Journal of Medical Microbiology Papers in Press. Published July 7, 2014 as doi:10.1099/jmm.0.075796-0

1

Category: Clinical microbiology and virology

2

3

D-amino acids inhibit biofilm formation in Staphylococcus epidermidis

4

strains from ocular infections

5

Miriam L. Ramón-Peréza, Francisco Diaz-Cedillob, J. Antonio Ibarraa, Azael

6

Torales-Cardeñac, Sandra Rodríguez-Martínezc, Janet Jan-Robleroa, Mario E.

7

Cancino-Diazc, Juan C. Cancino Diaza*.

8

Departments of Microbiologya, Organic Chemestryb and Immunologyc, Escuela

9

Nacional de Ciencias Biológicas-Instituto Politécnico Nacional, Carpio y Plan de

10

Ayala S/N. Col. Santo Tomas. Deleg. Miguel Hidalgo. C.P. 11340 Mexico City,

11

Mexico.

12

13

*Corresponding Author:

14

Dr. Juan Carlos Cancino-Diaz

15

Carpio y Plan de Ayala S/N

16

Col. Santo Tomas.

17

Delegación Miguel Hidalgo, 11340

18

Mexico City, Mexico

19

Phone: 55-57296000 Ext: 46209

20

Email: [email protected]

21

22

Running title: D-amino acids inhibit S. epidermidis biofilm

23

Abbreviations: OI, ocular infection; HC, healthy conjunctiva; HS, healthy skin.

24

25

Abstract. Biofilm formation on medical and surgical devices is a major virulence

26

determinant for Staphylococcus epidermidis. S. epidermidis is able to produce

27

biofilms on biotic and abiotic surfaces and is the cause of ocular infections (OI).

28

Recent studies have shown that D-amino acids inhibit and disrupt biofilm formation

29

in the prototype strains Bacillus subtilis NCBI3610 and S. aureus SCO1. The effect

30

of D-amino acids on S. epidermidis biofilm formation has yet to be tested for

31

clinical or commensal isolates. S. epidermidis strains isolated from healthy skin (n=

32

3), conjunctiva (n=9), and OI (n=19) were treated with D-leucine (D-Leu), D-

33

tyrosine (D-Tyr), D-proline (D-Pro), D-phenylalanine (D-Phe), D-methionine (D-

34

Met), or D-alanine (D-Ala) and tested for biofilm formation. The presence of D-

35

amino acids during biofilm formation resulted in a variety of patterns: Some strains

36

were sensitive to all amino acids tested, while others were sensitive to one or

37

more, and one strain was resistant to all of them when added individually; in this

38

way D-Met inhibited most of the strains (26/31), followed by D-Phe (21/31).

39

Additionally, the use of D-Met inhibited biofilm formation on a contact lens. The use

40

of L-isomers caused no defect in biofilm formation in all strains tested. In contrast,

41

when biofilms were already formed D-Met, D-Phe, and D-Pro were able to disrupt

42

it. In summary, here we demonstrated the inhibitory effect of D-amino acids on

43

biofilm formation in S. epidermidis. Moreover, we showed, for the first time, that S.

44

epidermidis clinical strains have different sensitivity to these compounds during

45

biofilm formation.

46

Keywords: Staphylococcus epidermidis, D-amino acids, Biofilm.

47

48

Introduction As for many other compounds, two chiral amino acid variants exist in nature.

49

The L- version is used for protein synthesis by ribosomes in living organisms, while

50

the D- version is not involved in this pathway. Roles for D-amino acids have been

51

described in both eukaryotic and bacterial cells (Kim et al., 2010; Kleckner et al.,

52

1998; Wolosker et al., 2002, 2007). In bacteria, some peptides have D-amino acids

53

that could be incorporated through one of two mechanisms (Bodanszky et al.,

54

1969): 1) Post-translational conversion from L- to D- inside the peptide, or 2) by the

55

activity of non-ribosomal peptide synthetases (NRP), this being the most common

56

pathway in bacteria. D-amino acids have multiple roles in bacterial homeostasis;

57

they can be used as nutrients for growth (Chang et al., 1974; Conrad et al., 1974;

58

Pioli et al., 1976; Roesch et al., 2003), allow germination of bacterial spores, as in

59

Bacillus anthracis with D-alanine (D-Ala) (Hills et al., 1949; Halvorson &

60

Spiegelman 1952), and they could also be a component of the cell wall (Veiga et

61

al., 2006; Bellais et al., 2006). Moreover, D-amino acids can be secreted and have

62

an inhibitory effect in cell division. This effect is caused by the incorporation of

63

amino acids into the peptidoglycan layer causing destabilization of the whole

64

structure (Izaki et al., 1968; Hammes et al., 1978; Caparros et al., 1991, 1992).

