Synergism between hydrogen peroxide and seventeen acids against five agri-food-borne fungi and one yeast strain

Authors


  • Research paper on improvement of fungicidal efficacy of hydrogen peroxide in terms of synergy when associated with various acids.

Correspondence

Hélène Martin, Anses, Fougères Laboratory, La Haute Marche, Javené, BP 90203, 35302 Fougères Cedex, France. E-mail: helene.martin-lesaout@anses.fr

Abstract

Aims

The objective of this study was to evaluate fungicidal efficacy of hydrogen peroxide administered in combination with 17 mineral and organic acids authorized for use in the food industry.

Methods and Results

The assays were performed on a 96-well microplate using a microdilution technique based on the checkerboard titration method. The six selected strains (one yeast and five fungi) were reference strains and strains representative of contaminating fungi found in the food industry. Each synergistic hydrogen peroxide/acid combination found after fifteen minutes contact time at 20°C in distilled water was then tested in conditions simulating four different use conditions. Twelve combinations were synergistic in distilled water, eleven of these remained synergistic with one or more of the four mineral and organic interfering substances selected. Hydrogen peroxide/formic acid combination remained effective against four strains and was never antagonistic against the other two fungi. Combinations with propionic acid and acetic acid stayed synergistic against two strains. Those with oxalic acid and lactic acid kept their synergism only against Candida albicans. No synergism was detected against Penicillium cyclopium.

Conclusions

Synergistic combinations of disinfectants were revealed, among them the promising hydrogen peroxide/formic acid combination.

Significance and Impact of the Study

A rapid screening method developed in our laboratory for bacteria was adapted to fungi and used to reveal the synergistic potential of disinfectants and/or sanitizers combinations.

Introduction

Moulds and yeasts are ubiquitous in water, soils, plants and animals and consequently in a range of fresh and processed food, in raw materials and in numerous other products (Wirtanen and Juvonen 2002). Spoilage fungi also proliferate on process surfaces as complexes of yeast cells and excretory substances that protect them from sanitizing agents. Food industries present all the conditions necessary (temperature, hygrometry and nutritive conditions) to favour development of moulds and yeasts brought mostly by air, water or contact. At least 200 species of fungi exist in food (Deak 1991; Kurtzman and Fell 1998; Barnett et al. 2000). We previously showed the synergistic interaction of hydrogen peroxide with acids on bacterial contaminants (Martin and Maris 2012). Few reports have investigated the susceptibility of food-borne fungi to chemicals as disinfectants used in the food industry, and they focused on only a few spoilage species, food processes and sanitizing agents. Two papers tried to complete our knowledge in this field. The first one using a suspension method studied the resistance to 15 commercial disinfectants in 19 moulds and six yeast species found in bread and cheese factories (Bundgaard-Nielsen and Nielsen 1996). In that work, a 3·0% hydrogen peroxide solution with peracetic acid showed low fungicidal effects (0 to 1-log10 reduction) against 18 strains after 10-min contact time at 20°C. However, hydrogen peroxide damaged the conidia in seven of the moulds. According to the authors, the poor killing effects probably resulted from the low solution concentration and the short contact time used. The second work evaluated the efficacy of 17 disinfectants and foam cleaners on 25 yeast strains isolated from dairy, fermented functional food, bakery, praline, jam and sugar processes; all strains were tested in suspension tests, and 16 strains were tested by disinfection tests based on yeasts biofilms (Salo and Wirtanen 2005). A disinfectant with peracetic acid and hydrogen peroxide was effective at 1% on vegetative cells, except for some Candida spp. In fact, the efficacy of hydrogen peroxide against yeasts, including Candida albicans and some specific fungi, is already established in ophthalmology and dentistry (Rosenthal et al. 1999; Szymanska 2006). These results and the synergistic combinations between hydrogen peroxide and some acids that we reported against bacteria in a previous work (Martin and Maris 2012) prompted us to test the disinfectant efficacy of hydrogen peroxide with different acids (four mineral and 13 organic acids) towards a panel of fungal strains including reference strains selected in the French standards as model for yeasts and fungi representative of contaminants found in the food industry (Splittstoesser et al. 1980; Anon 1987a,b; Moreau 1988; Hamdy et al. 1990).

