Inhibition of microbial production of the malodorous substance isovaleric acid by 4,4ʹ dichloro 2‐hydroxydiphenyl ether (DCPP)

Abstract Human body malodour is a complex phenomenon. Several types of sweat glands produce odorless secretions that are metabolized by a consortium of skin‐resident microorganisms to a diverse set of malodorous substances. Isovaleric acid, a sweaty‐smelling compound, is one major malodorous component produced by staphylococci with the skin‐derived amino acid L‐leucine as a substrate. During wearing, fabrics are contaminated with sweat and microorganisms and high humidity propagates growth and microbial malodour production. Incomplete removal of sweat residues and microorganisms from fabrics during laundry with bleach‐free detergents and at low temperatures elevate the problem of textile malodour. This study aimed to analyze the inhibitory effect of the antimicrobial 4,4ʹ dichloro 2‐hydroxydiphenyl ether (DCPP) on the formation of isovaleric acid on fabrics. Therefore, GC‐FID‐ and GC–MS‐based methods for the analysis of isovaleric acid in an artificial human sweat‐mimicking medium and in textile extracts were established. Here, we show that antimicrobials capable to deposit on fabrics during laundry, such as DCPP, are effective in growth inhibition of typical malodour‐generating bacteria and prevent the staphylococcal formation of isovaleric acid on fabrics in a simple experimental setup. This can contribute to increased hygiene for mild laundry care approaches, where bacterial contamination and malodour production represent a considerable consumer problem.


| INTRODUC TI ON
In humans, the generation of body malodour is attributed to microbial activity on the skin. Various sites of the body harbor diverse glands secreting a wide variety of odorless compounds. This setting represents an attractive environment for a unique composition of microorganisms that transform odorless secretions into malodorous substances (Shelley et al., 1953).
The human body bears three types of sweat glands: The eccrine sweat glands are responsible for thermoregulation and mainly secrete a diluted salt solution and organic compounds such as lactic acid, vitamins, glucose, urea, and amino acids (Huang et al., 2002).
The sebaceous glands release lipids and esterified fatty acids, lubricating and waterproofing skin and hair. Finally, apocrine glands found in the human axillae, areolae, genitalia, and ear canal, but not on feet, produce lipid-and protein-rich secretions and odorless steroids (Leyden et al., 1981). These secretions, as well as degradation products of skin-derived keratin (e.g. from callus; Holland et al., 1990), represent a source of nutrients and water for microbial propagation and serve as a substrate for the production of malodorous compounds. Key species in body malodour are represented by the genera Staphylococcus, Corynebacterium, Propionibacterium, Micrococcus, and Brevibacterium (Costello et al., 2009;Leyden et al., 1981;Marshall et al., 1988;Minhas et al., 2018;Shehadeh & Kligman, 1963;Taylor et al., 2003).

