectrum of the chitosan, 892.61 660.57 orange oil and four Orange oil different microparticle 1147.91 1016.92 758.92 2965.87 2834.88 formulations (T3, T4, T5 1644.85 1376.86 1051.92 914.86 2919.81 1435.80 956.93 797.80 and T6) T3 885.61 3295.79 2922.80 1556.77 1511.79 1151.73 888.72 1407.75 1067.53

T, % T, T4 1028.53 2874.81 1376.76 3283.79 1557.79 1510.80 1151.72 887.69 752.71 T5 1028.49 3288.76 2876.79 1557.76 1510.79 1151.70 1408.74 888.66 T6 1063.46 888.69 1027.45 2875.81 1593.82 1376.79 3351.78 1509.83 1151.72

4000 3600 3200 2800 2400 2000 1800 1600 1400 1200 1000 800 650 cm–1 vibration at 797 cm–1, which was presumably due formulations were determined as 79.41% ± 3.36%, to the fact of the D-limonene ring being covered 81.28% ± 1.69%, 83.56% ± 0.66% and with chitosan (14, 36). 86.60% ± 1.13%, respectively. A great deal of Figure 4 show that bands between encapsulated orange oil is preferred. These results 1535–1547 cm–1 attributed to the NH band of showed that the microparticles’ encapsulation chitosan, did not appear in the blank leather efficiency is affected by the core:shell ratio. Increasing sample (34). Similarly, it was determined that IR the chitosan weight resulted in more encapsulated band vibration at 1095 cm–1 was observed in the orange oil, i.e. high encapsulation efficiency. This is microparticle loaded leathers but absent from the an effect similar to the oil:polymer ratio given by Li blank leather. That was evidence of the presence of and associates in their 2013 study (11). terpenoid, a component in orange oil (37). 3.4 In Vitro Release Studies of 3.3 Encapsulation Efficiency Microparticles and Microparticle Coated Insock Leathers Orange oil loaded microparticles were produced with a high orange oil encapsulation efficiency. Figure 5 shows the in vitro release behaviours The encapsulation efficiency of T3, T4, T5 and T6 of orange oil released from microparticles in four

Blank leather Fig. 4. The FTIR 3309.92 852.85 spectrum of footwear 1643.83 1379.80 720.79 2849.72 1728.46 1552.83 1024.72 1464.76 1237.63 insock leather without 2917.62 microparticles (blank T3 1159.51 leather) and with four 3285.90 2850.79 1644.83 different microparticles 1541.81 1379.80 1241.78 2916.72 1470.81 formulations (T3, T4, T5 1730.66 1155.66 T4 1024.55 and T6) 3309.93 1644.80 1380.81 851.84 1728.54 T, % T, 2850.76 1547.81 717.78 1464.78 1095.68 2916.66 1237.66 1024.67 T5 1159.59 3307.93 1644.81 1380.79 851.83 2850.77 1728.49 1544.80 718.78 2916.68 1464.78 1095.66 1237.64 1024.65 1158.55 T6 1023.64 3309.94 1236.61 1095.65 2850.82 1535.83 1157.52 851.83 2917.74 1463.79 732.78 1380.80

4000 3600 3200 2800 2400 2000 1800 1600 1400 1200 1000 800 650 cm–1

different formulations. The quantity of released 90 orange oil was measured at 202 nm in PBS at 80 different times. Previous experiments used PBS 70 as an in vitro release and diffusion medium for 60 T3 topical applications (38, 39). Oil release from 50 40 T4 microparticulate systems occurs via different C, % 30 T5 mechanisms including diffusion, desorption, 20 T6 disintegration and surface erosion (40). 10 The typical release pattern of the spray dried 0 microparticles is characterised by a small initial 30 60 90 120 150 180 210 240 270 300 Time, min burst release and a sustained release rate following Fig. 5. In vitro release of orange oil loaded that. It can be seen that orange oil release from microparticles microparticles gradually increased over time with exposure to PBS, which indicates that the orange oil disintegrated swiftly in PBS. This circumstance 60 is presumably owed to the fact that PBS is slightly 50 alkaline; chitosan is inclined to dissolve in slightly alkaline solution. Nonetheless, it can be seen that 40 T3 the release rate was not affected by chitosan 30 T4 concentration in the formulations. Figure 5 C, % 20 T5 graphs release behaviour as a function of orange T6 oil concentration, which was independent from 10 chitosan concentration. The in vitro release results of the leathers 0 500 1000 1500 Time, min impregnated with orange oil loaded microparticles in pH 7.4 PBS at 37°C are presented in Figure 6. Fig. 6. In vitro release of microparticle-coated This line graph shows controlled release behaviour insock leathers from leather treated with all formulations. Orange oil trapped inside the microparticles caused sustained release up to 24 h. When the 3.5 Microbiologic Studies on formulations are compared to each other, we Microparticle Coated Insock see the oil release ratio of insock leathers was Leathers affected by polymer concentration. High polymer concentration caused a slow release ratio of Table III shows microbiologic test results of orange oil. insock leathers treated with four microparticle

Table III Microbiologic Test Results of the Microparticle-Loaded Insock Leathers Candida Escherichia Klebsiella Staphylococcus Formulation Bacillus subtilis albicans ATCC® coli ATCC® pneumoniae aureus ATCC® codes ATCC® 6633TM 10231TM 25922TM ATCC® 4352TM 6538TM

formulations. An important revelation is that the 4. Conclusion test microorganisms did not grow on these leather samples. However, in some test groups, a meagre During the usage of footwear, perspiration and antimicrobial inhibition zone around the insock bacterial activity negatively impact foot health and leather samples meant that orange oil diffusion generate bad odours from both the feet and footwear. did not occur. There is a visible zone on Candida Shoe production using natural and non-toxic materials albicans in all formulations. T3 and T4 formulations, that prevent or inhibit bad odours and bacterial activity whose orange oil releases are higher in 24 h, is one solution to this hygienic problem. Likewise, the look more effective against Escherichia coli. The successful application of microparticles that release for antimicrobial activity is dependent on chitosan’s a long time on footwear insock leather is an important inherent behaviour and orange oil present on alternative to existing toxic products. leather samples. When the inhibition zones in the Our research found that emulsions with orange T6 formulation are examined, it can be seen that oil and chitosan have natural antibacterial activity. orange oil found in the insock leather samples These emulsions, when successfully converted into is more effective than the natural behaviour of encapsulated powders by a spray drying method, chitosan on antimicrobial activity. The antimicrobial produce a core-shell material. SEM images showed effect can be considered as proliferation or non- how an effective finishing process was used to proliferation in the area under the insock leather apply laboratory produced microparticles to the samples. This effect is also expressed as contact surface of footwear insock leather. Microbiological inhibition. No proliferation was observed on the tests performed on microparticle coated leathers contact surface of the insock leathers, that is, proved that footwear insock leathers were fortified on the surface where it touches the medium and with antibacterial properties. microorganism. Also, any proliferation on the These findings demonstrate that application of surfaces or edges of insock leather samples was orange oil-chitosan microparticles onto footwear not observed. There was no difference between insock leather surfaces is an alternative natural leather formulations on the antimicrobial test. method to control hygiene and eliminate bad

449 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15901340190139 Johnson Matthey Technol. Rev., 2020, 64, (4) odours. Non-toxic, functional, leather shoes J. Scapinello, C. N. Nesi, J. Dal Magro and can incorporate such natural materials in their S. L. Castellví, Braz. J. Biol., 2018, 78, (2), 311 manufacture and maintenance. This production 11. Y. Li, L. Ai, W. Yokoyama, C. F. Shoemaker, D. Wei, improvement would thus contribute to people’s J. Ma and F. Zhong, J. Agric. Food Chem., 2013, foot health, hygiene and comfort. 61, (13), 3311 12. I. Gönülşen, M. Sariişik, G. Erkan and S. Okur, Acknowledgements Tekst. ve Mühendis, 2016, 23, (101), 21 13. W. Rossi, M. Bonet-Aracil, E. Bou-Belda, J. Gisbert- The authors would like to thank the Scientific Payá, K. Wilson and L. Roldo, IOP Conf. Ser. Mater. Research Projects Coordination Unit, Ege University, Sci. Eng., 2017, 254, (2), 022007 Turkey (Project No: 17FBE006) for financial 14. P. Velmurugan, V. Ganeshan, N. F. Nishter and support and the Turkish Prime Ministry’s State R. R. Jonnalagadda, Surf. Interfac., 2017, 9, 124 Planning Organisation (Project No: 07DPT001) for 15. I. P. Fernandes, J. S. Amaral, V. Pinto, M. J. Ferreira equipment provision. and M. F. Barreiro, Carbohydr. Polym., 2013, 98, (1), 1229 References 16. M. M. Sánchez-Navarro, M. A. Pérez-Limiñana, F. Arán-Aís and C. Orgilés-Barceló, Polymer Int., 1. V. K. Özkan, M. T. Uzun and M. T. Gündoğan, 2015, 64, (10), 1458 J. Fungus, 2018, 9, (2), 182 17. C. Torres-Alvarez, A. Núñez González, J. Rodríguez, 2. H. Yalçın, B. Özkalp, ‘Importance of Body Hygiene S. Castillo, C. Leos-Rivas and J. G. Báez-González, and New Developments in Wound Care’, 4th CyTA-J. Food, 2017, 15, (1), 129 National Sterilization Disinfection Congress, 20th– 18. N. Suwannateep, S. Wanichwecharungruang, 24th April, 2005, Samsun, Turkey, Bilimsel Tıp S. F. Haag, S. Devahastin, N. Groth, J. W. Fluhr, Publishing House, Ankara, Turkey, 287–308 J. Lademann and M. C. Meinke, Eur. J. Pharm. 3. Z. Majidnia, A. Idris and P. Valipour, J. Teknol., Biopharm., 2012, 82, (3), 485 2013, 60, (1), 5 19. M. Kleban, J. Weisser, F. Koch and W. Schwaiger, 4. P. Velmurugan, M. Cho, S.-M. Lee, J.-H. Park, S. Bae Bayer Corp, ‘Leather Finished with Scent- and B.-T. Oh, Carbohydr. Polym., 2014, 106, 319 Containing Microcapsules’, US Patent Appl. 5. M. M. Sánchez-Navarro, M. A. Pérez-Limiñana, 2002/198,392 N. Cuesta-Garrote, M. I. Maestre-López, 20. F. Yalcin, H. A. Karavana, S. Rencber and M. Bertazzo, M. A. Martínez-Sánchez and F. Arán- S. Y. Karavana, J. Am. Leather Chem. Assoc., Aís, ‘Latest Developments in Antimicrobial 2020, 115, (3), 79 Functional Materials for Footwear’, in “Microbial 21. G. M. Caputo, P. R. Cavanagh, J. S. Ulbrecht, Pathogens and Strategies for Combating G. W. Gibbons and A. W. Karchmer, N. Engl. Them: Science, Technology and Education”, ed. J. Med., 1994, 331, (13), 854 A. Méndez-Vilas, Vol. 1, Formatex, New York City, 22. J. H. Calhoun, K. A. Overgaard, C. M. Stevens, USA, 2013, pp. 102–113 J. P. F. Dowling and J. T. Mader, Adv. Skin Wound 6. Z. Tülek, ‘A Research Over Side Industry Products Care, 2002, 15, (1), 31 in the Shoe Manufacturing’, Master Thesis, Social 23. B. Örmen, N. Türker, I. Vardar, N.A. Coşkun, Science Institute, Istanbul Arel University, Turkey, F. Kaptan, S. Ural, S. El and M. Türker, Turkish 2016, 126 pp J. Infect., 2007, 21, (2), 65 7. M. Gohel, R. K. Parikh, S. A. Nagori, A. V. Gandhi, 24. İ. Yaşa, N. Lkhagvajav, M. Koizhaiganova, E. Çelik M. S. Shroff, P. K. Patel, C. S. Gandhi, V. Patel, and Ö. Sarı, World J. Microbiol. Biotechnol., 2012, N. Y. Bhagat, S. D. Poptani, S. R. Kharadi, 28, (7), 2531 R. Pandya and T. C. Patel, Pharma. Rev., 2009, 25. H. A. Karavana, S. Rencber, S. Y. Karavana and 7, (5), 1 F. Yalcin, ‘Encapsulated Chlorhexidine Digluconate 8. B. N. Estevinho, F. Rocha, L. Santos and A. Alves, Usage on the Diabetic Footwear Lining Leathers’, Trends Food Sci. Technol., 2013, 31, (2), 138 6th International Conference on Advanced 9. D. Santos, A. C. Maurício, V. Sencadas, Materials and Systems (ICAMS), 20th–22nd J. D. Santos, M. H. Fernandes and P. S. Gomes, October, 2016, Bucharest, Romania, CERTEX, ‘Spray Drying: An Overview’, in “Biomaterials: Bucharest, Romania, 257–262 Physics and Chemistry”, ed. R. Pignatello, Ch. 2, 26. A. Soottitantawat, H. Yoshii, T. Furuta, InTechOpen, London, UK, 2018, pp. 9–35 M. Ohgawara, P. Forssell, R. Partanen, K. Poutanen 10. O. Tomazelli Júnior, F. Kuhn, P. J. M. Padilha, and P. Linko, J. Agric. Food Chem., 2004, 52, (5), L. R. M. Vicente, S. W. Costa, A. A. Boligon, 1269

27. A. Javid, Z. A. Raza, T. Hussain and A. Rehman, 34. A. K. T. Chang, R. R. Frias, L. V Alvarez, U. G. Bigol J. Microencapsul., 2014, 31, (5), 461 and J. P. M. D. Guzman, Biocatal. Agric. Biotechnol., 28. F.-L. Mi, H.-W. Sung, S.-S. Shyu, C.-C. Su and C.- 2019, 17, 189 K. Peng, Polymer, 2003, 44, (21), 6521 35. D. Cava, J. M. Lagarón, A. López-Rubio, R. Catalá and R. Gavara, Polym. Test., 2004, 23, (5), 551 29. D. R. Bhumkar and V. B. Pokharkar, AAPS 36. D. Li, H. Wu, W. Huang, L. Guo and H. Dou, Eur. PharmSciTech, 2006, 7, (2), E138 J. Lipid Sci. Technol., 2018, 120, (9), 1700521 30. R. Yoksan, J. Jirawutthiwongchai and K. Arpo, Coll. 37. H. Boughendjioua and S. Djeddi, Am. J. Opt. Surf. B: Biointer., 2010, 76, (1), 292 Photonics, 2017, 5, (3), 30 31. J. Jingou, H. Shilei, L. Weiqi, W. Danjun, W. Tengfei 38. T. Şenyiğit, F. Sonvico, S. Barbieri, Ö. Özer, and X. Yi, Coll. Surf. B: Biointer., 2011, 83, (1), P. Santi and P. Colombo, J. Control. Release, 2010, 103 142, (3), 368 32. L. Keawchaoon and R. Yoksan, Coll. Surf. B: 39. S. T. Tanrıverdi and Ö. Özer, Eur. J. Pharm. Sci., Biointer., 2011, 84, (1), 163 2013, 48, (4–5), 628 33. S. F. Hosseini, M. Zandi, M. Rezaei and 40. S. Hariharan, V. Bhardwaj, I. Bala, J. Sitterberg, F. Farahmandghavi, Carbohydr. Polym., 2013, U. Bakowsky and M. N. V. Ravi Kumar, Pharm. 95, (1), 50 Res., 2006, 23, (1), 184

The Authors

Buket Yılmaz graduated from Chemical Engineering, Faculty of Engineering, Anadolu University, Turkey, in 2015. For a period during her undergraduate education, she benefited from the FARABİ exchange programme for further chemical engineering studies at Ege University. Yılmaz won and completed a competitive internship at Turkey’s two leading companies involved in polymers and food production. In 2016, she started her master’s degree at Ege University’s Institute of Science, Materials Science and Engineering. Her scientific expertise has also been employed by the private sector in sales and in the quality control unit of a food production enterprise.

Hüseyin Ata Karavana graduated from the Leather Technology Department, Faculty of Agriculture, Ege University, Turkey. He earned his MSc degree in Leather Technology in 2001 from that institution’s Graduate School of Natural and Applied Science. From 2006 to 2007 he continued his studies as an Erasmus student in the Department of Footwear Engineering and Hygiene at the Tomas Bata University’s Faculty of Technology (Zlin, Czech Republic). Karavana completed his PhD degree in Leather Engineering at Ege University in 2008. Karavana currently serves as Associate Professor in the Department of Leather Engineering at Ege University’s Faculty of Engineering. His research interests are in all manner of leather and footwear engineering including plastic composites, microencapsulation, leather quality and control, footwear quality and control.

www.technology.matthey.com

Bacterial Community Composition in Produced Water of Diyarbakır Oil Fields in Turkey Bacterial communities in produced waters of south-eastern Turkey reported in detail for the first time

Tuğçe Tüccar* Among the classified bacteria, Proteobacteria Department of Biology, Institute of Graduate (29.2%), Firmicutes (8.3%), Bacteroidetes (8.3%) Studies in Sciences, Istanbul University, 34134, and Actinobacteria (4.2%) groups were identified. Vezneciler, Istanbul, Turkey Pseudomonas was the dominant genus detected in the produced water samples. The results of Esra Ilhan-Sungur this research provide, for the first time, insight Department of Biology, Faculty of Science, into the complexity of microbial communities in Istanbul University, 34134, Vezneciler, Istanbul, the Diyarbakır oil reservoirs and their dominant Turkey constituents.

Gerard Muyzer 1. Introduction Department of Biotechnology, Delft University of Technology, van der Maasweg 9, 2629 HZ Although much progress has been made in the use Delft, The Netherlands; Microbial Systems of renewable energy in recent years, fossil fuels Ecology, Department of Freshwater and Marine (especially oil and gas) still meet most of the global Microbiology, Institute for Biodiversity and energy demand, and they will continue to be the Ecosystem Dynamics, University of Amsterdam, dominant source of energy worldwide over the PO Box 94240, 1090 GE, Amsterdam, The next few decades (1). Netherlands Petroleum is a naturally occurring material found in various geological formations (reservoirs) *Email: [email protected] worldwide. Crude oil, the liquid part of petroleum, is primarily composed of hydrocarbons (2). However, it may also include compounds of nitrogen, sulfur, Oil fields harbour a wide variety of microorganisms oxygen and metals (3). Because crude oil in with different metabolic capabilities. To examine reservoirs is found as a mixture containing varying the microbial ecology of petroleum reservoirs, a constituents and proportions, each crude oil has molecular-based approach was used to assess the its own unique properties. The most important composition of bacterial communities in produced specified properties are density and sulfur water of Diyarbakır oil fields in Turkey. Denaturing content (4). The density of crude oil is reported gradient gel electrophoresis (DGGE) of polymerase in terms of American Petroleum Institute (API) chain reaction (PCR)-amplified 16S rRNA gene gravity (specific gravity). Based on the API gravity, fragments was performed to characterise the crude oils can be classified into light, medium, bacterial community structure of produced water heavy and extra heavy oils (3). Depending on the samples and to identify predominant community amount of sulfur content (elemental sulfur or sulfur members after sequencing of separated DGGE compounds such as hydrogen sulfide), the crude bands. The majority of bacterial sequences oil is categorised as ‘sweet’ or ‘sour’. In addition retrieved from DGGE analysis of produced water to chemical composition and physical properties, samples belonged to unclassified bacteria (50%). crude oil typically is also identified by underground

452 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15911723486216 Johnson Matthey Technol. Rev., 2020, 64, (4) reservoir (4). Reservoir characteristics (depth, reservoir souring and microbial corrosion. Reservoir temperature, pressure and other factors) vary souring, which is characterised by an increase significantly from one location to another, even in in production of H2S in the reservoir fluids, most the same geologic formation (5, 6). The fact that commonly occurs when sulfidogenic microorganisms microbial community composition and reservoir reduce sulfate to sulfide, a toxic and corrosive conditions vary dramatically not only between product (20). Undesirable accumulation of sulfide the different geographical areas, but also among minerals in reservoirs is one of the major challenging different oil fields in the same region, makes each problems in oil production because it causes plugging oil reservoir ecosystem unique. of reservoirs, decreasing the oil quality and value and Despite the extreme environmental conditions increasing the refining costs. Moreover, exposure to in the oil-bearing formations (i.e. anoxic, high H2S can be dangerous in terms of worker health temperature, high salinity), many microorganisms and safety due to its high toxicity. Additionally, the are capable of surviving in the oil and water produced H2S promotes corrosion of the metallic phases of the oil wells (7, 8). Oil fields harbour equipment and structures used for oil production and mainly facultative aerobic and strictly anaerobic processing (21). Another destructive phenomenon microorganisms due to the low redox potential is biocorrosion, which is defined as microbial attack in the reservoirs (8). These ecosystems contain on the surface of the metal infrastructure leading different types of microbial communities (such as to disruption of the material (22). In addition to mesophiles, thermophiles and halophiles) which sulfate-reducing bacteria, which play a major role adapt to the reservoir conditions (9). Bacterial in biocorrosion, other corrosive microbes, such as and archaeal groups identified in oil fields include acetogenic bacteria and methanogenic archaea, sulfate-reducing bacteria (10), sulfur-oxidising are also associated with corrosion failures (23). bacteria (11), methanogens (12), fermentative Biocorrosion is a great concern because it leads to loss microorganisms (13), acetogens (14), nitrate of material, large economic losses and safety issues reducers (15), manganese and iron reducers (16) in the oil industry (24). In contrast, hydrocarbon- and hydrocarbon degraders (17). Among these degrading bacteria may be used for environmental microbes, sulfate-reducing bacteria have attracted clean-up processes (6). Bacterial degradation of much attention due to their detrimental effects hydrocarbons was carried out by both aerobic such as reservoir souring and biocorrosion (7). In (for example, Rhodococcus sp., Sphingomonas addition, different members of the oil microbial sp., Pseudomonas putida, Pseudomonas stutzeri, community are involved in syntrophic interactions. Acinetobacter sp.) and anaerobic bacteria (such Fermenting bacteria and methanogenic archaea as Fe(III)-reducing bacteria, sulfate-reducing are involved in methanogenic hydrocarbon bacteria) (6, 17). Furthermore, microbial products biodegradation through their close syntrophic such as biopolymers and biosurfactants can be associations (18). This microbial process is used for facilitating oil movement in a widely used undesirable in oil reservoirs because it causes a technology, known as microbial enhanced oil recovery decrease in oil quality and value (19). Syntrophic (MEOR) (1). Compared with other conventional oil microorganisms in oil reservoirs also play important recovery techniques, MEOR has advantages such as roles in the global biogeochemical cycling of sulfur, low cost, wide application, high efficiency and low carbon and nitrogen. For instance, sulfate-reducing environmental pollution (25). Therefore, diversity, bacteria and sulfur-oxidising bacteria, the key metabolic processes and habitat conditions of drivers in sulfur transformations, are involved microbial communities in oil reservoirs should be in the sulfur cycle (11). Thus, knowledge of the investigated, so that their negative effects can be microbial groups and microbial dynamics in oil decreased and their positive effects can be exploited. fields enable us to obtain detailed insights into the This study aimed to determine the bacterial microbial ecology of oil associated environments. community composition and to identify the Understanding the microbial ecology of oil reservoirs predominant community members in produced is crucial to the petroleum industry because the water from oil fields located in the Diyarbakır success of oilfield operations is strongly influenced region in Turkey. To this end, we used PCR-DGGE by the activity of microorganisms. Oil microbes to analyse 20 produced water samples from the with different metabolic capabilities have significant Diyarbakır region. There are limited studies on negative and positive impacts on the petroleum produced water from the Diyarbakır region and this resources and the extraction processes (7). Microbial paper represents the only in situ study available. activity may lead to severe problems such as The results of this study provide not only new

453 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15911723486216 Johnson Matthey Technol. Rev., 2020, 64, (4) data about the microbial ecology of the Diyarbakır 2.2 DNA Extraction oil fields, but also information on the bacterial populations which may have potential roles in Bacteria in the produced water samples were terms of increasing or decreasing the efficiency of collected by filtration over 0.20 μm pore size industrial applications. polyamide filters (Sartolon®, Sartorius AG, Germany). Genomic DNA was extracted with ® 2. Materials and Methods the UltraClean Microbial DNA isolation kit (MO BIO Laboratories Inc, USA) according to the 2.1 Sampling Procedure manufacturer’s protocol.

The sampling site, the Diyarbakır region, is located 2.3 Polymerase Chain Reaction at the boundary of the Anatolian plate and the Amplification Middle Eastern oil region in south-eastern Turkey. A total of 20 crude oil samples (B1, B6, B8, B14, Extracted DNA was used as the template for PCR B23, B32, B56, GK8, GS6, GS15, M3, K2, K3, amplification of partial 16S rRNA fragments. Primer K32, K35, K44, S4, S15, Y18 and Y30) consisting pair consisting of 341F with a GC clamp and 907R of an oil/water mixture were collected from the was used for DGGE analysis (26). A 40-base GC production wells of Diyarbakır oil fields (Figure 1). clamp was used to prevent complete denaturation These wells produced oils withdrawn from the of the fragment during DGGE (27). oil sandstone deposits (depths from 1600 m to Due to the low DNA yield, a two-step PCR strategy 2620 m, API gravity from 24.3° to 42.3°, water was used. At the first step, a real-time PCR content around 94%, an average pH of 7.0 and (quantitative PCR, qPCR) approach was applied to salinity from 2966 mg l−1 to 26,961 mg l−1). The the produced water samples. The reaction mixture samples were aseptically taken at the wellhead and in a final volume of 22.5 µl contained 0.2 µl of put into sterile 500 ml serum bottles sealed with each primer, 12.5 µl iQTM SYBR® Green Supermix rubber stoppers and aluminium caps. The samples (Bio-Rad Laboratories Inc, USA), 9.6 µl RNase- were shipped at ambient temperature. Upon arrival Free Water (Qiagen, Germany) and 0.5 µl DNA at the laboratory, the samples were immediately template. qPCR was performed in iCycler iQTM Real- analysed. All samples were treated within 48 h Time PCR Detection System (Bio-Rad Laboratories after collection. Decantation was used to separate Inc, USA) using the following conditions: 5 min at produced water from the oil/water mixture. 95°C; 40 cycles of 95°C for 30 s, 57°C for 40 s,

37º00' 40º20' 43º40' Black sea

Yerevan

39º40' B1

36º20' M3 Y18 S4 GS15 K44 Beirut K35 Y30 Damascus GS6 B32 K2 37º00' 40º20' 43º40' B32 K3 N B23 GK8 B14 S15 B56 K32B8 B6

0 10 20 km

Fig. 1. Sampling locations in Diyarbakır region. Produced water samples were collected from 20 different oil wells © Maphill / Creative Commons Attribution-NoDerivatives (CC BY-ND)

72°C for 40 s and 80°C for 25 s; and a final 72°C KF720820, KF720823, KF720825 - KF720826, for 10 min. In the qPCR method, after each cycle, KF720828, KF720830 - KF720832, KF720839, a signal was formed. By observing the signals for KF720844, KF720852, KF720855, KF720858, each sample, PCR products could be detected. The KF720872, KF720877, KF720882 - KF720884, reaction was terminated when the desired amount KF720886 - KF720889, KF720891, KF720893 - of product was reached. At the second step, a KF720894, KF720896 and KF720903. conventional PCR approach was applied to the qPCR products. Reaction mixture in a final volume of 25 µl 3. Results contained 0.2 µl of each primer, 12.5 µl Taq PCR Master Mix (Qiagen, Germany), 9.6 µl RNase-Free 3.1 Molecular Analysis of Bacterial Water (Qiagen, Germany) and 0.5 µl DNA template. Communities The PCR was performed in TGradient thermocycler (Biometra, Germany) using the following conditions: Bacterial DNA isolation could only be achieved for 16 5 min at 95°C; 12 cycles of 95°C for 30 s, 57°C for (B1, B8, B6, B14, B23, B32, B56, GS6, GK8, K35, 40 s and 72°C for 40 s; and a final 74°C for 30 min. K44, M3, S4, S15, Y18, Y30) of the 20 produced water samples. Because the water phase could not 2.4 Denaturing Gradient Gel be separated from the oil phase for the other four produced water samples, DNA could not be extracted Electrophoresis from these samples. The extracted DNA was used The DCodeTM system (Bio-Rad Laboratories, USA) as template DNA for the amplification of 16S rRNA was used for DGGE analysis. 25 µl of each PCR gene fragment. Unfortunately, direct PCR with product (200–300 ng) were loaded onto 6% bacterial primers did not yield a product from any polyacrylamide gels (w/v) containing gradients of the produced water samples. For this reason, a of 20% to 70% denaturants (urea/formamide). two-step PCR was applied: the first step was a qPCR The gels were run for 16 h at 100 V and 60°C in to increase the concentration of genetic material to 1× Tris-acetate-EDTA buffer. After completion of measurable amounts (30), while the second step was electrophoresis, the gels were stained with SYBR® a normal PCR to obtain enough material for DGGE Gold Nucleic Acid Gel Stain (InvitrogenTM, Thermo analysis. For produced water samples, a total of 113 Fisher Scientific, USA) for 20 min, visualised and DGGE gel bands were analysed, but only 69 bands photographed. Selected predominant DGGE bands yielded sequences of satisfactory quality (Figure 2). were excised, eluted in 40 µl of 1× Tris buffer (pH 8) for 2 d at 4°C and re-amplified with 25 cycles as described above. Reaction mixture in a B32 B6 B14 B23 S4 GK8 GS6 M3 B56 B1 B8 K35 Y30 Y18 S15 final volume of 25 µl contained 0.125 µl of primer 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 341F, 0.125 µl of primer 907R, 12.5 µl of Taq PCR (a) (b) (c) Master Mix, 9.75 µl of ultra-pure water and 0.5 µl of template. The PCR products were quantified on a 1.5% (w/v) agarose gel and then sequenced by Macrogen Inc (Seoul, South Korea).

2.5 Comparative Sequence Analysis

The resulting sequences were first aligned and edited using CodonCode Aligner software (CodonCode Corp, USA). Then they were compared to sequences stored in the database GenBank® using the National Center for Biotechnology Information (NCBI) Basic Local Alignment Search ® Tool (BLAST ) (28, 29). All obtained partial 16S Fig. 2. DGGE profiles of 16S rRNA gene fragments rRNA gene sequences were deposited in GenBank® amplified from produced water samples. See database under the following accession numbers: legend to Figure 1. (a) 1, B32; 2, B6; 3, B14; 4, KF720792 - KF720796, KF720798, KF720801 - B23; 5, S4, 6, GK8; (b) 7, GS6; 8, M3; (c) 9, B56; 10, B1; 11, B8; 12, K35; 13, Y30; 14, Y14; 15, KF720802, KF720804, KF720806 - KF720808, S15 KF720810 - KF720811, KF720814, KF720818,

Comparative sequence analysis of the DGGE Acinetobacter comprises important soil organisms bands indicated that 50% of the bacterial where they contribute to the mineralisation of sequences belonged to ‘unclassified bacteria’. aromatic compounds and they are suited to Among the classified bacteria, members of exploitation for biotechnological purposes, such as the phyla Proteobacteria, Bacteroidetes, biodegradation (33). B8_30 was related (96%) to Firmicutes and Actinobacteria, and the classes Marinobacter sp. Trimyema-2, a thermophilic strain Alphaproteobacteria, Betaproteobacteria, that was isolated from the hydrothermally heated Gammaproteobacteria, Sphingobacteriia, Bacilli sea floor at Vulcano Island, Italy (34). Members and Actinobacteria were identified Figure( 3). of the genus Marinobacter were also identified in the production water retrieved from a Dutch oil 3.1.1 Proteobacteria field (35). The sequence from B32_2 was distantly related (93%) to Thermithiobacillus sp. ParkerM Proteobacteria was the dominant phylum, (HM173631) that is moderately thermophilic comprising 29.2% of the total sequences and obligately chemolithoautotrophic on reduced retrieved from the produced water samples inorganic sulfur compounds (36). Another member (Figure 3). The sequences B6_19 and B14_37 of the class Gammaproteobacteria was close to shared 100% and 99% identity with uncultured the sequence of uncultured hydrocarbon seep bacteria (EU044497 and JF421153, respectively) bacterium (91% similarity) (AF154088) (Table I). (Table I). The sequence B32_3 was distantly (94%) related to a moderately thermophilic 3.1.2 Bacteroidetes bacterium Phenylobacterium lituiforme, a member of Alphaproteobacteria (31). Within 8.3% of the sequences detected among the Betaproteobacteria, the sequence represented produced water samples fell into Bacteroidetes by B32_55 was identified (98%) as Aquincola sp. (Figure 3). The sequence of band B6_14 was THE-49 (JN128637), isolated from water reservoir affiliated to unclassified Chitinophagaceae. It (published only in GenBank®). The produced shared 99% identity with Chitinophagaceae water contained different members of the class bacterium F1 (AB535716), isolated from compost Gammaproteobacteria. DGGE bands B32_4, (Table I). DGGE bands B6_16, S4_67 and B8_27 B14_35, S4_70, GS6_2 and GK8_79 were affiliated were identified (92% to 99% sequence identity) as (100%, 94%, 99%, 93%, 99%, respectively) to uncultured Bacteroidetes bacteria (Table I). The Pseudomonas stutzeri (Table I), a non-fluorescent sequences from S4_67 and B8_27 were related to denitrifying bacterium (32). The sequence uncultured bacteria that were taught as members from band B1_20 showed a 100% similarity to of biocorroding microbiota colonising on steel Acinetobacter sp. VKPM 2838 (Table I). The genus surfaces immerged in coastal seawater (37).