Materials and methods

Chemicals

Hydrogen peroxide and the 17 mineral and organic acids were purchased from three suppliers. Hydrogen peroxide (50% m/v), phosphoric acid (purity 84%), nitric acid (purity 65–69%), formic acid (purity 100%), acetic acid (purity 100%) and oxalic acid (purity 75%) were provided by Prolabo (VWR, Fontenay-sous-Bois, France). Boric acid (purity 99·8%), sulfuric acid (purity 95–97%), citric acid (purity 91·4%), benzoic acid (purity 91·4%), adipic acid (purity 100%), glutaric acid (purity >99%), propionic acid (purity 99%), succinic acid (purity 99·5%), tartaric acid (purity 99·5%) and mandelic acid (purity 99%) were provided by Merck (VWR). Sulfamic acid (purity 100%) and lactic acid (purity 85%) were provided by Sigma (Saint-Quentin-Fallavier, France). Stock solutions were prepared in sterile distilled water at 1% (m/v) and 2% (m/v) for adipic and boric acids, at 2% (m/v) for benzoic acid, at 5% (m/v) for oxalic and succinic acids, at 10% (m/v) for mandelic and sulfamic acids and at 20% (m/v) for citric, glutaric, tartaric acids or sulfuric and nitric acids. Stock solutions of the other acids (formic, acetic, phosphoric, propionic and lactic acids) were prepared at 20, 40 or 80% (v/v).

Yeast and fungal strains

We selected six fungal strains (one yeast and five fungi) including two reference strains and four strains representative of different contaminant species found in the agri-food industry. Candida albicans ATCC 2091 and Absidia corymbifera CIP 1129-75 were reference strains in the French A.F.NOR standards recommended by French regulation, before the edition of the European standards, to determine the fungicidal efficacy of commercial disinfectants. These fungi are cosmopolitan and ubiquitous fungi. Candida albicans is a yeast often found in fruit and vegetable products, and A. corymbifera is well-known for its resistance to products and grows easily in humid and warm environments. The four other fungi selected, Geotrichum candidum CIP 285-54, Scopulariopsis brevicaulis CIP 210-53, Aspergillus versicolor CIP 1187-79 and Penicillium cyclopium CIP 1231-80, were strains representative of meat, milk and cheese industries contaminants and/or strains found in fruit, vegetable and cereal transformation or storage processes.

Assay conditions

Combinations between hydrogen peroxide and acids were tested first using sterile distilled water. Then, synergistic combinations were tested using interfering substances. Four interfering substances were chosen among those recommended in the French NF T 72-170 standard (Anon 1988). Two hard water levels (300 and 600 mg l−1 CaCO3) were tested to evaluate the impact of water hardness on fungicidal activity of combinations. Two organic substances were chosen: one to simulate clean conditions (0·3% bovine albumin with hard water at 300 mg l−1 CaCO3), the other to simulate dirty conditions (1% bovine albumin + 1% yeast extract) according to the conditions of use of products in food industry area. The concentrations indicated were the final concentrations in the assays.