have demonstrated how
IVA is microbially produced on human skin: Amino acids originating either from sweat secretions or from microbial keratin degradation are used as a substrate for IVA formation. Keratin degradation is accomplished by microorganisms, including Staphylococcus epidermidis, propionibacteria (Holland, 1993), or Kytococcus sedentarius (Holland et al., 1990(Holland et al., , 1992Longshaw et al., 2002;Nordstrom et al., 1987).
The latter was shown to degrade keratin-containing callus by several keratinases into peptides and free amino acids, for example, L-leucine. Numerous staphylococci isolated from human axillae and feet metabolized L-leucine in a semisynthetic medium that mimics human sweat, to 2-oxoisocaprioic acid, isovaleryl-CoA and eventually IVA. Besides L-leucine, also L-isoleucine and L-valine are microbially degraded to malodorous VFA, namely, 2-methylbutyric acid and isobutyric acid, respectively, but to a much lesser extent. .
Besides body malodour that is present on the skin itself, secretions and skin-degradation products, skin-borne bacteria, and malodorous substances are also transferred to clothes while wearing. Propagated by high humidity, microbial malodour production then occurs on the fabrics during wearing but also afterwards in the laundry basket. The use of mild, bleach-free detergents and low-temperature washing prevents the efficient removal of body soils, bacteria, and malodorous substances. Over several wearing/ washing cycles, this promotes microbial propagation and malodour production on fabrics and in the washing machine (Hammer et al., 2011;Van Herreweghen et al., 2020;Riley et al., 2017). As one of the possible ways to overcome such problems, the use of antimicrobial substances formulated in the liquid laundry detergent has been proposed (Hazenkamp & Ochs, 2011;Ochs et al., 1999). 4,4ʹ dichloro 2-hydroxydiphenyl ether (DCPP) is such an antibacterial compound ( Figure 1). DCPP, available as Tinosan ® HP 100 (30% DCPP in 1,2-propylene glycol), represents a non-ionic substance that is compatible with liquid and powder detergents and exhibits broadspectrum antibacterial properties (Ochs et al., 1999). Applied during laundry, DCPP prevents microbial growth on washed textiles (Ochs et al., 1999). Whether-besides its antimicrobial activity-DCPP also affects microbial malodour formation on textiles is so far unknown.
The aim of this study was therefore to investigate the potential inhibitory effect of the antimicrobial DCPP on the microbial formation of IVA as a model malodorous substance in a relatively simple experimental setup. A medium adapted from James et al.  was incubated with the bacterium Staphylococcus aureus under various in vitro setups including shake flask cultivation and cultivation on cotton fabrics that were washed with a DCPP-containing detergent. Physiological parameters such as growth, extracellular pH, and eventually production of malodorous IVA were followed. A GC-FID and GC-MS method was established to analyze IVA amounts produced by S. aureus on fabrics.
For in vitro growth studies in shake flasks, the cell pellet was re- For growth studies on textiles, a procedure adapted from the US standard AATCC 100-2012 test method (assessment of antibacterial finishes on textile materials (AATCC, 2012)) was conducted. S. aureus cells were pre-cultivated as described above, except that the pre-cultivation was performed in 10 ml TSBT in 100 ml Erlenmeyer shake flasks.
Disks (Ø 4 cm) were punched out from treated cotton textiles and inoculated with 125 µl cell suspension (in SSM, as described above) per textile swatch. Cultivation was performed in sterile Petri dishes (Ø 5.5 cm) at 35°C under humid conditions (>90% humidity) for up to 24 h. Immediately after inoculation of the fabrics and after 24 h of incubation, cells were resuspended from the fabrics in sterile Stomacher ® bags 80 (Seward Ltd.) containing 10 ml buffer (1.7 g/L KH 2 PO 4 , 9.6 g/L Na 2 HPO 4 xH 2 O, 10 g/L Tween ® 80, 3 g/L lecithin, pH 7.4) and were agitated in Seward Stomacher ® 80 (Seward Ltd.) for 1 min at normal speed. Dilutions were prepared in sterile deionized water, and colony-forming units per ml (cfu/ml) were determined on TSAT via the pour plate method after incubation of the plates at 35°C for 24 h. Cfu data presented as log 10 values.

| Treatment of cotton textiles
Standard white cotton fabrics (Renforcé-1-3005, Spoerri & Co. AG, 130 g/m 2 ) were used. To remove any production-related contamination from the fabrics, they were rinsed in 90°C hot tap water for 20 min (10 g fabric/75 ml water) under agitation in a LiniTest ® machine (Atlas), then rinsed by hand in cold tap water for 15 s, dried and sterilized by autoclavation. 10 g of pre-rinsed, sterile cotton textile was then laundered with 50 g of washing solution (0.487% standard European liquid laundry detergent bought in a German supermarket, optionally supplemented with 0.3% or 0.6% Tinosan ® HP 100 (corresponding to 0.09% or 0.18% DCPP), dissolved in tap water (water hardness: 1.75 mM calcium/magnesium ions = 9.8°dH) at room temperature (RT) in a LiniTest ® machine for 45 min. Rinsing of the textiles was performed with 3 x 1 L of sterile tap water at RT for 2 min. Textiles were wrung out and dried.
Treatment was conducted with sterile equipment under low-germ conditions.