65

Recently, it has been found that diverse bacterial phyla produce different types of

66

D-amino acids in the stationary growth phase and this correlates with the

67

expression of specific racemaces (Lam et al., 2009). Thus, it seems that bacteria

68

have evolved to produce and secrete different D-amino acids in stationary phase,

69

which, in turn, have multiple roles in bacterial homeostasis.

70

Recently, D-amino acids have been shown to disrupt biofilms (Kolodkin-Gal

71

et al., 2010). Biofilms are sessile microbial communities attached to a solid surface

72

and surrounded by an extracellular polysaccharide matrix (EPS). The EPS consists

73

mainly of polysaccharides, proteins, and extracellular DNA (Spormann et al., 2008;

74

Stewart & Franklin 2008). Biofilm formation is a three-step process: Bacterial

75

attachment to either an abiotic or biotic surface, intercellular adhesion or

76

aggregation, and disassembly. For S. epidermidis biofilm formation different

77

proteins have been described depending on the surface the biofilm is formed:

78

autolysin E or AtlE (Heilmann et al., 1997) and the biofilm-associated protein or

79

Bap (Tormo et al., 2005) for abiotic surfaces; while adherence to biotic surfaces is

80

led by SdrG (Arrecubieta et al., 2007) and SdrF (Hartford et al., 2001). Poly-N-

81

acetylglucosamine (PNAG; Mack et al., 1996), and two cell-wall proteins, the

82

accumulation-associated protein or Aap (Rohde et al., 2005) and the extracellular

83

matrix-binding protein or Embp (Christner et al., 2010) have been shown to be

84

involved in intercellular aggregation. The process of biofilm disassembly has yet to

85

be fully understood. Recently, Kolodkin-Gal et al. (2010), reported that

86

disassembly of B. subtilis biofilms is induced by a mixture of D-amino acids (D-Leu,

87

D-Met, D-Trp, and D-Tyr) that were produced by an aged culture (Kolodkin-Gal et

88

al., 2010). These amino acids are produced by their corresponding racemases and

89

accumulate in the mature biofilm. It has been proposed that biofilm inhibition by D-

90

amino acids occurs when protein fibers in the biofilm are disassembled by these

91

compounds (Kolodkin-Gal et al., 2010). A similar effect has also been observed in

92

S. aureus (Hochbaum et al., 2011) and P. aeruginosa (Kolodkin-Gal et al., 2010)

93

biofilms, suggesting this may be a universal property.

94

In Staphylococci, biofilm production is an important virulence factor.

95

Specifically, biofilms have been shown to have a key role in S. epidermidis

96

infections (Guo et al., 2007). S. epidermidis strains isolated from infectious

97

processes have been shown to produce biofilms when compared to commensal

98

isolates (Duggirala et al., 2007). Biofilms enable the constituent bacteria to resist

99

antibiotics and defenses by the host immune system. Moreover, attachment of

100

these bacteria to medical devices serves as the main source of infection for

101

humans (Uçkay et al., 2009). As previously mentioned, D-amino acids (D-Phe, D-

102

Pro, and D-Tyr) has been shown to inhibit biofilm formation in S. aureus SC01

103

(Hochbaum et al., 2011). To our knowledge, the effect of these compounds has yet

104

to be tested in S. epidermidis. Additionally, there is no data demonstrating the

105

effect of D-amino acids in biofilm formation in clinical strains. Although S. aureus

106

and S. epidermidis belong to the same genus, it has been shown that they have

107

different virulence mechanisms and that they have a high genomic diversity, which

108

in turn, might generate different responses to multiple stimuli (Conlan et al., 2012).

109

Therefore, the aim of this work was to study the effect of D-amino acids in biofilm

110

formation by S. epidermidis strains isolated from ocular infections. Our results

111

show differences in the sensitivity to the D-amino acids tested for biofilm formation.

112

Moreover, we propose that the use of a combination of D-amino acids can be used

113

on medical devices to prevent S. epidermidis infections.