Test procedure

The ND/1500 checkerboard titration method was used also to study the efficacy of each component of the combination against each strain. The minimal fungicidal concentrations (MFCs) were thus obtained under basic conditions (with distilled water) or conditions simulating the practical use (with interfering substances). The ND/1500 method was based on the checkerboard titration method that has long been used for antibiotics (Krogstad and Moellering 1980). The ND/1500 method was used to prevent insufficient neutralization of the products. This method combined the use of a 1 : 1500 dilution with the use of a chemical neutralizer to ensure sufficient neutralization of the products. This method was developed in our laboratory to test efficacy of disinfectant combinations against bacteria (Martin and Maris 1993) and was already applied for bacteria in a similar study (see descriptions of the ND/1500 test procedure, the neutralization control and the results validation methods in Martin and Maris 2012). In this work, ND/1500 method was adapted for yeast and fungi following the recommendations of the French A.F.NOR NF T 72-200 (Anon 1987b) and NF T 72-170 (Anon 1988). These modifications concerned the level of inocula in contact with products (1 or 3×107 CFU ml−1 instead of 1 or 3×108 CFU ml−1 for bacteria), the contact time between combinations and yeast or fungus (15 min at 20°C instead of 5 min at 20°C), the composition of the liquid for preparing spore suspensions of fungi (0·05% (v/v) aqueous solution of polysorbate 80) and the culture medium used to measure the survival of yeast and fungi (yeast extract 5 g, glucose 20 g, Bacto agar 15 g and distilled water 1 l). The following temperatures and incubation times for agar plates were selected: 30°C during 72 h for C. albicans, 30°C during 40–70 h for A. corymbifera and 25°C during 4 days for the other strains. As for bacteria, MFC value of a product was the lowest product concentration for which no vegetative fungal cell was detected on agar plates. Synergistic combinations were detected and their strength calculated (see below) when no fungal growth was detected on agar plates, corresponding to the combinations of products at the lowest concentrations. Neutralization of products was considered sufficient if results of numerations collected on agar plates for neutralized product concentrations were higher or equal to the half of the control numeration. To definitely validate our results, the consistency of MFC values was checked for each product as described in the study by Martin and Maris (2012) for bacteria using MFC values previously collected with a home-made test (Maris et al. 1982) adapted and routinely used in our laboratory to control the efficacy of disinfectants against fungi. A ±1 dilution difference was accepted between results of ND/1500 method and the home-made test, when a MFC value was reached with the two methods (for hydrogen peroxide and some acids). For acids that were not effective on a fungal strain (MFC not reached, even with the highest product concentration), ND/1500 assay was performed using serial dilutions made from the highest concentration of acid that it is possible to prepare (depending on the acid dissolution capacity in distilled water). Acid serial dilutions were then mixed in the microplate with hydrogen peroxide serial dilutions as described in the study by Martin and Maris (2012). For each MFC and each synergistic combination, pH value in assay conditions was measured.

Exploitation of results

Minimal Fungicidal Concentration (MFC): for the microplate row or column corresponding to each product alone, fungal growth or total microbial destruction was noted. First dilutions without fungal growth were considered to be the minimal fungicidal concentrations (MFCs). At these concentrations, a 3-log10 reduction in viable count was theoretically reached. Results were expressed in acid or hydrogen peroxide percentages (v/v or m/v).

Fractional Fungicidal Concentration index (FFC): each result was expressed as a fractional fungicidal concentration index (FFC), representing the degree of interaction of products when used in combination. FFC values were expressed as a fraction of the MFC (FFC = MFC product in combination/MFC product alone). We then calculated the sum of the FFC values (Σ FFC = MFC product A in association/MFC product A + MFC product B in association/MFC product B) for interpretation of the results based on the following criteria:

  1. Σ FFC ≤ 0·50: synergism
  2. 0·50 < Σ FFC < 2: addition
  3. Σ FFC ≥ 2: antagonism

Interference index: to facilitate interpretation of the interference data on the effect of hard water and organic substances on the efficacy of the combinations, MFC values representing optimal synergy were converted into interference indices. These interference indices corresponded to the ratio of the MFC value of each product when tested in combination in the presence of an interfering substance (hard water or organic substance) to the MFC value of the same product tested in combination in the absence of interfering substance. We used the interference classification system previously described (Guiraud-Dauriac and Crémieux 1984; Martin and Maris 2012). The intervals, chosen arbitrarily, were as follows:

  1. Class 1: no effect, interference index ≤1.
  2. Class 2: weak effect, interference index between 1·1 and 5.
  3. Class 3: intermediate effect, interference index between 5·1 and 10.
  4. Class 4: strong effect, interference index between 10·1 and 25.
  5. Class 5: very strong effect, interference index >25.

Results

Fungicidal efficacy of hydrogen peroxide in distilled water

Median MFC values (± 1 dilution) obtained in basic assay conditions using the ND/1500 micromethod showed that hydrogen peroxide was fungicidal at concentrations varying from 0·39% (m/v) for the more sensitive fungus (G. candidum) to 12·5% for the less sensitive fungus (A. corymbifera). The means and the standard errors calculated from the 16 or 18 results of MFC collected for each strain in distilled water gave a more precise result for these MFC values and confirmed the results furnished by median values (Table 1). The hydrogen peroxide MFC values collected for each synergistic combination detected in basic assay conditions during each ND/1500 checkerboard titration assay are indicated in Table 2.

