Colony aggregation and biofilm formation in xylem chemistry-based media for Xylella fastidiosa

Abstract

Two chemically defined media based on xylem fluid chemistry were developed for Xylella fastidiosa. These media were tested and compared to chemically defined media XDM₂, XDM₄ and XF-26. New media were evaluated for the Pierce's disease (PD) strain UCLA-PD. Our media either was similar to the concentration of some amino acids found in the xylem fluid of the PD-susceptible Vitis vinifera cv. Chardonnay (medium CHARD2) or incorporated the tripeptide glutathione found in xylem fluid composition (medium 3G10-R). CHARD2 and 3G10-R are among the simplest chemically defined media available. Xylem fluid chemistry-based media supported X. fastidiosa growth and especially stimulated aggregation and biofilm formation.

Xylem fluid chemistry, Glutathione, Chemically defined medium, Biofilm, Aggregation, Xylella fastidiosa

1 Introduction

Xylella fastidiosa is a Gram-negative xylem limited bacterium, which is pathogenic to a wide variety of plants. In grapevines X. fastidiosa, the causal agent of Pierce's disease (PD) [1], precludes the culture of Vitis vinifera in the southeastern USA and has greatly impacted the grape industry in California [2]. The wine institute (http://www.wineinstitute.org) estimates that total value added economic impact of grapes to the state of California is 33 billion dollars. PD is characterized by a drying or ‘scorching’ of leaves, which become chlorotic along the margins before drying [3,4]. Leafhoppers, including the glassy-winged sharpshooter (Homalodisca coagulata), are vectors of PD [4]. The genome of X. fastidiosa from citrus [5] and grapevines [6] has recently been sequenced, allowing an overview of the metabolic functions of these strains and a comparative analysis of genes that may be involved in pathogenicity.

Periwinkle wilt (PW⁺) medium [7] often supports vigorous growth of X. fastidiosa; however, the presence of numerous undefined constituents renders this medium less amenable for nutritional requirement studies. The XF-26 medium is chemically defined and composed of three inorganic salts, two tricarboxylic acids, 17 amino acids, phenol red and potato starch [8]. Recently, genomics-based defined media (XDM series) with fewer components also supported the growth of X. fastidiosa[9].

Simple chemically defined media are required to address the essentiality of compounds and the nutritional basis for growth of X. fastidiosa. Also, experimental studies concerning gene induction and gene expression require simplified chemical conditions. Nutritional information may help to reveal aspects of growth (colony number, protein content, aggregation and biofilm formation) as well as how bacteria interact with xylem wall (adhesion and biofilm formation in planta). This study is also important because it help us understand how X. fastidiosa can survive in xylem fluid (a nutritionally poor environment) and form biofilm [10].

Three media were used as a starting point: (1) NFbHPN, a nitrogen-fixing bacteria medium [11], (2) PW⁺, an undefined medium [7], and (3) XF-26, a defined medium [8]. We also incorporated components present in the xylem fluid chemistry of V. vinifera cv. Chardonnay that we suspected could affect the growth of X. fastidiosa. The capacity of X. fastidiosa to grow, form biofilm and alter cell aggregation was evaluated in these different media.

2 Materials and methods

2.1 Bacterial strain and growth parameters

The strain of X. fastidiosa utilized in the work was UCLA-PD (16S gene sequence at GenBank accession number AF073231), isolated from infected grape plants in California (supplied by Dr. Alexander H. Purcell, University of California at Berkeley). The media were inoculated with X. fastidiosa cultures incubated at 28°C and 150 rpm, using an orbital shaker (New Brunswick Scientific, Edison, NJ, USA). A cell suspension in PW⁺ medium (OD_600nm=0.6) was utilized as inoculum for 30 ml of medium in a 50 ml conical tube (60% of the flask capacity) (Falcon, Becton&Dickson Labware). Cell pellets were resuspended in each medium evaluated before constituting the inoculum. Each medium was inoculated (1% v/v) with a cell suspension, giving the final OD_600nm=0.02.