(a) Actinobacteria (b) 4.2% Firmicutes Actinobacteria Gammaproteobacteria 8.3% 15% 38% Bacteroidetes 8.3% Sphingobacteriia 8% Unclassified bacteria Bacilli 50% 8%

Betaproteobacteria Proteobacteria 8% 29.2% Alphaproteobacteria 23%

Fig. 3. Phylogenetic distribution of the 16S rRNA sequences of produced water samples from the Diyarbakır oil wells at: (a) the phylum level; and (b) the class level

456 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15911723486216 Johnson Matthey Technol. Rev., 2020, 64, (4) (Continued) Isolation source — reservoir Petroleum Produced fluid Compost copepod Total extracts Bioreactor Soil Charleston Harbor sediment Crude oil contaminated soil Steel surfaces immerged in marine water from water Brackish fjord Nitinat anoxic at depth of Lake 50 m The hydrothermally heated sea floor Class Gamma proteobacteria Bacilli — Sphingobacteriia — — Alpha proteobacteria — Actinobacteria — — Gamma proteobacteria Phylum Proteobacteria Firmicutes Firmicutes Bacteroidetes Bacteroidetes — Proteobacteria Firmicutes Actinobacteria Bacteroidetes — Proteobacteria 100 89 98 99 92 98 100 96 100 99 96 96 Similarity, %

® JF891390 JN701188 HM041942 AB535716 FR871413 GQ259593 EU044497 EU194836 HQ133029 EF491430 FJ628289 AJ292528 BLAST accession number sp.

bacterium bacterium strain MCM strain match bacterium bacterium F1 ® sp. VKPM 2838 sp. sp. Trimyema-2 sp. Firmicutes Bacteroidetes Sphingomonas Firmicutes Bacteroidetes Closest BLAST Acinetobacter Aeribacillus pallidus B-886 Uncultured Chitinophagaceae Uncultured bacterium Uncultured bacterium Uncultured Uncultured Coriobacteriaceae enrichment culture clone B3113 Uncultured bacterium Uncultured bacterium Marinobacter Accession Numbers Assigned to these Sequences Accession Numbers Assigned to ® Accession number KF720883 KF720891 KF720893 KF720818 KF720830 KF720793 KF720802 KF720808 KF720798 KF720882 KF720887 KF720888 GenBank Phylogenetic Affiliations of Bacterial Sequences Retrieved from Produced Water Samples Based on 16S rRNA Analysis and Water Samples Based Produced from Sequences Retrieved of Bacterial Affiliations Phylogenetic DGGE band B1_20 B1_23 B1_24 B6_14 B6_16 B6_17 B6_19 B6_20 B6_26 B8_27 B8_29 B8_30 Table I Well no. B1 B6 B8

457 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15911723486216 Johnson Matthey Technol. Rev., 2020, 64, (4) (Continued) Isolation source mix pile Water/soil of samples from oil wells Petroleum- contaminated soil with saline-alkali phytoremediation of oil Wastewater refinery treatment plant sludge Activated from industrial wastewater treatment Rhizosphere Domestic toilet biofilm reservoir Water from Groundwater drinking water treatment plant Metal and hydrocarbon contaminated soil — Subsurface aquifer Area contaminated crude oil and by chemicals Produced fluid Class Gamma proteobacteria Alpha proteobacteria — Actinobacteria — — proteobacteria Beta — — Gamma proteobacteria Alpha proteobacteria Gamma proteobacteria — Phylum Proteobacteria Proteobacteria — Actinobacteria — — Proteobacteria — — Proteobacteria Proteobacteria Proteobacteria Firmicutes Similarity, % 94 99 98 100 96 99 98 99 99 93 94 100 97

® BLAST accession number HQ189755 JF421153 FN429535 HQ246163 GQ457025 FN401244 JN128637 HM921144 AY221598 HM173631 AY534887 FJ345693 HM041942 sp. sp. bacterium match sp. ParkerM sp. ® Caenispirillum Firmicutes sp. THE-49 sp. Closest BLAST Pseudomonas stutzeri Uncultured clone Ppss_Ma27 Uncultured bacterium Georgenia daeguensis Uncultured bacterium Uncultured bacterium Aquincola Uncultured bacterium Uncultured soil bacterium Thermithiobacillus Phenylobacterium lituiforme Pseudomonas stutzeri Uncultured Accession number KF720804 KF720814 KF720820 KF720825 KF720794 KF720810 KF720826 KF720831 KF720792 KF720796 KF720801 KF720807 KF720903 DGGE band B14_35 B14_37 B14_38 B14_39 B14_41 B23_52 B23_55 B23_56 B32_1 B32_2 B32_3 B32_4 B56_17 Well no. B14 B23 B32 B56

458 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15911723486216 Johnson Matthey Technol. Rev., 2020, 64, (4) (Continued) Isolation source Petroleum- contaminated soils saline-alkali Fissure water collected from a borehole — sludge Activated Microcosm experiment Batch reactor from Groundwater drinking water treatment plant Bulking activated sludge Enhanced biological phosphorus removal (EBPR) sludge seep Hydrocarbon sediment from Groundwater drinking water treatment plant Batch reactor EBPR sludge Class Gamma proteobacteria — Gamma proteobacteria — — — — — — Gamma proteobacteria — — — Phylum Proteobacteria — Proteobacteria — — — — — — Proteobacteria — — — Similarity, % 99 99 93 90 90 99 98 99 99 91 98 97 99

® BLAST accession number JF727663 JN030519 JN228329 JF497820 FM211087 AF314430 HM921144 HQ538639 AB231448 AF154088 HM921144 FJ623379 AB231448 match ® Closest BLAST Pseudomonas stutzeri Uncultured bacterium Pseudomonas stutzeri Uncultured bacterium Uncultured marine bacterium Uncultured bacterium PHOS- HE31 Uncultured bacterium Uncultured bacterium Uncultured bacterium seep Uncultured hydrocarbon bacterium BPC028 Uncultured bacterium Uncultured bacterium Uncultured bacterium Accession number KF720828 KF720832 KF720839 KF720852 KF720858 KF720855 KF720872 KF720877 KF720844 KF720884 KF720889 KF720894 KF720896 DGGE band GK8_79 GS6_1 GS6_2 GS6_4 GS6_5 M3_28 M3_31 M3_32 M3_34 K35_44 K35_46 K35_48 K35_49 Well no. GK8 GS6 M3 K35

3.1.3 Firmicutes

Sequences belonging to members of Firmicutes accounted for 8.3% of the bacteria in the produced water (Figure 3). DGGE band B1_23 was distantly related (89%) to Aeribacillus Isolation source Steel surfaces immersed in marine water from water Brackish fjord anoxic — Sea seep Hydrocarbon sediment from Groundwater drinking water treatment plant from Groundwater drinking water treatment plant Batch reactor EBPR sludge pallidus strain MCM B-886 (JN701188), isolated from petroleum reservoir (published only in GenBank®) (Table I). In addition, different strains of Aeribacillus pallidus (with sequence similarity values from 98% to 99.6%) were isolated previously from various Class — — Gamma proteobacteria — Gamma proteobacteria — — — — geothermal sites of Turkey (38). DGGE band B6_20 was distantly related (96%) to an uncultured Firmicutes bacterium, isolated from marine sediment in Charleston, South Carolina, USA (39). B56_17 and B1_24 were affiliated (97% Phylum Bacteroidetes — Proteobacteria — Proteobacteria — — — — and 98%, respectively) to an uncultured Firmicutes bacterium (Table I), detected in produced fluid from non-water-flooded high-temperature reservoir of the Niibori oilfield, Japan (40). Similarity, % 92 96 99 100 91 98 98 97 99

3.1.4 Actinobacteria

® The phylum Actinobacteria comprised 4.2% of the bacterial community recovered from BLAST accession number EF491430 FJ628289 JN228329 JF514265 AF154088 HM921144 HM921144 FJ623379 AB231448 the produced water (Figure 3). DGGE band B6_26 displayed 100% sequence similarity to Coriobacteriaceae bacterium enrichment

culture clone B3113 (HQ133029) isolated from crude oil contaminated soil of Shengli oil fields, China (41). The sequence match

® B14_39 was closely related (100% similarity) to an aerobic bacterial strain Bacteroidetes Georgenia daeguensis 2C6-43, isolated from an activated sludge sample collected from an industrial wastewater treatment plant in Daegu, South Korea (42). Closest BLAST Uncultured bacterium Uncultured bacterium Pseudomonas stutzeri Uncultured bacterium seep Uncultured hydrocarbon bacterium BPC028 Uncultured bacterium Uncultured bacterium Uncultured bacterium Uncultured bacterium Although little is known about the presence of G. daeguensis in oil associated environments, it was reported that different strains of G. daeguensis were

Accession number KF720806 KF720811 KF720823 KF720795 KF720886 KF720889 KF720889 KF720894 KF720896 isolated from hydrocarbon contaminated soil of an industrial zone and oil-saturated soil under laboratory conditions (43–45). DGGE band S4_67 S4_68 S4_70 S4_73 S15_60 Y18_70 Y30_66 Y30_68 Y30_69 4. Discussion

In order to increase our knowledge about

Well no. S4 S15 Y18 Y30 microbial diversity, culture-dependent

460 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15911723486216 Johnson Matthey Technol. Rev., 2020, 64, (4) and molecular-based approaches are used for Ppss_Ma27 were detected in our study. This describing the diversity of microbes. Molecular- result is consistent with the fact that the vast based approaches such as PCR-DGGE methodology, majority of microorganisms are uncultured and which is a useful tool for monitoring the genetic do not grow under laboratory conditions as stated diversity of complex microbial populations (26), by Lewis et al. (52). In order to isolate more provide valuable information about the microbial microbes, an appropriate identification laboratory community structure and dynamics in nature. For protocol should be followed. At this point, different these reasons, PCR-DGGE fingerprinting analysis strategies such as mimicking natural conditions via of environmental samples was used in this study. decreased nutrient, extended incubation times, The choice of appropriate primers for PCR the modification of isolating media formulations amplification is a crucial step to accurately and different incubation parameters (for example, characterise the microbial communities. In this temperature) were suggested for the cultivation study, primer pair (341F-GC/907R), targeting the of microorganisms (53). For instance, pollutant V3-V5 region of the 16S rRNA gene fragment, degrader Sphingomonas, which seemed to be was selected due to its suitability for DGGE previously uncultured by nutrient-rich methods, analysis of bacterial populations in environmental could be isolated from crude oil contaminated samples (26). This primer pair designed by Muyzer soil by using an in situ method that mimics the et al. (27, 28) has been used predominantly for original environment (54). In addition, culture- microbial community analysis (26). dependent investigation should also be supported The DNA yield obtained from produced water by molecular techniques. samples was very low. It is known that crude oil Based on the sequences, organisms related to samples contain low amounts of biomass which known mesophilic bacteria were predominant makes DNA isolation difficult to achieve (46). In in the produced water samples. In addition, this study, the permit included taking up to 500 ml some organisms related to thermophilic bacteria of oil/water mixture from each sampling point so (Aeribacillus pallidus, Marinobacter sp. Trimyema-2, that only ca. 25 ml of each produced water sample Phenylobacterium lituiforme and Thermithiobacillus could be obtained. In this scope, the low sample sp.) were also identified. Bacteria having different volumes of produced water separated from the oil/ metabolic capabilities (denitrifying, biodegrading water mixture may be a reason for the low amount and sulfur removing bacteria) were also detected. of DNA. It was reported in other studies that higher In addition, bacteria which may cause biocorrosion sample volumes (100–4000 ml) of produced water on steel surfaces were detected. were used for DNA isolation (35, 47–50). The The dominant bacterial phylum was the low DNA yield affected the efficiency of the PCR Proteobacteria. The members of this phylum were technique and for this reason, a two-step PCR was also frequently found in many other studies on applied to the produced water samples. Thus, a microbial diversity of oil field produced waters sufficient amount of PCR product for DGGE for the (55–58). Moreover, it was stated that Proteobacteria produced water samples could be obtained. are ubiquitous in oil reservoirs over all temperature Bacterial communities associated with the ranges (59). produced waters was analysed by the PCR-DGGE In this study, among the detected genera in approach. Although numerous bands were visible produced water samples that potentially contain on the DGGE gel, only dominant bands could be hydrocarbon degrading bacteria were Aeribacillus, excised and sequenced. Most of the sequences Acinetobacter, Sphingomonas, Marinobacter and retrieved from produced water samples were Phenylobacterium. It has been known for years related to unclassified bacteria. Different studies that the species belonging to these genera are on oil reservoir microbiota have also shown that oil capable of degrading hydrocarbons (6, 17, 60, 61). fields harbour new and still unidentified microbial In addition, G. daeguensis, a hydrocarbonoclastic species. For example, Lenchi et al. described bacterium, was detected in produced water sample microbial communities in production and injection with a 100% sequence similarity. G. daeguensis waters from the Algerian oil fields. In their study, has also been demonstrated as a potential they detected that a large number of unclassified microbe for bioremediation due to its hydrocarbon bacterial and archaeal sequences were found in degradation ability (44). Further investigations the water samples (51). Furthermore, uncultured are needed because our current knowledge of the bacteria such as uncultured Sphingomonas metabolic capability of G. daeguensis is limited. sp. and uncultured Caenispirillum sp. clone Moreover, sulfur-oxidising Thermithiobacillus sp.

461 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15911723486216 Johnson Matthey Technol. Rev., 2020, 64, (4) was also identified in produced water sample. oil or rock surfaces (59). In addition, biofilm may Sulfur-oxidising bacteria, which oxidise the sulfur form on the metal surfaces of the pipes in the oil- compounds produced by the activity of sulfate- producing wells (70). Oil microbiome studies focus reducing bacteria in oil reservoirs, may play a key mainly on the analysis of the water phase due to role in the oil industry because they can be utilised its easy sampling. However, it should be noted that to resolve processing problems such as reservoir the water phase itself contains only a minor portion souring (11). of the microbes found in the oil reservoir (59). On Pseudomonas was the dominant genus the other hand, the sampling of sessile microbes is detected among the produced water samples. likely to be more challenging (70). Pseudomonas stutzeri was the species identified in five produced water samples. P. stutzeri was 5. Conclusion previously isolated not only from formation water, produced from the petroleum wells in This study reported for the first time the bacterial Adıyaman (62), but also oil-contaminated soils in community composition of produced water Batman petroleum refinery, Turkey (63). These from Diyarbakır oil reservoirs as obtained by two areas are close to the Diyarbakır region from DGGE analysis of PCR-amplified 16S rRNA gene where the samples in this study were collected and fragments. DGGE analysis of produced water these findings show that P. stutzeri is distributed samples demonstrated that the majority of the widely in south-eastern Turkey. In other different bacterial sequences belonged to unclassified geographical areas, this species was also isolated bacteria, indicating that oil reservoirs harbour from oil-associated environments, such as oil still undescribed microbial species. Among the field production water (64), oil sludge (65) and classified bacteria, the members of Proteobacteria oil contaminated soil (66). However, although were more abundant. Pseudomonas was the P. stutzeri is often isolated from oil reservoirs, dominant genus detected in the produced water. the origin of P. stutzeri in oil reservoirs is a Although the members of Pseudomonas were debatable issue. Because oil reservoirs have known as exogenous organisms inoculated into oil low redox potentials and contain little oxygen, reservoirs, Pseudomonas stutzeri was found in five anaerobic microorganisms are considered as truly produced water samples. Bacteria having different indigenous to oil reservoirs (67). In this regard, metabolic capabilities (denitrifying, biodegrading it is believed that P. stutzeri, most of whose and sulfur removing bacteria) were also detected. strains are aerobes, is an exogenous organism It can be stated that the metabolic capacities of inoculated into oil reservoirs during the oil these bacteria make them potential candidates production processes. Even if strains of P. stutzeri for utilising in biodegradation, bioremediation, are introduced into oil reservoirs with injected the improvement of oil quality and oil recovery fluids, they should adapt to the physicochemical processes. The knowledge of the bacterial characteristics of the reservoir to survive. At this community composition in oil reservoirs of the point, it has been proposed that extreme reservoir Diyarbakır region obtained in this study will be conditions may act as special factors for the of great interest for both scientific research and evolution of P. stutzeri, thereby forming mutant applications in the oil industry. To build on the data strains (68). Furthermore, P. stutzeri, being found presented in this study, metagenomic analyses in a wide variety of habitats, is known for its should be performed to explore the undescribed diverse metabolism. Some strains of P. stutzeri microbes. are capable of denitrification, degradation of aromatic compounds and nitrogen fixation (32). Acknowledgements These metabolic features make P. stutzeri highly attractive for biotechnological processes, such as This work was supported by ‘Research Fund of reservoir souring control (69), microbial enhanced Istanbul University’ (Project number: 28699). oil recovery (64) and bioremediation of oil-polluted Tuğçe Tüccar was awarded an Erasmus LLP environments (65). Scholarship. Esra Ilhan-Sungur was awarded a In undisturbed oil reservoirs, microorganisms Post-doctoral Research Scholarship by the Scientific are found in different phases such as reservoir and Technological Research Council of Turkey fluid containing crude oil and formation water, and (TUBITAK-BIDEB). We thank the Turkish Petroleum rock surfaces. While planktonic microbes thrive in Corporation for permission to collect samples, and the water phase, sessile microbes may attach to Ender Taptık and Hasan Kaya for their assistance

462 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15911723486216 Johnson Matthey Technol. Rev., 2020, 64, (4) with the sample collection. We thank Ben Abbas Y. Kamagata and H. Tamaki, Int. J. Syst. Evol. for his technical assistance. We acknowledge Microbiol., 2017, 67, (10), 3982 Hakan Hosgormez for his helpful comments and 17. J. D. Van Hamme, A. Singh and O. P. Ward, suggestions. Microbiol. Mol. Biol. Rev., 2003, 67, (4), 503 18. S. Che and Y. Men, J. Ind. Microbiol. Biotechnol., References 2019, 46, (9–10), 1343 19. N. Jiménez, H. H. Richnow, C. Vogt, T. Treude and 1. C. Nikolova and T. Gutierrez, Front. Microbiol., M. Krüger, J. Mol. Microbiol. Biotechnol., 2016, 2020, 10, 2996 26, (1–3), 227 2. H. Dembicki, ‘Introduction’, in “Practical Petroleum 20. Y. Xue and G. Voordouw, Front. Microbiol., 2015, Geochemistry for Exploration and Production”, Ch. 6, 1387 1, Elsevier Inc, Amsterdam, The Netherlands, 21. D. Enning and J. Garrelfs, Appl. Environ. Microbiol., 2017, pp. 1–17 2014, 80, (4), 1226 3. J. G. Speight, “Handbook of Industrial Hydrocarbon 22. J. Telegdi, A. Shaban and L. Trif, ‘Corrosion Processes”, 2nd Edn., Elsevier Inc, Cambridge, Mechanisms: Current Knowledge, Gaps and Future USA, 2020, 786 pp Research: Microbiologically Influenced Corrosion 4. S. Romanow-Garcia and H. L. Hoffman, ‘Petroleum (MIC)’, in “Trends in Oil and Gas Corrosion Research and Its Products’, in “Kent and Riegel’s Handbook and Technologies: Production and Transmission”, of Industrial Chemistry and Biotechnology”, Vol. 1, ed. A. M. El-Sherik, Part 3, Ch. 8, Elsevier Ltd, 11th Edn., Springer Science and Business Media Duxford, UK, 2017, pp. 191–214 LLC, New York, USA, 2007, pp. 801–842 23. S. Kato, Microb. Biotechnol., 2016, 9, (2), 141 5. A. Satter and G. M. Iqbal, “Reservoir Engineering: 24. R. F. Wright, P. Lu, J. Devkota, F. Lu, M. Ziomek- The Fundamentals, Simulation, and Management Moroz and P. R. Ohodnicki, Sensors, 2019, of Conventional and Unconventional Recoveries”, 19, (18), 3964 Elsevier Inc, Waltham, USA, 2016, 472 pp 25. M. Safdel, M. A. Anbaz, A. Daryasafar and 6. S. J. Varjani, Bioresour. Technol., 2017, 223, 277 M. Jamialahmadi, Renew. Sustain. Energy Rev., 7. S. J. Varjani and E. Gnansounou, Bioresour. 2017, 74, 159 Technol., 2017, 245, (A), 1258 26. S. J. Green, M. B. Leigh and J. D. Neufeld, 8. N. Youssef, M. S. Elshahed and M. J. McInerney, ‘Denaturing Gradient Gel Electrophoresis (DGGE) ‘Microbial Processes in Oil Fields: Culprits, for Microbial Community Analysis’, in “Hydrocarbon Problems, and Opportunities’, in “Advances and Lipid Microbiology Protocols: Microbial in Applied Microbiology”, eds. A. I. Laskin, Quantitation, Community Profiling and Array S. Sariaslani and G. M. Gadd, Ch. 6, Elsevier Inc, Approaches”, eds. T. J. McGenity, K. N. Timmis San Diego, USA, 2009, pp. 141–251 and B. Nogales, Springer-Verlag, Berlin, Germany, 9. M. Magot, O. Basso, C. Tardy-Jacquenod and 2017, pp. 77–100 P. Caumette, Int. J. Syst. Evol. Microbiol., 2004, 27. G. Muyzer, E. C. de Waal and A. G. Uitterlinden, 54, (5), 1693 Appl. Environ. Microbiol., 1993, 59, (3), 695 10. A. Hussain, A. Hasan, A. Javid and J. I. Qazi, 3 28. G. Muyzer, A. Teske, C. O. Wirsen and Biotech, 2016, 6, (2), 119 H. W. Jannasch, Arch. Microbiol., 1995, 164, 165 11. H. Tian, P. Gao, Z. Chen, Y. Li, Y. Li, Y. Wang, 29. Basic Local Alignment Search Tool (BLAST®), J. Zhou, G. Li and T. Ma, Front. Microbiol., 2017, National Center for Biotechnology Information, US 8, 143 National Library of Medicine, Bethesda, MD, USA, 12. C. Berdugo-Clavijo and L. M. Gieg, Front. 18th June, 2020 Microbiol., 2014, 5, 197 30. T. L. Skovhus, N. B. Ramsing, C. Holmström, 13. S. Kh. Bidzhieva, D. Sh. Sokolova, T. P. Tourova S. Kjelleberg and I. Dahllöf, Appl. Environ. and T. N. Nazina, Microbiology, 2018, 87, (6), 757 Microbiol., 2004, 70, (4), 2373 14. J.-F. Liu, X.-B. Sun, G.-C. Yang, S. M. Mbadinga, 31. S. Kanso and B. K. C. Patel, Int. J. Syst. Evol. J.-D. Gu and B.-Z. Mu, Front. Microbiol., 2015, 6, Microbiol., 2004, 54, (6), 2141 236 32. J. Lalucat, A. Bennasar, R. Bosch, E. García-Valdes 15. C. C. Okoro and O. O. Amund, Petrol. Sci. Technol., and N. J. Palleroni, Microbiol. Mol. Biol. Rev., 2018, 36, (4), 293 2006, 70, (2), 510 16. S. Tamazawa, D. Mayumi, H. Mochimaru, 33. D. L. Gutnick and H. Bach, ‘Potential Application of S. Sakata, H. Maeda, T. Wakayama, M. Ikarashi, Acinetobacter in Biotechnology’, in “Acinetobacter:

Molecular Biology”, ed. U. Gerischer, Ch. 9, Caister 52. K. Lewis, S. Epstein, A. D’Onofrio and L. L. Ling, Academic Press, Caister, UK, 2008, pp. 231–264 J. Antibiot., 2010, 63, (8), 468 34. M. Baumgartner, K. O. Stetter and W. Foissner, 53. A. Bodor, N. Bounedjoum, G. E. Vincze, Á. Erdeiné J. Eukaryot. Microbiol., 2002, 49, (3), 227 Kis, K. Laczi, G. Bende, Á. Szilágyi, T. Kovács, 35. G. M. van der Kraan, J. Bruining, B. P. Lomans, K. Perei and G. Rákhely, Rev. Environ. Sci. Bio/ M. C. M. van Loosdrecht and G. Muyzer, FEMS Technol., 2020, 19, (1), 1 Microbiol. Ecol., 2010, 71, (3), 428 54. H. Zhao, Y. Zhang, X. Xiao, G. Li, Y. Zhao and 36. D. P. Kelly and A. P. Wood, Int. J. Syst. Evol. Y. Liang, Int. Biodeterior. Biodegrad., 2017, 117, Microbiol., 2000, 50, (2), 511 269 55. J. You, G. Wu, F. Ren, Q. Chang, B. Yu, Y. Xue 37. H. Dang, R. Chen, L. Wang, S. Shao, L. Dai, and B. Mu, Appl. Microbiol. Biotechnol., 2016, Y. Ye, L. Guo, G. Huang and M. G. Klotz, Environ. 100, (3), 1469 Microbiol., 2011, 13, (11), 3059 56. W.-F. Song, J.-W. Wang, Y.-C. Yan, L.-Y. An, 38. A. C. Cihan, B. Ozcan, N. Tekin and C. Cokmus, F. Zhang, L. Wang, Y. Xu, M.-Z. Tian, Y. Nie and X.- World J. Microbiol. Biotechnol., 2011, 27, (11), L. Wu, Int. Biodet. Biodeg., 2018, 132, 18 2683 57. X. Wang, X. Li, L. Yu, L. Huang, J. Xiu, W. Lin and 39. B. J. Mathis, C. W. Marshall, C. E. Milliken, Y. Zhang, Sci. Total Environ., 2019, 653, 872 R. S. Makkar, S. E. Creager and H. D. May, Appl. 58. T. Tüccar, E. Ilhan-Sungur, B. Abbas and G. Muyzer, Microbiol. Biotechnol., 2008, 78, (1), 147 Anaerobe, 2019, 59, 19 40. H. Kobayashi, K. Endo, S. Sakata, D. Mayumi, 59. M. Pannekens, L. Kroll, H. Müller, F. T. Mbow and H. Kawaguchi,, M. Ikarashi, Y. Miyagawa, H. Maeda R. U. Meckenstock, New Biotechnol., 2019, 49, 1 and K. Sato, J. Biosci. Bioeng., 2012, 113, (2), 204 60. J. Eberspächer and F. Lingens, ‘The Genus Phenylobacterium’, in “The Prokaryotes”, 41. L. Cheng, J. Rui, Q. Li, H. Zhang and Y. Lu, FEMS eds. M. Dworkin, S. Falkow, E. Rosenberg, Microbiol. Ecol., 2013, 83, (3), 757 K. H. Schleifer and E. Stackebrandt, Vol. 5, 3rd 42. S. G. Woo, Y. Cui, M.-S. Kang, L. Jin, K. K. Kim, Edn., Springer Science and Business Media LLC, S. T. Lee, M. Lee and J. Park, Int. J. Syst. Evol. New York, USA, 2006, pp. 250–256 Microbiol., 2012, 62, (7), 1703 61. G. T. Mehetre, S. G. Dastager and M. S. Dharne, 43. Y. Toptaş, M. Çelikdemir, C. Tuncer, Y. B. Şahin, Sci. Total Environ., 2019, 679, 52 P. A. Çelik, N. Burnak, A. Çabuk and V. Bütün, 62. Ç. Babaarslan, A. Tekeli and T. Mehmetoğlu, Turkish J. Biochem., 2016, 41, (5), 338 Energy Sources, 2003, 25, (7), 733 44. N. Ali, N. Dashti, M. Khanafer, H. Al-Awadhi and 63. T. Kaya, ‘Çeşitli endüstriyel atık maddelerde bazı S. Radwan, Sci. Rep., 2020, 10, 1116 mikroorganızmaların yüzey aktif özelliklerinin 45. S. S. Radwan, D. M. Al-Mailem and M. K. Kansour, incelenmesi’ [‘Research of Surface Active Sci. Rep., 2019, 9, 19508 Properties of Some Microorganisms in Various Industrial Wastes’], Masters Thesis, Biology 46. T. B. P. Oldenburg, S. R. Larter, J. J. Adams, Department, Gazi University, Ankara, Turkey, M. Clements, C. Hubert, A. K. Rowan, A. Brown, 28th April, 2008, 126 pp I. M. Head, A. A. Grigoriyan, G. Voordouw and 64. F. Zhao, C. Guo, Q. Cui, Q. Hao, J. Xiu, S. Han and M. Fustic, Anal. Chem., 2009, 81, (10), 4130 Y. Zhang, Carbohyd. Polym., 2018, 199, 375 47. V. J. Orphan, L. T. Taylor, D. Hafenbradl and 65. A. Afifi, H. Motamedi, B. Alizadeh and H. Leilavi, E. F. Delong, Appl. Environ. Microbiol., 2000, Environ. Experi. Biol., 2015, 13, (1), 13 66, (2), 700 66. Y. Anwar, A. A. El-Hanafy, J. S. M. Sabir, S. M. S. Al- 48. H. Li, S.-Z. Yang, B.-Z. Mu, Z.-F. Rong and J. Zhang, Garni, K. Al-Ghamdi, H. Almehdar and M. Waqas, FEMS Microbiol. Ecol., 2007, 60, (1), 74 Polycyc. Aromat. Comp., 2020, 40, (1), 135 49. J. Wang, T. Ma, L. Zhao, J. Lv, G. Li, F. Liang and 67. M. Magot, ‘Indigenous Microbial Communities R. Liu, World J. Microbiol. Biotechnol., 2008, in Oil Fields’, in “Petroleum Microbiology”, eds. 24, (9), 1981 B. Ollivier and M. Magot, ASM Press, Washington, 50. R. Kumaraswamy, S. Ebert, M. R. Gray, DC, USA, 2005, pp. 21–34 P. M. Fedorak and J. M. Foght, Appl. Microbiol. 68. F. Zhang, Y.-H. She, I. M. Banat, L.-J. Chai, L.- Biotechnol., 2010, 89, (6), 2027 Q. Huang, S.-J. Yi, Z.-L. Wang, H.-L. Dong and D.- 51. N. Lenchi, Ö. İnceoğlu, S. Kebbouche-Gana, J. Hou, MicrobiologyOpen, 2014, 3, (4), 446 M. L. Gana, M. Llirós, P. Servais and T. García- 69. F. Fan, B. Zhang, P. L. Morrill and T. Husain, RSC Armisen, PLoS One, 2013, 8, (6), e66588 Adv., 2018, 8, (47), 26596

70. X. Zhu and M. A. Al-Moniee, ‘Corrosion Inhibitors – Research and Technologies: Production and Advancements in Testing: Molecular Microbiology Transmission”, ed. A. M. El-Sherik, Part 4, Ch. Techniques’, in “Trends in Oil and Gas Corrosion 19, Elsevier Ltd, Duxford, UK, 2017, pp. 513–536

The Authors

Tuğçe Tüccar is a PhD candidate in Fundamental and Industrial Microbiology at Istanbul University, Turkey. She received her Bachelor’s degree in Biology from Middle East Technical University, Turkey, in 2009. She obtained her Master’s degree in Fundamental and Industrial Microbiology from Istanbul University, Turkey, in 2011. Her dissertation was on investigation of sulfate-reducing bacteria in petroleum samples. She was awarded an Erasmus LLP Scholarship and conducted her research work at Delft University of Technology, The Netherlands. Areas of interest are microbial ecology, microbial genetics, petroleum microbiology and microbial corrosion.

Esra Ilhan-Sungur is professor in the Biology Department at Istanbul University, Turkey, since 2018. A key focus of her research is microbiologically induced corrosion and its prevention. Further research interests lie in the area of anaerobic bacteria (especially sulfate-reducing bacteria), petroleum microbiology, microbial diversity and ecology, microbial genetics and biofilm. She was awarded a postdoctoral research scholarship by the Scientific and Technological Research Council of Turkey (TUBITAK-BIDEB) and worked as a guest researcher at Delft University of Technology.

Gerard Muyzer is Professor in Microbial Systems Ecology at the University of Amsterdam, The Netherlands. He is studying the structure, function and dynamics of microbial communities, their role in biogeochemical cycles and their application in biotechnological processes. For this he is using a systems biology approach in which he combines experimental work, the use of state-of-the-art omics techniques, and mathematical modelling. He is mainly focusing on the microbial sulfur cycle, and in particular on sulfur bacteria that are present in natural ecosystems (such as soda lakes, stratified lakes, rhizosphere of seagrasses) as well as man-made ecosystems, such as full-scale bioreactors removing toxic sulfur compounds from wastewater.

www.technology.matthey.com

The Biotechnological Potentials of Bacteria Isolated from Parsık Cave, Turkey Measuring the enzyme profiles, antibiotic resistance and antimicrobial activity in bacteria

Begüm Çandiroğlu** 1. Introduction Institute of Graduate Studies in Sciences, Istanbul University, Balabanaga Mah. Caves are dark environments with high humidity, Sehzadebasi Cd., 34134 Vezneciler, low nutrients, stable temperature and high Fatih-Istanbul, Turkey mineral diversity. They are natural geological formations constituting ecological niches for Nihal Doğruöz Güngör* microorganisms (1). Each cave is singular in Department of Biology, Faculty of Science, its physical, chemical, biological and ecological Istanbul University, Balabanaga Mah. factors. These conditions contribute to the Sehzadebasi Cd., 34134 Vezneciler, formation of unique microbial communities in Fatih-Istanbul, Turkey every cave. Moreover, caves contain some unique microorganisms which lead to rock weathering Email: *[email protected]; process and biomineralisation by carrying out **[email protected] various enzymatic reactions as a result of their metabolism. These microorganisms play an important and major role in the formation of cave Since cave ecosystems have extraordinary structures such as stalactites, stalagmites, cave environmental conditions, these ecosystems offer pearls and curtains (2–5). Studies have shown that opportunities for microbiological studies. In this cave isolates have biotechnological and industrial study, cultivable bacteria isolated from Parsık cave, applications such as microplastic degradation (6), Turkey, were investigated regarding enzyme profiles, biological treatment of metal contaminated soil antibiotic resistance and potential for production of and groundwater (7) and use in self-healing antimicrobial agents. The metabolic properties of concrete (8). 321 bacterial isolates were determined. The most The insufficient nutrient levels in caves stimulate produced enzyme by the isolates was found to be competition among microorganisms by forcing them tyrosine arylamidase. The enzymatic reactions of to develop survival strategies such as producing the bacteria showed that Parsık cave isolates have high amounts of exopolymeric substances, enzymes high aminopeptidase activity. The highest antibiotic and antimicrobial metabolites. Hence, caves could resistance frequency of the isolates was 38.6% be considered as incomparable environments for against ampicillin. While the isolates displayed the discovery of new antibiotics and production of variable inhibition rates against tested pathogenic novel enzymes (9–11). microorganisms, they showed the highest inhibition Since microorganisms have the capacity to against Candida albicans. The results show that the produce a high quantity of stable enzymes in a bacteria isolated from Parsık cave have potential short period of time, they become the preferred for further studies related to biotechnological source of industrial enzymes. Microbial enzymes are applications. The study findings contribute increased used in the clinical field for diagnosis, treatment, knowledge on metabolic peculiarities of bacteria biochemical tests and monitoring of various isolated from cave ecosystems. diseases. Furthermore, cave microbial enzymes are

466 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15923194903811 Johnson Matthey Technol. Rev., 2020, 64, (4) used in biotechnological and industrial fields such as which they can produce neutralising or detoxifying biodegradation, recycling of waste (12), purification products which act against microorganisms in the and dirt or waste-dissolving products. It is reported same environment. This explains the imperative that enzymes from microorganisms isolated from production of antibiotics in these bacteria. Since the cold cave or ocean environments offer economic resistance and antimicrobial biosynthesis genes are benefits and contribute to energy conservation due often linked and coregulated, antibiotic resistance to their activation at low temperatures (13, 14). in environmental bacteria remains a major indicator Apart from the importance of enzymes isolated of antibiotic production, as is the case of bacteria from cave microorganisms, it is interesting isolated from soil (19, 20). Therefore, it is important to investigate the potential of producing new to establish antibiotic resistance profiles as well as antimicrobial agents. Since the World Health the antibacterial properties of bacteria. Organization pointed out the need for new antibiotics This study has two main goals: because of increasing microbial resistance (15), • Detection of enzyme profiles of the isolates and studies in this field are multiplying and many cave determination of isolates that have potential isolates producing antimicrobial substances have uses in biotechnology been discovered. Cervimycin A, B, C and D from • Investigation of antimicrobial agents and Streptomyces tendae strain HKI 0179 isolated from antibiotic resistance of cave bacteria. Grotta dei Cervi in Italy (16), Xiakemycin A from Streptomyces sp. CC8-201 isolated from Chongqing 2. Experimental City karst soil in China (17), and Hypogeamicin A, B, C and D from Nonomuraea specus isolated from 2.1 Studying Area and Sampling Hardin’s cave system in Tennessee, USA (18) were the first produced and purified bioactive substances Parsık cave is located in Izmit-Aksığın village from microorganisms of caves situated in different (Global Positioning System (GPS) coordinates 40° geographical regions. 37’ 50.1060”N, 29° 57’ 56.5056”E), in the north- Bacteria in environments far away from human west of Turkey. It is a horizontal cave with a length influence are not expected to have antibiotic of 778 m and a depth of 166 m. There is an intense resistance. However, studies have shown that bacteria water inlet in Parsık cave throughout four seasons. isolated from such environments do have antibiotic Samples were taken from water, soil and surface resistance. Some bacteria have resistance genes by formations (‘moonmilk’) (Figure 1). The selected

Plan

MANYETIK North KUZEY 0 m 6 m 12 m

0 cm 5 cm 10 cm

B C

Fig. 1. Map of Parsık cave (red dots show the sampling areas) from the Anatolian Speleology Association, Turkey

467 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15923194903811 Johnson Matthey Technol. Rev., 2020, 64, (4) sampling zones are the sole area away from the isolation from soil samples was on SCA, ISP4, entrance area, trip and running water pathway. AIA-G, SEA and 1/2 tryptic soy agar (TSA) media, Although Parsık cave is not a show cave, it is open and that from water samples was on 1/2 TSA only. to cavers and researchers. Plates were incubated aerobically for a period of Surface formation samples were collected 5–30 days at 20°C (21). At the end of incubation, by sterile swabs under aseptic conditions and plates which contained between 30 and 300 cultivated on starch casein agar (SCA), inorganic colonies were considered for both soil and water salt-starch agar (ISP4), soil extract agar (SEA) and samples. Colonies which appeared different were Actinomycetes isolation agar (AIA-G) in duplicate selected for identification, then stored at –86°C for for each region. Once the plates reached the subsequent uses. laboratory, they were incubated aerobically for a period of 5–30 days at 20°C (21). All water and soil 2.5 Identification of Cave Isolates samples were taken in sterile sample containers. and Their Enzymatic Reactions

2.2 Physicochemical Measurements Cave isolates were identified through biochemical tests performed in the VITEK® 2 system (bioMérieux of Sampling Areas SA, France). One of the three formats of this system Humidity and temperature values of the sampling is the VITEK® 2 Compact 30 which focuses mainly areas were measured by a portable temperature/ on the industrial microbiology-testing environment. humidity meter. In addition, the temperature, Based on this industrial software, three reagent conductivity, amount of dissolved substances and cards of VITEK® 2 Compact 30, named Gram- pH values of the sampled water sources were negative fermenting and non-fermenting bacilli measured during sampling and recorded by a (GN), Gram-positive cocci and non-spore-forming HQ40D digital two channel multimeter (Hach Lange bacilli (GP) and Gram-positive spore-forming bacilli GmbH, Germany). (BCL), were used to characterise the isolated bacteria following the procedure and data given 2.3 Total (Live/Dead) Bacteria by the system manufacturers. Reagent cards are based on established biochemical methods Number and developed substrates (23). The results of The redox dye 5-cyano-2,3-ditolyl-tetrazolium biochemical reactions were interpreted to establish chloride (CTC) was used together with the enzymatic profiles of isolates. DNA-binding fluorescent dye’ 4 ,6-diamidino-2- phenylindole (DAPI) to determine the total number 2.6 Ability of Cave Bacteria to of bacterial cells and the viable count of bacteria Produce Antimicrobial Materials which actively respire. The concept is to distinguish between the metabolically active cells and the dead The ability of Bacilli or Actinobacteria to produce cells present in each of the water and soil samples. antimicrobial agents was tested on standard The experimentation procedure is the same as strains of fungi species of Candida albicans (ATCC® previously described by Güngör and Yurudu (22). 10231TM) and bacterial species of Escherichia coli (ATCC® 8739TM), Pseudomonas aeruginosa (ATCC® TM ® TM 2.4 Enumeration and Isolation of 9027 ), Staphylococcus aureus (ATCC 6538 ), Bacillus subtilis (ATCC® 6633TM), Staphylococcus Culturable Aerobic Heterotrophic epidermidis (ATCC® 12228TM), Klebsiella Bacteria pneumoniae (ATCC® 4352TM), Enterococcus 1 l of water samples were condensed by using hirae (ATCC® 10541TM), vancomycin-resistant polyamide filters of 0.22 µm pore size. Filters were Enterococcus faecalis (VRE) (ATCC® 51299TM) re-suspended in 20 ml of sterile physiological saline and methicillin-resistant Staphylococcus aureus water. 1 g of the soil samples was homogenised (MRSA) (ATCC® 33591TM). in 9 ml of sterile physiological saline water. All Bacterial suspensions containing 3 × 108 cells ml–1 samples were cultivated using the 10-fold serial of the selected isolates were prepared. 2.5 μl of each dilution method. Diluted samples were cultured suspension were incubated on Mueller Hinton Agar on tap water agar (TWA) and Reasoner’s 2A agar (MHA) plates at 20°C for 24 h. After incubation, all (R2A) for enumeration and isolation of bacteria media in which bacterial colonies were observed, from water and soil samples. In addition, bacterial were exposed to ultraviolet (UV) radiation in an