Table 1. MFC values of hydrogen peroxide in distilled water with ND/1500 method (median and mean values)
StrainsMFCa% median value (m/v or v/v) (± 1 dilution)MFCa% mean value ± standard error (m/v or v/v)
  1. a

    Minimum number of assays per result: 16–18.

Absidia corymbifera CIP 1129-7512·512·50 ± 0
Candida albicans ATCC 20913·123·90 ± 1·54
Scopulariopsis brevicaulis CIP 210-533·125·51 ± 3·03
Aspergillus versicolor CIP 1187-791·561·71 ± 0·59
Penicillium cyclopium CIP 1231-801·562·34 ± 1·28
Geotrichum candidum CIP 285-540·390·50 ± 0·34
Table 2. Values of acid and hydrogen peroxide concentrations (%/%) and pH measured in combinations staying synergistic with interfering substances
AcidCombinationAcid/H2O2Acid/H2O2Acid/H2O2Acid/H2O2Acid/H2O2MFC of acida alone (%)MFC of H2O2b alone (%)
In combination with H2O2Assay conditionsDistilled water300 mg l−1 CaCO3600 mg l−1 CaCO3Bovine albumin (0·3%)Bovine albumin yeast extract (1%)Distilled waterDistilled water
Strain%/%PH%/% PH %/% PH %/% PH %/% PH % PH % PH
  1. ND, not determined.

  2. a

    Value of MFC (%) measured for acid alone during ND/1500 assay realized in distilled water conditions.

  3. b

    Value of MFC (%) measured for H2O2 alone during ND/1500 assay realized in distilled water conditions.

  4. c

    Combination not found synergistic in the assay(s) condition(s) (FFC >0·50).

  5. d

    MFC value was determined in assay conditions using liquid for preparing spore suspensions.

  6. e

    MFC value determined in assay conditions using liquid for preparing yeast suspensions.

FormicAbsidia sp.0·31/1·56 2·80 0·31/0·78 2·75 0·31/0·78 2·65 0·31/0·78 2·95 0·31/1·56 3·36 2·5 2·05 12·5 3·35 d
Candida sp.0·039/0·195 3·22

0·15/0·19

0·078/0·39

2·76

2·91

0·078/0·39 2·88

0·15/0·39

0·078/0·78

3·11

3·37

0·078/0·78 4·19 1·25 2·19 3·12 6·14 e
Scopulariopsis sp.0·078/0·39 3·02 0·15/0·78 2·82

0·15/0·195

0·078/1·56

2·81

2·91

0·156/0·19

0·078/1·56

2·95

3·42

c 0·31 2·50 3·12 4·29
Geotrichum sp.0·019/0·095 3·35 0·039/0·19 3·13 0·039/0·095 3·10 0·078/0·19 3·43 0·078/0·78 4·08 0·15 >2·50 0·39 4·91
PropionicAbsidia sp.2·5/3·12 2·87 2·5/3·12 2·84 2·5/1·56 2·87 2·5/3·12 3·23 2·5/3·12 3·64 20 2·54 12·5 3·35
Aspergillus sp.0·62/0·39 3·22 c c c 1·25/0·195 3·89 5 2·79 1·56 4·69
AceticAbsidia sp.2·5/0·78 2·90 1·25/0·78 2·95 2·5/0·78 2·76 2·5/1·56 3·09 5/0·78 3·25 20 2·30 12·5 3·35
Candida sp.

1·25/0·195

0·31/0·78

3·00

3·19

2·5/0·78

1·25/1·56

2·74

2·84

1·25/0·39 2·85 1·25/0·39 3·25 1·25/0·39 3·69 10 2·36 3·12 6·14
OxalicCandida sp.0·62/0·39 1·75 0·62/1·56 1·64 1·25/1·56 1·40 1·25/1·56 1·42 0·62/1·56 2·15 2·5 1·17 3·12 6·14
LacticCandida sp.2·5/0·39 2·26 5/1·56 2·08 c

5/0·78

2·5/1·56

2·22

2·40

5/3·12 2·52 10 2·03 3·12 6·14
SulfuricCandida sp.0·78/0·78 0·99 c 2·5/3·12ND c c 5 0·60 3·12 6·14