2.2 Colony size assay

A drop (50 μl) from each bacterial culture in PW⁺, XF-26, CHARD2 and 3G10-R (OD_600nm=0.15 or 10⁵ colony forming units CFU ml⁻¹) was placed onto PW⁺ medium without dilution. To avoid the destruction of X. fastidiosa aggregates, drops were allowed to naturally settle down on the medium surface, without being disturbed. Colonies were maintained for 10 days at 28°C. Plating was performed in triplicate for each medium. Digital images of the colonies were taken with a digital camera (Sony). For a pictorial analysis of colony size of X. fastidiosa cultivated in different chemically defined media, we utilized the program ImageJ version 1.7 from National Institutes of Health, USA (http://rsb.info.nih.gov/ij/). Colonies were categorized by diameter as small (<0.1 mm), medium (0.1–0.5 mm), and large (0.5–1.0 mm).

2.3 Growth curve

Four samples of each medium (30 ml) were collected 0, 96, 192 and 336 h after inoculation. Bacterial cells were pelleted after centrifugation at 2600×g for 15 min. The treatment of bacterial pellets and the measurement of cell turbidity at absorbance of 600 nm were assayed according to Lemos and collaborators [9]. The remaining cells were used for protein determination [12]. Biofilm formation and percentage of aggregation were also evaluated at each time after inoculation. The purity of X. fastidiosa during the experiment was evaluated by polymerase chain reaction (PCR) using specific primers [13], and by light microscopy using Gram staining [14].

2.4 Protein content

The remaining cell sample from each medium was transferred to an 1.5 ml Eppendorf tube. Each sample was centrifuged 7.4×g for 5 min and the pellet was resuspended in 100 μl of 0.1 M NaOH for 30 min for cell lysis. The protein content was measured in 1 ml of Bradford reagent (Sigma Scientific Co) after addition to the lysate [12]. The absorbance was recorded at 595 nm. Protein concentration was determined by comparison to a bovine serum albumin (BSA) standard curve.

2.5 Biofilm formation

Biofilm formation on the surface walls of conical polypropylene tubes (Falcon, Becton&Dickison Labware) was assayed by the crystal violet method [15]. After the incubation period, the tubes were rinsed with deionized water, and 1% (wt/vol) solution of crystal violet was added to each tube. After 15 min at room temperature, crystal violet was eluted from the biofilm by adding ethanol to each tube. The absorbance of the solution was measured at 600 nm.

2.6 Aggregation assay

This assay was based on the fact that the optical density (540 nm) is correlated with bacterial cell aggregation state [16]. The initial cell turbidity was measured at 540 nm (OD_i). Then, cells were dispersed in a tissue homogenizer for 1 min, and the total optical density (OD_t) was determined. Percentage of aggregation was estimated as follows: % aggregation=(OD_t−OD_i)×100/OD_t[16].

2.7 Defined media formulations

The base of all defined media is KH₂PO₄ (1 g l⁻¹), K₂HPO₄ (1.5 g l⁻¹), MgSO₄ (0.2 g l⁻¹), d-biotin (1×10⁻⁴ g l⁻¹), phenol red (0.02 g l⁻¹), ferric pyrophosphate (0.25 g l⁻¹) and l-glutamine (4 g l⁻¹), 3G10-R contains the reduced form of glutathione (20 μM) and glucose (2 g l⁻¹); CHARD2 has l-alanine (0.17 g l⁻¹), l-aspartic acid (0.58 g l⁻¹), l-glutamine (1.8 g l⁻¹), l-arginine (1.05 g l⁻¹), l-cysteine (0.01 g l⁻¹). The criteria to select the amino acids to be incorporated in CHARD2 medium were those occurring in high concentration found in V. vinifera cv. Chardonnay xylem fluid and those able to support growth as determined by Chang and Donaldson [17]. The media were prepared in 1 l flasks and autoclaved for 20 min prior to the addition of temperature-sensitive components. l-glutamine (4 g) was dissolved in 50 ml of deionized water and ferric pyrophosphate (0.25 g) was dissolved in 60 ml of deionized water. For 3G10-R, glutathione (62 mg) was dissolved in 10 ml of deionized water. 1 ml of glutathione solution was mixed with the glutamine solution prior to filtration. The volumes of filtered solution were subtracted from the medium volume before autoclaving. Note that l-glutamine in CHARD2 medium is added twice, once before (1.8 g l⁻¹) and once after (4 g l⁻¹) autoclaving. The pH was adjusted to the range between 6.6 and 6.7.