468 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15923194903811 Johnson Matthey Technol. Rev., 2020, 64, (4) open laminar flow cabinet. Therefore, the vitality These details highlight the differences in chemical of the bacteria was destroyed. 1.5 × 108 cells ml–1 environment that may exist within the cave areas. of 24 h fresh cultures of the standard strains were prepared. 100 μl of each suspension was mixed 3.2 Number of Determined Total with TSA medium at 45°C. Subsequently, it was (Live/Dead) Bacteria poured into the previously UV exposed plates, then incubated for 24 hours at 37°C after solidification. The highest vitality percentage of bacteria isolated in At the end of the incubation period, the growth of soil samples was found in samples from point B with the standard bacteria in the TSA was investigated 38.7%, whereas the highest vitality percentage in and the zone diameters were measured (24). the water samples was found in samples from point C with 26.3% (Table II). In cave environments, it 2.7 Susceptibility to Antibiotics is observed that bacteria can survive metabolically but cannot be cultured. This is because bacteria The sensitivity of 101 selected isolates to antibiotics enter a viable but nonculturable cell form under was examined by using the disc diffusion method extreme environmental conditions such as low or of Kirby-Bauer (21) in which 10 antibiotics were high temperature, nutrient deficiency, osmolarity used: piperacillin (100 µg), erythromycin (15 µg), and light. In addition, cave microorganisms obtain vancomycin (30 µg), ampicillin (10 µg), their energy from the cave atmosphere or the cave neomycin (10 µg), gentamycin (10 µg), surfaces to which they are attached (28, 29). chloramphenicol (30 µg), tetracycline (10 µg), rifampicin (30 µg) and ofloxacin (10 µg). The 3.3 Number and Classification of incubation conditions were 24 h at 20°C. Culturable Aerobic Heterotrophic Escherichia coli (ATCC® 8739TM), Pseudomonas Bacteria aeruginosa (ATCC® 9027TM) and Staphylococcus aureus (ATCC® 6538TM) were tested against the SCA, ISP4, SEA and AIA-G have been used same antibiotics as control microorganisms (25). especially in surface and soil samples to increase the probability of isolating bacteria belonging to 3. Results and Discussion phylum Actinobacteria, which have an extremely high potentials in terms of antimicrobial 3.1 Physicochemical Measurements production (30). TWA and R2A medium were used of Sampling Areas for both isolation and counting of other bacterial groups. Apart from these media, 1/2 TSA was Temperature, pH, conductivity and hardness values used for isolation of other bacterial groups from of water samples are shown in Table I. The air all samples. The cave environment in general is temperature of the sample areas A, B, C (Figure 1) oligotrophic and these media provide a similar was determined. The temperature of area A was environment to the culturable cave bacteria. The 9.8°C and that of B and C were determined as number of culturable aerobic heterotrophic bacteria 9.4°C. The moisture value was evaluated as 93% from water and soil samples obtained from R2A in all these areas. The Parsık cave resembles most and TWA media is given in Figure 2. cave systems with its high level of humidity and When the bacterial counts of water and soil stable air temperature (26, 27). It was determined samples in R2A and TWA media were examined, that the pH and hardness values of the waters at the highest bacterial numbers were found in R2A points B and C were higher than those at point A. medium. These results were evaluated statistically using the Kruskal-Wallis test. The p value was found to be 0.09 and no significant difference was Table I Physicochemical Measurements of found between the numbers of bacteria grown on Water Samples the R2A and TWA media. In a study conducted in Measured values Water sample areas 2014 (31), the efficiency of various media (SEA, TWA, SCA, TSA) was compared to their suitability A point B point C point for bacterial counting. Efficient results for both pH 8.2 9.8 9.8 isolation and counting were obtained in TWA. Hardness, ppt 0.107 0.145 0.145 A total of 372 bacteria were isolated from all Conductivity, mS 0.22 0.30 0.30 samples. VITEK® analysis was applied to only Temperature, °C 10 9.2 9.2 321 bacteria which had different characteristics in

Table II Number of Bacteria in Water and Soil Samples According to DAPI/CTC Method Samples Total number of signals, cm2 Vitality, % CTC Total SA 406,505,880 2,947,167,630 13.8 SB 135,501,960 643,634,310 21 SC 508,132,350 1,930,902,930 26.3 TA 1,084,015,680 5,318,451,930 20.3

TB 8,130,117,600 21,002,803,800 38.7

TC 6,097,588,200 20,664,048,900 29.5

and water isolates was Actinobacteria with 36% 7 and 35% respectively. Considering all the samples, 6 R2A at the class level, Actinobacteria was the most TWA 5 dominant with 33%, while Bacilli with 23% was detected as the second dominant class. It was

log(10) 4 log, 10

–1 demonstrated through previous studies that 3 Actinobacteria existed mainly in cave walls, soil, 2 sediment and on speleothem surfaces, which might CFU ml 1 have considerably contributed to the formation of 0 cave structures and the biomineralisation in the TA TB TC SA SB SC cave ecosystems (4–37). Actinobacteria as well Points of samples as Firmicutes are frequent among the microbial Fig. 2. Number of aerobic heterotrophic bacteria population inhabiting the caves. Compared to that can be cultured from water and soil samples the Proteobacteria group, Firmicutes are more (TA = soil sample A; TB = soil sample B; TC = resistant to stress caused by dehydration as soil sample C; SA = water sample A; SB = water well as restriction of nutrients (37). Contrary sample B; SC = water sample C) to our findings for Parsık cave, Proteobacteria are a dominant group in heterotrophic bacterial culture-based analyses. The results of the systematic communities in many caves (33, 34, 38–40). In classification of the bacteria were compiled by the current study, Proteobacteria were determined biochemical analysis using the VITEK® 2 Compact respectively as 10%, 21% and 17% in the surface, 30 automated system. Actinobacteria (33%) was water and soil samples. The dominant classes of this determined to be the dominant phylum in this phylum were found to be Gammaproteobacteria study while the other determined phyla were and Alphaproteobacteria with 9.2% and 6.4%, Firmicutes (25%) and Proteobacteria (16%). In respectively. In our previous study in Kadıini our previous work in Kadıini cave in Turkey, the cave, Alphaproteobacteria were detected at 2%, dominant phylum was Firmicutes (86%), followed while Gammaproteobacteria were at 9% (32). by Proteobacteria (12%) and Actinobacteria (2%) The phylum Proteobacteria, having a key role in respectively (32). In addition, in the study done biogeochemical cycles, and being abundant in by Tomova et al. (33), Proteobacteria (51.45%) samples from cave sediment, soil, dripping water were found to be the dominant phylum in the and cave surface, is a cosmopolitan bacterial samples taken from the Magura cave, Bulgaria, group (37). followed by Actinobacteria (43.68%) and Bacteroidetes (3.88%). Although the bacterial 3.4 Enzymatic Reactions of Parsık habitat of each cave is specific, Proteobacteria, Cave Bacteria Actinobacteria, Bacteroidetes and Firmicutes are the most identified groups in culture-based Enzymatic reactions of microorganisms give microbiological studies in caves (34–36). us ideas of their metabolic activities which are In our study, Firmicutes was the most common related to their environment. The biochemical phylum in soil samples with a rate of 33%, while tests of our isolates in the VITEK® system were the most common phylum determined in surface not only useful for bacterial identification but

470 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15923194903811 Johnson Matthey Technol. Rev., 2020, 64, (4) also to provide more information about nutrients positive and Gram-positive spore forming bacilli in Parsık cave. In addition, results of these tests revealed tyrosine-arylamidase activity. Tyrosine is were used to evaluate the potentials of the isolates a non-essential amino acid which is synthesised for biotechnological uses in terms of their enzyme through phenylalanine hydrolysis. It plays a production. 76 Gram-negative bacteria, 194 Gram- major role in most enzyme synthesis as reported positive and 51 Gram-positive spore forming by Kalkan and Altuğ (42), since it is the phosphate bacteria have been tested using the GN, GP and and sulfate receptor of protein kinase during protein BCL cards respectively in the VITEK® 2 compact synthesis. It is also used to reinforce the activity device, and results are given in Figure 3, Figure 4 of proteins as demonstrated in a study conducted and Figure 5 respectively. Most of the isolates in thrombin inhibitors showing that tyrosine displayed peptidase (arylamidase) while only sulfation could open a way for the development Gram-negative bacteria (less than 10%) showed of an anti-thrombotic drug (43). Hence, tyrosine lipolytic activity. In the study conducted in Gumki arylamidase has a valuable role in biotechnology cave, India, 75.5% of bacteria produced lipase, since it contributes to the liberation of the amino 47% were amylase producers and 24% produced acid tyrosine. protease (41). Another study screening cave Enzymes like leucine arylamidase have been bacteria for enzyme production found 40% lipase reported to be important in food processing and 87% protease producers (33). This variation in industries and the treatment of waste products (44, enzymatic profiles in cave bacteria reinforces the 45). The degradation of leucine and other amino idea that every cave is unique. acids results in volatile molecules responsible for The high activity of amino acids arylamidase the flavours of some foods like meat products as determined in our tested isolates indicates their reported by Papamanoli et al. (46) and Lee et potential for protein catabolism (42). The phyla al. (44). In addition, a study showed the roles Firmicutes (31%) and Actinobacteria (30.7%) of bacteria in conversion of paper mill sludge produced the highest amounts of arylamidases demonstrating the important contribution of amino identified among the tested isolates. 85.52%, acid peptidase with leucine arylamidase (45). In 65.97% and 82.35% of Gram-negative, Gram- our study, 81.95% and 88.23% of Gram-positive

Gr– bacteria 70

0 k a P p p A C C O AG LI yrA LT CI T PL E AT OF F LD SA UR E 5K G CM T GGT MN T T AD OD C PyrA ProA Glv IARL dCEL BXYL AP PA dTRE ELLM dT dGLU AGAL BGAL dMAL SUCT AGLU BGLU dSOR IL PHOS IM LT dMNE BAla IHIS a NAGA BNAG dMAN GGA A AG 0129R O

Fig. 3. Biochemical properties of Gram-negative bacteria. Tests for Gram-negative bacteria by VITEK® 2 Compact 30 micro device. See Glossary in main text for explanation of terms

Gr+ bacteria 180

160

140

120

100

0 r s L k Y B C C yrA AT PU L SA LAC SA UR E AM NA G T AlaA PyrA ProA dRIB LeuA dXYL BACI AspA AP PA dTRE dRAF dGAL AGAL BGAL dMAL AGLU OP TO CDEX dSOR IL PHOS ADH1 BGAR dMNE PIPL BGUR dMAN MBdG AMAN NOVO NC6. 5 PO LY 0129R BGUR O ADH2 POLYB_R Fig. 4. Biochemical properties of Gram-positive bacteria. Tests for Gram-positive bacteria by VITEK® 2 Compact 30 micro device. See Glossary in main text for explanation of terms

Gr(+) bacilli bacteria 60

0 E A Z N sA YG AG yrA B R PL E AT TT IN U IN O ES C PH C MT E OL D KA NA G T AlaA Ly PyrA ProA Gly dRIB LeuA AspA PheA BXYL AP PA IRHA dMIZ ELLM dGAL dT GL dGLU AGAL BGAL AGLU BGLU CDEX LY dMNE BNAG dMAN AMAN NC6.5 PV NC6.5% PO POLYB_R

Fig. 5. Biochemical properties of Gram-positive Bacilli bacteria. Tests for Gram-positive Bacilli bacteria by VITEK® 2 Compact 30 micro device. See Glossary in main text for explanation of terms

bacteria and Gram-positive bacilli showed leucine produce this enzyme could be used directly or arylamidase activity. This enzyme was the indirectly by using their enzymes in both composting second most produced enzyme, after the tyrosine of sludge and fermentation of food products such arylamidase, by our isolates. Bacteria which can as meat and dairy products.

VITEK® results have showed that some Parsık cave the highest inhibition rate was observed against isolates exhibit beta-galactosidase activity which is Candida albicans. Some of our cave isolates have the more expressed carbohydrate hydrolase in this been found to have inhibitory effects against S. study. Considering the whole of the tested isolates, aureus, S. epidermidis, VRE and P. aeruginosa. The most of the bacteria producing beta-galactosidase zone diameters of cave bacteria with antimicrobial belong to the Firmicutes phylum with 40.6%, while properties against tested microorganisms are only 10.9% of beta-galactosidase producers were shown in Table III. classified under the phylum Proteobacteria. In our study, the isolate which affected S. The main role of the beta-galactosidase enzyme is epidermidis belongs to the Bacilli class and those to convert lactose into monosaccharides. Glucose which inhibit VRE and S. aureus belong to the and galactose resulting from this reaction not only Actinobacteria class. Some studies have shown contribute to the development of the cell but can that bacteria with antimicrobial activity inhabiting also be used in dairy product processing (44, 47). karst caves are often from the Actinobacteria This enzyme is important since it solves the problem class (30, 31). However, cave bacteria belonging of human lactose intolerance. The hydrolysation of to phyla Proteobacteria, Firmicutes (especially lactose by this enzyme results in molecules like Bacilli class) and Bacteroides were determined to galactooligosaccharides which have health benefits have antimicrobial and bioactive substances. Thus, as prebiotics (47). Moreover, breakdown of some approximately 50% of the bacteria isolated from sugars like D-mannose, D-mannitol and D-glucose the Magura cave, Bulgaria were detected to inhibit by fermentation was reported, especially in Gram- the increase of P. aeruginosa (33). Cave bacteria negative bacteria. inhibiting MRSA and VRE clinical strains were Lipolytic activity was also observed in some of our determined in a study on Actinomycetes isolated isolated Gram-negative bacteria (less than 10%). from 19 different caves in Turkey (30). Certainly the Even if it was produced by a minimum number of bacteria belonging to the class Actinobacteria are isolates, the activity of lipase was fully expressed the best known in terms of antimicrobial material by bacteria belonging to the phylum Proteobacteria. synthesis, but the isolation of bacteria belonging This class of enzymes which is used in hydrolysation to the other classes is very important especially in of lipids could be important in bioremediation since karst environments. it could participate in oil degradation. Sharma et al. reported that microbial lipases are best for 3.6 Determination of Antibiotic biodiesel production (48). Since they can use all Resistance Profiles of Isolated types of free fatty acids and glycerides, they exhibit Bacteria a high activity, thermostability, alcohol resistance, less reaction time as well as less production Antibiotic resistant bacteria are widespread inhibition (48). Other enzymes were produced in several environments. In this study, resistance by some of the bacteria in Parsık caves. Further to 10 different antibiotics of 101 bacteria (76% studies should be carried out to clarify them and Gram-positive; 25% Gram-negative) selected from assess their biotechnological uses. the cave isolates was investigated. Isolates with a metabolic reaction rate of at least 95% similarities ® 3.5 Antimicrobial Agent Production to the data in the VITEK database were selected. When the antibiotic resistance profiles of the Capability isolates were examined, 7% of the bacteria Microorganisms with broad-spectrum bioactive belonging to the cave were resistant to all antibiotics. components, antifungal and antibacterial agents in The highest number of bacteria showed resistance cave-specific habitats are common in these extreme against ampicillin with a rate of 38.6%. In addition, environments (17). In our study, a total of 129 35.6% of the isolates showed resistance against cave bacteria were tested for their antimicrobial two or more antibiotics. effect against nine different standard bacterial Antibiotic resistance profiles of Gram-positive and strains and one fungal strain. Experiments have Gram-negative cave isolates are shown in Figure 6. shown that 10 of the selected bacteria (six from The lowest resistance was observed to rifampicin (9% Actinobacteria class, four from Bacilli class) have for Gram-positive and 8% for Gram-negative). antimicrobial effects against the standard strains. In parallel with our study, it was determined that Parsık cave isolates displayed variable inhibition all the Pajsarjeva jama, Slovenia, isolates were rates against the tested microorganisms and sensitive to rifampicin (49). Likewise, low levels

473 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15923194903811 Johnson Matthey Technol. Rev., 2020, 64, (4) Candida albicans – – – – 13 30 24 – 30 13 ND ND ND ND ND ND K. pneumoniae – – – – – – – – – – 22 – 22 19 15 35 P. aeruginosa – 16 13.5 – – – – – – – 16 – 11 5 8 23 VRE 9 – – – – – – – – – 31 18 – 14 28 30 MRSA – – – – – – – – – – 9 21 20 – 35 31 S. epidermidis – – – – – – – 15 – – 20 10 13 14 18 15 S. aureus – – – 13 – – – – – – 13 8 8 8 10 13 B. subtilis – – – – – – – – – – 19 13 16 16 17 21 E. faecalis – – – – – – – – – – 24 16 6.5 – 16 18 Resulting zone diameters, mm E. coli – – – – – – – – – – 11 – 11 10 7 15 Antimicrobial Agent Production AbilityAntimicrobial Agent Production Table III Isolates/classes of bacteria SA22/ Actinobacteria TA44/ Actinobacteria TA12/ Actinobacteria SA56/ Actinomycetes TB48/ Bacilli SB1/ Bacilli TA62/ Actinomycetes TB27/ Bacilli SC3/ Bacilli TB64/ Actinobacteria Antibiotics Piperacillin Vancomycin Gentamicin Tetracycline Rifampicin Ofloxacin ND = not determined

(a) Gram-positive isolates Cell wall synthesis 35 inhibitors 30S ribosome synthesis 30 inhibitors e 25 50S ribosome synthesis inhibitors 20 DNA/RNA synthesis inhibitors 15

10 showing resistanc

Percentage of bacteria 5

0 n n n l

Ampicillin Ofloxacin Piperacillin Neomycin Rifampicin Vancomyci Gentamyci Tetracycline Erythromyci Chloramphenico Antibiotics

(b) Gram-negative isolates 70

e 50

20 showing resistanc Percentage of bacteria 10

0 n n l

Ampicillin Ofloxacin Piperacillin Neomycin Rifampicin Vancomyci GentamycinTetracycline Erythromyci Chloramphenico Antibiotics

Fig. 6. Resistance levels of Parsık cave bacteria against various antibiotics which are grouped according to their mode of action: (a) Gram-positive isolates; (b) Gram-negative isolates

of resistance to ofloxacin, which is a DNA/RNA and Gram-negative bacteria. Contrary to our study, synthesis inhibitor like rifampicin, were observed in Lavoie et al. (50) showed that cave isolates were Parsık cave isolates (11% in Gram-positive and 12% highly resistant to gentamycin, neomycin and in Gram-negative). The resistance rate of Pajsarjeva chloramphenicol antibiotics (33–66% for Gram- jama isolates to erythromycin was 73.6% for Gram- negative bacteria and 61–83% for Gram-positive negative and 39% for Gram-positive bacteria. The bacteria). resistance of Parsik cave isolates to erythromycin was Furthermore, the lowest resistance to cell wall determined at lower levels of 20% and 21% for Gram- synthesis inhibitors was observed in piperacillin negative and Gram-positive bacteria respectively. for both Gram-positive (12%) and Gram- The levels of resistance to protein synthesis inhibitors negative (16%) bacteria. In the study conducted other than erythromycin (gentamycin, neomycin, by Lavoie et al. (50), the resistance to piperacillin, tetracycline and chloramphenicol) were determined compared to other antibiotic resistance, was found to range from 12% to 20% for both Gram-positive to be lower (average 37.5%).

In our study, considering the cell wall synthesis produced by these isolates. Apart from the potential inhibitors (vancomycin and ampicillin), Gram- of bacteria isolated from Parsık cave to produce negative bacteria were found to be more resistant enzymes and antimicrobial agents, it is planned than the Gram-positive ones. Similar to our study, to determine their potential use in biodegradation, Avguštin et al. (49) revealed that cave Gram- self-healing concrete and production of negative isolates showed higher resistance to antimicrobial drugs against phytopathogens and ampicillin. entomopathogens. According to VITEK® results, except ampicillin and vancomycin, Actinobacteria were determined 5. Conclusion to be the most resistant (47–70%) phylum to all tested antibiotics. The highest resistance The microorganisms attached to the specific to ampicillin and vancomycin was observed in environmental conditions of caves are important the phylum Proteobacteria. Like the microbial in terms of exploring their uses and specific diversity of caves, antibiotic resistance is also features. This study was the first microbiological variable. While the antibiotic resistance rates study in Parsık cave. It has been demonstrated were high, no isolate producing antimicrobial that enzymes such as arylamidases, carbohydrate agent was detected in the study conducted by hydrolases and lipases found in bacteria isolated Lavoie et al. (50). However, one of the antibiotic from Parsık cave can be important in industrial resistance hypotheses in caves is that there is a as well as clinical fields. In addition, some of our high rate of antibiotic resistance in the presence of isolates have shown antimicrobial activity and can microorganisms producing antimicrobial agents. contribute to the development of new antibiotics. Studies have shown that bacteria having antibiotic Antibiotic resistance profiles of Parsık cave isolates genes can also produce antimicrobial agents (51, should be clarified in further studies through 52). In our study, it was found that 50% of the studies of their genes. isolates producing antimicrobial agents were resistant to at least two antibiotics. Therefore, Acknowledgements study of bacterial antibiotic resistance may contribute to the development of new antibiotics. The author would like to thank the Anatolian To clarify this issue, studies in this issue should be Speleology Association, Turkey, and Nahdhoit continued. Ahamada Rachid, Department of Fundamental and Industrial Microbiology of Istanbul University, for 4. Further Work their contributions. The author also thanks Istanbul University Scientific Project Unit, Turkey, (BAP In our future studies, we are planning to purify Project No FYL-2016-20759 and FHZ-2017‑26457) and use the enzymes and antimicrobial substances for financial support.

Glossary

Term Definition AspA L-aspartate arylamidase 5KG 5-keto-D-gluconate BACI bacitracin resistance ADH1 arginine dihydrolase 1 BAlap beta-alanine arylamidase pNA ADH2s arginine dihydrolase 2 BCL Gram-positive spore-forming bacilli ADO Adonitol BGAL beta-galactosidase AGAL alpha-galactosidase BGAR beta-galactopyranosidase AGLTp glutamyl arylamidase pNA BGLU beta-glucosidase BGUR/ AGLU alpha-glucosidase beta-glucuronidase BGURr AIA-G Actinomycetes isolation agar BNAG beta-N-acetyl-glucosaminidase AlaA alanine arylamidase BXYL beta-xylosidase AMAN alpha-mannosidase CDEX cyclodextrin AMY D-amygdalin CIT citrate (sodium) APPA Ala-Phe-Pro-arylamidase

CMT coumarate LIP lipase 5-cyano-2,3-ditolyl-tetrazolium LysA L-lysine-arylamidase CTC chloride MBdG methyl-beta-D-glucopyranoside DAPI 4’,6-diamidino-2-phenylindole MHA Mueller Hinton Agar dCEL D-cellobiose MNT malonate dGAL D-galactose methicillin-resistant Staphylococcus MRSA dGLU D-glucose aureus dMAL D-maltose MTE maltotriose dMAN D-mannitol NAG N-acetyl-D-glucosamine NAGA beta-N-acetyl-galactosaminidase dMLZ D-melezitose NC6.5 growth in 6.5% NaCl dMNE D-mannose NOVO novobiocin resistance dRAF D-raffinose O129R O/129 resistance (comp. vibrio.) dRIB D-ribose ODC ornithine decarboxylase dSOR D-sorbitol OFF fermentation/glucose dTAG D-tagatose OLD oleandomycin resistance dTRE D-trehalose OPTO optochin resistance dXYL D-xylose PHC phosphoryl choline ELLM Ellman PheA phenylalanine arylamidase ESC esculin hydrolysis PHOS phosphatase GGAA Glu-Gly-Arg-arylamidase PIPLC phosphatidylinositol phospholipase C GGT gamma-glutamyl-transferase PLE PalatinoseTM GlyA glycine arylamidase POLYB_R polymyxin B resistance GLYG glycogen ProA L-proline arylamidase Gram-negative fermenting and non- GN PUL pullulan fermenting bacilli PVATE pyruvate Gram-positive cocci and non-spore- GP forming bacilli PyrA L-pyrrolidonyl-arylamidase GPS Global Positioning System R2A Reasoner’s 2A agar IARL L-arabitol SAC saccharose/sucrose SAL salicin IHISa L-histidine assimilation SCA starch casein agar ILATk L-lactate alkalinisation SEA soil extract agar IMLTa L-malate assimilation SUCT succinate alkalinisation INO myo-inositol TSA 1/2 tryptic soy agar INU inulin TTZ tetrazolium red IRHA L-rhamnose TWA tap water agar ISP4 inorganic salt-starch agar TyrA tyrosine arylamidase KAN kanamycin resistance URE urease LAC lactose UV ultraviolet LDC lysine decarboxylase vancomycin-resistant Enterococcus VRE LeuA leucine arylamidase faecalis

References 3. L. C. Kelly, C. S. Cockell, A. Herrera-Belaroussi, Y. Piceno, G. Andersen, T. DeSantis, E. Brodie, 1. C. Schabereiter-Gurtner, C. Saiz-Jimenez, G. Piñar, T. Thorsteinsson, V. Marteinsson, F. Poly and W. Lubitz and S. Rölleke, FEMS Microbiol. Ecol., X. LeRoux, Microb. Ecol., 2011, 62, (1), 69 2004, 47, (2), 235 4. S. Cuezva, A. Fernandez-Cortes, E. Porca, L. Pašić, 2. H. A. Barton, J. Caves Karst Stud., 2006, 68, (2), V. Jurado, M. Hernandez-Marine, P. Serrano-Ortiz, 43 B. Hermosin, J. C. Cañaveras, S. Sanchez-Moral

and C. Saiz-Jimenez, FEMS Microbiol. Ecol., 2012, 19. V. M. D’Costa, C. E. King, L. Kalan, M. Morar, 81, (1), 281 W. W. L. Sung, C. Schwartz, D. Froese, G. Zazula, 5. A. Rusznyák, D. M. Akob, S. Nietzsche, F. Calmels, R. Debruyne, G. B. Golding, H. N. Poinar K. Eusterhues, K. U. Totsche, T. R. Neu, T. Frosch, and G. D. Wright, Nature, 2011, 477, (7365), 457 J. Popp, R. Keiner, J. Geletneky, L. Katzschmann, 20. J. R. Nodwell, J. Bacteriol., 2007, 189, (10), 3683 E.-D. Schulze and K. Küsel, Appl. Environ. 21. S. Massa, M. Caruso, F. Trovatelli and M. Tosques, Microbiol., 2012, 78, (4), 1157 World J. Microbiol. Biotechnol., 1998, 14, (5), 727 6. S. U. U. Jamil, S. Zada, I. Khan, W. Sajjad, 22. N. Doğruöz Güngör and N. Ö. Şanlı Yürüdü, M. Rafiq, A. A. Shah and F. Hasan, J. Caves Karst “The Battle Against Microbial Pathogens: Basic Stud., 2017, 79, (1), 73 Science, Technological Advances and Educational 7. N. K. Dhami, M. E. C. Quirin and A. Mukherjee, Programs”, ed. A. Méndez-Vilas, Vol. 5, Formatex, Ecol. Eng., 2017, 103, (A), 106 Badajoz, Spain, 2015, pp 923–929 8. A. Nugroho, A. Sumarno, L. N. Ngeljaratan, 23. D. H. Pincus, ‘Microbial Identification Using the D. Zulfiana, N. P. R. A. Krishanti, T. Triastutil and bioMérieux VITEK® 2 System’, in “Encyclopedia E. Widodo, J. Kim. Terap. Indones., 2019, 21, (1), of Rapid Microbiological Methods”, ed. M. J. Miller, 7 Parenteral Drug Association, Bethesda, USA, 9. A. A. Elmanama, M. T. Alhour, J. Adv. Sci. Eng. 2006, pp 1–32 Res., 2013, 3, (4), 388 24. K. B. Ritchie, M. Schwarz, J. Mueller, V. A. Lapacek, 10. S. Krishnapriya, D. L. Venkatesh Babu, G. Prince D. Merselis, C. J. Walsh and C. A. Luer, Front. Arulraj, Microbiol Res., 2015, 174, 48 Microbiol., 2017, 8, 1050 11. A. I. Omoregie, ‘Isolation, Identification and 25. M. J. Ferraro, M. A. Wikler, W. A. Craig, M. N. Dudley, Characterisation of Urease-Producing Bacteria from G. M. Eliopoulos, D. W. Hecht, J. Hindler, L. Barth Limestone Caves of Sarawak’, in ‘Characterization Reller, A. T. Sheldon, J. M. Swenson, F. C. Tenover, of Ureolytic Bacteria Isolated from Limestone R. T. Testa and M. P. Weinstein, ‘Performance Caves of Sarawak and Evaluation of their Efficiency Standards for Antimicrobial Disk Susceptibility in Biocementation’, Master of Science (Research) Tests: Approved Standard’, 8th Edn., M2-A8, Thesis, Chapter 2, Faculty of Engineering, Vol. 23, No. 1, National Committee for Clinical Computing and Science, Swinburne University of Laboratory Standards (NCCLS), Wayne, USA, Technology, Melbourne, Australia, 2016, pp 41–92 2003, 63 pp 12. S. Mondal and D. Palit, ‘Effective Role of 26. S. Leuko, K. Koskinen, L. Sanna, I. M. D’Angeli, Microorganism in Waste Management and J. De Waele, P. Marcia, C. Moissl-Eichinger and Environmental Sustainability’, in “Sustainable P. Rettberg, PLoS One, 2017, 12, (7), e0180700 Agriculture, Forest and Environmental 27. K. H. Lavoie, A. S. Winter, K. J. H. Read, Management”, Eds. M. K. Jhariya, A. Banerjee, E. M. Hughes, M. N. Spilde and D. E. Northup, PloS R. S. Meena and D. K. Yadav, Springer Nature One, 2017, 12, (2), e0169339 Singapore Pte Ltd, Singapore, 2019, pp 485–515 28. H. A. Barton, M. R. Taylor and N. R. Pace, 13. C. Gerday, M. Aittaleb, M. Bentahir, J. P. Chessa, Geomicrobiol. J., 2004, 21, (1), 11 P. Claverie, T. Collins, S. D’Amico, J. Dumont, 29. M. K. Chelius and J. C. Moore, Geomicrobiol. J., G. Garsoux, D. Georlette, A. Hoyoux, T. Lonhienne, 2004, 21, (2), 123 M.-A. Meuwis and G. Feller, Trends Biotechnol., 30. S. Yücel and M. Yamaç, Pak. J. Pharm. Sci., 2010, 2000, 18, (3), 103 23, (1), 1 14. M. S. Cabeza, F. L. Baca, E. M. Puntes, F. Loto, 31. B. H. Velikonja, R. Tkavc and L. Pašić, Int. M. D. Baigorí and V. I. Morata, Food Technol. J. Speleol., 2014, 43, (1), 45 Biotech., 2011, 49, (2), 187 15. Z. Kmietowicz, Brit. Med. J., 2017, 358, j4430 32. N. Doğruöz-Güngör, B. Çandıroğlu, G. Altuğ, J. Caves Karst Stud., 2020, 82, (2), 106 16. K. Herold, F. A. Gollmick, I. Groth, M. Roth, K.- D. Menzel, U. Möllmann, U. Gräfe and C. Hertweck, 33. I. Tomova, I. Lazarkevich, A. Tomova, Chem. Eur. J., 2005, 11, (19), 5523 M. Kambourova and E. Vasileva-Tonkova, Int. J. Speleol., 2013, 42, (1), 65 17. Z. Jiang, L. Guo, C. Chen, S. Liu, L. Zhang, S. Dai, Q. He, X. You, X. Hu, L. Tuo, W. Jiang and C. Sun, 34. J. P. Zhou, Y. Q. Gu, C. S. Zou and M. H. Mo, J. Antibiot., 2015, 68, (12), 771 J. Microbiol., 2007, 45, (2), 105 18. D. K. Derewacz, C. R. McNees, G. Scalmani, 35. H. A. Barton and V. Jurado, Microbe, 2007, 2, 132 C. L. Covington, G. Shanmugam, L. J. Marnett, 36. V. Jurado, E. Porca, S. Cuezva, A. Fernandez- P. L. Polavarapu and B. O. Bachmann, J. Nat. Cortes, S. Sanches-Moral and C. Saiz-Jimenez, Prod., 2014, 77, (8), 1759 Sci. Total Environ., 2010, 408, (17), 3632

37. K. Tomczyk-Żak and U. Zielenkiewicz, Geomicrobiol 46. E. Papamanoli, N. Tzanetakis, E. Litopoulou- J., 2016, 33, (1), 20 Tzanetaki and P. Kotzekidou, Meat Sci., 2003, 38. M. Yasir, Braz. J. Microbiol., 2018, 49, (2), 248 65, (2), 859 39. V. Ivanova, I. Tomova, A. Kamburov, A. Tomova, 47. J. R. Xavier, K. V. Ramana, R. K. Sharma, J. Food Biochem., 2018, 42, (5), e12564 E. Vasileva-Tonkova, M. Kambourova, J. Caves Karst Stud., 2013, 75, (3), 218 48. A. Sharma, Shadiya, T. Sharma, R. Kumar, K. Meena, S. S. Kanwar, ‘Biodiesel and the Potential 40. C. Schabereiter-Gurtner, C. Saiz-Jimenez, G. Piñar, Role of Microbial Lipases in Its Production’, W. Lubitz and S. Rölleke, FEMS Microbiol. Lett., in “Microbial Technology for the Welfare of 2002, 211, 7 Society: Microorganisms for Sustainability”, ed. 41. R. Rautela, S. Rawat, R. Rawat, P. Verma, P. K. Arora, Vol. 17, Springer Nature Singapore A. B. Bhatt, Environ. Conserv. J., 2017, 18, (3), Pte Ltd, Singapore, 2019, pp. 83–89 115 49. J. A. Avguštin, P. Petrič and L. Pašić, Int. J. Speleol., 42. S. Kalkan, and G. Altuğ, Environ. Monit. Assess., 2019, 48, (3), 295 2020, 192, (6), 356 50. K. Lavoie, T. Ruhumbika, A. Bawa, A. Whitney and 43. R. E. Thompson, X. Liu, J. Ripoll-Rozada, J. De Ondarza, Diversity, 2017, 9, (4), 42 N. Alonso-García, B. L. Parker, P. J. B. Pereira and 51. M. K. Gibson, B. Wang, S. Ahmadi, C.- R. J. Payne, Nature Chem., 2017, 9, (9), 909 A. D. Burnham, P. I. Tarr, B. B. Warner and 44. N. K. Lee, J. Y. Hong, S. H. Yi, S. P. Hong, J. E. Lee G. Dantas, Nature Microbiol., 2016, 1, (4), 16024 and H. D. Paik, J. Funct. Foods, 2019, 58, 324 52. A. C. Pawlowski, W. Wang, K. Koteva, H. A. Barton, 45. R. K. Ganguly, and S. K. Chakraborty, J. Environ. A. G. McArthur and G. D. Wright, Nature Commun., Health Sci. Eng., 2018, 16, (2), 205 2016, 7, 13803

The Authors

Nihal Doğruöz Güngör is an Associate Professor in the Department of Fundamental and Industrial Microbiology at Istanbul University, Turkey. She obtained her doctorate at Istanbul University in 2008, focusing on microbiological corrosion of copper. Her research interests include cave microbiology, antimicrobial activities of bacteria, microbial corrosion and biotechnology.

Begüm Çandıroğlu is a biologist from Istanbul University. In 2014 she obtained her MSc in microbiology in the Department of Fundamental and Industrial Microbiology at the same university. She studied cave microbiology and worked with cave bacteria that live in soil, water and cave walls. She is interested in biotechnology, enzymatic reactions of bacteria and antibiotic resistance.

www.technology.matthey.com

Antibacterial Potential of Six Lichen Species against Enterococcus durans from Leather Industry Evaluation of acetone extracts obtained from several lichen species as alternative natural antibacterial agents

Didem Berber* effect. The most successful lichen extract was Department of Biology, Faculty of Arts and determined to be Usnea sp. at the concentrations Sciences, Marmara University, Istanbul, Turkey; of 240 µg ml–1, 120 µg ml–1, 60 µg ml–1, 30 µg ml–1 Gastronomy and Culinary Arts Department, and 15 µg ml–1. In conclusion, lichen extracts seem Faculty of Fine Arts, Maltepe University, to have potential antibacterial efficacies against Marmara Eğitim Köyü, Istanbul, Turkey E. durans.