114

115

116

Materials and methods

117

Strains

118

Strains used in this study have been described previously (Juárez-Verdayes

119

et al., 2012). Briefly, three collections of S. epidermidis strains with the ability to

120

form biofilms (Juárez-Verdayes et al., 2012) were isolated from healthy skin (HS;

121

n=3), healthy conjunctiva (HC; n=9), and from ocular infections (OI; n=19). All

122

strains were grown in tryptic-soy broth (TSB; Becton Dickinson, NJ, USA).

123

Biofilm determination

124

Semi-quantitative determination of biofilm formation was performed in 96-

125

well tissue culture plates (Nunc, Roskilde, Denmark) based on the method reported

126

by Christensen et al. (1985). A volume of 100 μl of 1:200 diluted overnight

127

bacterial culture was added to a 96-well plate and bacteria were grown at 37°C in

128

TSB medium. After 24 h of growth, the plates were washed vigorously with 1X

129

PBS, dried for 30 min at 55°C, and stained with 0.5 % (w/v) crystal violet solution.

130

After staining, the plates were washed with 1X PBS. The A490 nm of the adhered,

131

stained cells was measured using a Multiskan EX Microplate Photometer (Thermo

132

Scientific). According to the criterion outlined by Christensen et al. (1985), a strain

133

was considered as non-adherent and therefore biofilm-negative if the A490 nm was ≤

134

0.12; biofilm forming strains were those with an A490 nm > 0.12. According to

135

Christensen et al. (1985) an strain was considered as strongly biofilm-positive if the

136

A490 nm was higher than 0.240 and an strain was classified as weak biofilm-positive

137

if A490nm was greater than 0.12 but less than or equal to 0.240. Assays were

138

repeated six times, and the mean biofilm absorbance value was obtained.

139

Kinetics of biofilm formation

140

In order to determine the kinetics of biofilm formation, an overnight culture

141

was diluted 1:200 with fresh media and 100 μl were deposited in a 96-well plate.

142

The plate was incubated at 37°C without agitation and the biofilm formation was

143

determined every 6 h for a total of 54 h. In order to determine whether a secreted

144

component had inhibitory properties on biofilm formation, a similar experiment was

145

performed with the following modifications: After 24 h of incubation at 37°C, the

146

culture was discarded from each well without disturbing the biofilm, and washed

147

three times with 1X PBS. The media was replaced with fresh media and the

148

biofilm formation was determined every 6 h. Biofilm formation was measured as

149

described above.

150

Determination of D-amino acids in bacterial supernatants.

151

Bacteria were grown as mentioned before the kinetics experiments and

152

centrifuged at 2000 x g for 10 min. Supernatants were recovered, filtered through a

153

0.22 μm and processed as follows: Analytical HPLC was performed using a Varian

154

Pro Star 330 fitted with an ELSD detector, a solvent delivery system and an Astec

155

Chirobiotic T column (25 cm, 4.6 mm I.D., 5 μm particles). Column temperature

156

was 30°C, the mobile phase was water/methanol/formic acid (30:40:02); the flow

157

rate was 10 ml min-1 using a sample size of 10 μl for the injection. Stock solutions

158

of 1.0 mM D- and L-Ser, D and L-Thr, D- and L-Ala, D- and L-Tyr, D- and L-Val, D-

159

and L-Trp and D- and L-Leu (Sigma, Toluca, Mexico) were prepared and injected

160

to determine retention times. The sample from the supernatant was injected and D-

161

amino acids were detected based on correlation of retention times compared to the

162

standards previously described.

163

Biofilm formation in presence of D-amino acids

164

A volume of 100 μl of 1:200 diluted overnight bacterial culture was added to

165

a 96-well plate with TSB supplemented with either of the following D-amino acids:

166

15 to 100 mM D-Ala, 17 to 50 mM D-Leu, 6 to 50 μM D-Tyr, 3 to 10 μM D-Pro, 3 to

167

20 μM D-Phe or with 15 to 50 mM D-Met (Sigma, Toluca, Mexico). The plate was

168

incubated at 37ºC for 24 h and the biofilm formation was determined as previously

169

described. For the mature biofilm disassembly determination, the mature biofilm

170

was allowed to form for 24 h, washed and D-amino acids (different concentrations)

171

were added. The bacteria were incubated for 6 h and biofilm was measured.