| Analysis of isovaleric acid
A gas chromatography (GC) method for the quantification of IVA was adapted from . For the analysis of shake flask cultivation, cultures were harvested by centrifugation after distinct incubation periods, the supernatant was sterile-filtered (pore size: 0.2 µm) and stored at −20°C until use. After thawing, samples were analyzed through GC using a flame ionization detector (GC-FID). GC-FID was performed using a 7890 gas chromatograph equipped with a split/splitless injector and FID (Agilent Technologies).
Chromatographic separation was achieved using an AT-1000 column (100% polyethylene glycol modified with nitroterephthalate; 30 m × 320 µm × 0.25 µm; Thermo Fisher Scientific) with a temperature program of 80°C for 2.5 min; 15 K/min to 155°C for 10 min. 1 µl of the sample was injected at 240°C using a split ratio of 20:1.
For the analysis of textiles, samples were stored at −20°C after specific incubation periods and before extraction. The samples were supplemented with 0.5 ml 2 M HCl/textile swatch, 0.125 ml internal standard (0.012 mg/ml hexanoic acid in ethyl acetate), and 1.125 ml ethyl acetate. Samples were gently agitated on a shaker for 2 h; reaction tubes were rotated by 90° every 30 min. After phase separation, the upper organic phase (containing IVA and the internal standard) was removed and analyzed by the means of gas chromatography-mass spectrometry (GC-MS), since some residues from the laundry detergent on the fabrics interfered with the detection of IVA via FID. GC-MS was performed using a 7890 gas chromatograph (Agilent Technologies) coupled to a 5977A mass selective detector (MSD; Agilent Technologies) using the MassHunter software (Agilent Technologies) for instrument control and data analysis. Chromatographic separation was achieved using an AT-1000 column (100% polyethylene glycol modified with nitroterephthalate; 30 m × 320 µm × 0.25 µm; Thermo Fisher Scientific), with a temperature program of 80°C for 2.5 min; 15 K/min to 155°C; and 155°C for 30 min. 1 µl of the sample was injected at 250°C using a split ratio of 20:1. The MSD was operated in selected ion-monitoring mode acquiring m/z 60, 87, and 43 for IVA and m/z 60, 73, and 87 for hexanoic acid (internal standard). Successful extraction of IVA from TA B L E 1 Bacterial strains used in this study

Staphylococcus aureus ATCC 6538
American Type Culture Collection (ATCC)

Corynebacterium xerosis DSM 20170
Leibniz Institute DSMZ-German Collection of Microorganisms and Cell Cultures textiles was confirmed by analysis of a water-treated textile that was abiotically incubated with IVA.

| Statistical analysis
Data are presented as arithmetic mean values and standard deviations of 3 independent biological replicates, if not stated differently.
Statistical analyses were carried out as a one-way analysis of variance (ANOVA). If significant differences were found, ANOVA was followed by Tukey's honestly significant difference (HSD) test, results presented as p values (p). Analyses were performed assuming independent data and standard weighted means. For statistical analysis of cfu data, Student's t-test was performed for paired values with equal variances. The significance level α = 0.05 was chosen.

| DCPP inhibits the formation of the VFA isovaleric acid by S. aureus in a semisynthetic medium
In the body malodour formation, staphylococci represent key microorganisms for the production of IVA in human sweat. In a semisyn-

| Inhibition of isovaleric acid formation in S. aureus on DCPP-treated textiles
In addition to the shake flask model, a different test system based on fabrics was established to investigate the effect of DCPP on IVA production in a laundry setting. 0.09% or 0.18% DCPP was added to a commercially available liquid laundry detergent (LLD). Cotton fabrics that were washed with the test detergent were inoculated with S. aureus in SSM as described above. Cotton represents a suitable material for our studies as Staphylococcus species tend to grow better on cotton than on polyester (Callewaert, de Maeseneire, et al., 2014). Moreover, Munk et al. demonstrated that odor generated on cotton during wet storage was significantly greater than on polyester (Munk et al., 2001). This may be due to greater water absorbency of the hydrophilic and porous cellulosic cotton fibers.
A stronger odor that is sometimes encountered for polyester fibers might be caused by organic odorous substances that tend to stick stronger to polyester than to cotton and thus are more difficult to remove in washing. (Munk et al., 2001). This phenomenon might be of importance for the field of malodour in clothes but is not in the scope of this study. The growth of the bacterium on the textiles was determined directly after inoculation and after 24 h of incubation ( Figure 4a). For the fabrics treated with either water or LLD w/o DCPP, significant growth of S. aureus could be observed over 24 h of incubation (from 5.4 log 10 cfu/ml immediately after inoculation to 8.2 log 10 cfu/ml after 24 h), demonstrating that potential detergent residues present on the fabrics did not exhibit antimicrobial activity under the test conditions applied. In contrast, on fabrics treated with LLD containing DCPP, no significant growth occurred during