İpek Türkmenoğlu 1. Introduction Biology Department, Institute of Pure and Applied Sciences, Marmara University, Istanbul, The leather industry produces and exports high- Turkey quality products with high added value to the world market. However, several bacterial problems during Nüzhet Cenk Sesal leather-making processes are reflected in finished Department of Biology, Faculty of Arts and products and lead to economic losses. After the Sciences, Marmara University, Istanbul, Turkey flaying process in slaughterhouses, microflora on hide or skin surfaces change due to bacterial *Email: [email protected] contamination originating from faeces, air, dust or the animal skin itself and some bacteria easily colonise (1–4). Antibacterial resistant bacteria are a significant The soaking process is the first tannery operation problem in the hide or skin soaking process due that recovers water loss during raw hide or skin to their destructive properties on finished leather. curing applications. There are some criteria to be Lichens may be a solution to overcome this taken into consideration during the soaking process resistance problem. Enterococcus durans (99.86%) of raw hides or skins. Especially prolonged soaking was isolated from soak liquor samples. For screening provides a convenient milieu for bacterial activity and of possible antibacterial effects of lichen acetone damage to hides or skins may occur. Due to reduced extracts, six lichen species (Hypogymnia tubulosa, salt content and high protein and lipid constituents, H. physodes, Evernia divaricata, Pseudevernia hides or skins become defenceless against bacterial furfuracea, Parmelia sulcata and Usnea sp.) were attacks in the soaking process (5–8). It has been examined by nine-fold dilution against E. durans. reported that the number of bacterial populations H. tubulosa, H. physodes and E. divaricata extracts in soak liquors may be up to 105 colony forming showed antibacterial effects at the concentrations unit (CFU) ml–1 (5). But in a previous study, it was of 240 µg ml–1, 120 µg ml–1 and 60 µg ml–1 demonstrated that total bacterial numbers were whereas the extracts of P. furfuracea had an considerably higher than 105 CFU ml–1 in soak antibacterial effect at 240 µg ml–1 and 120 µg ml–1. liquor samples (9). The adverse effects of the On the other hand, P. sulcata had no antibacterial soaking process on the hide quality originate from

480 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15942856494595 Johnson Matthey Technol. Rev., 2020, 64, (4) degradative enzymatic properties of bacteria such from numerous lichen extracts have been reported as protease and lipase activities. These enzymatic to have biological activities such as antibacterial activities can irreversibly affect the structure of activity against Gram-positive and Gram-negative hide or skin substances that cannot be fixed at bacteria (24–27). It has been reported that the subsequent stages of hide processing (10). approximately 2000 of the 20,000 lichen species High numbers of bacteria with protease and lipase in the world are in Turkish lichen mycota. There activities cause unwanted defects such as hair-slip, are many studies evaluating the bioactivities of putrefaction, grain peeling, loose grain, holes on lichen species in Turkey against different bacterial the hides or skins and light stains on the suede species (25–27). In the previous study, the acetone surface (1, 3, 11–15). extracts of H. physodes, E. divaricata, P. furfuracea Antibiotics are used in various industries as and Usnea sp. at different concentrations were well as in the treatment of diseases. The World tested on some Bacillus species which were isolated Health Organization declared that antimicrobial from soak liquor samples. These extracts were resistance in most countries and industrial detected to have potential antibacterial effects (28). sectors has increased dramatically (16, 17). From this point, lichen species may have potential The emergence of antibiotic-resistant bacteria antibacterial efficacies against various antibacterial- due to improperly used antibiotics in humans, resistant bacterial strains in the soaking process animals and agriculture has been reported in the which cannot be exterminated by antimicrobial literature (17). In the leather industry, to control agents. Therefore, the antibacterial effects of bacterial numbers and their degradative properties acetone extracts of lichen species H. tubulosa, on hides or skins, various antibacterial agents are H. physodes, E. divaricata, P. furfuracea, P. sulcata utilised during the soaking process of beam house and Usnea sp. against Isolate 1 (E. durans), which operations. The normal microflora in animals has protease and lipase acitivities, was evaluated comprises many harmless bacteria but any of in the present study. them may become resistant to commonly utilised antibacterial agents due to intrinsic or acquired 2. Materials and Methods resistance (17, 18). The resistant bacteria may survive despite bactericides and may transfer their 2.1 Sample Collection resistance properties to others through horizontal gene transfer (5, 9, 18). Bactericides may remain Three soak liquor samples were collected from ineffective against proteolytic and lipolytic bacteria Istanbul Leather Organized Industrial Zone, Tuzla, in soak liquors because of high organic content Istanbul, Turkey. These samples were immediately in soak liquors (9, 19). The existence of many placed into sterile sample bags and carried on ice non-halophilic bacteria was demonstrated in the during transportation. Direct and serial dilutions presence of an antimicrobial agent at twofold were spread onto nutrient agar plates. The increased concentration (0.8 g l–1) (19). This morphologically different colony was picked up finding emphasises the antibacterial resistance to obtain the pure culture of the isolate and was of bacteria in the soaking process. More recently, numbered as Isolate 1. it was reported that antimicrobial agents used in the soaking process could not control multidrug- 2.2 Biochemical and Molecular resistant Enterobacteriaceae from soaked Analyses sheepskins and cattle hides treated with an antibacterial agent (20). Gram staining, catalase, oxidase, lipase and Over the past decades, it has been suggested protease activities were examined. Protease that alternative compounds from natural resources activity of Isolate 1 was examined on gelatin agar may overcome the antimicrobial resistance of many medium containing 2% gelatin (w/v). The agar bacteria. Previously, the potential of lichen derived plates were flooded with Frazier solution following extracts from P. furfuracea (L.) Zopf was reported 24 h incubation. Clear zones around the colonies in the leather industry (21). Lichens are symbiotic were evaluated as positive for protease activity. organisms between a fungus and one or more Lipase activity was tested on Tween® 80 agar algae or cyanobacteria. They synthesise unique medium containing 1% (w/v) Tween® 80. After secondary metabolites that cannot be synthesised incubation, opaque zones around the colonies were by higher plants (22, 23). Secondary metabolites accepted as evidence of lipase activity (29, 30).

Genomic DNA of Isolate 1 which was determined 2.5 Determination of Antibacterial to have protease and lipase activities were Efficacies of Lichen Samples extracted by phenol/chloroform extraction and ethanol precipitation. DNA isolation was The test isolate was grown on Tryptic soy agar confirmed by agarose gel electrophoresis. DNA media at 37°C for 24 h. The tests were performed samples were stored at −20°C until use. The in 96-well CELLSTAR®, F-bottom microplates with 16S rRNA gene was amplified by polymerase lid (Greiner Bio-One GmbH, Austria). Tryptic soy chain reaction (PCR) with the universal bacterial broth was added to each well and nine-fold serial primers 27F (5-AGAGTTTGATCMTGGCTCAG) dilutions of the acetone extracts of H. tubulosa, and 1492R (5-TACCTTGTTACGACTT). Negative H. physodes, E. divaricata, P. furfuracea, P. sulcata control was included in PCR amplifications. and Usnea sp. were made. Final concentrations of PCR amplification was carried out by an initial all lichen extracts were 240 µg ml–1, 120 µg ml– 1, denaturation at 95°C for 4 min, followed by 30 60 µg ml–1, 30 µg ml–1, 15 µg ml–1, 7.5 µg ml– 1, cycles at 95°C for 1 min, 57°C for 1 min and 73°C 3.75 µg ml–1, 1.9 µg ml–1 and 0.9 µg ml–1. for 1 min. The reactions were finished by a final Overnight culture of the isolate was added to extension at 73°C for 7 min. The PCR products were obtain a total volume of 100 µl with an optical also monitored by agarose gel electrophoresis. density (OD) 600 nm of 0.01. The experiments These products were purified by GeneJETTM Gel included untreated and blank controls. The tests Extraction Kit (Thermo ScientificTM, Thermo were performed in three replicates. Bacterial Fisher Scientific, USA). These purified samples growth ratios at an OD 600 nm were measured were analysed by Medsantek Ltd Co, Istanbul, using CytationTM 3 Multi-Mode microplate reader Turkey. The 16S rRNA sequence contigs were (BioTek Instruments Inc, USA). generated by the software ChromasPro version 2.1.8 (Technelysium Pty Ltd, Australia). Then, 3. Results and Discussion consensus sequences were exported in FASTA format for each sample for data analysis. These In the present study, Isolate 1, which was obtained sequences were compared with sequences in the from soak liquor samples collected from different National Center for Biotechnology Information tanneries in Istanbul Leather Organized Industrial (NCBI) using the Basic Local Alignment Search Zone, Turkey, was identified by biochemical and Tool (BLAST®) search program. molecular techniques. To our knowledge, there is no study on the antibacterial efficacies of lichen 2.3 Lichen Samples extracts against E. durans from soak liquor samples. For the first time, H. tubulosa, H. physodes, The lichen samples belonging to H. tubulosa, E. divaricata, P. furfuracea, P. sulcata and Usnea H. physodes, E. divaricata, P. furfuracea, P. sulcata sp. acetone extracts were examined against and Usnea sp. were collected from fir trees of E. durans isolated from soak liquor samples. Kastamonu province in the north-west of Turkey. Isolate 1 was Gram-positive, oxidase and They were identified through classical taxonomical catalase-negative, protease and lipase positive. methods by microscopic examination. The degradative protease and lipase activities of H. tubulosa, H. physodes, E. divaricata, bacteria have an important role in the production P. furfuracea, P. sulcata and Usnea sp: Turkey, of high-quality leather. There are many studies Kastamonu province, Kapaklı Village, 41.24492, focused on protease and lipase activities of 34.18330, G. Çobanoğlu. halophilic, extremely halophilic and non-halophilic bacteria on hides or skins in the literature. 2.4 Extraction of Lichen Samples McLaughlin and Highberger reported that bacterial strains with proteolytic activity were present in The experiment steps included washing, drying in high percentages on salt-cured goat skins (31). The air, weighing, pulverising by liquid nitrogen, adding proteolytic and lipolytic activities of halophilic and acetone (ACS, ISO, Reag. Ph. Eur.), keeping in a extremely halophilic bacteria were also reported dark place for 24 h followed by filtration through in previous studies. Birbir reported that 91% of filter paper. Then, the evaporation of acetone in a 35 salt-cured skins had halophilic bacteria and rotary evaporator was performed and crude lichen 67% of 85 extremely halophilic bacterial strains acetone extracts were obtained (27). had proteolytic activities (32). Bailey and Birbir

482 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15942856494595 Johnson Matthey Technol. Rev., 2020, 64, (4) detected that 98% of 131 brine-cured skin samples several unwanted defects on finished products had extremely halophilic microorganisms and due to its enzymatic activities. 94% of 332 isolates from these samples showed Antibacterial agents that are commonly used in proteolytic activity (12). Bitlisli et al. demonstrated the soaking process seem to be ineffective due that 53–74% of halophilic bacteria from salt- to random or insufficient application and lead to cured sheepskins had proteolytic activity and antimicrobial-resistant bacteria in soak liquors (12, 47–62% of them had lipolytic activity (33). There 19). From this point, we can suggest that E. durans are also several studies revealing the proteolytic from salted hides or skins could not be exterminated and lipolytic activities of non-halophilic bacteria by curing methods and also in the soaking process from soak liquor samples. Veyselova et al. showed despite the use of antibacterial agents. There are proteolytic activity of some bacteria belonging to the several studies focused on the determination of genera Enterobacter, Pseudomonas, Enterococcus, effective concentrations of several antimicrobial Lactococcus, Aerococcus, Vibrio, Kocuria, agents against various species of bacteria. Both Staphylococcus and Micrococcus and lipolytic the ineffectiveness of antibacterial agents in some activity of B. licheniformis, B. pumilus, P. luteola cases and possible harmful and toxic effects for the and E. cloacae from soak liquor samples (10). environment and human health of some synthetic In molecular analyses, the tested isolate was antimicrobial agents were emphasised in the identified by comparative partial 16S rRNA gene literature (19, 21). In this respect, the need for sequence analysis with the sequences deposited in safer, more ecological and effective materials has the GenBank® database via the BLAST® program. come into prominence for the leather industry. In The Isolate 1 had similarities with E. durans the previous study, the potential antibacterial effects CMGB-120 (99.86%, GenBank® accession number of acetone extracts of H. physodes, E. divaricata, MF348232.1). The existence of Enterococcus P. furfuracea and Usnea sp. at the concentrations of species was previously reported from hides or skins 240 µg ml–1, 120 µg ml–1, 60 µg ml–1 and 30 µg ml–1 in the leather industry (6, 34). It is well known were demonstrated against Bacillus toyonensis, that Enterococcus species are common in surface B. mojavensis, B. subtilis, B. amyloliquefaciens, B. water, soil, vegetables and animal products and velezensis, B. cereus and B. licheniformis which were they are naturally commensal members of gut isolated from soak liquor samples (28). In respect microflora of human and warm-blooded animals. to these findings, we suggested that H. tubulosa, Enterococcus avium, E. casseliflavus, E. durans, H. physodes, E. divaricata, P. furfuracea, P. sulcata E. faecalis, E. faecium and E. gallinarum have and Usnea sp. acetone extracts may have been isolated from salted hide samples (34). antibacterial potential against E. durans which has Furthermore, despite increasing the concentration protease and lipase activities. of antimicrobial agents containing didecyl According to our results, the acetone extracts of dimethyl ammonium chloride from 0.4 g l–1 to P. sulcata had no antibacterial effect at all tested 0.8 g l–1, several bacteria including E. avium concentrations against E. durans (Figure 1). and E. faecium were reported from soak liquor On the other hand, we observed a considerable samples (19). These results suggest that some antibacterial effect for the acetone extracts of Enterococcus species may come from salted hides H. tubulosa and H. physodes against E. durans. and can survive in soak liquor samples even in the High inhibitory effects of these tested extracts presence of antibacterial agents. Fluckey et al. for the growth of E. durans (above 50% isolated 279 Enterococcus isolates from faecal inhibition) were detected at the concentrations and hide samples. Among them, 169 isolates were of 240 µg ml–1, 120 µg ml–1 and 60 µg ml–1 with detected to be E. durans by biochemical tests (35). inhibition ratios of 82.54%, 79.53% and 79.98% E. durans is mostly found in pre-ruminant calves for H. tubulosa, and 86.8%, 78.2%, 77.75% for and young chickens and can survive in moderately H. physodes, respectively (Figures 2 and 3). harsh conditions such as various temperature The acetone extracts of P. furfuracea also had ranges, pH degrees and salt concentrations as well antibacterial effect against E. durans at the as detergents (36–38). Similarly to our results, concentrations of 240 µg ml–1 and 120 µg ml–1 the proteolytic and lipolytic activities of E. durans by the inhibition percentages of 80.63% and were also demonstrated in previous studies. Aslan 85.2%. The other tested concentrations had and Birbir detected that six E. durans isolates had also inhibitory effects on the tested bacteria but proteolytic and lipolytic activities (34). In this the inhibition ratios recorded were below 50% regard, Isolate 1 may have the potential to cause (Figure 4).

2.5

2.0 OD

y, 1.5

1.0 Optical densit 0.5

0 d 5 9 9 60 30 15 7. 1. 0. 240 120 3.75

Untreate

Concentration, µg ml–1

Fig. 1. Antibacterial effect of acetone extracts ofP. sulcata against E. durans from soak liquor samples

2.0

1.5 OD y,

1.0

0.5 Optical densit

0 d 5 9 9 60 30 15 7. 1. 0. 240 120 3.75

Untreate

Concentration, µg ml–1

Fig. 2. Antibacterial effect of acetone extracts ofH. tubulosa against E. durans from soak liquor samples

Potential antibacterial efficacy was also detected and 89.5% respectively. Furthermore, a 58.1% for the acetone extracts of E. divaricata against inhibition ratio was noted for the concentration of E. durans. At the concentration of 240 µg ml–1, we 7.5 µg ml–1 (Figure 6). detected 91% inhibition on the bacterial growth. All data showed that the acetone extracts of Antibacterial effects were observed at the H. tubulosa, H. physodes, P. furfuracea, E. divaricata concentrations of 120 µg ml–1 and 60 µg ml–1 with and Usnea sp. had potential antibacterial efficacies inhibition ratios of 81% and 79% (Figure 5). at varying concentrations against E. durans. Usnea sp. acetone extract was determined to be the Usnea sp. acetone extracts were found to have a most successful among the tested lichen extracts. stronger inhibitory effect on the bacterial growth 240 µg ml–1, 120 µg ml–1, 60 µg ml–1, 30 µg ml–1 of E. durans, even at a low concentration of and 15 µg ml–1 of the extracts belonging to Usnea 15 µg ml–1 (89.5% inhibition) compared to other sp. had an antibacterial effect above 80% inhibition. extracts. These results emphasise the potential of The inhibition ratios at these concentrations were lichens to be utilised as an antibacterial agent in similar and recorded as 88.7%, 84.2%, 92%, 87.8% the leather industry. Further studies are needed

2.0

1.5

1.0

0.5 Optical density, OD

d 5 9 9 60 30 15 7. 1. 0. 240 120 3.75

Untreate

Concentration, µg ml–1

Fig. 3. Antibacterial effect of acetone extracts ofH. physodes against E. durans from soak liquor samples

2.0

1.5 OD y,

1.0

Optical densit 0.5

0 d 5 9 9 60 30 15 7. 1. 0. 240 120 3.75

Untreate

Concentration, µg ml–1

Fig. 4. Antibacterial effect of acetone extracts ofP. furfuracea against E. durans from soak liquor samples

to detect potential compounds of these lichen P. furfuracea, E. divaricata and Usnea sp.) have species and then these compounds may be used in antibacterial effects againstE. durans with protease formulations in the industry. and lipase properties. Whereas P. sulcata did not have any antibacterial efficacy against E. durans, 4. Conclusions other tested extracts were successful depending on the lichen species and concentrations applied. In the leather industry, bacteria with proteolytic and The acetone extracts of Usnea sp. had the highest lipolytic activities are important in terms of finished antibacterial efficacy. The potential antibacterial product quality. In this study, we tried to answer the efficacies of several lichen species suggest that question of whether acetone extracts of six lichen compound(s) extracted from lichens as natural species (H. tubulosa, H. physodes, P. sulcata, resources may be used in the leather industry. We

2.0

1.5 OD y,

1.0

Optical densit 0.5

0 d 5 9 9 60 30 15 7. 1. 0. 240 120 3.75

Untreate

Concentration, µg ml–1

Fig. 5. Antibacterial effect of acetone extracts ofE. divaricata against E. durans from soak liquor samples

2.0

1.5

1.0

Optical density, OD 0.5

d 5 9 9 60 30 15 7. 1. 0. 240 120 3.75

Untreate

Concentration, µg ml–1

Fig. 6. Antibacterial effect of acetone extracts ofUsnea sp. against E. durans from soak liquor samples

believe that more comprehensive studies about (Marmara University) for sharing their experiences their unique chemical compounds will provide new about the experiments. insight to utilise them in this sector. References Acknowledgement 1. S. Dahl, J. Am. Leather Chem. Assoc., 1956, 51, The authors are grateful to Gülşah Çobanoğlu 103 Özyiğitoğlu for taxonomic identification of lichen 2. D. Solaiman, R. Ashby, M. Birbir and P. Caglayan, species tested. The authors would like to thank J. Am. Leather Chem. Assoc., 2016, 111, (10), Arhun Ali Balkan, Ayla Yıldız and Barış Gökalsın 358

3. M. Birbir and A. Ilgaz, J. Soc. Leather Technol. 22. T. H. Nash III, “Lichen Biology”, 2nd Edn., Chem., 1996, 80, (5), 147 Cambridge University Press, Cambridge, UK, 4. Y. Birbir, N. Dolek, M. Birbir and P. Caglayan, Rom. 2008, p. 303 Biotechnol. Lett., 2015, 20, (1), 10123 23. K. Molnár and E. Farkas, Z. Naturforsch. C., 2010, 5. R. Rangarajan, T. D. Didato and S. Bryant, J. Am. 65, (3–4), 157 Leather Chem. Assoc., 2003, 98, (12), 477 24. B. Paudel, H. D. Bhattarai, J. S. Lee, S. G. Hong, 6. A. Orlita, Int. Biodet. Biodeg., 2004, 53, (3), 157 H. W. Shin and J. H. Yim, Phytother. Res., 2008, 22, (9), 1269 7. M. Birbir, Y. Birbir, E. Yilmaz and P. Caglayan, Int. J. Biosci. Biochem. Bioinform., 2016, 6, (4), 25. G. Çobanoğlu, C. Sesal, B. Gökmen and S. Çakar, 121 South West. J. Hortic. Biol. Environ., 2010, 1, (2), 153 8. J. Wu, L. Zhao, X. Liu, W. Chen and H. Gu, J. Clean. 26. G. Çobanoğlu, C. Sesal, B. Açıkgöz and İ. Karaltı, Prod., 2017, 148, 158 Mod. Phytomorphol., 2016, 10, 19 9. D. Berber and M. Birbir, J. Am. Leather Chem. 27. B. Gökalsın, D. Berber, G. Ç. Özyiğitoğlu, Assoc., 2010, 105, (10), 320 E. Yeşilada and N. C. Sesal, Plant Biosyst., 2019 10. C. Veyselova, M. Birbir and D. Berber, J. Soc. 28. D. Berber, J. Am. Leather Chem. Assoc., 2020, Leather Technol. Chem., 2013, 97, (4), 166 115, (3), 96 11. P. Caglayan, M. Birbir, C. Sánchez-Porro, A. Ventosa 29. P. Caglayan, M. Birbir, C. Sánchez-Porro and and Y. Birbir, J. Am. Leather Chem. Assoc., 2018, A. Ventosa, Turk. J. Biochem., 2018, 43, (3), 312 113, (2), 41 30. P. Caglayan, M. Birbir, C. Sánchez-Porro and 12. D. G. Bailey and M. Birbir, J. Am. Leather Chem. A. Ventosa, J. Am. Leather Chem. Assoc., 2017, Assoc., 1993, 88, 285 112, (6), 207 13. H. Anderson, J. Soc. Leather Trade. Chem., 1949, 31. G. D. McLaughlin and J. H. Highberger, J. Am. 33, 250 Leather Chem. Assoc., 1926, 21, 280 14. B. M. Haines, J. Am. Leather Chem. Assoc., 1984, 32. M. Birbir, J. Turk. Microbiol. Soc., 1997, 27, 68 79, (8), 319 33. B. O. Bitlisli, H. A. Karavana, B. Basaran, O. Sarı, 15. J. J. Tancous, W. T. Roddy and F. O’Flaherty, “Skin, I. Yasa and M. Birbir, J. Am. Leather Chem. Assoc., Hide and Leather Defects”, The Western Hills 2004, 99, (12), 494 Publishing Company, Ohio, USA, 1959 34. E. Aslan and M. Birbir, J. Am. Leather Chem. 16. ‘Antibiotic Resistance’, World Health Organization, Assoc., 2011, 106, (12), 372 Geneva, Switzerland, 31st July, 2020 35. W. M. Fluckey, G. H. Loneragan, R. D. Warner, 17. M. Birbir, K. Ulusoy and P. Caglayan, J. Am. Leather A. Echeverry and M. M. Brashears, J. Food Protect., Chem. Assoc., 2016, 111, (9), 334 2009, 72, (4), 766 18. Y. Birbir, G. Uğur and M. Birbir, J. Electrostat., 36. B. D. Shepard and M. S. Gilmore, Microb. Infect., 2008, 66, (7–8), 355 2002, 4, (2), 215 19. D. Berber, M. Birbir and H. Hacioglu, J. Am. Leather 37. D. M. F. Amaral, L. F. Silva, S. N. Casarotti, Chem. Assoc., 2010, 105, (11), 354 L. C. S. Nascimento and A. L. B. Penna, J. Dairy 20. M. Birbir, E. Yazici and P. Çağlayan, J. Soc. Leather Sci., 2017, 100, (2), 933 Technol. Chem., 2019, 103, 6 38. A. P. G. Frazzon, B. A. Gama, V. Hermes, 21. M. F. Türkan, A. Aslan, A. N. Yapıcı, B. Meriçli C. G. Bierhals, R. I. Pereira, A. G. Guedes, P. A. Yapıcı and S. T. Bilgi, Tekstil ve Konfeksiyon, 2013, d’Azevedo and J. Frazzon, World J. Microbiol. 23, (2), 176 Biotechnol., 2010, 26, (2), 365

The Authors

Didem Berber received her MSc degree from the Pediatric Allergy-Immunology Department, School of Medicine, Marmara University, Turkey, in 2003 and PhD from the Department of Biology, Faculty of Arts and Sciences, Marmara University in 2010. She has been studying as postdoctoral researcher in the same department from 2016 up to date. She contributed to projects (European Cooperation in Science and Technology (COST) and other bilateral collaboration projects) on bacterial quorum sensing and biofilm inhibition. Her research topics are hide microbiology, environmental microbiology, antimicrobial agents, fungi, quorum sensing and biofilm formation.

İpek Türkmenoğlu graduated from the Biology Department, Atatürk Faculty of Education, Marmara University in 2012. She is continuing to the master’s programme and she is studying as a scholarship researcher with the support of Scientific and Technological Research Council of Turkey (TÜBİTAK) on the determination and utilisation of species- specific allosteric inhibition zones in glycolytic enzymes in pharmaceutical design. Her research topics are hide microbiology, environmental microbiology, antimicrobial agents, quorum sensing and biofilm formation.

Nüzhet Cenk Sesal graduated from the Biology Department, Atatürk Faculty of Education, Marmara University. He has been working at the Department of Biology, Faculty of Arts and Sciences, Marmara University since 2001. His research area is molecular microbiology. He has been working as a principal investigator, researcher, and consultant in national and international projects, especially about molecular diversity, environmental microbiology, antimicrobial agents, quorum sensing and biofilm formation.

www.technology.matthey.com

The Destructive Effects of Extremely Halophilic Archaeal Strains on Sheepskins, and Proposals for Remedial Curing Processes Use of sterile brine or direct electric current to prevent red heat damage on salted sheepskins

Meral Birbir*, Pinar Caglayan brine using DC to prevent archaeal damage on Division of Plant Diseases and Microbiology, cured hides and skins in the leather industry. Department of Biology, Faculty of Arts and Sciences, Marmara University, Göztepe Campus, 1. Introduction 34722 Kadıköy, Istanbul, Turkey Extremely halophilic archaea have been found in Yasar Birbir hypersaline salt lakes, salterns, salt mines, salted Department of Electrical and Electronics foods and salted hides. There have been numerous Engineering, Faculty of Technology, Marmara studies on the presence of extremely halophilic archaea University, Göztepe Campus, 34722 Kadıköy, in these hypersaline environments (1–12). Due to Istanbul, Turkey the high salt requirements of extreme halophiles (15–30% NaCl), these microorganisms have been *Email: [email protected] denominated as extremely halophilic archaea (13, 14). Cells of Haloarchaea staining Gram-negatively are irregular rods, cocci, pleomorphic rods, cups, Proteolytic and lipolytic extremely halophilic irregular disks, flattened disks, irregular triangles, archaea found in curing salt may contaminate skins rectangles and squares (2, 5, 15). Chemoorganotroph during the brine curing process and damage skin extremely halophilic archaea, which can be motile or structure. In the present study, three proteolytic non-motile, grow aerobically and use different amino and lipolytic extremely halophilic archaea were acids. Colonies of these microorganisms are pink, red isolated from deteriorated salted sheepskins and and orange due to C50‑carotenoid pigments called characterised using conventional and molecular bacterioruberins (15, 16). methods. Each test strain (Haloarcula salaria Observation of red or violet discolorations on AT1, Halobacterium salinarum 22T6, Haloarcula the flesh side of salted hides and skins is the key tradensis 7T3), a mixed culture of these strains and for detecting extremely halophilic archaea in the the mixed culture treated with 1.5 A direct current leather industry. These discolorations are a sign of (DC) were used for brine curing processes of fresh bacterial deterioration of hides and skins (17, 18). sheepskins and examined during 47 days of storage Previous experiments reported that microorganisms to evaluate the degree of destruction wreaked by in curing salts and raceway brines contaminated these microorganisms. Both organoleptic properties hides and skins and caused red heat (10). The and scanning electron microscopy (SEM) images of brine cured hides and skins were often stored in sheepskins proved that each separate test strain hot warehouses, trucks or ships, and these high and the mixed culture caused serious damage. temperature conditions, combined with moisture, However, the mixed culture of strains treated with offer an ideal medium for proteolytic extremely electric current did not damage sheepskin structure. halophilic archaea to grow and potentially digest Therefore, we highly recommend sterilisation of collagen fibres in the hides and skins (10).

Extremely halophilic archaea (102–105 colony The adverse effects of extremely halophilic forming units (CFU) g–1), proteolytic (102–104 archaeal hide isolates and ATCC strains of extremely CFU g–1) and lipolytic (102–104 CFU g–1) extremely halophilic archaea on brine cured hides have been halophilic archaea were detected in 40 curing reported in these studies, respectively (25, 26). salt samples collected from different tanneries However, the destructive effects of salted sheepskin in Turkey (19). Almost all salted hides and skins strains of extremely halophilic Haloarcula salaria, contained extremely halophilic archaea, proteolytic Halobacterium salinarum and Haloarcula tradensis and lipolytic extremely halophilic archaea originating on brine cured sheepskins have not been examined in the curing salt. Extremely halophilic archaea were yet. Therefore, the aim of this study was to also detected on 94% of 131 brine-cured cattle examine adverse effects of proteolytic and lipolytic hides collected from USA, 91% of 35 salted hides archaeal sheepskin strains (Haloarcula salaria cured in France and Russia and all salted hides cured AT1, Halobacterium salinarum 22T6, Haloarcula in Turkey, Greece, the UK, USA, Serbia, Bulgaria, tradensis 7T3) and the mixed culture of these Russia, South Africa and Australia (20–22). Five strains on sheepskins during a 47-day storage extremely halophilic archaeal species, Halorubrum period at 33°C. We also investigated effective saccharovorum, Halorubrum tebenquichense, curing methods to prevent the destructive Halorubrum lacusprofundi, Natrinema pallidum effects of these microorganisms on sheepskins. and Natrinema gari were isolated from five salted Additionally, we evaluated pH values, ash contents, hides originating in England and Australia (22). moisture contents, salt saturations, total counts Also, 101 extremely halophilic archaeal strains of extremely halophilic archaea and organoleptic (Halorubrum tebenquichense, Halorubrum properties of the brine cured sheepskin samples saccharovorum, Halorubrum kocurii, Halorubrum during different storage periods to determine the terrestre, Halorubrum lipolyticum, Halococcus brine curing procedure’s efficiency and the test dombrowskii, Halococcus qingdaonensis, microorganisms’ adverse effects of on sheepskins. Halococcus morrhuae, Natrinema pellirubrum, Natrinema versiforme, Halostagnicola larsenii 2. Materials and Methods and Haloterrigena saccharevitans) were isolated from four salted sheepskin samples (Spain) 2.1 Isolation of Extremely Halophilic exhibiting bad odour, a slimy layer, hair slip, red Archaeal Strains from Deteriorated and yellow discolorations (23). Moreover, 28 Salted Sheepskins extremely halophilic archaeal strains (Natrialba aegyptia, Halovivax asiaticus, Halococcus Two deteriorated salted sheepskins containing red morrhuae, Halococcus thailandensis, Natrinema discolorations were collected from two tanneries pallidum, Halococcus dombrowskii, Halomicrobium in the Istanbul Leather Organized Industrial zhouii, Natronococcus jeotgali, Haloterrigena Zone (40°52′39.7″N,29°20′25.3″E) in Tuzla, thermotolerans, Natrinema versiforme and Turkey. The samples were immediately placed Halobacterium noricense) were isolated from eight into sterile sample bags and transported on ice to salted hide and skin samples from Turkey, Iraq, the laboratory. Then, 20 g of the salt-pack cured Turkmenistan and Kazakhstan (24). sheepskin samples were weighed and separately While there are many reports that detect the soaked in flasks containing 180 ml 30% NaCl presence of extremely halophilic archaea on salted (Merck KGaA, Germany) solution. The flasks were hides and skins (10, 17, 20–25), the destructive placed into a shaking incubator at 90 rpm, 24°C effects of these microorganisms on salted hides for 3 h. The suspension of the skin was diluted have been studied much less (25, 26). In our with sterile physiological saline water (30% NaCl). previous investigation, we found that extremely An aliquot of 100 µl each of direct and serial skin halophilic archaeal strains, isolated from hides suspension dilutions was spread onto the surface brine cured in the USA, damaged grain the surface of modified Brown agar media containing (per of hides at 41°C after 49 days (25). An experiment litre): 1 g CaCl2·H2O, 2 g KCl, 20 g MgSO4·7H2O, with extremely halophilic Haloferax gibbonsii 3 g trisodium citrate, 250 g NaCl, 5 g yeast extract, (ATCC 33959TM) and Haloarcula hispanica (ATCC 20 g agar, pH 7 (5, 27). The plates were incubated 33960TM) obtained from American Type Culture at 39°C for 10 days. Following incubation, red Collection (ATCC), USA, demonstrated that pigmented colonies on the agar media were Haloferax gibbonsii caused hair slip, loss of hide selected and restreaked several times to obtain substance and deterioration of brine cured hide pure cultures. A total of 22 isolates were obtained after 45 days at 40°C (26). from the sheepskins and then, these strains were

490 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15943793010464 Johnson Matthey Technol. Rev., 2020, 64, (4) examined the proteolytic and lipolytic activities. and oxidase activities, indole production, methyl

Proteolytic activity of each strain was detected red test, H2S and NH3 productions of each strain on gelatin agar medium containing 2% gelatin. were investigated according to the procedures After incubation, clear zones around the colonies described previously (4, 28, 31). Furthermore, on the gelatin agar medium indicated protease each strain’s caseinase activity was determined on production (5, 10). Lipolytic activity of each the agar medium containing 2% skim milk. After strain was screened on Tween® 80 agar medium incubation, clear zones around the colonies were containing 1% Tween® 80. After growth was evidence of positive caseinase activity (4). Urease obtained, opaque zones around the colonies were production was investigated on Christensen urea interpreted as positive lipase activity (5). In the agar medium. The tubes were examined for pink present study three red pigmented proteolytic and or red colour change in the medium after seven lipolytic strains (AT1, 22T6 and 7T3) were obtained days of incubation (28, 31). β-galactosidase from two salted sheepskins and these strains were activity was screened in test tubes containing used in the present study. ortho-nitrophenyl-β-galactoside (ONPG) discs and 1 ml of sterile saline water (30% NaCl). The yellow 2.2 Phenotypic Characteristics of colour formation in the test tube was accepted as positive β-galactosidase activity (5, 31). Amino Test Strains acid utilisation of each strain was examined in Exponentially growing pure cultures of three strains the test medium containing 1% amino acid, 0.5% designated as AT1, 22T6 and 7T3 were used in all beef extract, 0.5% peptone, 0.05% dextrose, experiments. First, the strains’ salt requirement and 0.0005% cresol red, 0.001% bromocresol purple, salt tolerance were examined on Brown agar plates 0.0005% pyridoxal and saline water (30% NaCl). containing different salt concentrations (0%, 0.5%, Purple colour formation in the test tube containing 3%, 5%, 7.5%, 10%, 12.5%, 15%, 20%, 25% archaeal culture was accepted as a positive test and 30%) (27). After detection of optimum salt after 10 days incubation period at 39°C (31). concentration for each strain, pH and temperature ranges for growth of each strain (AT1, 22T6, 7T3) 2.3 Amplification and Sequencing of were respectively examined at Brown agar plates 16S rRNA Genes of Test Strains with different pH values (pH 4, pH 5, pH 6, pH 7, pH 7.5, pH 8, pH 9, pH 10, pH 11 and pH 12) and Chromosomal DNA was isolated by QIAamp DNA different temperatures (4°C, 10°C, 15°C, 24°C, Mini Kit (Qiagen, Germany) and purified by QIAquick 28°C, 35°C, 37°C, 39°C, 45°C, 50°C, 55°C, 60°C) PCR Purification Kit (Qiagen, Germany) according according to the methods described in Proposed to the manufacturer’s directions. The 16S rRNA Minimal Standards for Description of New Taxa in genes of the strains were amplified by polymerase the Order Halobacteriales (28). Based on the pH, chain reaction (PCR) using forward primer 21F and and temperature range of each test strain, the reverse primer 1492R (32). The 16S rRNA gene optimal pH and growth temperature of each test sequences of three strains (AT1, 22T6 and 7T3) strain were determined. were determined by IONTEK Laboratory (Turkey). Pigmentation, size, margin, elevation and The sequences of these strains were analysed opacity of colonies of the strains grown on using ChromasPro v.2.1.8 software (Technelysium, Brown agar media were examined under optimal Australia) and then compared with the sequence growth conditions (28). Cell morphology, cell on the EZBioCloud Database (ChunLab, South length, cell width and motility of each strain Korea) (33). were examined using both light microscopy and electron microscopy. Microscopic observation of 2.4 Preparation of Test Strains and each strain was made by using freshly prepared Sheepskin Samples for Brine Curing wet mount (28). For SEM observations, 20 ml of Treatments each test strain were separately passed through 0.2 μm pore size cellulose nitrate membrane filter 2.4.1 Preparation of Strains and placed in the stainless steel funnel via vacuum Cultures Used in Brine Curing pump (Sartorius AG, Germany). The archaeal Processes cells of each strain trapped on the membrane filters were observed under SEM (QuantaTM 450 Pure cultures of each test strain (AT1, 22T6, FEG (FEI, USA)). Gram staining was performed 7T3) were separately grown in liquid Brown test with acetic acid-fixed slides (28–30). Catalase medium containing 30% NaCl for 10 days at 39°C.

Each archaeal cell suspension’s turbidity was culture in the electrolysis cell before the electric adjusted to 0.5 McFarland standard (108 CFU ml–1) current application, 100 µl of the test medium was using densitometer (DEN-1, BIOSAN, Latvia). removed from the electrolysis cell and diluted to Each cell suspension was diluted in sterile saline 10–2–10–4 using sterile 30% NaCl solution. The solution (30% NaCl) to adjust the cell suspension diluted archaeal suspensions were spread over the to 107 CFU ml–1. In addition, mixed cultures of Brown agar media. Then, 1.5 A DC was applied these strains (107 CFU ml–1) were prepared. Then, to the electrolysis cell for 22 min (Figure 1). A 20 ml of each test strain, 20 ml of the mixed culture 100 µl quantity of test medium was removed from were used in the brine curing solutions of T1–T4 the cell at intervals of 1 min, 4 min, 7 min, 10 min, (Table I). 13 min, 16 min, 19 min and 22 min of electric To prepare brine curing solution containing current application. Direct and diluted suspensions electrically inactivated mixed culture (T5), 20 ml of electrically inactivated the mixed culture were of the mixed culture containing AT1, 22T6, spread over Brown agar media. All inoculated 7T3 strains (107 CFU ml–1) were placed into the Brown media were incubated for 10 days at 39°C, electrolysis cell consisting of a glass beaker having and colonies on the agar plates were counted. two internally attached platinum wire electrodes This test medium was used for curing process of and 180 ml of sterile brine solution (30% NaCl) (34, the sheepskin (T5) after 22 min of electric current 35). To detect the archaeal numbers of the mixed application on the mixed culture (Table I).