Fungicidal efficacy of the 17 mineral and organic acids

MFC values obtained for the 17 acids were already published (Martin and Maris 2005). To facilitate comparisons between acids, MFC values were expressed in acid percentages (v/v or m/v) and in molar concentrations (mol l−1). For each MFC, the pH was measured and the ratio between undissociated and dissociated acid was calculated. A MFC value was reached for nine of the 17 acids in water. The number of effective acids varied from 5 (for A. versicolor) to 8 (for C. albicans and G. candidum). Four acids (nitric, formic, propionic and acetic acids) were fungicidal on all six strains tested, two acids (mandelic and lactic acids) on five strains, one acid (sulfuric acid) on four strains and two acids (oxalic and phosphoric acids) on only one strain (either C. albicans or S. brevicaulis). As seen for hydrogen peroxide, A. corymbifera was the most resistant strain and G. candidum was the most sensitive. The MFC values of acids collected for each synergistic combination detected in basic assay conditions during each ND/1500 checkerboard titration assay are indicated in Table 2.

Characteristics of the products combinations: detection of synergies

Among the 102 assays realized in distilled water, 12 pairs of product combination/fungal strain had a synergistic effect with Σ FFC between 0·09 and 0·50 (Table 3). In most cases (92%), Σ FFC was ≤0·37, a sign of strong synergy, indicating that concentrations <25% of compound MFCs alone were fungicidal. Six synergistic combinations were detected for C. albicans, three for A. corymbifera, one for S. brevicaulis, A. versicolor and G. candidum and none for P. cyclopium. A synergy between hydrogen peroxide and formic acid was noted for four of the six strains, and combinations with propionic acid, acetic acid, or oxalic acid were synergistic for three or two strains, including the resistant strain, A. corymbifera. All the four acids cited plus lactic and sulfuric acids were synergistic in combination against C. albicans in distilled water conditions. Oxalic, lactic and sulfuric acids were effective in combination only against C. albicans. Concerning P. cyclopium and G. candidum, no synergistic combination was detected in distilled water conditions.

Table 3. Recapitulative table of the synergistic combinations detected in distilled water. The synergy is expressed as the sum of the fractional fungicidal concentrations (Σ FFC)
Nature and characteristics of the acid in combination with hydrogen peroxideΣ FFC per fungus 
AcidMWg mol−1pK1;2;3123456Number of synergistic combinations per acid
  1. 1: Absidia corymbifera; 2: Candida albicans; 3: Scopulariopsis brevicaulis; 4: Aspergillus versicolor; 5: Penicillium cyclopium; 6: Geotrichum candidum.

  2. MW: molecular weight.

  3. pK1;2;3: pK values found coming from http://sakura.cpe.fr/pka.html.

  4. a

    Strong acid.

  5. b

    Combination not synergistic (ΣFBC > 0·50).

Formic46·023·740·250·090·37 b b 0·374
Propionic74·084·870·370·50 b 0·37 b b 3
Acetic60·054·740·180·19 b b b b 2
Oxalic90·041·27; 4·27 b 0·37 b b b b 1
Lactic90·083·83 b 0·37 b b b b 1
Sulfuric98·08a; 1·92 b 0·37 b b b b 1
Nitric63·02 a b b b b b b 0
Succinic118·094·16; 5·61 b b b b b b 0
Benzoic122·14·2 b b b b b b 0
Tartaric150·092·95; 4·25 b b b b b b 0
Citric192·123·14; 4·77; 6·39 b b b b b b 0
Sulfamic971 b b b b b b 0
Phosphoric982·15; 7·21; 12·67 b b b b b b 0
Mandelic152·143·40 b b b b b b 0
Adipic146·144·42; 5·42 b b b b b b 0
Glutaric132·114·34 b b b b b b 0
Boric61·839·24 b b b b b b 0
Total        12

Maintenance of synergism between hydrogen peroxide and acids in the presence of interfering substances

The 12 synergistic combinations (Σ FBC ≤ 0·50) were then tested in the presence of four interfering substances. Synergism was maintained for 11 combinations (Table 2). The respective concentration of each component of the hydrogen peroxide/acid combination and pH value at the optimal point of synergy (OPS) are given in Table 2. Of these 11 pairs (product combination/fungal strain), only seven stayed synergistic with the four interfering substances: three couples with formic acid, two with acetic acid, one with propionic acid and one with oxalic acid. One couple with formic acid and one with lactic acid maintained synergy with three interfering substances. In Table 4, pH values reached at OPS measured for all the synergistic combinations found for C. albicans were gathered as example of pH values evolution in the presence of various assay conditions simulating use conditions.