h incubation.
Extraction and GC-MS analysis of IVA that was produced on the fabrics during staphylococcal growth revealed significant microbial IVA production on water-treated and LLD w/o DCPP-treated textiles (18.3 ± 2.9 µg/ml) during 24 h (Figure 4b).
IVA could not be detected on textiles treated with LLD containing DCPP, which is consistent with the finding that these fabrics revealed strong growth inhibition of S. aureus. This outlines the antimicrobial effect of DCPP deposited on the fabrics during laundry that impairs S. aureus' growth and simultaneously prevents the microbial formation of malodorous IVA in a sweat model system on fabrics.  This study is, to our knowledge, the first one demonstrating the inhibition of the microbial production of IVA, one of the key body malodour substances (Caroprese et al., 2009;James, Austin et al., 2013;Kanda et al., 1990;Leyden et al., 1981), by an antimicrobial, DCPP, deposited on fabrics.

F I G U R E 3 Characterization of
SSM, a growth medium adapted from James et al.  mimicking human sweat and containing L-leucine, which is the substrate for IVA synthesis, was used for cultivation studies.
In SSM, 20.1 mg/L DCPP strongly impaired the growth of S. aureus, a representative of the genera of staphylococci. Staphylococci are known as key skin-resident bacteria involved in body malodour formation (Bawdon et al., 2015;. On clothes, predominantly Gram-positive bacteria such as Staphylococcus epidermidis and Staphylococcus hominis and corynebacteria were observed (Callewaert, de Maeseneire et al., 2014). In this study, S. aureus ATCC 6538 was used. This organism might not be the ideal representative of body malodourproducing staphylococci. However, it is a common organism found in many international microbiological standard test methods and was used as a representative Gram-positive bacterium able to metabolize branched aliphatic amino acids, such as L-leucine, to volatile degradation products, like IVA (James Cox, & Worrall, 2013;. For this model study, the relative abundance of a certain bacterial species is not of key importance, but rather its ability to perform these metabolic pathways. SSM was shown to be suitable for the cultivation of body malodour-producing staphylococci and the analysis of the resulting malodorous substances. or SCIN (Callewaert, Buysschaert et al., 2014) contain considerable amounts of sodium chloride, lactic acid and urea but no protein/ peptides, as summarized by Kulthong or Callewaert (Callewaert, Buysschaert et al., 2014;Kulthong et al., 2010).
Under these conditions, we observed bacterial growth inhibition rather than bactericidal activity (Figure 2a). DCPP's bacteriostatic activity was further confirmed by a very mild and delayed pH drop in the DCPP-containing sample that reflected the impaired metabolism. In contrast, a characteristic pH drop in the sample w/o DCPP revealed active metabolism and formation of fermentation products (Somerville et al., 2002). Surprisingly, little is known about the antimicrobial mechanism of DCPP. While DCPP provides bacteriostatic activity at low concentrations as found in typical in-use dilutions of LLDs (Figures 2 and 3), it exhibits rapid bactericidal activity at higher concentrations (Ochs et al., 1999). As observed for other phenolic compounds including diphenylethers, it can be assumed that the li-