Table I Protocol for Brine Curing Treatments of Sheepskins Brine Curing Compositions Control 59.5 g sheepskin sample + 200 ml sterile brine solution Treatments T1 59.5 g sheepskin sample + 180 ml sterile brine solution + 20 ml strain AT1 (107 CFU ml–1) T2 59.5 g sheepskin sample + 180 ml sterile brine solution + 20 ml strain 22T6 (107 CFU ml–1) T3 59.5 g sheepskin sample + 180 ml sterile brine solution + 20 ml strain 7T3 (107 CFU ml–1) T4 59.5 g sheepskin sample + 180 ml sterile brine solution + 20 ml mixed culture (107 CFU ml–1) 59.5 g sheepskin sample + 180 ml sterile brine solution containing 20 ml electrically inactivated T5 mixed culture

DC ammeter AC ammeter Voltmeter 0 DC 1 2 AC A A

AC-DC Platinum electrodes Main switch Anode Cathode Variac 100 0

220 V Fuse

50 Hz AC Test 0–220 V DC output medium Input AC output + – R Mp Insulated layer

Inverter switch Electrolysis cell

Fig. 1. Electrolysis cell system used 1.5 A DC treatment in this study (R: phase, Mp: ground)

2.4.2 Preparation of Sheepskin 2.6 Determination of pH, Moisture Samples for Brine Curing Treatments Content, Ash Content and Salt Saturation of Cured Sheepskin One freshly slaughtered, de-fleshed whole sheepskin Samples sample was obtained from a slaughterhouse in Istanbul, Turkey. Then, the sheepskin sample was After curing processes, 5 g of the sheepskins were immediately placed into sterile sample bag and cut and placed into flasks containing 100 ml of transported on ice to the laboratory. The sheepskin sterile distilled water. The flasks were placed in a was cut into six pieces perpendicular to backbone, shaking incubator for 1 h at 100 rpm and then pH from backbone to belly. Next, we carried out the was measured with a pH meter. Hairs and dirt on following six treatments for brine curing of the the samples were removed to properly determine sheepskin samples. In each treatment, sterile the samples’ moisture content. 3 g of the samples 30% NaCl (Merck KGaA) solution was used. In all were placed into an oven at 102°C for 6 h. The treatments, a 400% float of the brine solution (238 g dried samples were weighed, returned to the of the brines without test strain, with each test oven for 1 h, and then were weighed again. The strain or mixed culture/59.5 g of sheepskin) was drying procedure was repeated until the first dry used (25). Sterile 30% NaCl solution containing weight was equal to the second dry weight. The the sheepskin sample was used as Control. The samples were put into a desiccator for 30 min sheepskin samples (T1–T4) were separately placed to cool. Next, we calculated the skins’ moisture in a glass beaker containing the brine solution, each contents (20, 21). The dry sheepskins samples test strain or mixed culture (T1–T4, Table I). In were placed in ceramic crucibles and ashed in a the Treatment 5, the sheepskin sample was placed muffle furnace at 600°C for 8 h. After cooling, the in a glass beaker containing the brine solution with samples were weighed to determine ash content. electrically inactivated mixed culture (T5, Table I). Moisture content, ash content and salt saturations The curing processes of all sheepskins were of skin samples were calculated according to the carried out the protocol described in Table I. The aforementioned methods (30, 36). The pH value, sheepskin samples were separately cured in the ash content, moisture content and salt saturation brine solutions at 90 rpm for 18 h at 24°C. After of all cured sheepskin samples were examined at the curing processes, all sheepskins were taken different storage periods. from the brine solutions and stored for 47 days at 33°C. 2.7 Organoleptic Examination of Brine Cured Sheepskin Samples 2.5 Determination of Extremely During Storage Periods Halophilic Archaeal Counts in Curing All cured sheepskin samples were examined Solutions and Cured Sheepskin organoleptically (hair slip, deterioration of skins, Samples bad odour, sticky appearance, red heat, hole To determine total counts of extremely halophilic formation) during different storage periods. archaea in the curing solutions before the curing processes, 100 µl of the test medium was removed 2.8 Preparation of Sheepskin from the each curing solution and diluted to Samples for Scanning Electron 10–2–10–4 using sterile 30% NaCl solution. The Microscopy Observation diluted archaeal suspensions were spread over the Brown agar media. In addition, subsequent After a 47-day storage period, the sheepskin to each brine curing process detailed above samples were prepared for SEM observation. The (T1–T5), the suspensions of cured sheepskin samples were fixed in 4% glutaraldehyde solution samples were prepared at intervals of 5 days, prepared in 0.1 M phosphate buffer (pH 7.2) for 16 days, 28 days and 47 days of storage. 2 g of 30 min. The samples were washed three times each skin sample were put into a flask containing with 0.1 M phosphate buffer for 10 min and

18 ml sterile 30% NaCl solution and incubated were treated with 1% OsO4 prepared in 0.1 M for 1 h at 24°C and 100 rpm. Direct and serial phosphate buffer at room temperature for 1 h. dilutions of the suspensions were spread onto the The samples were washed two times in sterile surface of Brown agar media. All inoculated Brown distilled water for 10 min. Then, the water in the media were incubated at 39°C for 10 days and the sheepskins was gradually removed by 35%, 50%, colonies grown on the test media were counted. 75%, 95% and absolute ethanol. The mixtures

493 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15943793010464 Johnson Matthey Technol. Rev., 2020, 64, (4) of ethanol-hexamethyldisilazane (ethanol‑HMDS) lipase enzymes were selected and used as test [1:1 (v/v)] (1 × 30 min), ethanol-HMDS strains (AT1, 22T6 and 7T3) in the present study. [1:2 (v/v)] (1 × 30 min) and HMDS (2 × 30 min) were used for air drying process. After drying, 3.2 Phenotypic Characteristics of HMDS was poured from petri dishes and the Test Strains samples were placed in a desiccator for 12 h. Later, the sheepskin samples were examined under Strains AT1, 22T6 and 7T3 grew at 15–30% NaCl, SEM (QuantaTM 450 FEG) using sample stub with 15–30% NaCl, 20–30% NaCl concentrations, double-sided sticky tape (37). respectively. Optimum salt concentrations of strains AT1, 22T6 and 7T3 were determined as 25% NaCl. 3. Results and Discussion Hence, these strains were accepted as extremely halophilic archaea. The pH and temperature 3.1 Isolation and Selection of Test ranges for growth of strains AT1, 22T6 and 7T3 Strains from Sheepskins were respectively found as pH 6–11 and 20–50°C, pH 6–11 and 15–55°C, pH 5–11 and 15–55°C. All A total of 22 red coloured strains were isolated extremely halophilic archaeal strains optimally grew from two deteriorated salted sheepskin samples at 39°C and pH 7. The colony pigmentation, size, obtained from two tanneries in the Istanbul Leather margin, elevation and opacity of strains AT1, 22T6, Organized Industrial Zone in Tuzla, Turkey. While 7T3 were respectively observed as: red, 0.6– 2 mm, nine, seven and three strains respectively produced entire, convex, translucent; red, 1–2 mm, entire, protease, lipase, both protease and lipase, three convex, translucent; red, 0.8–1.9 mm, entire, strains did not produce either lipase or protease convex, translucent. The cells of strains AT1 enzymes. The red coloured three extremely (Figure 2(a)) and 7T3 (Figure 2(c)) were non- halophilic strains producing both protease and motile, extremely pleomorphic (triangle, square,

(a) (b) Fig. 2. SEM micrographs of pleomorphic test strains of (a) Haloarcula salaria (AT1) cells; (b) Halobacterium salinarum (22T6) cells; (c) Haloarcula tradensis (7T3) cells trapped on the membrane filter

5 μm 10 μm

(c)

5 μm

494 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15943793010464 Johnson Matthey Technol. Rev., 2020, 64, (4) irregular disk, short rod). The cells of strains AT1 and of sheepskin causing loss of skin substance. When 7T3 were approximately 0.4–1.3 µm × 0.4–2.0 µm and the protein structure of salted skins is broken down 0.3–0.7 µm × 0.3–4 µm, respectively. The cells of by proteolytic extremely halophilic archaea, these strain 22T6 (Figure 2(b)) were motile, pleomorphic microorganisms can utilise some amino acids as a rods, approximately 0.5–1.2 µm × 3.2–6.6 µm. source of carbon, nitrogen and energy. Haloarcula All strains were Gram-negative (Table II). While salaria AT1 and Halobacterium salinarum 22T6 all strains showed positive catalase, oxidase, utilised most of the amino acids examined. While protease, lipase activities, indole production, the Haloarcula salaria AT1, Halobacterium salinarum methyl red, caseinase, urease and β-galactosidase 22T6 utilised 17 amino acids, Haloarcula tradensis reactions of all strains were negative. The strains 7T3 used only three amino acids (Table III). In did not produce H2S and NH3 (Table II). another study, the liquid test media containing Our experimental results showed that Haloarcula calfskin samples, 30% NaCl and proteolytic red and salaria (AT1), Halobacterium salinarum (22T6), pink strains of the extremely halophilic archaea Haloarcula tradensis (7T3) strains have protease were separately prepared to show disintegration activities which can breakdown proteins in corium of the skin proteins. After an incubation period,

Table II Phenotypic Characteristics of Haloarcula salaria, Halobacterium salinarum, Haloarcula tradensis Characteristics Haloarcula salaria Halobacterium salinarum Haloarcula tradensis Strain code AT1 22T6 7T3 Motility Non-motile Motile Non-motile Extremely Cell morphology Pleomorphic rods Extremely pleomorphic pleomorphic Cell width, µm 0.4–1.3 0.5–1.2 0.3–0.7 Cell length, µm 0.4–2 3.2–6.6 0.3–4 Gram staining Negative Negative Negative Pigmentation Red Red Red Colony size, mm 0.6–2 1–2 0.8–1.9 Colony margin Entire Entire Entire Colony elevation Convex Convex Convex Colony opacity Translucent Translucent Translucent NaCl concentration, % 15–30 15–30 20–30 pH range 6–11 6–11 5–11 Temperature range, °C 20–50 15–55 15–55 Optimum NaCl 25 25 25 Optimum Temperature, 39 39 39 °C Optimum pH range 7 7 7 Catalase activity + + + Oxidase activity + + + Methyl red reaction – – – Caseinase activity – – – Urease activity – – – β-galactosidase activity – – – Indole production – – –

H2S production – – –

NH3 production – – – Protease activity + + +a Lipase activity + + + a Haloarcula tradensis (7T3) showed weak protease activity

Table III Utilisation of Amino Acids by Strains Halobacterium Haloarcula Amino acids Haloarcula salaria (AT1) salinarum (22T6) tradensis (7T3) L-arginine + + + L-cysteine – – – L-glycine + + – L-alanine + + – L-tyrosine + + – L-proline + + – L-hydroxyproline + + – L-glutamic acid – – – L-methionine + + – L-serine + + – L-isoleucine + + – myo-inositol + + – L-lysine + + + L-phenylalanine + + – L-leucine + – – L-valine + + – L-threonine + + – L-ornithine + + – L-histidine + + + L-aspartic acid – – – L-cystine – + –

decomposition of the skin samples in the media Center for Biotechnology Information, USA) under was detected by visual observation. While contents accession numbers as MN585896, MN585803, of asparagine, threonine, serine, glutamine, MN585804. proline, glycine, alanine, valine, isoleucine, In our previous study, extremely halophilic leucine, phenylalanine, lysine and arginine in the archaeal strains were isolated from Tuz Lake and test tubes were detected at high levels, contents of its salterns (5). In Turkish leather industry, curing methionine, tyrosine and histidine were low (10). salt is mostly obtained from Tuz Lake and its Phenotypic features of extremely halophilic AT1, salterns. Hence, we suspect that contaminations of 7T3 and 22T6 strains detected in this study were our sheepskin samples with Haloarcula salaria AT1, fairly similar to phenotypic features of Haloarcula Halobacterium salinarum 22T6 and Haloarcula salaria, Haloarcula tradensis and Halobacterium tradensis 7T3 were due to the curing salt obtained salinarum isolated by other researchers (15, 38, 39). from Tuz Lake and its salterns.

3.3 16S rRNA Gene Sequences of 3.4 Extremely Halophilic Archaeal Test Strains Counts in Curing Solutions Before Curing The phylogenetic analysis revealed that three strains shared highly similar identities with their In the study carried out with 25 salted sheepskin closest phylogenetic relatives. Strains AT1, 22T6, samples (Australia, Bulgaria, Dubai, Greece, 7T3 were respectively assigned to Haloarcula Israel, Kuwait, South Africa, Turkey, USA) and salaria (98.36%-1344 base pairs), Halobacterium 25 salted goat skin samples (Australia, Turkey, salinarum (99.78%-1345 base pairs), Haloarcula Bulgaria, Israel, South Africa, Russia, China, tradensis (98.37%-1355 base pairs). The gene France), proteolytic extremely halophilic archaea sequence data of the strains AT1, 22T6, 7T3 were and lipolytic extremely halophilic archaea were respectively deposited in GenBank® (National detected as 102–105 CFU g–1; 102–106 CFU g–1

496 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15943793010464 Johnson Matthey Technol. Rev., 2020, 64, (4) and 102–106 CFU g–1; 102–106 CFU g–1 on salted from 2.1 × 106 CFU ml–1 to 3.2 × 105 CFU ml–1 sheepskins and goat skins, respectively (40). after 1 min of DC treatment, the cell numbers of The highest number of proteolytic and lipolytic 1.24 × 102 CFU ml–1 was detected after 4 min extremely halophilic archaea on the salted skins of DC treatment. All archaeal cells in the mixed was found as 106 CFU g–1 (40). Therefore, the culture were completely killed in 7 min of DC archaeal cell numbers of test strains in the brine treatment. In the present study, log10 value of the curing solutions were adjusted to 106 CFU ml–1. mixed culture of extremely halophilic archaea in Before the curing processes of sheepskins, while the brine solution before the DC treatment was the archaeal cell numbers in the brine solutions 6.32. After 1 min, 4 min and 7 min of 1.5 A DC of Treatments 1, 3 and 4 were detected as treatment; 0.82, 4.23 and 6.32 log10 reduction 2.1 × 106 CFU ml–1, the archaeal cell numbers in values (CFU ml–1) of the mixed culture in the brine the brine solution of Treatment 2 was detected were detected, respectively. as 2.2 × 106 CFU ml–1. Temperature and pH of the electrolysis cell were The archaeal cell numbers in the mixed culture was respectively measured as 31°C and pH 6 prior to the detected as 2.1 × 106 CFU ml–1 in the electrolysis electric current treatment. After treating the brine cell before 1.5 A DC application. While the archaeal solution with the electric current, the temperature cell numbers in the mixed culture were reduced of the brine was adjusted to 24°C for using in

Table IV pH, Ash Content, Moisture Content and Salt Saturation Values, Total Extremely Halophilic Archaeal Counts of the Sheepskin Samples After Different Storage Periods Ash Moisture Salt Total count of extremely Experiment pH content, % content, % saturation, % halophilic archaea After 5 days Control 7.55 20 55 >100 0 T1 6.72 24 50 >100 2.0 × 107 T2 6.59 23 50 >100 3.4 × 107 T3 6.65 21 57 >100 2.2 × 107 T4 6.53 26 52 >100 3.8 × 107 T5 7.80 21 57 >100 0 After 16 days Control 7.43 25 50 >100 0 T1 6.52 30 47 >100 3.0 × 107 T2 6.70 27 51 >100 6.0 × 107 T3 6.65 22 50 >100 3.4 × 107 T4 6.85 32 46 >100 8.4 × 107 T5 7.32 23 55 >100 0 After 28 days Control 7.40 28 45 >100 0 T1 7.70 29 40 >100 1.2 × 107 T2 7.52 29 43 >100 2.0 × 107 T3 7.36 33 44 >100 2.0 × 107 T4 7.51 32 39 >100 3.4 × 107 T5 7.81 29 46 >100 0 After 47 days Control 7.26 41 30 >100 0 T1 7.58 34 26 >100 1.0 × 107 T2 7.47 34 35 >100 1.8 × 107 T3 7.31 44 24 >100 1.7 × 107 T4 7.60 37 38 >100 2.0 × 107 T5 7.64 33 33 >100 0

497 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15943793010464 Johnson Matthey Technol. Rev., 2020, 64, (4) curing process of sheepskin in the Treatment 5. 3.6 pH, Moisture Content, Ash While the temperature and pH of the test medium Content and Salt Saturation Values respectively increased from 31°C to 41°C and of Cured Sheepskin Samples from pH 6 to pH 8.5 during the electric current treatment, voltage values slightly decreased from After the curing processes of skins, pH values of the 4.7 V to 4.3 V. sheepskin samples were measured as pH 7.35 for We also demonstrated the inactivation Control; pH 6.89 for T1; pH 7.09 for T2; pH 7.05 of extremely halophilic strains via DC and for T3; pH 7.16 for T4; pH 8.05 for T5. While salt alternating electric current (AC) in our previous saturation values of all cured sheepskins were studies (35, 41, 42). A 0.5 A DC was applied for higher than 100% during all storage periods, pH, 30 min to several strains of extremely halophilic ash content and moisture content values changed archaea (107 CFU ml–1) isolated from Tuz Lake, during different storage periods. pH, ash content Kaldırım and Kayacık salterns (35). While the and moisture content values of the cured skins mixed culture of extremely halophilic archaea were detected between pH 6.52–7.81, 20–44%, was exterminated in 10 min, protease producing 24–57%, respectively (Table IV). extremely halophilic archaea were killed in 5 min. Moisture, minimum and maximum ash contents, However, lipase or lipase and protease producing salt saturation values of adequately cured salted extremely halophilic archaea were exterminated hides were suggested as 40–48%, 14–48%, higher in 20 min (35). In another experiment, lipase than 85%, respectively (36). Due to detection of and protease producing extremely halophilic high moisture content in all samples (between 50– strains (105–106 CFU ml–1), separately grown 57%) after five days storage, sterile salt was added in liquid Brown media, were inactivated by a to all sheepskins to reduce their moisture contents 10 min treatment with 0.5 A DC (41). It was according to curing procedure described in the also detected that 1 min of 2 A AC treatment previous study (43). While all skin samples reached was enough to kill extremely halophilic archaea the suggested moisture content values (39–46%) found in brine solution (102–104 CFU ml–1). When after 28 days, the suggested saturation values 2 A AC was applied to lipolytic extremely halophilic were detected after five days. The samples’ archaea, proteolytic extremely halophilic lowest moisture content values were detected archaea, both proteolytic and lipolytic extremely after 47 days. Ash contents of all skins (20–44%) halophilic archaea, and a mixed culture of these were close to suggested values (36). While the strains (106 CFU ml–1), all test microorganisms skins’ pH values changed during storage periods, found in 25% NaCl solution were exterminated all values were found sufficient to support the in 5 min (42). growth of extremely halophilic strains (Table IV). The pH, moisture content, ash content and salt 3.5 Extremely Halophilic Archaeal saturation values detected in this study were also consistent with pH range (pH 6.53–8.01), moisture Counts on Cured Sheepskin Samples content (32–68%), ash content (12–30%) and During Storage salt saturation (58–100%) values of 25 salted After the curing processes of sheepskins, we did sheepskin samples determined in the previous not detect any extremely halophilic archaea on experiment (40). the sheepskin sample cured with the sterile brine solution (Control) and the sheepskin sample cured 3.7 Organoleptic Characteristics with the brine solution containing electrically of Brine Cured Sheepskin Samples inactivated mixed culture (T5) during the all During Storage storage periods. While extremely halophilic archaeal numbers on While hair slip and bad odour were detected on both skin samples cured with each strain and the the sheepskin samples cured with each strain and skin sample cured with mixed cultures of the strains the mixed culture after five days at 33°C, sticky slowly increased from 106 CFU ml–1 to 107 CFU appearance and red heat were observed on the cured during five days and 16 days storage periods, the sheepskin samples after 16 days (T1–T4, Figure numbers of these strains slowly decreased 28 days 3). In addition to the aforementioned organoleptic and 47 days storage periods due to attachment of properties, hole formations were observed on these cells to sheepskins (Table IV). these sheepskin samples after 28 days. However,

498 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15943793010464 Johnson Matthey Technol. Rev., 2020, 64, (4) we did not detect any organoleptic properties on significantly debilitated structural integrity of the sheepskin samples cured with sterile brine and cells in the mixed culture trapped on the filter the brine treated with 1.5 A DC (Control and T5, (Figure 5). The SEM images clearly showed that Figure 3). electric current application damaged cell structures In another study, the commercially cured of each strain in the mixed culture (Figure 5). As hides stored one year in the USA were also seen in Figure 6, the sterile brine curing process examined for proteolytic activity of extremely protected the sheepskin against microbial damage halophilic archaea. Experimental results of during 47 days of storage. that study showed that the flesh side of hides Attachment of Haloarcula salaria AT1, containing extremely halophilic archaea had pink Halobacterium salinarum 22T6 and Haloarcula discolorations called red heat. When these hides tradensis 7T3 to corium fibres and the consequent were incubated at 35°C–40°C, bad odour, hair destructive effects on sheepskins are seen slip and severe grain damage were detected. in Figures 7–10. Haloarcula salaria AT1, Damaged grain surfaces were observed on Halobacterium salinarum 22T6 and the mixed leather made from these hides (10). In another culture of the strains caused fibres in the corium to experiment researchers emphasised that split and weaken (Figures 7, 8 and 10). In contrast temperatures of the brines and hides should be with the skin samples treated with Haloarcula maintained below 20°C to prevent growth of salaria AT1, Halobacterium salinarum 22T6, skin extremely halophilic archaea (44). sample treated with Haloarcula tradensis 7T3 had compact appearance, although the shredding of 3.8 Scanning Electron Microscopy the fibres was still present in corium (Figure 9). That damage was due to the proteolytic activities Observation of Mixed Culture and of these microorganisms. Treated Sheepskin Samples Figure 11 clearly shows that the curing process of Figure 4 shows extremely halophilic archaeal cells sheepskin with the brine containing mixed culture of the mixed culture on 0.2 µm pore-size cellulose treated with DC prevented extremely halophilic nitrate membrane filter in pleomorphic shapes archaea from contaminating the sheepskin and such as triangle, square, irregular disk and rod. As furthermore protected the skin very well against seen in the SEM micrograph, 1.5 A DC treatment microbial damage during a long storage period.

(a) (b) (c) Fig. 3. Organoleptic characteristics of brine cured sheepskin samples after 16 days storage period: (a) Control, sheepskin sample cured with sterile brine (30% NaCl); (b) T1, sheepskin sample cured with brine containing Haloarcula salaria AT1; (c) T2, sheepskin sample cured (d) (e) (f) with brine containing Halobacterium salinarum 22T6; (d) T3, sheepskin sample cured with brine containing Haloarcula tradensis 7T3; (e) T4, sheepskin sample cured with brine containing mixed culture; (f) T5, sheepskin sample cured with brine containing electrically inactivated mixed culture

20 μm 30 μm

Fig. 4. SEM micrograph of mixed culture of Fig. 7. SEM micrograph of the longitudinal section undamaged pleomorphic cells of Haloarcula of damaged corium layer of sheepskin treated with salaria (AT1), Halobacterium salinarum (22T6) Haloarcula salaria (AT1) stored for 47 days at 33°C and Haloarcula tradensis (7T3) trapped on the membrane filter

10 μm 40 μm

Fig. 5. SEM micrograph of mixed culture of Fig. 8. SEM micrograph of the longitudinal section damaged Haloarcula salaria (AT1), Halobacterium of damaged corium layer of sheepskin treated with salinarum (22T6) and Haloarcula tradensis (7T3) Halobacterium salinarum (22T6) stored for 47 days cells treated with 1.5 A DC trapped on the at 33°C membrane filter

50 μm 50 μm Fig. 9. SEM micrograph of the longitudinal section Fig. 6. SEM micrograph of the longitudinal section of damaged corium layer of sheepskin treated with of undamaged sheepskin structure treated with Haloarcula tradensis (7T3) stored for 47 days at sterile brine (Control) stored for 47 days at 33°C 33°C

The present study proved that organoleptic Electron micrographs also showed that each test changes detected in the sheepskins were closely isolate and a mixed culture of extremely halophilic related to proteolytic and lipolytic activities of strains destroyed the skins’ collagen fibres. We extremely halophilic archaeal strains on the skin. did not detect any difference when assessing the

50 μm 100 μm

Fig. 10. SEM micrograph of the longitudinal section Fig. 11. SEM micrograph of the longitudinal section of damaged corium layer of sheepskin treated of undamaged sheepskin structure treated with with mixed culture of Haloarcula salaria (AT1), electrically inactivated mixed culture stored for 47 Halobacterium salinarum (22T6), Haloarcula days at 33°C tradensis (7T3) stored for 47 days at 33°C efficacy of sterile brine and electrically treated brine application of 1.5 A DC treatment in 7 min. Our curing processes of sheepskin samples throughout experimental results proved that a curing process 47 days. We did not observe any damage to the using sterile brine or brine treated with electric compactness of sheepskin structure cured with current prevented red heat and deterioration of both the sterile brine and electrically treated brine sheepskins during long storage periods. Therefore, containing the mixed culture. Both methods were an environmentally friendly, easy, cheap, very found very effective for preventing archaeal growth simple electrolysis system is a very attractive and damage on the brine cured sheepskins. alternative for brine disinfection: (a) it kills different Our results were consistent with those of other species of microorganisms including proteolytic experimental studies on the extremely halophilic and lipolytic extremely halophilic archaea; (b) it strains and culture collection strains of extremely prevents development of resistant strains in leather halophilic archaea (25, 26). In our previous factories; (c) it kills very effectively the aggregated experiment, SEM images showed that hides cured microorganisms found in the brine containing high with proteolytic extremely halophilic archaeal organic substances; (d) it can achieve a reduction strains had red heat and severe grain damage after factor of more than 6 log10 for proteolytic and 49 days of storage at 41°C (25). In another study, lipolytic extremely halophilic archaea; and (e) it the cured hides with extremely halophilic Haloferax has irreversibly lethal action. Hence, we suggest gibbonsii (ATCC 33959TM) exhibited hair loss, using this effective brine disinfection system in thinner and flaccid structure; these consequences of the leather industry after sufficient insulation and deterioration and loss of hide substance. The open grounding are provided by an electrical engineer. fibre structure was also detected in the corium of the hide inoculated with Haloferax gibbonsii (27). Acknowledgements The SEM images showed that the fibre structures of hide were broken down into the smaller fibres This study is dedicated to our wonderful teachers after 43 days (27). who have contributed so much to science and taught us to love this discipline. We thank Yeşim 4. Conclusion Müge Şahin, Demet Sezgin Mansuroğlu (Istanbul Arel University, Turkey) and Aslıhan Çetinbaş Genç This is the first study that detects adverse effects (Marmara University) for the SEM micrographs. We of characterised extremely halophilic archaeal are also grateful to Martin Louis Duncan (Virginia strains on brine cured sheepskins with SEM. Union University, USA) for his critical reading of the SEM images proved that proteolytic and lipolytic manuscript. Haloarcula salaria AT1, Halobacterium salinarum 22T6, Haloarcula tradensis 7T3 caused corium References fibres to split apart in cured sheepskins after 47 days in storage. The mixed culture of proteolytic 1. F. Rodriguez-Valera, F. Ruiz-Berraquero and and lipolytic extremely halophilic archaea A. Ramos-Cormenzana, Microb. Ecol., 1981, originating in curing salt can be exterminated with 7, (3), 235

2. D. R. Arahal, F. E. Dewhirst, B. J. Paster, of Microorganisms”, 15th Edn., Global Edition, B. E. Volcani and A. Ventosa, Appl. Environ. Pearson Education Ltd, London, UK, 2019 Microbiol., 1996, 62, (10), 3779 17. M. Bergmann, J. Soc. Leather Technol. Chem., 3. A. Oren, A. Ventosa, M. C. Gutiérrez and 1929, 11, 599 M. Kamekura, Int. J. Syst. Bacteriol., 1999, 18. L. S. Stuart, R. W. Frey and L. H. James, 49, (3), 1149 ‘Microbiological Studies of Salt in Relation to the 4. M. Birbir, A. Ogan, B. Calli and B. Mertoglu, World Reddening of Salted Hides’, Technical Bulletin No. J. Microbiol. Biotechnol., 2004, 20, (6), 613 383, United States Department of Agriculture, 5. M. Birbir, B. Calli, B. Mertoglu, R. E. Bardavid, Washington, DC, USA, September, 1933, 32 pp A. Oren, M. N. Ogmen and A. Ogan, World 19. D. Berber and M. Birbir, J. Am. Leather Chem. J. Microbiol. Biotechnol., 2007, 23, (3), 309 Assoc., 2010, 105, (10), 320 6. R. Ghai, L. Pašić, A. B. Fernández, A.-B. Martin- 20. D. G. Bailey and M. Birbir, J. Am. Leather Chem. Cuadrado, C. M. Mizuno, K. D. McMahon, Assoc., 1993, 88 (8), 285 R. T. Papke, R. Stepanauskas, B. Rodriguez-Brito, 21. M. Birbir, J. Turk. Microbiol. Soc., 1997, 27, 68 F. Rohwer, C. Sánchez-Porro, A. Ventosa and 22. P. Caglayan, C. Sánchez Porro, A. Ventosa and F. Rodríguez-Valera, Sci. Rep., 2011, 1, 135 M. Birbir, “Characterization of Moderately and 7. A. Ventosa, A. B. Fernández, M. J. León, Extremely Halophilic Microorganisms from Salt- C. Sánchez-Porro and F. Rodriguez-Valera, pack Cured Hides”, 9th International Congress Extremophiles, 2014, 18, (5), 811 Extremophiles, Seville, Spain, 10th–13th 8. A. B. Fernández, R. Ghai, A.-B. Martin-Cuadrado, September, 2012, pp. 128 C. Sánchez-Porro, F. Rodriguez-Valera and 23. C. Akpolat, A. Ventosa, M. Birbir, C. Sánchez-Porro A. Ventosa, FEMS Microbiol. Ecol., 2014, 88, (3), and P. Caglayan, J. Am. Leather Chem. Assoc., 623 2015, 110, (7), 211 9. A. Ventosa, R. R. de la Haba, C. Sánchez-Porro 24. S. T. Bilgi, B. Meriçli Yapici and İ. Karaboz, J. Am. and R. T. Papke, Curr. Opin. Microbiol., 2015, 25, Leather Chem. Assoc., 2015, 110, (2), 33 80 25. D. G. Bailey and M. Birbir, J. Am. Leather Chem. 10. W. E. Kallenberger, ‘Halophilic Bacteria in Hide Assoc., 1996, 91, (2), 47 Curing’, PhD Thesis, Division of Graduate Studies 26. R. H. Vreeland, S. Angelini and D. G. Bailey, J. Am. and Research of the University of Cincinnati, Leather Chem. Assoc., 1998, 93 (4), 121 Department of Basic Science Tanning Research of 27. A. D. Brown, Biochim. Biophys. Acta, 1963, 75, the Collage of Arts and Science, Cincinnati, USA, 425 6th August, 1985, p. 79 28. A. Oren, A. Ventosa and W. D. Grant, Int. J. Syst. 11. M. A. Amoozegar, M. Siroosi, S. Atashgahi, Bacteriol., 1997, 47, (1) 233 H. Smidt and A. Ventosa, Microbiology, 2017, 29. H. P. Dussault, J. Bacteriol., 1955, 70, (4), 484 163, (5), 623 30. M. Birbir and A. Ilgaz, J. Soc. Leather Technol. 12. P. Corral, M. A. Amoozegar and A. Ventosa, Mar. Chem., 1996, 80, (5), 147 Drugs, 2020, 18, (1), 33 31. T. R. Johnson and C. L. Case, “Laboratory 13. D. J. Kushner, ‘Life in High Salt and Solute Experiments in Microbiology”, 9th Edn., Pearson Concentrations’, in “Microbial Life in Extreme Benjamin Cummings, San Francisco, CA, USA, Environments”, ed. D. J. Kushner, Academic Press, 2010 London, UK, 1978, p. 318 32. E. F. Delong, Proc. Natl. Acad. Sci. USA, 1992, 14. D. J. Kushner, ‘Growth and Nutrition of Halophilic 89, (12), 5685 Bacteria’, in “The Biology of Halophilic Bacteria”, 33. S.-H. Yoon, S.-M. Ha, S. Kwon, J. Lim, Y. Kim, eds. R. H. Vreeland, L. I. Hochstein, 1st Edn, CRC H. Seo and J. Chun, Int. J. Syst. Evol. Microbiol., Press Inc, Boca Raton, Florida, USA, 1992, p. 87 2017, 67, (5), 1613 15. W. D. Grant, M. Kamekura, T. J. McGenity and 34. J.-C. Park, M. S. Lee, D. H. Lee, B. J. Park, D.- A. Ventosa, ‘Order I. Halobacteriales Grant and W. Han, M. Uzawa and K. Takatori, Appl. Environ. Larsen 1989b, 495VP’, in “Bergey’s Manual® of Microbiol., 2003, 69, (4), 2405 Systematic Bacteriology: The Archaea and the Deeply Branching and Phototrophic Bacteria”, eds. 35. Y. Birbir and M. Birbir, J. Electrostat., 2006, D. R. Boone, R. W. Castenholz and G. M. Garrity, 64, (12), 791 2nd Edn., Springer-Verlag, New York, USA, 2001, 36. D. G. Bailey, J. Am. Leather Chem. Assoc., 2003, Vol. 1, p. 294 98, 308 16. M. T. Madigan, K. S. Bender, D. H. Buckley, 37. M. Das Murtey and P. Ramasamy, ‘Sample W. M. Sattley and D. A. Stahl, “Brock Biology Preparations for Scanning Electron Microscopy – Life

Sciences’, in “Modern Electron Microscopy in Physical 41. Y. Birbir, D. Degirmenci and M. Birbir, J. Electrostat., and Life Sciences”, ed. M. Janecek, InTech Open Ltd, 2008, 66, (7–8), 388 London, UK, 2016, pp. 162–185 42. Y. Birbir, S. Anik, M. Birbir and P. Caglayan, Johnson 38. S. Namwong, S. Tanasupawat, T. Kudo and T. Itoh, Matthey Technol. Rev., 2015, 59, (2), 109 Int. J. Syst. Evol. Microbiol., 2011, 61, (2), 231 43. C. A. Money and U. Adminis, ‘Curing of Hides and 39. L. Ducharme, A. T. Matheson, M. Yaguchi and Skins: Alternative Methods’, Meat Technology L. P. Visentin, Can. J. Microbiol., 1972, 18, (8), 1349 Update, 94/3, Australian Meat Technology (AMT), Queensland, Australia, 1994 40. P. Caglayan, M. Birbir, C. Sánchez-Porro, A. Ventosa and Y. Birbir, J. Am. Leather Chem. Assoc., 2018, 44. W. E. Kallenberger and R. M. Lollar, J. Am. Leather 113, (2), 41 Chem. Assoc., 1986, 81, 248

The Authors

Meral Birbir graduated from the Biology Department, Ataturk Faculty of Education, Marmara University, Turkey. She received her MSc and PhD Degrees in Biology (especially microbiology) from the Institute of Pure and Applied Sciences, Marmara University. Professor Birbir has been working in the Biology Department of Marmara University since 1985. She was a research scientist at the Department of Pathology and Microbiology, Veterinary Medical School, Purdue University, USA (1990) and Hides and Leather Department of the United States Department of Agriculture (1992–1993). Her research interests are halophilic microorganisms, hide/skin microbiology, antimicrobial agents, electric current applications on microorganisms and microbial communities in hypersaline environments.

Pinar Caglayan graduated from the Biology Department, Ataturk Faculty of Education, Marmara University. She received her MSc and PhD Degrees in Biology from the Institute of Pure and Applied Sciences, Marmara University. She was an Erasmus student in the Department of Microbiology and Parasitology, Faculty of Pharmacy, Sevilla University, Spain (2008–2009). She has been working as a research and teaching assistant at the Division of Plant Diseases and Microbiology, Marmara University since 2011. Her research interests are moderately halophilic bacteria, extremely halophilic archaea, antimicrobial agents, hide microbiology and electric current applications on microorganisms.