Table 4. PH values measured in fungicidal mixtures (acid or combination solutions + yeast suspension + distilled water or interfering substances) at CMF of acid or at the optimal point of synergy (OPS) found in the different assay conditions tested for Candida albicans ATCC 2091
Acid in association with hydrogen peroxidepH
MFC of acid aloneOPS of combination distilled waterOPS of combination 300 mg l−1 CaCO3OPS of combination 600 mg l−1 CaCO3OPS of combination OM1bOPS of combination OM2c
  1. a

    Combination not found synergistic.

  2. b

    0·3% bovine albumin (in hard water at 300 mg l−1 CaCO3).

  3. c

    1% bovine albumin + 1% yeast extract.

  4. ND, Not determined.

Formic2·193·222·75; 2·912·883·11; 3·374·19
Acetic2·363·00; 3·192·74; 2·842·853·253·69
Oxalic1·171·751·641·401·422·15
Lactic2·032·262·08 a 2·22; 2·402·52
Sulfuric0·600·99 a ND a a

Interference indices: the effect of mineral or organic substances on MFC values of acid and hydrogen peroxide in a synergistic combination

Interference indices calculated for hydrogen peroxide (see Table 5) in the presence of hard water and organic interfering substances were class 1 interference indices (interference index values equal to 0·25, 0·50 and 1). One exception was for G. candidum and hydrogen peroxide/formic acid combination. In this case, we found interference indices of class 2 (interference index equal to 2) in hard water conditions and class 3 with a high organic load (interference index equal to 8). Candida albicans showed class 2 (with interference indices between 2 and 4) or class 3 interference indices (interference index equal to 8) with lactic acid and bovine albumin plus yeast extract. Acids generally showed similar interference indices (see Table 5) to hydrogen peroxide (class 1 and 2, with interference indices usually between 1 and 2, and not more than 4). No class 3 interference was detected. These interference indices were stable under all assay conditions. Absidia corymbifera showed interference values of 1 or less.

Table 5. Interference indices (FFC) calculated for acid and hydrogen peroxide for each combination synergistic in the presence of interfering substances
StrainAcidHydrogen peroxide
Nature of acid in combination with H2O2300 mg l−1 CaCO3a600 mg l−1 CaCO3bOM1cOM2d300 mg l−1 CaCO3600 mg l−1 CaCO3OM1OM2
  1. a

    hard water at 300 mg l−1 CaCO3.

  2. b

    hard water at 600 mg l−1 CaCO3.

  3. c

    0·3% bovine albumin (in hard water at 300 mg l−1 CaCO3).

  4. d

    1% bovine albumin + 1% yeast extract.

  5. e

    combination not synergistic (ΣFBC > 0·50).

Absidia corymbifera CIP 1129-75Formic11110·500·500·501
Acetic0·501121121
Propionic111110·5011

Candida albicans

ATCC 2091

Formic4-224-221-222-44
Acetic2-1; 8-41-41-41-44-8; 1-22-0·502-0·502-0·50
Oxalic11214444
Lactic222-124 e 2-48
Sulfuric e 3 e e e 4 e e
Scopulariopsis brevicaulis CIP 210-53Formic222 a 20·500·50 e