172

Detection of inhibition and biofilm formation by confocal scanning laser

173

microscopy (CSLM)

174

In order to detect biofilm formation and the effect of D-methionine on the surface of

175

contact lenses, we used silicone-hydrogel contact lenses containing 42 % (v/v)

176

Filcon 1B and 58 % (v/v) water (Acuvue, Johnson & Johnson Vision Care). The

177

lenses were washed twice with cold sterile phosphate buffered saline (PBS) and

178

placed in a 24-well plate, one contact lens per well. Later, 2 ml of TSB or TSB

179

containing 50 mM D-methionine were added to each well; followed by the addition

180

of 100 μl of 1:200 diluted overnight bacterial culture. The plate was incubated at

181

37°C for 24 h. After that, the contact lenses were washed vigorously with 1X PBS,

182

and stained with Calcein AM (Invitrogen, Carlsbad, CA, USA) for 15 min. The

183

biofilm was observed using a Zeiss Axiovert 200M with an argon laser for

184

excitation at 488 nm (green fluorescence). Images were captured and processed

185

by using LSM5 Pascal software.

186

187

188

189

190

191

192

193

194

195

196

197

198

199

Results

200

Kinetics of the biofilm formation by S. epidermidis

201

In order to determine the time that it takes a mature biofilm to form and the

202

time it takes for its disruption, we analyzed the biofilm formation kinetics through a

203

window of 54 h. We analyzed 3 biofilm-forming S. epidermidis strains isolated from

204

HS, HC, and OI. As shown in Fig. 1(a), all three strains have a similar biofilm

205

formation and biofilm disruption time. Specifically, biofilm formation began after 12

206

h of incubation, reached a maximum after 24 h and mature biofilm disruption

207

occurred after 30 h. The OI strain formed a thicker biofilm, as evident by its ability

208

to retain more dye than the other two. This result shows that after a long incubation

209

S. epidermidis biofilms dissolve and suggests that a compound produced by the

210

bacteria is responsible for this phenotype.

211

To test our hypothesis, we repeated the experiment using the OI-derived

212

strain and removed the media after 24 h from each well, washed them, and

213

replaced it with fresh media every 6 hours. We observed that by replacing media

214

with fresh media the mature biofilm disruption was delayed for up to 42 h when

215

compared to a control in which the media was not replaced (Fig. 1b). We also

216

conducted a similar experiment that involved diluting the fresh media to 50% with a

217

48 h cell-free supernatant obtained in a parallel experiment. As seen in Fig. 1(c)

218

and 1(d), the presence of aged media (48 h cell-free supernatant), either diluted or

219

undiluted, disrupted the biofilm in two of the strains. These results suggest a

220

component present in the aged media is able to induce biofilm disruption in S.

221

epidermidis.

222

D-amino acids are present in bacterial biofilm supernatants

223

Given the results described above and recent evidence supporting the

224

disruptive effect of D-amino acids in biofilms, we aimed to determine the presence

225

of some of these compounds by analytical HPLC chromatography in 48 h cell-free

226

supernatant originated from S. epidermidis strains. We observed the presence of

227

D-Tyr, D-Leu, D-Trp, and D-Ala in the analyzed sample (data not shown). These

228

results demonstrate that S. epidermidis produces D-amino acids after a long

229

growth period. Therefore, it may be suggested that these compounds have a role

230

in the inhibition and disruption of S. epidermidis biofilms.

231

Effect of D-amino acids in S. epidermidis biofilm formation

232

To evaluate the inhibitory role of D-amino acids in biofilm formation, we

233

tested multiple S. epidermidis strains isolated from HS, HC (hereon called

234

commensals), and from OI (pathogenic). As shown in Fig. 2 (a-f), when 6 different

235

D-amino acids (D-Ala, D-Leu, D-Tyr, D-Pro, D-Phe and D-Met) were tested we

236

observed statistical significant (p0.12) effects in biofilm formation. Here we followed Christensen et al.

238

(1985) guidelines for classification of biofilm formers and determined that in some

239

strong biofilm former strains a treatment with D-amino acids reduced significantly

240

biofilm formation (A490nm >0.240 to A490nm

D-Amino acids inhibit biofilm formation in Staphylococcus epidermidis strains from ocular infections.

Biofilm formation on medical and surgical devices is a major virulence determinant for Staphylococcus epidermidis. The bacterium S. epidermidis is abl...
365KB Sizes 0 Downloads 4 Views