Yasar Birbir received his BSc Degree from Gazi University, Turkey, and MSc and PhD Degrees in Electrical Education from Marmara University. He has been working at Marmara University since 1983. He attended the World Bank Industrial Training Project at Indiana and Purdue Universities, USA (1989–1990). He worked as a visiting research scientist at the Electrical and Computer Engineering Department of Drexel University, USA (1992–1993). He has been working as a Professor at the Technology Faculty, Department of Electrical Engineering, Marmara University. His current interests are power electronic converters and drivers, electromagnetic filtering processes in industry and applications of electric currents for inactivation of different microorganisms.

www.technology.matthey.com

Johnson Matthey Highlights A selection of recent publications by Johnson Matthey R&D staff and collaborators

A Stochastic Approach to Model Chemical Looping propylene selectivity had a decrease in response Combustion to the metals. M. A. Schnellmann, G. Williams and J. S. Dennis, Ga V O (OH) , A Synthetic Member of the Powder Technol., 2020, 365, 39 2.52 2.48 7.33 0.67 Nolanite/Akdalaite-Type Family of Oxyhydroxides Using two coupled fluidised-bed reactors, a Containing Trivalent Vanadium stochastic model for reactor-regenerator systems D. S. Cook, M. R. Lees, J. M. Fisher, D. Thompsett was established. As a result the stochastic model and R. I. Walton, J. Solid State Chem., 2020, 288, was able to achieve the simulation of the circulating 121396 fluidised bed with acceptable precision. This Powder neutron diffraction shows oxyhydroxide model had been used before to comprehend how Ga V O (OH) prepared by reaction sensitive a chemical looping combustion (CLC) 2.52 2.48 7.33 0.67 between gallium metal and Na VO in a 1:1 process is when other factors, for example the 3 4 monoethanolamine:water mixture at 240ºC nature of gas-solid reactions, are included. To demonstrates the material is isostructural with show the stochastic model has value for simulation nolanite and akdalaite (Figure 1). Rietveld and optimisation formations of CLC it is applied refinement was undertaken against the data here with methane fuel gas in a laboratory-scale showing all vanadium is octahedrally coordinated. circulating fluidised bed. Vanadium’s oxidation state is close to V3+ when ZSM-5 Additive Deactivation with Nickel and vanadium K-edge XANES spectroscopy is used. Vanadium Metals in the Fluid Catalytic Cracking There is dehydration around 300ºC (oxide

(FCC) Process Ga2.52V2.48O8 is produced and has a larger amount 4+ A. A. Gusev, A. C. Psarras, K. S. Triantafyllidis, A. of V ) followed by decomposition at 500ºC. While A. Lappas, P. A. Diddams and I. A. Vasalos, Ind. both materials seem to follow the Curie-Weiss Eng. Chem. Res., 2020, 59, (6), 2631 law at high temperatures, this is not so at low temperature. No reducing gas atmospheres are This article explores properties of ZSM-5 additives required in the preparation of V(III) oxides. and the role of nickel and vanadium in fluid catalytic cracking (FCC). Loadings of 4000 ppm and 12,000 ppm nickel and vanadium were found. There was deactivation of ZSM-5 in a cyclic deactivation unit and vacuum gas oil (VGO) reacted with nickel and vanadium naphthenates when cracking-regeneration reactions were undertaken. There was an even distribution of nickel across a particle where characterisation of deactivated ZSM-5 additives by nitrogen physiosorption, SEM and pyridine FTIR techniques were used. The 1 µm disparity was small in the Brønsted acidity, but Fig. 1. Reprinted from D. S. Cook et al., J. Solid when nickel and vanadium were added there was State Chem., 2020, 288, 121396, Copyright a rise in Lewis acidity. Further tests were carried (2020), with permission from Elsevier out with VGO and where butylene increased,

Hierarchical ZSM-5 Catalysts: The Effect of Different with an impregnation of metal precursors onto Intracrystalline Pore Dimensions on Catalyst activated carbon from a low boiling point solvent. Deactivation Behaviour in the MTO Reaction Single-site gold, palladium, ruthenium and platinum T. Weissenberger, A. G. F. Machoke, J. Bauer, R. catalysts supported on carbon were prepared in Dotzel, J. L. Casci, M. Hartmann and W. Schwieger, a facile method. In addition, it is shown that a ChemCatChem, 2020, 12, (9), 2461 single-site gold on carbon catalyst for acetylene hydrochlorination can be produced by this method. ZSM-5 zeolites used as methanol-to-olefins (MTO) catalysts were studied to determine Efficient and Selective Solvent-Free Homogeneous the effect of intracrystalline pore systems in Hydrogenation of Aldehydes Under Mild Reaction different combinations. Intracrystalline mesopores, Conditions Using [RuCl2(dppb)(ampy)] intracrystalline macropores and a novel ZSM-5 A. Zanotti-Gerosa, T. Angelini and S. Roseblade, type zeolite with both intracrystalline meso and Tetrahedron Lett., 2020, 61, (13), 151677 macropores were used. There was a prolonged catalyst lifetime with the hierarchical catalysts Using commercial grade aldehydes, effective, unlike microporous only ZSM-5 catalyst. Using solvent-free homogeneous hydrogenation of the intracrystalline mesopores and intracrystalline aldehydes was undertaken with catalysts macropores as catalysts resulted in the ZSM-5 [RuCl2(dppb)(ampy)] and [RuCl2(dppf)(ampy)]. catalyst lasting up to three times longer. There were This gave high conversion to the related alcohols a number of important outcomes which were also using molar catalyst loadings of 10,000/1– 50,000/1. noted including how mesopores and macropores Aldehydes can be reduced preventing byproducts effect catalyst deactivation. Overall, the study being formed with the minimum of waste which has demonstrates intracrystalline macropores (alone led to a solvent-free protocol being established. This or in combination with mesopores) significantly gives a straightforward hydrogenation technique for enhance the ZSM-5 catalytic performance in the reduction of aldehydes to alcohols and commercial MTO reaction. grade aldehydes require no further purification.

Accelerating Pharmaceutical Development via Innovation in Fischer–Tropsch: Developing Metal-Mediated Bond Formation Fundamental Understanding to Support Commercial Opportunities C. J. Borths and S. D. Walker, Israel J. Chem., 2020, 60, (3–4), 340 M. Peacock, J. Paterson, L. Reed, S. Davies, S. Carter, A. Coe and J. Clarkson, Top. Catal., 2020, This article presents work being undertaken related 63, (3–4), 328 to progress with metal-mediated carbon–carbon and carbon–heteroatom bond forming processes. The BP-Johnson Matthey proprietary Fischer‑Tropsch TM The viewpoint of a current drug portfolio is used. technology and advanced CANS reactor and Several case studies are discussed from the authors’ catalyst system are detailed. It provides improved laboratories looking at synthetic challenges plus heat transfer, reduced pressure drop and higher prospects offered by pharmaceutically pertinent productivity and subsequently less financial platforms. In several instances, available synthetic expenditure. A clear understanding of how catalysts methods are challenged by target structures. This behave is crucial to finding a catalyst stable during drives development permitting the acceleration of its use and life. This report presents a study on technology for drug development. There is also catalyst activation on different catalyst supports a discussion about metal-mediated processes at and combines in situ techniques and reactor testing. large scale. Logical and systematic catalyst programmes are crucial for their development and are discussed Facile Synthesis of Precious-Metal Single-Site in the results. Also, catalyst understanding, Catalysts Using Organic Solvents optimisation and development in combination with X. Sun, S. R. Dawson, T. E. Parmentier, G. Malta, the novel CANSTM reactor design can maximise T. E. Davies, Q. He, L. Lu, D. J. Morgan, N. Carthey, potential. P. Johnston, S. A. Kondrat, S. J. Freakley, C. J. Kiely and G. J. Hutchings, Nature Chem., 2020, 12, (6), Cu/M:ZnO (M = Mg, Al, Cu) Colloidal Nanocatalysts 560 for the Solution Hydrogenation of Carbon Dioxide to Methanol In many catalytic reactions, high activity and A. H. M. Leung, A. García-Trenco, A. Phanopoulos, A. selectivity can be demonstrated by single-site Regoutz, M. E. Schuster, S. D. Pike, M. S. P. Shaffer catalysts. Creation of low metal loadings or a and C. K. Williams, J. Mater. Chem. A, 2020, 8, variety of metal species can be achieved by (22), 11282 impregnation, for example, from strongly oxidising aqueous solutions. This study shows an atomic A synthesis is undertaken using controlled distribution of cationic metal species is achieved hydrolysis of a mixture of organometallic precursors

505 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15983488324308 Johnson Matthey Technol. Rev., 2020, 64, (4) for doped-ZnO NPs capped with dioctylphosphinate Chem. Sci., 2020, 11, (27), 7040 ligands. Following substitutional doping and after The environmental impact of acetylene hydrolysis, colloidal nanoparticles (2–3 nm) were hydrochlorination was substantially reduced by characterised. Doped-ZnO nanoparticles and replacing HgCl /C with Au/C as a catalyst. Atomically colloidal Cu(0) nanoparticles in solution were 2 dispersed cationic gold species are the catalytically applied for hydrogenation catalysis of CO to 2 active site. There have been limited studies which methanol in a liquid-phase continuous flow stirred look at the ligand environment around the metal tank reactor under the following conditions: 210ºC, centre. This study uses K-edge soft XAS. Three 50 bar, CO :H = 1:3, 150 ml min−1, mesitylene, 2 2 separate chlorine species are identified and how 20 h. Higher rates are displayed for all catalyst they evolve in the reaction is demonstrated. Au–S systems with respect to methanol production interactions are established in catalysts prepared compared to a benchmark catalyst. There is better using thiosulfate precursors. The catalysts display stability. There was around double the activity evidence of high stability towards reduction to for Al(III)‑doped nanocatalyst. Mg(II) doping inactive metal NPs. Gas switching experiments outperforms the benchmark catalyst but was worse made clear this stability. C H on its own did not compared to undoped ZnO. There is an implication 2 2 particularly change the gold electronic structure that Al(III) migrates to the catalyst surface, and and the thiosulfate catalyst was not deactivated. is proposed to enable stabilisation of the catalytic ZnO/Cu interfaces. Optimization of Biomass Pyrolysis Vapor Upgrading N-Functionalised Imidazoles as Stabilisers for Metal Using a Laminar Entrained-Flow Reactor System Nanoparticles in Catalysis and Anion Binding B. Peterson, C. Engtrakul, T. J. Evans, K. Iisa, C. J. Serpell, J. Cookson and P. D. Beer, M. J. Watson, M. W. Jarvis, D. J. Robichaud, ChemistryOpen, 2020, 9, (6), 683 C. Mukarakate and M. R. Nimlos, Energy Fuels, 2020, 34, (5), 6030 The physicochemical properties of metal NPs are discrete from bulk and molecular metal species. To obtain understanding of commercial scale Consequently, this delivers opportunities in areas ex situ catalytic fast-pyrolysis (CFP) a customised like catalysis and sensing, for example. The surface bench‑scale continuous-flow catalytic fast- of the NPs usually need to be protected to hamper pyrolysis CFP reactor system was built. The study aggregation. However, access to the surface can successfully carried out CFP of pine over two also be blocked by these coatings preventing the commercial zeolite catalysts. The transmission of ability to benefit from their uncommon properties. pyrolysis vapours to the vapour-phase upgrader The article shows that palladium, platinum, gold was optimised to limit secondary thermal cracking and silver NPs can be stabilised by alkyl imidazoles. and preserve carbon in the ex situ CFP process. It also outlines the limits of their synthesis. Products attained were comparable to those from Proof‑of-principle in catalysis and anion binding is fixed bed and fluidised bed reactor systems and established showing that the ligands deliver a level entrained-flow riser reactor systems. Experiments of surface protection. that duplicated the process provided a good average mass balance closure and comparable In Situ K-edge X-ray Absorption Spectroscopy of the trends in deactivation of catalyst were seen in the Ligand Environment of Single-Site Au/C Catalysts laminar entrained-flow reactor system. There was During Acetylene Hydrochlorination a decreasing catalyst‑to-biomass ratio. Optimised G. Malta, S. A. Kondrat, S. J. Freakley, D. J. Morgan, conditions suggest a feasible option for CFP of E. K. Gibson, P. P. Wells, M. Aramini, D. Gianolio, pine. For the two catalysts tested, minor variances P. B. J. Thompson, P. Johnston and G. J. Hutchings, were detected.

www.technology.matthey.com

Antibiotic and Heavy Metal Resistant Bacteria Isolated from Aegean Sea Water and Sediment in Güllük Bay, Turkey Quantifying the resistance of identified bacteria species with potential for environmental remediation applications

Gülşen Altuğ* in sediment and seawater samples taken from Department of Marine Biology, Faculty the Aegean Sea, Turkey, between 2011 and 2013. of Aquatic Sciences, Istanbul University, Bioindicator bacteria in seawater samples were Balabanağa Mahallesi Ordu Caddesi No 8, Laleli, tested using the membrane filtration technique. The Fatih, Istanbul, 34134, Turkey spread plate technique and VITEK® 2 Compact 30 micro identification system were used for Mine Çardak heterotrophic aerobic bacteria in the samples. The Department of Fisheries Technology, Faculty of minimum inhibition concentration method was Çanakkale Applied Sciences, Çanakkale Onsekiz used for heavy metal-resistant bacteria. Antibiotic- Mart University, Terzioğlu Campus, Çanakkale, resistant bacteria were tested using the disk diffusion 17020, Turkey method. All bacteria isolated from sediment samples showed 100% resistance to rifampicin, sulfonamide, Pelin Saliha Çiftçi Türetken tetracycline and ampicillin. 98% of isolates were Department of Marine Biology, Faculty resistant against nitrofurantoin and oxytetracycline. of Aquatic Sciences, Istanbul University, Higher antibiotic and heavy metal resistance was Balabanağa Mahallesi Ordu Caddesi No 8, Laleli, recorded in bacteria isolated from sediment than Fatih, Istanbul, 34134, Turkey seawater samples. The highest levels of bacterial metal resistance were recorded against copper Samet Kalkan (58.3%), zinc (33.8%), lead (32.1%), chromium Department of Marine Biology, Faculty of (31%) and iron (25.2%). The results show that Fisheries and Aquatic Sciences, Recep Tayyip antibiotic and heavy metal resistance in bacteria Erdoğan University, Zihni Derin Campus, Rize, from sediment and seawater can be observed as 53100, Turkey responses to environmental influences including pollution in marine areas. Sevan Gürün Department of Marine Biology, Faculty 1. Introduction of Aquatic Sciences, Istanbul University, Balabanağa Mahallesi Ordu Caddesi No 8, Laleli, In the era of Industry 4.0, with global climate change, Fatih, Istanbul, 34134, Turkey increasing population and developing technology, the spread of heavy metal pollutants in aquatic *Email: [email protected] areas is increasing. Bacterial resistance and metal accumulation capability are common phenomena that can be exploited for the bioremediation of Heavy metal and antibiotic-resistant bacteria the environment, hence these resistant bacteria have potential for environmental bioremediation may be potential candidates for biotechnological applications. Resistant bacteria were investigated applications. Despite the risks caused by antibiotic-

507 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15953337767424 Johnson Matthey Technol. Rev., 2020, 64, (4) resistant bacteria, heavy metal-resistant bacteria persistence of antibiotic resistance in bacterial can be used in detoxification processes to convert pathogens is a threat and a source of considerable a toxic form to a harmless form of a substance concern to public health (8–15). It is known that by developing biotransformation mechanisms. environmental factors such as overpopulation, Bioremediation studies have been carried out to livestock farming, insufficient drainage and identify candidate species (1–4). sanitation infrastructure may provide hotspots In recent years, the increase in pollution by toxic for environmental antibiotic-resistant bacteria compounds and heavy metals in marine areas makes transmission (16). it increasingly important to study the relationships In aquatic environments, antibiotic-resistant between bacteria and toxic compounds. Studies bacteria can be accompanied by heavy metal- related to the transformation of compounds into resistant bacteria that are often induced by the different forms via bacterial metabolic processes presence of metal caused by anthropogenic activities for the removal of toxic substances from the and environmental factors (16, 17). Heavy metals environment have gained importance. Detection are introduced into the marine environment in of bacteria that are resistant to heavy metals in different ways. Accumulation in sediment can affect natural environments constitutes the first step to aquatic life negatively for a long time. Bacteria provide data for remediation studies. that will take part in the transformation of heavy Bacteria are some of the most important metal salts into harmless forms must be resistant components in marine ecosystems. Since bacteria to the heavy metals. Bacteria that cannot adapt to adapt to new conditions created by environmental the changes metabolically will be eliminated and variables around them, knowledge of bacteria therefore various pollution inputs accumulated provides data in terms of defining environmental in the sediment will affect the composition of factors, public health status and ecosystem microbial diversity. Sediments containing harmful, function. Marine areas are exposed to domestic and inorganic or organic particles are relatively industrial wastes depending on local technology heterogeneous in terms of physical, chemical and levels and population. Many xenobiotic micro biological properties and are an important source pollutants, antibiotic derivatives and metabolites of heavy metal contamination (11). It has been reach the sea from human activity. This concerning reported that microplastics mediate the spread of issue is considered an important factor for global metal- and antibiotic-resistant pathogens due to health with respect to the evolution and detection their ability to adsorb various pollutants (18, 19). of antibiotic resistance in bacterial pathogens (5). Bacteria resistant to heavy metals in marine areas Since the spread of antimicrobial resistance is not have developed various resistance mechanisms restricted by phylogenetic, geographic or ecological to counteract heavy metal stress. Only bacteria borders, studies describing regional status of that can withstand the current heavy metal bacterial resistance in natural areas are important. concentration can survive in these areas. Antibiotic resistance can spread rapidly among Heavy metals accumulate in biota via food chains bacterial species (6). It is known that the and are transferred between organisms in marine occurrence of antibiotic-resistant bacteria in environments. This cumulative process, named natural environments reduces the effectiveness of biomagnification, is higher in the sea than in antibiotics in the treatment of infectious diseases. terrestrial environments (15, 20) and this implies Due to the increasing global resistance of bacteria significant effects of heavy metal pollution in against antibiotics, humanity is constantly being marine areas. On the other hand, heavy metal- forced to develop new antibiotic derivatives. resistant bacteria can play a role in detoxification by This vicious circle is one of the most important converting a toxic form into a harmless form through problems of our age and poses a threat for the biotransformation mechanisms that develop in future. Thus, it is important to know the resistance natural environments. These mechanisms include levels of bacteria and to produce regional antibiotic the formation and sequestration of heavy metals resistance profiles in natural areas. Aquatic in complexes and the reduction of a metal to a less environments constitute a way to disseminate toxic species (21, 22). not only antibiotic-resistant bacteria but also the Metal-resistant bacteria have developed very resistant genes in natural bacterial habitats (7). efficient and varying mechanisms for tolerating It has been well documented that the aquatic high levels of toxic metals and thus they carry an environment is a potential reservoir of antibiotic- important potential for controlling heavy metal resistant bacteria, furthermore the prevalence and pollution (23). In many prokaryotes, it has been

508 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15953337767424 Johnson Matthey Technol. Rev., 2020, 64, (4) shown that the mechanism for resisting heavy 2. Material and Methods metals develops over time. This process has 2.1 Sampling Area been studied in species such as Escherichia coli and Staphylococcus aureus. It is reported that Güllük Bay is an important location due to its many different species of Pseudomonas, Bacillus, natural resources. The region is open to different Enterobacter, Providencia and Chryseobacterium environmental influences and inputs due to are efficient for reducing heavy metals (1–4). tourism, port activities, marine transportation, It is known that the occurrence of bacteria domestic and industrial wastes and fish farms. resistant to antibiotics and heavy metal salts in The bay is also affected by the presence of Sarıçay the sea is related to the pollutants present in the Creek, Kazıklı Port, Güllük Port and Akbük Port environment. For the reasons highlighted above, it (24–26). Fish farms were operated in Güllük Bay is important to determine the profile of antibiotic until 2008. Although they have been relocated and heavy metal-resistant bacteria in marine away from the coastal regions to an offshore area, environments. Marine areas which have different the indirect effects of this long-time pollution may environmental inputs present novel media for have contributed to the sediment. bacterial studies. The export of feldspar and bauxite from the region For the present study, the Güllük Bay of the has been conducted from ports within the borders Aegean Sea, Turkey, was chosen since it is a of Güllük Bay. The port is mainly used by dry cargo dynamic area due to marine transportation, and other cargo-type ships. It is reported that an seasonal population growth depending on tourism, annual average of 800,000 tonnes of ballast water aquaculture, recreational and agricultural activities is transported to the bay from 157 different ports. and terrestrial pollution inputs transported from The amount of ballast water carried is reported rivers. as: 68% from the Mediterranean, 21% from the Probable faecal source analysis conducted in Aegean Sea, 7% from the Sea of Marmara, 2% Güllük Bay showed that the primary source of the from the Atlantic Ocean and 1% from the Black detected bacteriological pollution is anthropogenic Sea and Red Sea, respectively (37). The operation (24). A significant part of domestic wastewater of many tourism-oriented boats in Güllük Bay is in the region collects in sealed septic tanks. It also among the possible polluters of the bay due is possible for the wastewater to reach the sea to bilge water and wastewater. More than half of by mixing the sedimentary septic tanks with Turkey’s sea bream and sea bass production was groundwater. Chemical and biological studies in farms operating in the coastal areas of the (24– 33) confirm that regional pollutants have Güllük Bay for many years. These farms have reached Güllük Bay. been operating in the offshore areas of the region It is well known that sewage transported via for the past 10 years. The domestic wastewater domestic wastewater carries antibiotics to marine of the human population, reaching approximately environments. This has an effect on metabolic 50,000 around the region in the summer months, capabilities of bacteria in marine environments. and the wastes of small industrial establishments For example, β-lactam antibiotic derivatives used such as yogurt, yeast and olive oil producers that for human infection treatment may enter marine directly reach streams are the other main sources environments via domestic wastewater. Bacteria of pollution in Güllük Bay. The population of the may obtain resistance via intercellular contact Bodrum peninsula, which is 25,000 in winter, can mostly using a conjugation mechanism (34). reach 1,500,000 in summer (27). The change in The existence of antibiotic-resistant bacteria is the population between the seasons was among an indicator of domestic pollution. Furthermore, the biggest pollution sources according to the antibiotic-resistant bacteria may cause a vicious terrestrial bioindicator bacteria distribution in cycle. This problem has grown in recent years due coastal areas in the region (24, 26). to systematic use of antibiotics in animal husbandry Sampling stations were selected to represent and overuse of antibiotics (35, 36). tourist areas (G1, G5, G7, G8); harbours (G4, G6); The frequencies of heavy metal-resistant bacteria fresh water entry-exit points of the Sarıçay Creek and antibiotic-resistant bacteria were investigated (G9); fish farms (G11, G12, G13); and the deepest in seawater and sediment samples collected from point in the bay as a reference station (G14). Güllük Bay in the period between May 2011 and Figure 1 shows the location of Güllük Bay and the February 2013. sampling stations.

Code Mid point, m Max. depth, m

G1 25 50

G1 G4 25 50

G5 10 20 G4 G6 11 22 G7 14 28 G6 G14 G5 G8 G8 8 16

G7 G9 G9 4 8

G10 18 37 G10 Black Sea G11 5 10 G13 Marmara Sea G12 25 50 G11 Aegean TURKEY G12 Sea G13 12 25

G14 33 66 Mediterranean Sea

Fig. 1. Location of Güllük Bay and seawater (0–30 cm surface, mid-point and bottom-point) and sediment sampling stations

2.2 Sampling 2.3 Bacterial Isolation and Identification Seawater and sediment samples were collected from 12 different sampling stations in Güllük Bay Bacterial heavy metal and antibiotic resistance between May 2011 and February 2013. Three were tested in heterotrophic aerobic bacteria units of seawater samples were taken from each isolated from seawater and sediment samples. station at surface (0–30 cm), mid-point and Heavy metal and antibiotic resistance of indicator bottom-point water (Figure 1). In each sampling bacteria (faecal coliform, coliform and faecal process covering 12 stations, 36 seawater Streptococcus) isolated from the seawater samples samples were collected. In the spring and summer were also tested. months monthly, at other times seasonally, a total of 432 seawater samples were collected in the 2.3.1 Seawater Samples period between May 2011 and February 2013. The seawater samples were collected using a Indicator bacteria and heterotrophic aerobic bacteria Nansen bottle that was cleaned with acid (10% analyses were performed on the seawater samples. HCl in distilled water), sterilised with alcohol The membrane filtration technique was used to (50:50, v/v) and rinsed with sterile water. The detect indicator bacteria. A sample containing seawater samples were then transferred into 300 ml seawater was diluted serially (10–5 dilution) brown sterile glass bottles and transported to the and filtered through membrane filters (0.45 µm, laboratory as a cold chain. Sartorius AG, Göttingen, Germany). The filters Surface sediment samples were collected using were placed on m-Endo, m-FC and m-Azide media Ekman grab (HYDRO-BIOS Apparatebau GmbH, (Sartorius AG). The plates were incubated for 24 h Germany, 15 × 15) from the sampling stations (at 37 ± 0.1°C; at 44 ± 0.1°C for m-FC). Brown‑red which have various depths from 8 m to 66 m colonies growing on the azide medium were (Figure 1). A total of 144 units of sediment considered as suspicious faecal Streptococcus, blue samples were collected during the two-year study colonies growing on the m-FC medium as suspicious from 12 stations (one from each station). The faecal coliform and yellow-green colonies with yellow- sediment samples were transferred into sterile zip metallic gloss on the m-Endo medium as suspicious seal bags from Ekman grab and transferred in the coliform. Cytochrome oxidase test (API® 20 Strep, cold chain to the laboratory. bioMérieux, France) was performed on suspicious

510 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15953337767424 Johnson Matthey Technol. Rev., 2020, 64, (4) coliform colonies and oxidase negative colonies were Further processes related to heterotrophic bacteria evaluated numerically. Cytochrome oxidase (API® 20 identification were continued by using ®VITEK 2 Strep, bioMérieux) and indole tests were performed Compact 30 similarly to the seawater samples on the suspicious colonies of faecal coliform. described above. Colonies with oxidase negative and indole positive results were evaluated as faecal coliform. Suspicious 2.4 Bacterial Resistance Against Streptococcus colonies, to which the catalase test Antibiotics was applied (1 ml, 3% H2O2), were incubated on Bile Esculin Agar (BEA) for 18 h at 37°C for esculin The antibiotic resistance of the isolates was hydrolysis and 40% bile resistance control. Blackening examined by the Kirby–Bauer method with slight in the medium and the formation of black shadow modifications. Two or three colonies of each isolate around the colony, positive of esculin hydrolysis, were suspended with 5 ml of DifcoTM Marine Broth and the number of colonies showing growth in the 2216 and diluted with sterile water against the medium were evaluated as 40% bile resistant, and 0.5 McFarland turbidity standard to approximately catalase negative and breeding colonies in BEA 106 cells ml–1 and swabbed as 2 ml on DifcoTM were evaluated as faecal Streptococcus. Counted Marine Agar 2216. Antibiotic discs (Oxoid, UK) colonies were multiplied by the 10–5 dilution factor to containing ampicillin (10 µg), nitrofurantoin determine the number of colony forming units (CFU) (300 µg), oxytetracycline (30 µg), sulfonamide 100 ml–1 in the original sample (38). (300 µg), rifampicin (2 µg), tetracycline (10 µg) The spread plate technique was used for and tetracycline (30 µg) were incubated for two to heterotrophic aerobic bacteria analyses in seawater. three days at 37°C. The results were interpreted Seawater samples 0.1 ml with 10–5 dilution were according to the guidelines of the Clinical Laboratory used for duplicate spreading on the DifcoTM Marine Standard Institute (CLSI) (41). All isolates that Agar 2216 (Becton, Dickinson and Company, USA) showed resistance were classified as ‘resistant’. and the plates were incubated for five days at Other isolates that did not show resistance were 22 ± 0.1°C. At the end of the incubating period, classified as ‘sensitive’ or ‘susceptible’. counted colonies were multiplied by the 10–5 –1 dilution factor to determine the number of CFU ml 2.4.1 Multiple Antibiotic Resistance in the original sample. An average of 10 different colonies were picked and restreaked several times The multiple antibiotic resistance (MAR) index of to obtain pure cultures. The pure isolates were a given sample was calculated by the equation: Gram-stained. For identification of spore-forming a/ (bc), where a represents the aggregate antibiotic bacilli, the isolates were stained with Indian ink resistance score of all isolates from a sample; b is according to the negative staining technique the total number of isolates; and c is the number of and were evaluated using a light microscope isolates from a sample (42). Bacterial isolates that (Nikon E110, Nikon, Japan). The isolates were displayed resistance to three or more antibiotic then tested using Gram‑negative fermenting agents were designated as multiple antibiotic and non‑fermenting bacilli (GN), Gram-positive resistant (ranging from two to 10). cocci and non-spore-forming bacilli (GP) and Gram‑positive spore-forming bacilli (BCL) cards in 2.5 Bacterial Resistance Against the automated micro identification system VITEK® Heavy Metal Salts 2 Compact 30 (bioMérieux) (39). Different concentrations (50 μg ml–1, 100 μg ml–1, –1 –1 –1 2.3.2 Sediment Samples 150 μg ml , 200 μg ml and 250 μg ml ) of heavy metal salts (FeSO4, ZnSO4, CuSO4, Cr2(SO4)3 and

The spread plate technique was used for Pb(NO3)2) were used to test the bacteria resistivity heterotrophic aerobic bacteria analyses in sediment against iron, zinc, copper, chromium and lead. samples. Each sediment sample was mixed and The microdilution method was followed with minor homogenised. Then 1 g sample was taken from modifications to determine the resistance of isolates each and serially diluted with sterile commercial to heavy metals (43). Stock solutions of metal seawater. 0.1 ml samples of 10–5 dilutions were salts prepared in distilled water were sterilised by taken and spread on DifcoTM Marine Agar 2216. The filtration (0.20 μm). In U-well microtiter plates, plates were incubated for five days at 22 ± 0.1°C. serial dilutions of heavy metals were prepared Growing colonies were evaluated as CFU g–1 (40). and then each well was inoculated with bacteria

511 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15953337767424 Johnson Matthey Technol. Rev., 2020, 64, (4) inoculation. The OxoidTM Turbidometer (Thermo Shewanella algae and Vibrio parahaemolyticus Fisher Scientific Inc, USA) provides the inoculum isolates from the sediment samples showed density standardisation for 0.5 McFarland which is resistance to all antibiotics (Table I). necessary to ensure accurate reproducible results. The highest number of antibiotic-resistant Before the addition of bacterial inoculation, no bacteria were detected from the sediment precipitation was seen. The plates were incubated samples. The frequency of resistant bacteria (%) at 37°C for 24 h and then examined for visual to oxytetracycline (30 µg), nitrofurantoin (300 µg), turbidity. The lowest concentration of the metal rifampicin (2 µg), tetracycline (10 µg), tetracycline salt, at which growth was inhibited (indicated (30 µg), sulfonamide (300 µg) and ampicillin by lack of turbidity), was taken as the minimum (10 µg) from the seawater and sediment samples inhibitory concentration (MIC) (44) Samples of are shown in Figure 2. The frequencies of antibiotic 10 μl were drawn from each well without turbidity resistance in bacteria species from seawater and and were subcultured on agar plates to determine sediment samples are shown in Figure 3. bactericidal concentration. A total of 258 and 158 isolates were tested Reference strains of Escherichia coli (ATCC® against antibiotics from seawater and sediment 25922TM), Salmonella enterica (ATCC® 2577TM) samples, respectively. The frequencies of and Staphylococcus epidermidis (ATCC® 12228TM) resistance against seven antibiotics in bacteria which are susceptible to Cu2+, Zn2+, Pb2+, Cr2+ and species isolated from the seawater samples Fe3+ and metal-free plates were used in the control were recorded as 49% in Gammaproteobacteria, tests to evaluate the viability of the strains and 22% in Αlphaproteobacteria, 3% in culture media. All of the experiments were carried Betaproteobacteria, 14% in Bacilli, 8% in out in triplicate. Flavobacteriia and 4% in Actinomycetales. The resistance frequencies against seven antibiotics 3. Results in bacteria isolated from the sediment samples were recorded as 43% in Gammaproteobacteria, 3.1 Bacterial Resistance Against 34% in Bacilli, 7% in Αlphaproteobacteria, 7% Antibiotics in Betaproteobacteria, 7% in Flavobacteriia and 2% in Actinomycetales. Table I shows the antibiotic-resistant, intermediate or susceptible bacteria species isolated from the 3.2 Multiple Antibiotic Resistance seawater and sediment samples in this study. Indexes Bacterial species isolated from the seawater samples showed considerable resistance to rifampicin (98%), The MAR index was calculated for each of the sulfonamide (98%) and ampicillin (76%) and antibiotic-resistant bacteria. If the MAR index considerable sensitivity to tetracycline-30 µg (52%), is lower than 0.2, it shows a non-point based tetracycline-10 µg (39%) and oxytetracycline source of pollution and if it is higher than 0.2 it (33%). Almost all the bacterial species isolated from shows point‑based pollution and a high risk of sediment samples showed resistance to rifampicin contamination by excessive antibiotic presence (100%), sulfonamide (100%), ampicillin (100%), (23). Table II shows the MAR indexes. nitrofurantoin (98%), tetracycline-30 µg (100%), The MAR indexes of the study showed possible tetracycline-10 µg (100%) and oxytetracycline exposure of these bacterial isolates to the tested (98%) while they showed almost no sensitivity antibiotics. The MAR index of bacteria isolated to antibiotics except nitrofurantoin (2%) and from all stations around fish farm areas (0.0576) oxytetracycline (2%). Pseudomonas aeruginosa was 2.6 times greater than the MAR index for the (24%) and Sphingomonas paucimobilis (20%), combined non-fish farm areas (0.022). isolated from seawater samples, showed higher resistance to antibiotics than did Raoultella 3.3 Bacterial Resistance Against oxytica, Staphylococcus xylosus, Kocuria kristinae, Heavy Metals Aeromonas salmonicida and Proteus vulgaris strains. On the contrary, Aeromonas caviae, The frequencies of heavy metal resistance in the Alicyclobacillus acidoterrestris, Brevundimonas bacteria species isolated from the seawater samples diminuta, Chryseobacterium indologenes, were recorded as 76.72% in Gammaproteobacteria, Lactococcus garvieae, Neisseria animaloris, 71.82% in Αlphaproteobacteria, 80.01% in Pseudomonas aeruginosa, Serratia marcescens, Bacilli, 56.92% in Flavobacteriia and 75% in

512 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15953337767424 Johnson Matthey Technol. Rev., 2020, 64, (4) OT (30 µg) R: 66.7% I: 33.3% S: 0.0% R: 25% I: 0.0% S: 75% R: 26.31% I: 34.21% S: 39.47% R: 25% I: 25% S: 50% R: 0.0% I: 0.0% S: 100% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 0.0% I: 100% S: 0.0% R: 0.0% I: 50% S: 50% R: 50% I: 0.0% S: 50% R: 0.0% I: 100% S: 0.0% R: 0.0% I: 0.0% S: 100% F/M (300 µg) R: 66.7% I: 0.0% S: 33.3% R: 50% I: 0.0% S: 50% R: 60.52% I: 0.0% S: 39.47% R: 75% I: 0% S: 25% R: 0.0% I: 0.0% S: 100% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 50% I: 0.0% S: 50% R: 50% I: 0.0% S: 50% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% RD (2 µg) R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 97.36% I: 0.0% S: 2.63% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 50% I: 0.0% S: 50% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% TE (10 µg) R: 33.3% I: 33.3% S: 33.3% R: 50% I: 0.0% S: 50% R: 42.10% I: 7.89% S: 50% R: 0% I: 25% S: 75% R: 0.0% I: 0.0% S: 100% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 0.0% I: 0.0% S: 100% R: 50% I: 0.0% S: 50% R: 50% I: 0.0% S: 50% R: 0.0% I: 0.0% S: 100% R: 0.0% I: 0.0% S: 100% R: 97.38% I: 0.0% S: 2.63% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% S (300 µg) R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% TE (30 µg) R: 33.3% I: 0.0% S: 66.7% R: 50% I: 0.0% S: 50% R: 31.57% I: 5.26% S: 63.15% R: 0.0% I: 0.0% S: 100% R: 0.0% I: 0.0% S: 100% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 0.0% I: 0.0% S: 100% R: 50% I: 0% S: 50% R: 50% I: 0.0% S: 50% R: 0.0% I: 0.0% S: 100% R: 0.0% I: 0.0% S: 100% a R: 71.05% I: 2.63% S: 26.31% R: 50% I: 0.0% S: 50% R: 100% I: 0.0% S: 0.0% R: 50% I: 0.0% S: 50% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% Antibiotics AM (10 µg) R: 66.7% I: 33.3% S: 0.0% R: 75% I: 0.0% S: 25% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0%

(3) (3) (38) (4) (4) (4) (3) Bacterial isolates tested (n) Brevundimonas diminuta Brevundimonas vesicularis Sphingomonas paucimobilis Sphingomonas thalpophilum Burkholderia cepacia (3) Burkholderia mallei (3) Neisseria animaloris (3) Acinetobacter lwoffii (3) Aeromonas hydrophila (4) Aeromonas salmonicida Aeromonas sobria Aeromonas veronii Antibiotic Resistant, Intermediate or Susceptible Bacteria Species Isolated from Seawater and Sediment from Species Isolated Susceptible Bacteria or Resistant, Intermediate Antibiotic

Order/class tested (%) Proteobacteria/ Alpha proteobacteria (27%) Proteobacteria/ Beta proteobacteria (%) Proteobacteria/ Gamma proteobacteria (53%)

Sample r Seawate Table I

513 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15953337767424 Johnson Matthey Technol. Rev., 2020, 64, (4) R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 33.3% I: 66.6% S: 0.0% R: 0.0% I: 0.0% S: 100% R: 0.0% I: 0.0% S: 100% R: 50% I: 50% S: 0% R: 75% I: 0.0% S: 25% R: 100% I: 0.0% S: 0.0% R: 0.0% I: 0.0% S: 100% R: 100% I: 0.0% S: 0.0% R: 90.91% I: 0.0% S: 9.09% R: 0.0% I: 0.0% S: 100% R: 100% I: 0.0% S: 0.0% OT (30 µg) R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 33.3% I: 0.0% S: 66.6% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100.0% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 0.0% I: 0.0% S: 100% R: 100% I: 0.0% S: 0.0% R: 90.91% I: 0.0% S: 9.09% R: 0.0% I: 0.0% S: 100% R: 100% I: 0.0% S: 0.0% F/M (300 µg) R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100.0% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% RD (2 µg) R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 66.6% I: 0.0% S: 33.3% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 89.3% I: 0.0% S: 10.7% R: 100% I: 0.0% S: 0.0% R: 0.0% I: 0.0% S: 100% R: 100% I: 0.0% S: 0.0% R: 90.91% I: 0.0% S: 9.09% R: 0.0% I: 0.0% S: 100% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 0.0% I: 100% S: 0.0% R: 92.9% I: 3.5% S: 3.5% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% TE (10 µg) R: 100% I: 0.0% S: 0.0% R: 66.6% I: 0.0% S: 33.3% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% S (300 µg) R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 33.3% I: 0.0% S: 66.6% R: 0.0% I: 100% S: 0.0% R: 0.0% I: 0.0% S: 100% R: 50% I: 0.0% S: 50% R: 71.4% I: 10.7% S: 17.8% R: 100% I: 0.0% S: 0.0% R: 0.0% I: 0.0% S: 100% R: 100% I: 0.0% S: 0.0% R: 90.91% I: 0.0% S: 9.09% R: 0.0% I: 0.0% S: 100% R: 100% I: 0.0% S: 0.0% R: 0.0% I: 0.0% S: 100% R: 100% I: 0.0% S: 0.0% R: 0.0% I: 50% S: 50% R: 78.5% I: 10.7% S: 10.7% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% Antibiotics TE (30 µg) R: 100% I: 0.0% S: 0.0% R: 33.3% I: 33.3% S: 33.3% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0%