Aspergillus versicolor

CIP 1187-789

Propionic e e e 1 e e e 0·50
Geotrichum candidum CIP 285-54Formic22442128

Discussion

The sensitivity of A. corymbifera, C. albicans and S. brevicaulis to hydrogen peroxide was similar to Gram-positive bacteria (Martin and Maris 2012), with MFCs between 3·12 and 12·5%. Hydrogen peroxide MFCs for A. versicolor and P. cyclopium were identical to Gram-negative bacteria and equal to 1·56%. Geotrichum candidum was the most sensitive fungus to hydrogen peroxide with a MFC of 0·39%, 32-fold lower than the MFC for A. corymbifera. Bundgaard-Nielsen and Nielsen (1996) found that 10-min contact with a 3% hydrogen peroxide solution at 20°C had lower fungicidal effects (0 to 1-log10 reduction) on Penicillium, Cladosporium, Scopulariopsis, Aspergillus and Eurotium, but conidia damage occurred in seven moulds. We found a 3-log10 fungicidal reduction after 15 min at 20°C with 1·56 or 3·12% hydrogen peroxide solutions against S. brevicaulis, A. versicolor and P. cyclopium (Table 1). Similarly, Buchen and Marth (1977) noted times between 18·3 min and >120 min were necessary to destroy Aspergillus parasiticus and Aspergillus flavus conidia in 4% hydrogen peroxide. A 3% hydrogen peroxide system showed an average 0·4 to 4-log10 reduction for C. albicans, Fusarium solari and Aspergillus fumigatus after 2–6 h of contact lens disinfecting time (Rosenthal et al. 1999). In a study on dental unit waterline disinfection (DUWL), Szymanska (2006) confirmed the antimycotic effectiveness of a disinfecting procedure constituted of the use of a 0·25% hydrogen peroxide solution enhanced with silver ions for 30 min followed by the constant presence of 0·02% hydrogen peroxide in DUWL. This disinfecting method used on yeast-like and mould fungi caused a significant decrease both in the number of total fungi and in individual species as C. albicans (constituting from 31·2 to 85·7% of the total fungi). Application of hydrogen peroxide caused a reduction of total flora from 410 to 5·6 CFU ml−1 in reservoir, 578·5–6·4 CFU ml−1 in hand piece and 43·43 CFU mm−2–0·22 CFU mm−2 in biofilm. However, different isolates of one species may show different responses to the same disinfectant, resulting in an effective killing one case and almost no effect in others (Bundgaard-Nielsen and Nielsen 1996).

Generally, yeast and fungi grow better in acidic media, whereas bacteria grow better in neutral or slightly alkaline pH (Genigeorgis 1981). The MFCs for the 17 acids against the six fungi were previously presented (Martin and Maris 2005). In this last paper, a relatively low variability was shown between pH values reached at MFCs of acids determined for the six fungal strains tested. An example was the case of formic acid: a pH level of 2·86 was necessary to kill G. candidum when pH values between 2·05 and 2·50 were necessary against the five other fungal strains. The mineral acids were fungicidal at low pH (between 0·77 and 1·02 for nitric acid, 0·40 and 0·85 for sulfuric acid and phosphoric acids), and the organic acids were fungicidal at higher pH (between 1·75 and 2·92). Sensitivity gradients for formic acid, acetic acid, or propionic acid showed A. corymbifera as the most resistant fungus and G. candidum as the least resistant fungus. Strong acidity was always necessary to destroy fungal contamination. Indeed, Salo and Wirtanen (2005) confirmed that sulfamic acid was ineffective against strains growing in biofilm, even at a concentration fourfold higher than recommended by the manufacturer (one tablet of 15–30% maleic acid + 5–15% sulfamic acid in 2·5 and 1·25 l). Propionic, benzoic and sorbic acids require long treatment periods to be effective against acidophilic fungi, but can inhibit the growth of moulds and bacteria (Collins 1971; Thabib et al. 1982).