pophilic DCPP exerts perturbing effects on bacterial membranes.
This results in impacts on membrane functionality, and, esp. at higher concentrations, on membrane integrity (McDonnell & Russel, 1999;Villalaín et al., 2001). Furthermore, an inhibitory effect of DCPP on membrane lipid synthesis could be concluded from studies on DCPP complexes with the enoyl-acyl carrier protein reductase of Helicobacter pylori (Lee et al., 2007). Furthermore, the bacteriostatic effect of DCPP is accompanied by a reduction of the metabolization of L-leucine to IVA, caused by a lower number of bacteria and indicated by the finding that almost no IVA was produced in the DCPPcontaining sample, while significant amounts were found in the absence of DCPP (Figure 2b). Additionally, we could show the broadspectrum antimicrobial activity of DCPP in medium containing typical components of human sweat against the second important class of body malodour-producing microorganisms: corynebacteria ( Figure 3; Callewaert et al., 2013;Leyden et al., 1981;Shehadeh & Kligman, 1963)). Further studies are required to provide more evidence for this assumption. DCPP exhibited a lower impact on the growth of C. xerosis than observed for S. aureus. This observation is in consistence with previous results on minimal inhibitory concentration (MIC) values for DCPP, revealing a lower susceptibility of C. xerosis (MIC = 6 mg/L) compared to S. aureus and other skin-borne staphylococci, such as S. epidermidis (MIC = 0.06 mg/L) (Ochs et al., 1999). These findings clearly outline that our simplified experimental malodour setup is suitable for detecting the effectiveness of an antimicrobial, here DCPP, in the inhibition of one of the major body malodour producers, Staphylococcus. Furthermore, the data indicate that corynebacterial malodour production might also be affected.
We wondered whether these effects could be confirmed under laundry care conditions. Cotton fabrics washed with a standard LLD containing 0.09% to 0.18% DCPP, a realistic concentration used in laundry care, were inoculated with a suspension of S. aureus and SMM ( Figure 4). Confirming the data obtained in shake flask cultiva-  (Ochs et al., 1999). In the liquid culture experiment, the concentration of 20.1 mg/L is clearly above the MIC value and was chosen to demonstrate general principles of malodour prevention by DCPP.
To tackle increased microbial contamination of fabrics (Hammer et al., 2011;Riley et al., 2017) and emerging malodour formation caused by mild laundry processes, several technologies are available, as previously reviewed (Hazenkamp & Ochs, 2011  In vivo studies with fabrics that were washed with a DCPPcontaining detergent and which were worn by volunteers could shed more light on the relevance of these findings and might be the subject for future studies. Although biocides, unlike antibiotics, have a nonspecific mode of action, the use of broad-spectrum biocides might impact the composition of microbial communities, as recently shown . Therefore, the effect of DCPP on the microbiome and consequences of potential alterations thereof for malodour production remain interesting questions which may be addressed in future in vivo studies, as well. For all uses of biocides in consumer care, besides general toxicological and environmental hazards, the risk assessment for the products need to consider specific features of biocidal actives, such as resistance and cross-resistance development to other biocides or antibiotics (e.g., (Condell et al., 2012;Edgar & Bibi, 1997;Tandukar et al., 2013)), and their relevance under realistic use conditions (SCENIHR, 2009, Bloomfield, 2002, Smith et al., 2012.

| CON CLUS IONS
In this study, we have established GC-FID-and GC-MS-based methods to analyze microbially produced IVA in artificial human sweatmimicking medium and on fabrics. We demonstrated that IVA is produced in this medium on fabrics by the skin-resident bacterium S. aureus. Moreover, we show a relatively simple experimental setup for analyzing the effects of an antimicrobial on body malodour. The study provides evidence that staphylococcal growth, metabolism, and IVA production in a medium that functionally mimics sweat and on fabrics contaminated with the bacterium can be inhibited by an antimicrobial substance, DCPP, that is deposited on fabrics during laundry with a DCPP-containing detergent. Furthermore, the activity of the antimicrobial against corynebacteria, important body malodour producers, was shown, suggesting a broad-spectrum malodour inhibition on fabrics. Thus, antimicrobial technologies showing deposition on textiles during the laundry process represent an interesting approach to increase hygiene in laundry care where lowtemperature washing with mild detergents elevates the problem of microbial contamination and malodour formation.

ACK N OWLED G EM ENTS
The authors would like to thank Frank Metzger and Christian Rasch Writing-review & editing (supporting).

E TH I C S S TATEM ENT
None required.

DATA AVA I L A B I L I T Y S TAT E M E N T
The datasets generated and/or analyzed during the current study are provided in full in this publication.