(4) (3) group (3) (3) (3) (30) (3) (3) subsp. (13) (6) (4)

ozaenae dissolvens Bacterial isolates tested (n) AM (10 µg) Citrobacter sedlakii (3) Cronobacter dublinensis lausannensis Enterobacter aerogenes Enterobacter cloacae subsp. Enterobacter cloacae (4) Enterobacter cloacae complex Escherichia coli Klebsiella pneumoniae subsp. Pasteurella canis Proteus vulgaris Proteus penneri Pseudomonas aeruginosa Raoultella ornithinolytica Raoultella ytica

Order/class tested (%)

Sample Seawater

514 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15953337767424 Johnson Matthey Technol. Rev., 2020, 64, (4) OT (30 µg) R: 66.6% I: 0.0% S: 33.4% R: 54.54% I: 36.36% S: 9.06% R: 20% I: 0.0% S: 80% R: 0.0% I: 50% S: 50% R: 100% I: 0.0% S: 0.0% R: 0.0% I: 100% S: 0.0% R: 60% I: 20% S: 20% R: 66.7% I: 0.0% S: 33.3% R: 100% I: 0.0% S: 0.0% R: 0.0% I: 0.0% S: 0.0% R: 33.3% I: 0.0% S: 66.7% R: 45.45% I: 18.18% S: 36.36% R: 66.6% I: 0.0% S: 33.3% F/M (300 µg) R: 100% I: 0.0% S: 0.0% R: 63.63% I: 0.0% S: 36.36% R: 100% I: 0.0% S: 0.0% R: 0.0% I: 0.0% S: 100% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 66.7% I: 0.0% S: 33.3% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 54.54% I: 0.0% S: 45.45% R: 66.6% I: 0.0% S: 33.3% RD (2 µg) R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 33.3% I: 66.7% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% TE (10 µg) R: 66.6% I: 0.0% S: 33.4% R: 72.72% I: 0.0% S: 27.27% R: 20% I: 50% S: 50% R: 0.0% I: 0.0% S: 100% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 80% I: 20% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 66.7% I: 0.0% S: 33.3% R: 45.45% I: 0.0% S: 54.54% R: 66.6% I: 0.0% S: 33.3% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% S (300 µg) R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 0.0% I: 100% S: 100% R: 66.7% I: 0.0% S: 33.3% TE (30 µg) R: 66.6% I: 0.0% S: 33.4% R: 45.45% I: 18.18% S: 36.36% R: 40% I: 0.0% S: 60% R: 0.0% I: 0.0% S: 100% R: 100% I: 0.0% S: 0.0% R: 0.0% I: 0.0% S: 100% R: 60% I: 0.0% S: 40% R: 66.7% I: 0.0% S: 33.3% R: 100% I: 0.0% S: 0.0% R: 0.0% I: 0.0% S: 100% R: 33.4% I: 0.0% S: 66.7% R: 36.36% I: 0.0% S: 63.63% R: 66.6% I: 0.0% S: 33.3% R: 81.81% I: 0.0% S: 18.18% R: 50% I: 0.0% S: 50% R: 0.0% I: 100% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 66.7% I: 0.0% S: 33.3% R: 54.54% I: 18.18% S: 27.27% R: 100% I: 0.0% S: 0.0% Antibiotics AM (10 µg) R: 100% I: 0.0% S: 0.0% R: 80% I: 0.0% S: 20% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 0.0% I: 0.0% S: 100% R: 33.3% I: 0.0% S: 66.7% (5) (4) (7) (5) (3) . (13) (13) (7) spp (5) (3) (3) Bacterial isolates tested (n) Serratia marcescens (5) Shewanella putrefaciens Stenotrophomonas maltophilia Vibrio vulnificus Enterococcus faecium (3) Alicyclobacillus acidocaldarius Bacillus cereus Bacillus pumilus Staphylococcus xylosus Staphylococcus aureus Staphylococcus warneri Chryseobacterium indologenes Myroides

Order/class tested (%) Firmicutes/ Bacilli (9%) Bacteroidetes/ (8%) Flavobacteriia

Sample r Seawate

515 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15953337767424 Johnson Matthey Technol. Rev., 2020, 64, (4) OT (30 µg) R: 0.0% I: 0.0% S: 100% R: 50% I: 0.0% S: 50% R: 0.0% I: 1000% S: 0.0% R: 0.0% I: 100% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% F/M (300 µg) R: 0.0% I: 0.0% S: 100% R: 100% I: 0.0% S: 0.0% R: 0.0% I: 0.0% S: 100% R: 50% I: 0.0% S: 50% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% RD (2 µg) R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% TE (10 µg) R: 0.0% I: 0.0% S: 100% R: 50% I: 0.0% S: 50% R: 100% I: 0.0% S: 0.0% R: 50% I: 0.0% S: 50% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% S (300 µg) R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% TE (30 µg) R: 0.0% I: 0.0% S: 100% R: 100% I: 0.0% S: 0.0% R: 0.0% I: 0.0% S: 100% R: 50% I: 0.0% S: 50% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 0.0% I: 0.0% S: 100% R: 100% I: 0.0% S: 0.0% R: 0.0% I: 100% S: 0.0% R: 50% I: 0.0% S: 50% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% Antibiotics AM (10 µg) R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% (1) (1) (4) (3) (4) (3) (1) (1) (1) (1) (1) Bacterial isolates tested (n) Dermacoccus nishinomiyaensis Kocuria kristinae Kocuria varians Micrococcus luteus Brevundimonas diminuta Sphingomonas paucimobilis Sphingomonas thalpophilum Neisseria animaloris (3) Chromobacterium violaceum Aeromonas caviae Aeromonas sobria Pseudomonas aeruginosa Serratia marcescens (5)

Order/class tested (%) Actinobacteria/ Actinomycetales (3%) Proteobacteria/ Alpha proteobacteria (7%) Proteobacteria/ Beta proteobacteria (7%) Proteobacteria/ Gamma proteobacteria (43%)

Sample r Seawate Sediment

516 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15953337767424 Johnson Matthey Technol. Rev., 2020, 64, (4) OT (30 µg) R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 66.7% I: 0.0% S: 33.3% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% F/M (300 µg) R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 66.7% I: 0.0% S: 33.3% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% RD (2 µg) R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% TE (10 µg) R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% S (300 µg) R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% TE (30 µg) R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% Antibiotics AM (10 µg) R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% R: 100% I: 0.0% S: 0.0% (15) (11) (11) (11) (23) (11) (11) (12) . (11) (12) spp Bacterial isolates tested (n) Shewanella algae Shewanella putrefaciens Vibrio alginolyticus (14) Vibrio fluvialis Vibrio parahaemolyticus (13) Vibrio vulnificus Alicyclobacillus acidoterrestris Bacillus cereus Bacillus pumilus Lactococcus garvieae (13) Chryseobacterium indologenes Myroides Micrococcus lylae

Order/class tested (%) Firmicutes/ Bacilli (34%) Bacteroidetes/ Flavobacteriia (7%) Actinobacteria/ Actinomycetales (2%)

Sample t Sedimen Ampicillin (AM, 10 µg), nitrofurantoin (F/M, 300 µg), oxytetracycline (OT, 30 µg), sulfonamide (S, 300 µg), rifampicin (RD, 2 µg), tetracycline (TE, 10 µg) and tetracycline (TE, 30 µg). Resistant (TE, 30 µg). Resistant (TE, 10 µg) and tetracycline 2 µg), tetracycline 300 µg), rifampicin (RD, 30 µg), sulfonamide (S, (OT, (F/M, 300 µg), oxytetracycline Ampicillin (AM, 10 µg), nitrofurantoin a (R), intermediate (I) or susceptible (S)

Sediment Seawater Fig. 2. The frequency of bacteria resistant to Ampicillin, 10 µg specific antibiotics (%) Sulfonamide, 300 µg in the seawater and sediment samples Oxytetracycline, 30 µg Tetracycline, 30 µg Tetracycline, 10 µg Rifampicin, 2 µg Nitrofurantoin, 300 µg

0 20 40 60 80 100 Resistance, %

(a) (b) Betaproteobacteria 3% Actinomycetales 2% Actinomycetales 4% Flavobacteriia 7% Flavobacteriia 8% Betaproteobacteria 7% Bacilli 14% Alphaproteobacteria 7% Alphaproteobacteria 22% Bacilli 34%

Gammaproteobacteria 49% Gammaproteobacteria 43%

0 20 40 60 0 10 20 30 40 50 Resistance, % Resistance, % Fig 3. The frequencies of antibiotic resistance in bacteria species from (a) seawater samples and (b) sediment samples

Table II Multiple Antibiotic Resistance Indexes and Resistance Ratios Number of antibiotics to which bacteria show MAR Index Resistance, % p-value resistance 1 0.0052 0.75187 0.3635 2 0.0104 18.0451 0.6513 3 0.0022 12.7819 0.1323 4 0.0294 12.7819 0.1325 5 0.048 9.7744 0.3635 6 0.0576 14.2857 0.4084 7 0.0208 31.57894 0.1234

Actinomycetales. The frequencies of resistance to and Fe2+ were detected as an average of 33.3%, Cu2+, Zn2+, Pb2+, Cr2+ and Fe2+ were detected as 30.3%, 25.5%, 35.3% and 28.4% respectively in an average of 58.3%, 33.8%, 32.1%, 31.0% and 158 strains isolated from the sediment samples. 25.2% respectively in 258 bacterial strains isolated The frequencies of heavy metal-resistant bacteria from seawater samples. isolated from sediment samples were higher than The frequencies of heavy metal resistance in the frequencies of heavy metal-resistant bacteria bacteria species isolated from the sediment samples isolated from the seawater samples. Table III were recorded as 100% in Αlphaproteobacteria, shows the heavy metal resistance in bacteria 100% in Betaproteobacteria, 97.5% in isolates from seawater and sediment in Güllük Bay. Flavobacteriia, 95% in Gammaproteobacteria, The MICs of the isolates ranged from 0.004 mM 72.5% in Bacilli and 66.6% in Actinomycetales. The to 2.5 mM. The isolates from sediment samples frequencies of resistance to Cu2+, Zn2+, Pb2+, Cr2+ obtained from stations close to fish farms showed

Table III Heavy Metal Resistance in Bacteria Species from Seawater and Sediment in Güllük Bay, Turkey Resistant Metal concentrations, µg ml–1 Isolates Heavy Sampling isolates metals sides 0.8 1.6 3.1 6.5 12.5 25 50 100 200 >200 n n % Seawater 3 3 3 7 8 9 10 11 6 – 258 149 58.3 Cu2+ Sediment 7 4 9 3 6 17 17 29 11 – 158 53 33.3 Seawater 3 6 4 9 8 7 11 8 2 – 258 86 33.8 Zn2+ Sediment 4 2 2 8 4 19 36 22 6 – 158 48 30.3 Seawater 1 8 2 7 7 10 11 12 2 – 258 82 32.1 Pb2+ Sediment 1 9 4 4 3 18 13 23 17 – 158 40 25.5 Seawater 3 2 7 4 6 7 12 6 4 16 258 79 31.0 Cr2+ Sediment 3 4 5 8 4 9 11 10 27 32 158 56 35.3 Seawater 1 2 – – 9 24 15 6 – – 258 67 25.2 Fe2+ Sediment 3 13 3 5 9 15 24 29 – – 158 45 28.4 Total number of tested isolates 416 higher frequency of resistance against chromium, in another study (48). For example, there were copper and zinc than other stations. The highest significant increases in numbers of bacteria resistant resistance (MIC value: 2.5 mM) was displayed to oxytetracycline, oxolinic acid and florfenicol against Cr+ by all isolates. Bacillus isolates showed a in sediments from an aquaculture site compared higher resistance to chromium, lead and copper than with those from a non-aquaculture control site. Pseudomonas isolates, and Vibrio isolates showed Interestingly, in another study a similar number higher resistance to zinc, copper and chromium than of antibiotic-resistant bacteria were isolated from Escherichia coli. Tolerance to the maximum MIC aquaculture and non-aquaculture sites (49). Gram- (>2.5 mM) for chromium was 10.1% for Bacillus negative bacteria (predominantly Plesiomonas and 0.8% for Pseudomonas isolates. Bacillus isolates shigelloides and Aeromonas hydrophila) were from sediment samples showed higher resistance to isolated from aquaculture ponds in the south- chromium, lead, iron and copper than Klebsiella spp. eastern USA and it was reported that antibiotic and Escherichia coli strains from seawater samples. resistance to tetracycline, oxytetracycline, Similarly, Shewanella spp. and Serratia spp. strains chloramphenicol, ampicillin and nitrofurantoin from the sediment samples also showed higher were higher in antibiotic-treated ponds compared resistance than the species mentioned above. to non-treated rivers (50). It was determined that bacteria isolated from Sopot Beach, Poland, were 4. Discussion resistant to ampicillin (51). A high percentage of bacteria were reported as resistant to streptomycin Indicator bacteria levels reported in Güllük Bay (100%), cefazolin (89.8%), ampicillin (83.7%) and and the presence of pathogenic bacteria (25, 26) trimethoprim-sulfamethoxazole (69.4%), whereas support the relationship between the resistance data a low percentage of bacteria were resistant to detected in the current study with bacteriological cefepime (12.3%) and meropenem (14.3%) in the pollution levels. In the present study, bioindicator aquaculture region of İskenderun Bay, Turkey (52). bacteria showing human-induced pollution input In the current study, higher numbers of sulfonamide, isolated from seawater had the highest frequency rifampicin and ampicillin-resistant bacteria were of resistance against nitrofurantoin (100%) and recorded in the stations around aquaculture areas sulfonamide (95%). Sulfonamides were the first than other stations. Sphingomonas paucimobilis, antibiotics developed for clinical use. Sulfonamides Escherichia coli and Enterobacter cloacae isolated have been widely used to treat bacterial and from both seawater and sediment at the stations protozoan infections in humans, domestic animals around aquaculture areas had the highest levels and fish since their introduction to clinical practice of antibiotic resistance. The development of in 1935 (45–47). The results of higher resistance resistant pathogens in aquaculture environments against sulfonamide in the present study were is well documented (53, 54) and evidence of similar to the findings of sulfonamide resistance transfer of resistance encoding plasmids between

519 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X15953337767424 Johnson Matthey Technol. Rev., 2020, 64, (4) aquaculture environments and humans has been was found to be 2.6 times greater in the stations presented recently (55). It has been reported that around fish farm areas (0.057) than the other antibiotic‑resistant bacteria are present in a seafood stations (0.022). ecosystem where antibiotics have never been used Marine sediments offer more informative results (56). This is interesting in terms of showing that than seawater about environmental pollution aquaculture areas may be adversely affected by due to the accumulation of various pollutants the presence of environmental antibiotic-resistant at the bottom of the sea, therefore analysis of bacteria. sediments is widely used in tests. The association In the present study, a high percentage of the of microorganisms with sediment particles is one bacteria Sphingomonas paucimobilis were isolated, of the primary factors in assessing microbial fate which was especially prevalent in Güllük Bay. The in aquatic systems. In this study, the bacteria natural habitat of Sphingomonas has not been isolated from sediment in all samples showed a defined, but it is widely distributed in the natural higher resistance rate than bacteria isolated from environment especially in water and soil (57). The seawater. Detection of higher antibiotic resistance second most prevalent species were Escherichia in sediment bacteria than bacteria isolated from coli and Enterobacter cloacae. Escherichia coli is seawater showed that sediment bacteria were an indicator of faecal contamination in aquatic exposed to more antibiotics. Natural ecosystems environments. Enterobacter cloacae is the most containing high concentrations of heavy metals frequent species associated with nosocomial are also frequent. Heavy metal resistance infections along with Klebsiella pneumoniae that genes are commonly found in environmental is a growing problem in human healthcare. The bacteria (71). The resistance to seven highest number of Bacillus cereus was isolated heavy metals has been reported in the order from the sediment underneath fish farms. A few Cu > Mn > Ni > Zn > Pb > Cd > Fe for seawater Bacilli of marine origin have been reported to bacteria isolated from the Golden Horn, Istanbul, produce unusual metabolites different from those Turkey (17). Heavy metal resistance in bacteria isolated from terrestrial bacteria (58). Due to the found in seawater from the Mediterranean has ubiquity and ability of the Bacillus species to survive been reported as Cd > Cu > Cr = Pb > Mn; in under difficult circumstances, Bacillus strains Karataş, Turkey Cd > Cu > Cr = Mn > Pb; and are considered to be species of certain habitats İskenderun Bay, Cu > Cd > Mn > Cr > Pb (72). (59, 60). In the current study, Bacillus pumilus, In the present study, resistance to five different B. thuringiensis, B. mycoides and B. cereus were heavy metals (Zn2+, Pb2+, Cu2+, Cr3+ and isolated from the sediment samples of the stations Fe3+) were investigated for all isolates. Trends around fish farms. in heavy metal resistance vary depending on The high frequency of resistance among bacterial the sample sites: Güllük Bay, fish farm water isolates in the present study confirms the earlier column: Cu > Zn > Pb > Cr > Fe; sediment: reports regarding the role of antimicrobial use that Cr > Cu> Zn > Fe > Pb. Frequency of bacteria plays a role in selecting antibiotic-resistant bacteria resistance to heavy metals shows the direct in water and aquatic sediments (46–52). Many effects of metal pollution. Neisseria animaloris, previous studies have shown that the increases in Aeromonas caviae and Bacillus cereus isolated antibiotic resistance in human medicine, agriculture from sediment samples were the most tolerant and aquaculture are directly related to the amounts of all the heavy metal salts. Chryseobacterium of antimicrobials used (61–65). indologenes displayed the highest degree of Infections caused by antibiotic-resistant bacteria sensitivity to all metal salts while Lactococcus are one of the most important public health garvieae showed the highest degree of sensitivity concerns worldwide. Currently, MARs have been to Zn2+, Pb2+, Cu2+ and Fe3+. Kocuria kristinae, reported in a wide range of human pathogenic Escherichia coli and Acinetobacter lwoffii, which or opportunistic bacteria such as Vibrio sp. (66), were isolated from the seawater underneath the Klebsiella pneumoniae (67), Salmonella sp. (68), fish farm, displayed similar sensitivities to all Pseudomonas aeruginosa and also in pathogens tested heavy metal salts. Resistances to heavy (69, 70). Reservoirs of antibiotic resistance can metals for Aeromonas and Pseudomonas isolates interact between different ecological systems and were similar to those from İskenderun Bay, with potential transfer of resistant bacteria or resistant cadmium, 35.0% and 56.5%; copper, 98.3% and genes from animals to humans may occur through 75.4%; chromium, 38.3% and 31.9%; lead, 1.7% the food chain (70). In the current study, the and 7.2%; manganese, 43.3% and 44.9%; and MAR index of multiple antibiotic-resistant bacteria zinc 35.0% and 41.3%, respectively (72).

Both Gram-positive and Gram-negative bacteria pesticides and heavy metals might encourage can resist heavy metals (73). Resistance to selection and result in antibiotic and heavy metal toxic metals in bacteria probably reflects the resistance. Marine environmental conditions are level of environmental contamination with extremely dynamic compared to the terrestrial these substances and it may be related to the environment, allowing bacteria to bring resistance concentration of bacteria (74). The present project mechanisms they have developed together while found heavy metal pollution in Güllük Bay sediment being adapted to the varying conditions. This samples at all stations. In the sediment samples, makes the isolation of various bacteria useful to the heavy metal contents were reported at varying assess environmental pollution and provides a rates: between 1 μg g–1 and 209 μg g–1 for lead; pathway to possible solutions to remove pollution 10 μg g–1 and 259 μg g–1 for zinc; 1 μg g–1 and from marine environments. For bacteria to take 59 μg g–1 for copper; 0.1 μg g–1 and 46 μg g–1 part in the transformation of any heavy metal salt for chromium; <0.01 μg g–1 and 2.8 μg g–1 for into a harmless form, those bacteria must firstly cadmium; <0.01 μg g–1 and 0.4 μg g–1 for arsenic; be resistant to the heavy metal; thus the data and 0.6% and 5.9% for aluminium, respectively. related to frequency of metal resistant bacteria can The region was defined according to cadmium, lead provide knowledge on the continual accumulation and zinc levels as moderately polluted. Recorded or transformation of heavy metals in the marine high metal values were evaluated as an indicator of environment. domestic and industrial inputs, carried via Sarıçay The findings of the current study provide Creek, port operations and tourism activities data regarding the distribution of heavy metal- within Güllük Bay (75). In the current study, the and antibiotic-resistant bacteria in seawater high frequencies of heavy metal-resistant bacteria and sediment samples of Güllük Bay, Aegean detected in the sediment samples support this Sea, Turkey. As a result, preliminary data on data. Bacterial heavy metal resistance detected candidate bacteria will offer opportunities for in the study may depend on many factors. A further studies on the elimination of heavy metal possible explanation for differences in heavy metal contamination by the detection of heavy metal- resistance is the proximity of Güllük Bay to iron-steel resistant bacteria. factories. Additionally, Güllük Harbour is a serious pollution source. It was reported that 2862‑unit 5. Conclusions ships carried 4.8 million tonnes of ballast water to Güllük Harbour during 2007–2012 (37). Another Analyses of the presence of antibiotic resistance potential source of increased resistance may be the in bacteria provide knowledge on pollution sources discharge of thermal power plants located 107 km, such as septic systems on regional ecosystems. 46 km and 39 km away from Güllük Bay. The Since antibiotic-resistant bacteria can affect effects of thermal power plant discharge on the pathogen virulence, these pollution sources can accumulation of heavy metals have been reported induce pathogens and can create health risks in other studies (29, 75). for both humans and the ecosystem. In the The association between antibiotic resistance present study, bacteria resistant to antibiotics and resistance to heavy metals is quite common and heavy metals in seawater and sediment were in the same organism. The increasing numbers investigated. The bacterial information obtained of antibiotic and heavy metal-resistant bacteria provides essential data for identifying the regional could be a result of gene transfer activities distribution of resistant bacteria. Levels of demonstrating that industrial pollution most resistance against heavy metals and antibiotics in likely selects for antibiotic resistance and vice bacteria isolated from seawater and sediments of versa (58). In this study, similarly, the most the Aegean Sea were quantified. Bacteria isolated antibiotic-resistant bacteria such as Sphingomonas from Güllük Bay sediment were resistant to all paucimobilis, Escherichia coli and Enterobacter antibiotics tested and exhibited higher resistance cloacae were also resistant to heavy metals. than those isolated from seawater. The frequency Metal‑resistant isolates from Güllük Bay also of antibiotic-resistant bacteria was higher around showed high resistance to sulfonamide, rifampicin fish farms and near the exit of Sarıçay Creek. The and ampicillin. Bacteria from different sources widespread resistances of indicator bacteria to such as humans, animals and soil can transfer or antibiotics suggest the presence of anthropogenic exchange their resistance genes. At the same time, influences due to domestic waste and maritime water contaminated with antibiotics, disinfectants, transport.

In order for bacteria to take part in the 4. E. E. Bestawy, S. Helmy, H. Hussien, M. Fahmy transformation of heavy metal salts into harmless and R. Amer, Appl. Water Sci., 2012, 3, (1), 181 forms, they must initially be resistant to heavy 5. H. K. Allen, J. Donato, H. H. Wang, K. A. Cloud- metals. The frequency of resistance thus provides Hansen, J. Davies and J. Handelsman, Nat. Rev. Microbiol., 2010, 8, (4), 251 information regarding the continual accumulation or transformation of heavy metal salts in the 6. H. H. Wang and D. W. Schaffner, Appl. Environ. marine environment. The findings of the present Microbiol., 2011, 77, (20), 7093 research have shown the existing contamination 7. N. Rosenblatt-Farrell, Environ. Health Perspect., status of Güllük Bay via heavy metal and antibiotic 2009, 117, (6), A244 resistance tests. The study region is under pressure 8. K. Kümmerer, J. Antimicrob. Chemother., 2004, of pollution as stated in previous research (25, 26, 54, (2), 311 75) and the bacterial resistance data of the current 9. S. Kim and D. S. Aga, J. Toxicol. Environ. Health: study showed that there is a prevalence of resistant Part B, 2007, 10, (8), 559 bacteria in the region that may be due to indirect 10. A. J. Watkinson, G. B. Micalizzi, G. M. Graham, effects of environmental dynamics and pollution. J. B. Bates and S. D. Costanzo, Appl. Environ. In this study, the presence of higher levels Microbiol., 2007, 73, (17), 5667 of resistant bacteria in sediment compared to 11. J. L. Caplin, G. W. Hanlon and H. D. Taylor, Environ. seawater may indicate the presence of microplastics Microbiol., 2008, 10, (4), 885 in the sediment as well as the probability that the 12. L. Nonaka, K. Ikeno and S. Suzuki, Microbes sediment is a suitable medium for accumulation Environ., 2007, 22, (4), 355 of metals and antibiotics. Further studies on this 13. P. T. P. Hoa, L. Nonaka, P. Hung Viet and S. Suzuki, subject will provide detailed data on the spread of Sci. Total Environ., 2008, 405, (1–3), 377 antibiotic- and metal-resistant bacteria in marine 14. D. I. Andersson and D. Hughes, Nat. Rev. sediments. Microbiol., 2010, 8, (4), 260 The present study showed bacterial responses 15. S. Squadrone, Environ. Monit. Assess., 2020, to environmental stress and influences in terms 192, (4), 238 of antibiotic and heavy metal resistance both in sediment and seawater samples at Güllük Bay, 16. M. L. Nadimpalli, S. J. Marks, M. C. Montealegre, R. H. Gilman, M. J. Pajuelo, M. Saito, P. Tsukayama, Turkey. These findings highlight the necessity of S. M. Njenga, J. Kiiru, J. Swarthout, M. A. Islam, T. holistic assessments with a ‘one health’ approach R. Julian and A. J. Pickering, Nat. Microbiol., 2020, and the need to control bacteria entering marine 5, (6), 787 areas due to human activities, considering the 17. G. Altug and N. Balkis, Environ. Monit. Assess., contributions of resistant bacteria to global 2009, 149, (1–4), 61 distribution. The data may also provide a useful 18. M. Imran, K. R. Das and M. M. Naik, Chemosphere, resource to help identify strains of bacteria for 2019, 215, 846 environmental remediation applications. 19. P. Laganà, G. Caruso, I. Corsi, E. Bergami, V. Venuti, D. Majolino, R. La Ferla, M. Azzaro and S. Acknowledgments Cappello, Int. J. Hyg. Environ. Health, 2019, 222, (1), 89 The authors wish to thank the Scientific and Technical Research Council of Turkey (TÜBITAK, 20. G. Lunde, Environ. Health Perspect., 1977, 19, 47 project number: 110Y243, 2011) and Istanbul 21. A. Hernández, R. P. Mellado and J. L. Martínez, University Scientific Research Project Unit (İÜ BAP Appl. Environ. Microbiol., 1998, 64, (11), 4317 Project/19347) for their financial support. 22. L. D. Rasmussen and S. J. Sørensen, Curr. Microbiol., 1998, 36, (5), 291 References 23. D. H. Nies, Appl. Microbiol. Biotechnol., 1999, 51, (6), 730 1. D. A. Rouch, B. T. O. Lee and A. P. Morby, J. Ind. 24. G. Altuğ, M. Çardak and P. S. Ciftci, J. Fish. Aquat. Microbiol., 1995, 14, (2), 132 Sci., 2007, 22, (23), 39 (in Turkish) 2. S. Congeevaram, S. Dhanarani, J. Park, M. Dexilin 25. G. Altuğ, M. Çardak, P. S. Ciftci, S. Gürün and S. and K. Thamaraiselvi, J. Hazard. Mater., 2007, Kalkan, ‘Bacterial Diversity in Güllük Bay’, Tübitak 146, (1–2), 270 Project Workshop, 10th May, 2013, Güllük, Muğla, 3. M. R. Bruins, S. Kapil and F. W. Oehme, Ecotoxicol. Turkey, ed. G. Altuğ, Istanbul University, Turkey, Environ. Saf., 2000, 45, (3), 198 2013, pp. 3–7 (in Turkish)

26. S. Kalkan and G. Altuğ, Mar. Pollut. Bull., 2015, 42. P. H. Krumperman, Appl. Environ. Microbiol., 95, (1), 380 1983, 46, (1), 165 27. H. Yıldız, H. M. Doğan and Ö. Urla, Tarla Bit. Merk. 43. J. A. Washington and V. L. Sutter, ‘Dilution Test Arş. Ens. Derg., 2002, 11, (1–2), 142 (in Turkish) Procedures’, in “Manual of Clinical Microbiology”, 28. A. Demirak, A. Balci, Ö. Dalman and M. Tüfekçi, eds. E. H. Lennette A. Balows W. J. Hausler and Water Air Soil Pollut., 2005, 162, (1–4), 171 J. P. Truant, American Society for Microbiology, Washington, DC, USA, 1981, pp 549–555 29. A. Baba, A. Kaya and Y. K. Birsoy, Water Air Soil 44. A. D. Geiselbrecht, R. P. Herwig, J. W. Deming and Pollut., 2003, 149, (1–4), 93 J. T. Staley, Appl. Environ. Microbiol., 1996, 62, 30. P. Çiftçi G. Altuğ, M. Çardak and S. Gürün, (9), 3344 ‘Distribution of Indicator Bacteria in Recreational 45. V. Perreten and P. Boerlin, Antimicrob. Agents and Fish Farming Areas of Gulluk Bay, Turkey’, Chemother., 2003, 47, (3), 1169 Effective Utilization of Ocean Resources and Future Maritime Industries, Tokyo, Japan, 2nd– 11th 46. T. X. Le, Y. Munekage and S. Kato, Sci. Total November, 2011, Tokyo University of Marine Environ., 2005, 349, (1–3), 95 Science and Technology (TUMSAT), Japan, p. 25 47. M. T. Blahna, C. A. Zalewski, J. Reuer, G. Kahlmeter, 31. A. Demirak, A. Balci, and M. Tüfekçi, Environ. B. Foxman and C. F. Marrs, J. Antimicrob. Monit. Assess., 2006, 123, (1–3), 1 Chemother., 2006, 57, (4), 666 32. İ. Atılgan and Ö.Egemen, Ege J. Fish. Aquat. Sci., 48. Y. Agersø and A. Petersen, J. Antimicrob. 2001, 18, (1–2), 225 Chemother., 2007, 59, (1), 23 33. G. Yucel-Gier, I. Pazi and F. Kucuksezgin, Turk. J. 49. A. H. Buschmann, A. Tomova, A. López, M. A. Fish. Aquat. Sci., 2013, 13, (4), 737 Maldonado, L. A. Henríquez, L. Ivanova, F. Moy, H. P. Godfrey and F. C. Cabello, PLoS One, 2012, 34. E. Marti, E. Variatza and J. L. Balcazar, Trends 7, (8), e42724 Microbiol., 2014, 22, (1), 36 50. J. Kerry, R. Coyne, D. Gilroy, M. Hiney and P. 35. K. P. Acharya and R. T. Wilson, Front. Med., 2019, Smith, Aquaculture, 1996, 145, (1–4), 31 6, 105 51. F. Matyar, A. Kaya and S. Dinçer, Sci. Total 36. S. A. Kraemer, A. Ramachandran and G. G. Perron, Environ., 2008, 407, (1), 279 Microorganisms, 2019, 7, (6), 180 52. Z. J. Mudryk, Mar. Pollut. Bull., 2005, 50, (1), 80 37. A. Olgun, ‘Evaluation in Terms of Vessels and 53. H. Sørum, Acta Vet. Scand. Suppl., 1999, 92, 29 Ballast Waters entering Güllük Bay’, Bacteriology of Güllük Bay, Tübitak Project Workshop, 10th 54. V. Inglis, ‘Antibacterial Chemotherapy in May, 2013, Güllük, Muğla, Turkey, ed. G. Altuğ, Aquaculture: Review of Practice, Associated Istanbul University, Turkey, pp. 33–37 (in Turkish) Risks and Need for Action’, Use of Chemicals in Aquaculture in Asia, Tigbauan, Philippines, 20th– 38. “Standard Methods for the Examination of Water 22nd May, 1996, “Proceedings of the Meeting on and Waste Water”, eds. E. W. Rice, R. B. Baird, A. the Use of Chemicals in Aquaculture in Asia”, eds. J. D. Eaton and L. S. Clesceri, 22nd Edn., American R. Arthur, C. R. Larilla-Pitogo and R. P. Subasinghe, Public Health Association, American Water Works Southeast Asian Fisheries Development Center Association, Water Environment Federation, 2012 Aquaculture Department, Tigbauan, Philippines, 39. D. H. Pincus, ‘Microbial Identification Using the 2000, pp. 7–22 bioMérıeux VITEK® 2 System’, in “Encyclopedia 55. G. Rhodes, G. Huys, J. Swings, P. McGann, M. of Rapid Microbiological Methods”, ed. M. J. Miller, Hiney, P. Smith and R. W. Pickup, Appl. Environ. Vol. 2, Parenteral Drug Association, Bethesda, Microbiol., 2000, 66, (9), 3883 USA, 2006, pp. 1–32 56. Y. Huang, L. Zhang, L. Tiu and H. H. Wang, Front. 40. P. A. Sobecky, T. J. Mincer, M. C. Chang, A. Microbiol., 2015, 6, 914 Toukdarian and D. R. Helinski, Appl. Environ. 57. P.-R. Hsueh, L.-J. Teng, P.-C. Yang, Y.-C. Chen, H.- Microbiol., 1998, 64, (8), 2822 J. Pan, S.-W. Ho and K.-T. Luh, Clin. Infect. Dis., 41. M. A. Wikler, F. R. Cockerill, W. A. Craig, M. N. 1998, 26, (3), 676 Dudley, G. M. Eliopoulos, D. W. Hecht, J. F. 58. C. Baker-Austin, M. S. Wright, R. Stepanauskas Hindler, D. E. Low, D. J. Sheehan, F. C. Tenover, and J. V. McArthur, Trends Microbiol., 2006, 14, J. D. Turnidge, M. P. Weinstein, B. L. Zimmer, (4), 176 M. J. Ferraro and J. M. Swenson, ‘Performance Standards for Antimicrobial Susceptibility Testing: 59. P. R. Jensen and W. Fenical, Annu. Rev. Microbiol., Sixteenth Informational Supplement’, M100– S16, 1994, 48, 559 Clinical and Laboratory Standards Institute, 60. D. Claus and R. C. W. Berkeley, ‘Genus Bacillus, Wayne, USA, 2006, 192 pp Cohn 1872’, in “Bergey’s Manual of Systematic

Bacteriology” eds. P. H. A. Sneath , N. S. Mair, 68. L. P. Randall, S. W. Cooles, M. K. Osborn, L. J. M. E. Sharpe and J. G. Holt, Vol. 2, Williams and V. Piddock and M. J. Woodward, J. Antimicrob. Wilkins Co, Baltimore, USA, 1986, pp 1105–1139 Chemother., 2004, 53, (2), 208 61. S. B. Levy and B. Marshall, Nat. Med., 2004, 10, 69. A. S. Schmidt, M. S. Bruun, I. Dalsgaard and J. L. (12), S122 Larsen, Appl. Environ. Microbiol., 2001, 67, (12), 62. D. G. Capone, D. P. Weston, V. Miller and C. 5675 Shoemaker, Aquaculture, 1996, 145, (1–4), 55 70. M. Teuber, Curr. Opin. Microbiol., 2001, 4, (5), 63. F. M. Aarestrup, Basic Clin. Pharmacol. Toxicol., 493 2005, 96, (4), 271 71. S. Silver and L. T. Phung, Annu. Rev. Microbiol., 64. N. van de Sande-Bruinsma, H. Grundmann, D. 1996, 50, 753 Verloo, E. Tiemersma, J. Monen, H. Goossens and M. Ferech, Emerg. Infect. Dis., 2008, 14, (11), 1722 72. F. Matyar, T. Akkan, Y. Uçak and B. Eraslan, 65. E. Gullberg, S. Cao, O. G. Berg, C. Ilbäck, L. Environ. Monit. Assess., 2010, 167, (1–4), 309 Sandegren, D. Hughes and D. I. Andersson, PLoS 73. S. Silver and M. Walderhaug, Microbiol. Rev., Pathog., 2011, 7, (7), e1002158 1992, 56, (1), 195 66. C. P. Randall, A. Gupta, N. Jackson, D. Busse and A. J. O’Neill, J. Antimicrob. Chemother., 2015, 70, 74. H. Aiking, A. Stijnman, C. van Garderen, H. van (4), 1037 Heerikhuizen and J. van ’t Riet, Appl. Environ. Microbiol., 1984, 47, (2), 374 67. L. A. M. Carneiro, A. P. S. Silva, V. L. C. Merquior and M. L. P. Queiroz, FEMS Microbiol. Lett., 2003, 75. N. Çağlar, A. Aksu and G. Altuğ, Geol. Bull. Turkey, 228, (2), 175 2020, 63, (1), 117

The Authors

Gülşen Altuğ is a professor in the Department of Marine Biology of the Faculty of Aquatic Science at Istanbul University, Turkey. Her research focuses on marine bacteriology, including bacterial diversity and micro-geographical variations, clinical, industrial and ecological uses of marine isolates, bacterial pollution, epibiotic bacterial communities and anti-bacterial characteristics, bacterial remediation (oil degrading capacity of marine isolates) and resistant bacterial isolates against heavy metals and antibiotics. She is also the inventing founder of the biotechnology company named BIYOTEK15 R&D Training and Consulting Industry and Trade Ltd Company in Entertech of Istanbul University Technocity.

Mine Çardak is an associate professor at Çanakkale Onsekiz Mart University, School of Çanakkale Applied Sciences, Department of Fisheries Technology, Turkey. Her researches focus on marine bacteriology, bacterial resistance against heavy metals and antibiotics, bacterial pollution and biotechnology. She has worked as a scientist since 2000.