Only six acids were found synergistic with hydrogen peroxide against fungi, but these synergies were mostly stable regardless of the assay conditions (Table 2). Formic acid was the most effective pairing against A. corymbifera, C. albicans, S. brevicaulis and G. candidum, as well as for bacteria (Martin and Maris 2012). Fungicidal activity of these combinations is rarer than bactericidal activity, but the combination of hydrogen peroxide with formic acid keeps all its interest against fungi. Propionic acid also showed interesting efficacy in combination against A. corymbifera, but not against C. albicans and A. versicolor when interfering substances were added. Acetic acid showed stable synergy on A. corymbifera and C. albicans. Candida albicans was the most sensitive strain to synergistic interactions (six synergistic acids in water and five with interfering substances). In the work presented here, the six combinations of hydrogen peroxide with an acid were synergistic at pH conditions less acid than when the molecule of acid was used alone. So, in distilled water conditions, variations of pH values between +0·23 and +1·03 were measured at the optimal point of synergy of the combination comparing to pH value reached at MFC of acid (see Table 2). As for bacteria (Martin and Maris 2012), when tested in hard water conditions, the remaining synergistic combinations were effective at similar or slightly lower pH values than in distilled water conditions, related to protective effect on fungicidal cells of calcium ions. Organic interfering substances interfered with products, and fungicidal efficacy was maintained with higher concentrations of each pH and less acidic pH conditions (Tables 2 and 4). Assay conditions tested in our work were hard water and/or organic conditions imposed by the French standard institution (A.F.NOR) to simulate practical utilization planned for disinfectants. These interfering reference substances were used in France until European regulation came into force to determine the fungicidal efficacy of commercial disinfectants used in agriculture and food industry. However, the conditions of use of these combinations may be different, mainly if these combinations are used as fresh vegetables, meat or cereals sanitizers and they do not always reflect all the possibilities of application of the products. The impact of food pH, composition and texture and the impact of the temperature of storage are parameters that could influence favourably or not the performance of acid and hydrogen peroxide combinations. Each active substance constituting our synergistic combinations has yet been investigated individually for its capacity of decontamination, most often against bacteria, and its impact on organoleptic quality of produce has been sometimes investigated (Cai et al. 1995; Wagenaar and Snijders 2004; Rico et al. 2007). As the organoleptic consequences of the utilization of sanitizers are related to their level of concentration and as lower concentrations were necessary for each active substance in synergistic combinations, the negative organoleptic effect of each active substance on produce may be reduced. However, complementary studies are necessary to confirm this assumption.

As the objective of our work was only to screen and to place in a prominent position synergistic combinations, we were not interested in dealing about the additive or antagonistic combinations. However, it is important to notice that, contrarily to bacteria for which no antagonistic combination was determined (Martin and Maris 2012), we noticed such combinations for two of the six selected fungi, P. cyclopium and G. candidum. So, concerning P. cyclopium, the Σ FFC values were >2 or ≤2·25 (results not shown) for the combination of hydrogen peroxide with each of the seven following acids (acetic acid, nitric acid, propionic acid, sulfamic acid, succinic acid, benzoic acid and mandelic acid). For G. candidum, 11 antagonistic combinations were found, with Σ FFC values between 2·03 and 2·50 (acetic acid, boric acid, propionic acid, oxalic acid, sulfamic acid, orthophosphoric acid, sulfamic acid, succinic acid, glutaric acid and citric acid) or ≤4·12 (nitric acid) (results not shown). However, these antagonisms were weak, perhaps related to the composition of the fungal membranes that may contain degradation enzymes.

In conclusion, synergies between hydrogen peroxide and 17 different acids were found against one yeast and five fungi, and their stability in conditions similar to use conditions was confirmed. Only five acids, formic acid, acetic acid, propionic acid, oxalic acid and lactic acid, had a synergistic effect in the presence of mineral or organic matter. The combination of hydrogen peroxide with formic acid was synergistic against four strains (A. corymbifera, C. albicans, S. brevicaulis and G. candidum), those with acetic acid and propionic acid against two strains (A. corymbifera and C. albicans or A. corymbifera and A. versicolor). For oxalic acid or lactic acid, synergies were maintained with interfering substances for only one strain (C. albicans). Candida albicans was sensitive to more numerous combinations of hydrogen peroxide and acid than fungi. Less synergistic combinations were found for fungi than for bacteria. However, using the rapid screening micromethod described for bacteria and adapted in this work for yeast and fungi, we confirmed the interest of the hydrogen peroxide and formic acid combination. Its synergy was demonstrated in distilled water against fungal suspensions and was maintained with various interfering substances. Such combination could be promising to destroy acid-resistant food industry contaminants as disinfectant or sanitizer and may have interesting applications in ophthalmology, dentistry, food industry, and hospital or veterinary medicine. However, the maintenance of such synergy using surface tests simulating practical conditions of application of products would be demonstrated. The efficacy of such combinations against other microbes as viruses will be investigated in a next future.

Acknowledgements

I wish to thank the excellent technical assistance of Mrs G. Jehannin, A. Rault and R. Fresnel and the aid of Mrs MH Moreau for pH measures. We also thank Mrs Garnier for her help with the English translation.

Ancillary