Pelin Saliha Çiftçi Türetken is a researcher at İstanbul University, Faculty of Aquatic Sciences, Department of Marine Biology. Her research focus on marine bacteriology, bacterial remediation, bacterial resistance and biotechnology. She has a PhD degree in marine biology. She has worked as an academic at university since 2005.

Samet Kalkan has a PhD degree from Istanbul University, Institute of Graduate Studies in Science and Engineering, Department of Marine Biology. He currently works as a doctor scientist at Recep Tayyip Erdogan University- Faculty of Fisheries, Department of Marine Biology, Turkey. He has worked as academic at university since 2010. His main researches focus on marine bacteria, bacterial diversity, bacterial pollution, resistant bacteria against heavy metals-antibiotics, also marine biotechnology. He has scientific abroad experiences in Italy and Portugal.

Sevan Gürün graduated with a degree in Biology from Istanbul University. He has a PhD degree from Istanbul University, Institute of Graduate Studies in Science and Engineering, Department of Marine Biology. He worked as a researcher in various scientific projects. He has been working as a researcher in a private company since 2016. His expertise focuses on bacterial diversity, marine bacteria, bacterial pollution, bacterial biotechnology, resistant bacteria against heavy metals and antibiotics.

www.technology.matthey.com

“Nanomaterials and Environmental Biotechnology” Edited by Indu Bhushan (Shri Mata Vaishno Devi University, India), Vivek Kumar Singh (Shri Mata Vaishno Devi University, India), Durgesh Kumar Tripathi (Amity University, India), Nanotechnology in the Life Sciences Series, Springer Nature Switzerland AG, Cham, Switzerland, 2020, 434 pp, ISBN: 978-3-030-34543-3, £129.99, €155.99, US$179.99

Reviewed by Martin Hayes on aluminium oxide and zinc oxide nanoparticles Johnson Matthey, 28 Cambridge Science Park, hindering root growth rate. Milton Road, Cambridge, CB4 0FP, UK Chapter 2 considers an entirely separate but equally important area of use of nanomaterials in Email: [email protected] food packaging. It concludes that “it is essential to perform safety assessment of nanomaterials before their application in food packaging or processing” Introduction and provide a citation on how to do this using a “decision tree”. This book is a fascinating account of how nanoparticles and nanotechnology are increasingly Biosensors from Nanobiotechnology employed in a diverse array of applications ranging from plant growth to food packaging, biosensing, Chapter 7 introduces how biosensors derived enzyme immobilisation and more. The book is from nanobiotechnology can be used to monitor divided into 20 chapters, each dealing with a the environment and gain information relating specific application of nanotechnology and written to its health and the detrimental effects that by a different group of eminent academics from modernisation and industrialisation have had Indian universities and research institutes. on the planet. Biosensors need to be specific, Each chapter is a review of its own topic area and the rapid, sensitive and cost-effective. The advent literature citations at the end of each chapter make of nanotechnology and biosensors has made this it easy for the reader to use this book as a reference possible and the authors of this chapter (Gupta and volume from which further in-depth reading can be Kakkar) explore the different types of biosensors pursued by following the cited literature. that have been developed over recent years. The authors give a brief explanation of how different Safety of Nanoparticles in Plants and types of sensors work using a combination of bio- recognition components and different transduction Packaging principles. Types include: (a) immunosensors; Chapter 1 deals with the phytotoxicity of (b) enzymatic biosensors; (c) whole-cell based nanoparticles in plants which can lead to both sensors; (d) biosensors; (e) genosensors; positive and negative outcomes. For example, (f) aptasensors and (g) biomimetic biosensors. The improvement in germination rate and growth role of the transducer is to convert the biochemical have been reported in seeds of rice exposed to response into an analysable and measurable carbon nanotubes; on the other hand, toxicity signal. The outputs can be electrochemical, optical, has also been widely reported, including studies piezoelectric, thermometric or magnetic.

Enzyme Immobilisation to traditional drug delivery systems, offering multiple advantages in terms of drug delivery Chapter 10 tackles the interesting area of enzyme and bioavailability, as well as being economically immobilisation and the use of chitosan nanoparticles efficient and easy to produce on scale. therein. The biopolymer’s distinct physicochemical properties have been described to offer an excellent FDA-Approved Nanomedicines microenvironment for enzyme immobilisation through adsorption, covalent binding or cross- An extensive summary of FDA-approved linking, to achieve desirable enzymatic activity nanomedicines is included in Table 19.1 and stability. On the other hand, nanoparticles as (Chapter 19), which also summarises the materials of enhanced properties, owing to their advantages of these specific formulations. The high surface to volume ratio, have been introduced main types of nanocarriers described in Chapter 19 as attractive candidates for enzyme immobilisation. are vesicular carriers (liposomes and niosomes), The chapter briefly discussed various methods polymeric nanoparticles and inorganic carriers for the preparation of chitosan nanoparticles for (silica, gold and calcium nanoparticles). Liposomes enzyme immobilisation including reverse micelle, and solid lipid nanoparticles (see also Chapter 13) coprecipitation, ionotropic gelation and ionic or are suitable for the delivery of drugs by any emulsion cross-linking methods. Different methods route, either oral or parenteral and can be used for enzyme immobilisation such as support binding, with both hydrophilic and lipophilic drugs. Their cross-linking and entrapment, as well as different main advantages reside in protecting labile drugs, materials used as supports have been explained limited toxicity and a sustainable targeted release too. This section is then closed by presenting some of the drug. examples for immobilisation of different enzyme Inorganic nanocarriers exhibit higher stability and families (for example, α-amylase, β-galactosidase, resistance to microbial growth, while having a low cellulase, laccase, lipase or protease) through toxicity and allowing facile surface modifications. applying chitosan nanoparticles. Mesoporous silica nanoparticles allow encapsulation of the therapeutic agent and targeted delivery to Solid Lipid Nanoparticles tumour cells in cancer therapy. Gold nanoparticles are biocompatible and bio-inert and have been Chapter 13 offers an overview of solid lipid successfully used in covalent conjugation with nanoparticles (SLN) as pharmaceuticals delivery protein antigens in developing vaccines for cancer systems whereas Chapter 19 gives a review of the immunotherapy. Calcium phosphate nanoparticles most commonly used nanocarriers for drug delivery are excellent candidates for developing ceramic- systems, with a focus on vesicular, polymeric and based carriers for peptide drugs prone to inorganic carriers. degradation, such as insulin. SLN are lipid-based formulations, containing typically non-toxic biodegradable polymers Summary forming a solid hydrophobic core suspended in an aqueous phase, the whole structure being In conclusion, I consider this book to be a positive stabilised by surfactants. The therapeutic agent is contribution to the biotechnology literature, dissolved or dispersed in the solid lipid core, the although I do not recommend reading this book SLN being suitable for incorporation and delivery sequentially from Page 1 as the variety of topics of both hydrophilic and hydrophobic drugs. SLN introduced is too great and each individual topic present significant advantages over conventional is not explored in depth. It is best used (and drug delivery systems, including but not limited deserves recommendation) as a reference source to biocompatibility and bioavailability, reduced from which each chapter can be used as the drug leakage and increased physical stability starting point to a more in-depth study or review of the drugs. In addition, they have been used of a particular topic. There are some negative successfully in various drug delivery techniques. aspects of the presentation of this work which do, Novel applications of SLN as drug carriers are unfortunately, detract from its enjoyment. These described in the field of gene therapy, peptide drug are exemplified in the poor quality of the diagrams, delivery and vaccines. SLN production methods the grammatical errors and the somewhat odd use low mechanical force, allowing successful references of Chapter 7. incorporation and delivery of nucleic acids in gene Overall, though, this book is a positive addition to therapy. Overall, SLN are promising alternatives the biotechnology reference bookshelf.

"Nanomaterials and Environmental Biotechnology"

The Reviewer

Martin Hayes is Biotechnology Lead at Johnson Matthey based in Cambridge, UK. He has worked with Johnson Matthey since 1997 and has held multiple research, development and customer-facing technical roles across the company. He holds a PhD in heterogeneous catalysis and is interested in the application of biology as a technology to realise circular chemical processing and accelerate the transition to “Net Zero”.

www.technology.matthey.com

Biocatalytic Reduction of Activated Cinnamic Acid Derivatives Asymmetric reduction of C=C double bonds using Johnson Matthey enzymes

Samantha Staniland, Tommaso hydrogenation. Traditionally, whole cell Angelini, Ahir Pushpanath, Amin microorganisms were used for this purpose but Bornadel, Elina Siirola, Serena a recent increase in the number of isolated and Bisagni, Antonio Zanotti-Gerosa, characterised ENEs means that recombinantly- Beatriz Domínguez* expressed enzyme preparations are now generally Johnson Matthey, 260 Cambridge Science Park, favoured over whole cells, as a number of recent Milton Road, Cambridge, CB4 0WE, UK publications demonstrate (1–10). Double bond ‘activation’ to facilitate ENEs mediated *Email: [email protected] reduction can be achieved in many cases by alpha substituted functional groups including aldehydes, ketones or nitro moieties. Carboxylate derivatives The asymmetric reduction of C=C double bonds is (such as esters, lactones and anhydrides) can a sought-after chemical transformation to obtain also act as activating groups but their ability to chiral molecules used in the synthesis of fine sufficiently activate the C=C bond in the absence of chemicals. Biocatalytic C=C double bond reduction other groups is less evident (11, 12). The traditional is a particularly interesting transformation approach in these cases is to turn to chemocatalytic complementary to more established chemocatalytic hydrogenation (see (13–15) for reviews focused on methods. The enzymes capable of catalysing this industrial applications). Herein we describe a new reaction are called ene-reductases (ENEs). For approach to activate α,β-unsaturated carboxylic the reaction to take place, ENEs need an electron acids for the reduction with ENEs using a substrate withdrawing group (EWG) in conjugation with engineering approach. the double bond. Especially favourable EWGs are carbonyls and nitro groups; other EWGs, 2. Experimental such as carboxylic acids, esters or nitriles, often give poor results. In this work, a substrate 2.1 General engineering strategy is proposed whereby a simple transformation of the carboxylic acid into a All reagents and solvents were purchased from fluorinated ester or a cyclic imide allows to increase Sigma-Aldrich and Alfa Aesar, Thermo Fisher the ability of ENEs to reduce the conjugated double Scientific. They were of the highest available purity bond. Up to complete conversion of the substrates and were used without further purification. 1H tested was observed with enzymes ENE-105 and nuclear magnetic resonance (NMR) spectra were *ENE-69. recorded using a Bruker 400 MHz Avance III HD equipped with SMART probe (Bruker Corporation, 1. Introduction USA) where spectra are referenced to deuterated chloroform (CDCl3) 7.26 ppm, shifts are recorded The use of enzymes for the asymmetric reduction in parts per million and J values in hertz. The of activated C=C double bonds can be a viable NMR results can be found in the Supplementary and straightforward alternative to asymmetric Information.

2.2 Enzyme Preparations was then quenched by addition of saturated

aqueous NaHCO3 (20 ml) and extracted with

Genes coding for Johnson Matthey, ENEs (ENE-101, dichloromethane (2 × 20 ml), dried over MgSO4, ENE-102, ENE-103, ENE-104, ENE-105, *ENE-69 filtered and concentrated under reduced pressure and GDH-101) were ordered codon-optimised from to afford the corresponding fluorinated esters6a GeneArt (Thermo Fisher Scientific) and cloned and 7a in 95% to 99% yield. into T5 vector pJEx401 (ATUM). Enzymes were expressed recombinantly in Escherichia coli BL21 2.5 1-Cinnamoylpyrrolidin-2-one in both shake flasks and fed batch fermentations, (9a) whereby induction was carried out with isopropyl β-D thiogalactopyranoside (IPTG) at 30°C. Cinnamoyl chloride (5 g, 30.01 mmol), pyrrolidinone Harvested biomass was resuspended in 100 mM (2.3 ml, 36.01 mmol) and triethylamine (13 ml, potassium phosphate buffer (pH 7) and cells were 90.03 mmol) in dichloromethane (50 ml) were broken up either by sonication or homogenisation. stirred at room temperature overnight. The reaction The so-obtained cell lysate was clarified by was quenched by addition of water (20 ml), the centrifugation and filtrated prior to lyophilisation. organic layer was separated and washed with

Protein expression was assessed by sodium saturated aqueous NaCl (20 ml), dried over MgSO4, dodecyl sulfate polyacrylamide gel electrophoresis filtered and concentrated under reduced pressure (SDS-PAGE) and chromatographic activity assays. to afford9a in 81% yield. Enzymes ERED-103, ERED-110, ERED-112, ERED-207, ERED-P1-A04, ERED-P1-E04 and ERED- 2.6 3-Cinnamoyloxazolidin-2-one P1-H09 were purchased from Codexis. (8a)

2.3 2,2,2-Trifluoroethyl Cinnamate Cinnamic acid 1a (5 g, 33.56 mmol) and oxalyl chloride (2.85 ml, 33.56 mmol) in dichloromethane (3a) and 3-Phenyl-Acrylic Acid (5 ml) were stirred at room temperature overnight 2,2,2-Trifluoro-1-Trifluoromethyl- before removing the solvent under reduced Ethyl Ester (5a) pressure. The reaction crude was dissolved Cinnamic acid 1a (5 g, 33.75 mmol) and oxalyl in anhydrous tetrahydrofuran (THF) (20 ml) chloride (2.85 ml, 33.75 mmol) in dichloromethane and n-butyllithium (1.6 M in hexane, 21 ml, (5 ml) were stirred at 25°C for 2 h before adding the 33.56 mmol, one equivalent) was added dropwise fluorinated alcohol-trifluoro ethanol for3a (2.47 ml, over 30 min. The cinnamoyl chloride solution was 33.75 mmol) and 1,1,1,3,3,3-hexafluoropropan- then added dropwise to a solution of oxazolidinone 2-ol for 5a (3.50 ml, 33.75 mmol). The reaction (2.92 g, 33.56 mmol) in anhydrous THF (100 ml) at was then stirred at room temperature overnight 0°C before stirring at room temperature overnight. before being quenched by addition of saturated The reaction was quenched with water (50 ml), aqueous NaHCO3 (20 ml) and extracted with extracted with ethyl acetate (EtOAc) (2 × 100 ml), dichloromethane (2 × 20 ml), dried over MgSO4, washed with saturated aqueous NaHCO3 (20 filtered and concentrated under reduced pressure ml) and saturated aqueous NaCl (20 ml). The to afford the corresponding fluorinated esters 3a solvent was removed under reduced pressure and and 5a in quantitative yield. the solid was recrystallised from a 1:1 mixture EtOAc:heptane (20 ml). The solid was filtered and 2.4 3-Phenyl-Acrylic Acid washed with hexane (10 ml) to give crystals of 8a in 80% yield. 2,2,3,3,4,4,4-Heptafluoro-Butyl Ester (6a) and (Perfluorophenyl)Methyl Cinnamate (7a) 2.7 (E)-1-(2-Methyl-3- Phenylacryloyl)Pyrrolidin- Cinnamoyl chloride (0.75 g, 4.50 mmol) 2-one (10a) and (E)-1-(2,3- and the corresponding fluorinated alcohols – 2,2,3,3,4,4,4-heptafluorobutan-1-ol for Diphenylacryloyl)Pyrrolidin-2-one 6a (0.98 g, 4.50 mmol) and pentafluroro (11a) benzyl alcohol for 7a (0.89 g, 4.50 mmol) – in dichloromethane (2.5 ml) were stirred at (E)-2-methyl-3-phenylacrylic acid (5 g, room temperature overnight. The reaction 30.86 mmol) was converted to the corresponding

530 © 2020 Johnson Matthey https://doi.org/10.1595/205651320X16001815466116 Johnson Matthey Technol. Rev., 2020, 64, (4) acid chloride by addition of oxalyl chloride (1.4 ml, C18 SunFire Column (Waters Corporation, USA, 30.86 mmol) in dichloromethane (5 ml). The 150 × 4.6 mm, 3.5 µm) with an isocratic method reaction was stirred at room temperature for (MeCN:Water, 30:70 + 0.1% trifluoroacetic acid) 3 h. Pyrrolidinone (2.82 ml, 37.03 mmol) and and a flow rate of 1 ml min–1. triethylamine (13 ml, 92.58 mmol) were added Chiral HPLC analysis was performed on a Varian before stirring the reaction overnight. The reaction ProStar series (Agilent) with a CHIRALCEL® OD-H was quenched by addition of water (20 ml) and column (Chiral Technologies, USA, 250 × 4.6 mm, saturated aqueous NaCl (20 ml). The solvent 5 µm) with an isocratic method A (heptane:isopropyl was removed under reduced pressure and the alcohol (IPA), 88:12) and a flow rate of 1 ml min–1 solid was dissolved in EtOAc and treated with or isocratic method B (heptane:IPA, 98:2). activated charcoal (1 g), filtered through Celite® GC analysis of conversion was performed on and concentrated. The solid was recrystallised from a Varian CP-3800 (Agilent) using γ-DEX™ 225 heptane (10 ml) to give 10a in 55% yield. capillary column (Sigma-Aldrich, 30 m × 0.25 mm Following an identical procedure, 11a was × 0.25 μm) and using helium as carrier gas. synthesised in 53% yield from (E)-2,3- Percentage conversion was measured by integration diphenylacrylic acid (10 g, 44.64 mmol). of the product peak in the GC (uncorrected area under curve (AUC)), values below 100% indicate 2.8 Small Scale Screening Reactions that unreacted starting material was detected. No side products were detected in any of the reported Substrates 1a–9a (0.025 mmol) and enzymes reactions. GC program parameters: injector 250°C, ENE-101, ENE-102, ENE-103, ENE-104, ENE-105 flame ionization detector (FID) 250°C, 80°C for or *ENE-69 (2.5 mg), were added to reaction 3 min then 5°C min–1 up to 160°C, hold 1 min vials containing 500 µl of aqueous media at pH 7 (total time 20 min), constant flow 5 ml min–1. (250 mM potassium phosphate buffer pH 7, 1.1 mM + –1 NAD(P) , 100 mM D-glucose, 10 U ml GDH-101) 3. Results and Discussion to give a final concentration of substrate of 50 mM. The vials were shaken at 400 rpm, 30°C for 18 h. It has been found that a particular ENE in Johnson For high-performance liquid chromatography Matthey’s collection, a homologue from the tobacco (HPLC) analysis, the reactions were quenched with ENE reductase fold (16), ENE-105, was capable acetonitrile (MeCN) (1 ml), vortexed, centrifuged of reducing methyl ester 2a (Figure 1), albeit and aliquoted. For gas chromatography (GC) in a very low yield of 3% (Entry 2, Table I). By analysis, samples were extracted with EtOAc comparison, cinnamic acid 1a was a poor substrate

(2 × 0.5 ml), dried over MgSO4 and analysed and showed no conversion to the reduced product directly. For NMR analysis, the reactions were 1b at pH 7.0 (Entry 1, Table I). The pKa of extracted with CDCl3 and analysed directly. cinnamic acid 1a is 4.4 and therefore, at pH 7.0, the carboxylic acid should be deprotonated 2.9 Preparative Scale Screening affecting its ability to bind to the enzyme active site. This observation is in line with other literature Reactions examples where carboxylates were found to be Reactions were scaled up using three-neck round poor activating groups (17). Encouraged by this bottom flask equipped with stir bar and pH initial result, we turned our efforts towards the use titrator (10 M NaOH). To the flask was weighed of more activated esters. It was envisaged that 100–500 mg substrate (40–100 mM final converting the alkyl chain in the ester moiety to a concentration) and 5 mg ml–1 enzyme which was more EWG could lead to an increase in double bond suspended in aqueous media at pH 7 (250 mM activation. A similar approach has been reported potassium phosphate buffer pH 7, 1.1 mM previously by BASF SE for the lipase-catalysed NAD(P)+, 100–200 mM D-glucose (two equivalent), kinetic resolution of racemic amines and alcohols, 10 U ml–1 GDH-101) the reactions were stirred at where the choice of acylating agent proved critical 30°C, 400 rpm for 18 h. (18). We chose trifluoroethyl ester3a as a starting point which was reduced by ENE-105 and *ENE-69 2.10 Analytical Methods in 6% and 12% conversion respectively (Entry 3, Table I) suggesting that the addition of an EWG HPLC analysis of conversion was conducted on an had a positive activating-effect on the reduction. To 1260 Infinity II LC system (Agilent, USA) using a consolidate this theory, ethyl ester 4a was tested

Fig. 1. Reduction of O O cinnamic acid and [a] R R cinnamoyl esters. [a] = 1–7a (50 mM 1a R = OH 1b R = OH concentration), ENE- 2a R = OCH3 2b R = OCH3 105 or ENE-69 (5 mg –1 3a R = OCH2CF3 3b R = OCH2CF3 ml ), 500 µl buffer 4a R = OCH2CH3 4b R = OCH2CH3 (250 mM KPi, pH 7, 1.1 + 5a R = OCH(CF3)2 5b R = OCH(CF3)2 mM NAD(P) , 100 mM –1 6a R = OCH2CF2CF2CF3 6b R = OCH2CF2CF2CF3 D-glucose, 10 U ml 7a R = OCH2C6F5 7b R = OCH2C6F5 GDH-101), 400 rpm, 30°C, 18 h

Table I Reduction of Cinnamoyl Esters at Table II Reduction of Cinnamoyl Cyclic 50 mM Substrate Concentration, Imide Derivatives at 50 mM pH 7, 30°C, 18 h Substrate Concentration, pH 7, Conversion, %a 30°C, 18 h Entry Substrate ENE-105 *ENE-69 Conversion, %

1b 1a 0 0 Entry Substrate ENE-105 *ENE-69 a 2 2a 3 1 1 8a 51 39 b 3 3a 6 12 2 9a >95 >95 a 4 4a <0.5 <0.5 Integration of the product peak in the HPLC (chiral method, uncorrected AUC), values below 100% indicate that unreacted 5c 5a 0 0 starting material was detected; no side products were detected for these reactions 6d 6a <0.5 <0.5 bConversion calculated by 1H NMR 7d 7a 0 0 aIntegration of the product peak in the GC (uncorrected AUC), values below 100% indicate that unreacted starting material cyclic imides since activated substrates 8a and was detected; no side products were detected for these 9a have been shown to be highly activated reactions bIntegration of the product peak in the HPLC (achiral method, towards Michael addition reactions (19, 20, 21) uncorrected AUC), values below 100% indicate that unreacted (Figure 2). Compounds 8a and 9a were starting material was detected; no side products were detected for these reactions synthesised and tested with enzymes ENE-105 c10% cinnamic acid observed and *ENE-69. Pleasingly, oxazolidinone 8a was d Conversion calculated by 1H NMR successfully reduced by both ENEs (51% and 39% conversion to 8b, Entry 1, Table II) and with the novel ENEs; only a trace of reduction was pyrrolidinone 9a was reduced to 9b in >95% observed <0.5% (Entry 4, Table I). conversion (Entry 2, Table II), proving to be Other EWGs such as hexafluoroethyl in an excellent activating group. The 1H NMR shift compound 5a, heptafluorobutyl in 6a and of the alkene proton alpha to the carbonyl for pentafluorobenzyl in 7a could also activate the pyrrolidinone 9a is shifted down field (7.92 ppm) double bond in the same way, so 5a, 6a and 7a compared to cinnamic acid 1a (6.46 ppm), were prepared by reacting cinnamoyl chloride therefore supporting the electron-withdrawing with the corresponding fluorinated alcohols nature of the activating group. and these substrates were subsequently tested The enzymes were then tested for their ability to with the ENEs. Hexafluoro 5a was not reduced reduce α-substituted cinnamic acid derivatives such by ENE-105 or *ENE-69 (Entry 5, Table I), as α-methyl 10a and α-phenyl 11a (Figure 3). instead, a significant amount of hydrolysis Encouragingly, the tri-substituted double bond in product (cinnamic acid 1a, 10%) was observed. 10a was reduced to 10b in >95% conversion by Heptafluorobutyl 6a and pentafluoro 7a were 1H NMR analysis (Entry 2, Table III). However, poor activating groups with 6a showing only a bulkier substrate 11a, was not tolerated so well on trace amount of product 6b (Entry 6, Table I) an analytical scale due to solubility issues causing and 7a giving no conversion (Entry 7, Table I). mass-transfer limitations (Entry 3, Table III). With only limited success with the fluorinated The reaction was repeated on a larger scale with activating groups, our efforts turned towards stirring (Entry 4, Table III) and >95% conversion

O O Fig. 2. Cinnamoyl cyclic imide derivatives. [a] R R [a] = 8a–9a (50 mM concentration), ENE- 105 or ENE-69 (5 mg ml–1), 500 µl buffer O O O O (250 mM KPi, pH 7, 1.1 + N O N N O N mM NAD(P) , 100 mM R = R = D-glucose, 10 U ml–1 GDH-101), 400 rpm, 8a 9a 8b 9b 30°C, 18 h

O O O O Fig. 3. Reduction [a] of α-substituted N N cinnamoyl R R pyrrolidinones. [a] = 9a R = H 9b R = H 9a–11a (40–100 10a R = Me 10b R = Me mM concentration), 11a R = Ph 11b R = Ph ENE-105 or ENE-69 (5 mg ml–1), buffer (250 mM KPi, pH 7, 1.1 mM NAD(P)+, two equivalent D-glucose, 10 U ml–1 GDH-101), 400 rpm, 30°C, 18 h

Table III Reduction of α-Substituted carboxylic acids, other commercially available Cinnamoyl Pyrrolidinones at enzymes were tested as a comparison on the 50 mM Substrate Concentration, reduction of 10a (Table V). Six enzymes pH 7, 30°C, 18 h from Johnson Matthey collection (Entries 3 a Conversion, % to 6, Table V) and seven enzymes purchased Entry Substrate ENE-105 *ENE-69 from Codexis (Entries 7 to 13, Table V) were 1 9a >95 >95 compared with ENE-105 and ENE-69* (Entries 1 2 10a >95 >95 and 2, Table V). It was found that, despite the 3 11a 24 2 extra activation of the C=C double bond, none of the tested enzymes could reduce cinnamic acid 9b 11a >95 – derivative 10a, highlighting the unique ability of aConversion calculated by 1H NMR b100 mg scale with stirring ENE-105 and *ENE-69 within the focused library (13 enzymes) screened. In summary, we have shown that cinnamic acid was achieved. 10b and 11b were obtained as derivatives activated as fluorinated esters or racemic mixtures. as cyclic imides can be reduced using Johnson With a successful activating group found, the Matthey enzymes ENE-105 or *ENE-69. The reaction was repeated on a preparative scale to concept of ‘substrate engineering’ as opposed test reproducibility and scalability (Table IV). to ‘enzyme engineering’, offers a complimentary Pyrrolidinone 9a was successfully reduced using and faster approach to developing a bioprocess, enzyme ENE-105 at 130 mg scale with the desired making difficult transformations possible. The product 9b being obtained in 95% conversion by reduced products can be subsequently converted 1H NMR (Entry 1, Table IV). 72% conversion to to the parent carboxylic acids by LiOH hydrolysis 10b was achieved after 20 h (Entry 3, Table IV) (22, 23) and the potential re-use of these on the reduction of pyrrolidinone 10a at 500 mg activating groups will be investigated in the scale. future. It is envisaged that the work will lead Having found enzymes in Johnson Matthey’s to further examples of activated acids or esters collection that could successfully reduce masked being reduced by ENEs.

Table IV Reduction of Cinnamoyl Pyrrolidinones by ENE-105 at pH 7b and 30°C Entry Substrate Scale, mg Concentration, mM Time, h Conversion, %a 1 9a 130 40 16 >95 2 10a 500 100 4 33 3 10a 500 100 20 72 aConversion calculated by 1H NMR bpH controlled with NaOH titration (10 M)

Table V Reduction of Cinnamoyl procedure, and in combination with the right Pyrrolidinone 10a at 50 mM enzyme, it was possible to biocatalytically reduce Substrate Concentration, pH 7, the conjugated double bond of cinnamic acid and 30°C, 18 h substituted derivatives. Entry Enzyme Conversion, %a 1 ENE-105 >95 References 2 *ENE-69 >95 1. B. Dominguez, U. Schell, S. Bisagni and T. Kalthoff, 3 ENE-101 <0.5 Johnson Matthey Technol. Rev., 2016, 60, (4), 4 ENE-102 1 243 5 ENE-103 0 2. M. Hall, C. Stueckler, H. Ehammer, E. Pointner, G. 6 ENE-104 0 Oberdorfer, K. Gruber, B. Hauer, R. Stuermer, W. 7 ERED-103 0 Kroutil, P. Macheroux and K. Faber, Adv. Synth. 8 ERED-110 0.5 Catal., 2008, 350, (3), 411 9 ERED-112 0 3. M. Hall, C. Stueckler, W. Kroutil, P. Macheroux and K. Faber, Angew. Chem., Int. Ed., 2007, 46, (21), 10 ERED-207 <0.5 3934 11 ERED-P1-A04 1 4. J. F. Chaparro-Riggers, T. A. Rogers, E. Vazquez- 12 ERED-P1-E04 0 Figueroa, K. M. Polizzi and A. S. Bommarius, Adv. 13 ERED-P1-H09 0 Synth. Catal., 2007, 349, (8–9), 1521 a 1 Conversion calculated by H NMR 5. A. Müller, B. Hauer and B. Rosche, Biotechnol. Bioeng., 2007, 98, (1), 22 6. M. A. Swiderska and J. D. Stewart, J. Mol. Catal. 4. Conclusions B: Enzym., 2006, 42, (1–2), 52 7. D. Dobrijevic, L. Benhamou, A. E. Aliev, D. Méndez- The biocatalysed reduction of the double bond of Sánchez, N. Dawson, D. Baud, N. Tappertzhofen, cinnamic acid derivatives is strongly influenced by T. S. Moody, C. A. Orengo, H. C. Hailes and J. M. the nature of the EWG. While no conversion was Ward, RSC Adv., 2019, 9, (63), 36608 observed on the biocatalysed reduction of cinnamic 8. H. S. Toogood and N. S. Scrutton, ACS Catal., acid 1a, an enzyme in Johnson Matthey’s collection, 2018, 8, (4), 3532 ENE-105, was capable of reducing methyl ester 9. G. Brown, T. S. Moody, M. Smyth, S. J. C. derivative 2a in low conversion. By replacing the Taylor, ‘Almac: An Industrial Perspective of Ene alkyl chain in the ester moiety by a more EWG, Reductase (ERED) Biocatalysis’, in “Biocatalysis: such as fluorinated alkanes, and in the presence An Industrial Perspective”, eds. G. de Gonzalo and of enzymes ENE-105 and *ENE-69, we were able P. Domínguez de María, ch. 8, Royal Society of to significantly increase conversion to the reduced Chemistry, London, UK, 2018, pp. 229–256 product. Furthermore, other electronegative 10. D. Mangan, I. Miskelly and T. S. Moody, Adv. derivatives such as cyclic imides proved to be even Synth. Catal., 2012, 354, (11–12), 2185 better activating groups, allowing the reduction 11. R. Stuermer, B. Hauer, M. Hall and K. Faber, Curr. of challenging substituted double bonds such as Opin. Chem. Biol., 2007, 11, (2), 203 substrates 10a and 11a. 12. Y. Kawai, M. Hayashi, Y. Inaba, K. Saitou and A. In summary, by ‘masking’ the carboxylic acid Ohno, Tetrahedron Lett., 1998, 39, (29), 5225 moiety into a fluorinated alkyl ester or a cyclic 13. H. U. Blaser, B. Pugin and F. Spindler, ‘Asymmetric imide, following a straightforward synthetic Hydrogenation’, in “Topics in Organometallic

Chemistry: Organometallics as Catalysts in the 18. M. Breuer, K. Ditrich, T. Habicher, B. Hauer, M. Fine Chemical Industry”, eds., M. Beller, H. U. Keßeler, R. Stürmer and T. Zelinski, Angew. Blaser, Vol. 42, Springer-Verlag, Berlin, Germany, Chem., Int. Ed., 2004, 43, (7), 788 2012, pp. 65–102 19. D. Monge, H. Jiang and Y. Alvarez-Casao, Chem. 14. D. J. Ager, A. H. M. de Vries and J. G. de Vries, Eur. J., 2015, 21, (12), 4494 Chem. Soc. Rev., 2012, 41, (8), 3340 20. V. A. Soloshonok, C. Cai, V. J. Hruby, L. Van 15. C. S. G. Seo and R. H. Morris, Organometallics, Meervelt and T. Yamazaki, J. Org. Chem., 2000, 2018, 38, (1), 47 65, (20), 6688 16. D. J. Mansell, H. S. Toogood, J. Waller, J. M. X. 21. T. Inokuma, Y. Hoashi and Y. Takemoto, J. Am. Hughes, C. W. Levy, J. M. Gardiner and N. S. Chem. Soc., 2006, 128, (29), 9413 Scrutton, ACS Catal., 2013, 3, (3), 370 22. D. A. Evans, T. C. Britton, R. L. Dorow and J. F. 17. J. Waller, H. S. Toogood, V. Karuppiah, N. J. W. Dellaria, J. Am. Chem. Soc., 1986, 108, (20), Rattray, D. J. Mansell, D. Leys, J. M. Gardiner, A. Fryszkowska, S. T. Ahmed, R. Bandichhor, G. P. 6395 Reddy and N. S. Scrutton, Org. Biomol. Chem., 23. T. F. Woiwode and T. J. Wandless, J. Org. Chem., 2017, 15, (20), 4440 1999, 64, (20), 7670

The Authors

Samantha Staniland graduated from The University of Manchester, UK, in 2011 with an MChem in Chemistry with Industrial Experience, while carrying out her industrial placement at Pfizer, UK, in Medicinal Chemistry. In 2011–2015, Sam did a PhD in the groupsof Professor Jonathan Clayden and Professor Nicholas Turner on the biocatalytic asymmetric synthesis of atropisomers. Sam joined Johnson Matthey in 2015 as a research chemist in catalysis.

Tommaso Angelini completed his PhD in Chemical Science in 2010 from University of Perugia, Italy, working on the development of environmentally friendly synthetic protocols. During his postdoctoral studies, he finalised his work designing new continuous flow devices for the use of solid supported catalyst in low E-Factor transformations. Later, he gained experience in developing active pharmaceutical ingredient (API) production process at Procos (Italy). In 2015, he joined Johnson Matthey as Research Chemist, designing new enantioselective synthetic process for the preparation of APIs. He is now a Research Expert at Evotec Verona (Italy), working on the production of preclinical and Phase 1 API candidates.

Ahir Pushpanath obtained his PhD in Birkbeck College (University of London, UK) working on the engineering of enzymes for industrial biofuel production. With a biochemistry background, he specialises in the use of bioinformatics and computational biology in the rational design of new enzyme variants. Ahir joined Johnson Matthey in 2013 as a Senior Biologist and was instrumental in demonstrating the utility of computational techniques for rapid enzyme discovery through genome mining, in silico design and targeted enzyme engineering. He currently leads the enzyme development arm of biocatalysis, continuing to develop faster, more effective methods for ‘predictive biocatalysis’.

Amin Bornadel studied chemical engineering and received a PhD in biotechnology from Lund University in Sweden. For postdoctoral work, Amin went to Germany, where he carried out research within biocatalysis at University of Dresden and Technical University of Hamburg. In 2016, Amin joined Johnson Matthey to work as a biocatalysis researcher. He is currently a senior scientist working in the Biotech team.

Elina Siirola completed her PhD in 2012 from the University of Graz, Austria, where she worked on biocatalytic C=C bond hydrolysis. After a postdoctoral position in enzyme engineering at the Max Planck Institute for Coal Research, Germany, she joined Johnson Matthey in 2013, where she worked on biocatalysis research and development (R&D). Since 2017 she is a Principal Scientist in the Bioreactions group at Novartis Pharma in Basel, Switzerland.

Serena Bisagni completed her MSc in Industrial Biotechnology from the University of Pavia, Italy, in 2010 and then moved to Lund University, Sweden, for her postgraduate studies. In 2014 she obtained her PhD in Biotechnology in which she focused on the identification of new Baeyer-Villiger monooxygenases for fine chemicals synthesis within the Marie Curie Innovative Training Networks (ITN) ‘Biotrains’. In 2015 Serena joined Johnson Matthey. Her main interests are enzyme screening for synthesis of active pharmaceutical ingredients and fine chemicals and identification of novel biocatalysts.

Antonio Zanotti-Gerosa studied in Milano, Italy, completing his PhD in 1994 (organometallic chemistry). His academic experience include secondments to Imperial College, UK (Professor S. V. Ley), Nagoya University, Japan (Professor R. Noyori) and postdoctoral research at the University of Lausanne, Switzerland (Professor C. Floriani). Since 1997 he has been working on industrial applications of homogeneous catalysis. In 2003 he joined Johnson Matthey and, as R&D Director, he is leading the chemocatalysis group in the Cambridge laboratories.

Beatriz Domínguez gained her PhD in Synthetic Organic Chemistry from the University of Vigo, Spain, and then moved to the UK where she worked with Professor Tom Brown at the University of Southampton, UK, and with Professor Guy Lloyd-Jones at the University of Bristol, UK. In 2002 she joined Synetix, soon to become Johnson Matthey Catalysts and Chiral Technologies and has worked at Johnson Matthey’s facilities in Cambridge since. Beatriz has gained broad experience in the application of metal catalysis and biocatalysis, working closely with fine chemicals companies to deliver optimal catalysts for chemical processes.

536 © 2020 Johnson Matthey Johnson Matthey Technology Review is Johnson Matthey’s international journal of research exploring science and technology in industrial applications www.technology.matthey.com Editorial team

Manager Dan Carter Editor Sara Coles Editorial Assistant Yasmin Stephens Senior Information OfficerElisabeth Riley

Johnson Matthey Technology Review Johnson Matthey Plc Orchard Road Royston SG8 5HE UK Tel +44 (0)1763 253 000 Email [email protected] www.technology.matthey.com