The XDM₂ and XDM₄ were prepared according to Lemos and collaborators [9]. The modification of XDM₂ by decreasing the concentration of phenol red from 0.1 to 0.02%, generated the XDM₂* used in this work.

2.8 Statistics

Data were analyzed using SAS Institute System V 8. Least significant differences were determined for bacterial growth, protein concentration, biofilm formation and percentage of aggregation.

3 Results and discussion

X. fastidiosa is confined to a poor nutritive environment [18,19] and the genome of X. fastidiosa appears to contain the necessary biosynthetic pathways for growth on a chemically restricted diet [5]. Given the limited diversity of compounds in xylem fluid, we hypothesized that X. fastidiosa could grow on a minimum diet that contains a much smaller number of compounds. Our second hypothesis was that alterations in simple chemically defined media could change the characteristics of X. fastidiosa, as quantified by bacterial growth, aggregation, biofilm formation and colony size.

The two new media selected had six ingredients in common (Table 1). Major differences were noted between the XDM series (carbohydrate-based) and the amino acid-based media used in this work (CHARD2 and 3G10-R). l-glutamine, which is highest in concentration, is believed to be an indispensable medium component for X. fastidiosa[7,8]. Recently, new solid defined media were developed with l-glutamine and organic acids as main components (R.P.P. Almeida, R. Mann and A.H. Purcell, personal communication). It is not surprising that l-glutamine is the most abundant amino acid in Vitis xylem fluid [20,,22] (Table 1). By contrast, the genomic-based media (XDM series) exhibits 500-fold less l-glutamine than XF-26 and 1000-fold less l-glutamine than 3G10-R, CHARD2 and PW⁺ (Table 1). 10 g of glucose per liter was added to the XDM media. Glucose was not in the formulation of any other media, except 3G10-R, which had 5-fold less glucose (2 g l⁻¹) than the XDM media (Table 1). Since the organic constituents in xylem fluid are mainly amino acids and organic acids [18,,,,22], our formulation more closely approximates the composition in planta it appears to be CHARD2. Glucose occurs in low concentration (usually <50 μM) in xylem fluid of grapevine [21,22]. Thus, CHARD2 may be a particularly good chemically-defined diet for the assessment of nutritional requirements, metabolic pathways, and gene induction. Glucose was used in 3G10-R formulation, but did not appear to be essential since CHARD2 does not contain glucose (Table 1).

1

The composition of the media utilized in this work (PW⁺, NFbHPN, XF-26, XDM₂, XDM₄, XDM₂*, CHARD2 and 3G10-R) and their contribution to the development of the new media (CHARD2, 3G10-R) and the modification of XDM₂

	Media
	Undefined	Defined
	PW+[7]	NFbHPN [11]	XF-26 [8]	XDM2[9]	XDM4[9]	XDM2* (this work)	CHARD2 (this work)	3G10-R (this work)
Buffer system	K2HPO4, KH2PO4, (NH4)2HPO4	KH2PO4, K2HPO4	KH2PO4, (NH4)2HPO4	KH2PO4, K2HPO4	KH2PO4, K2HPO4	KH2PO4, K2HPO4	KH2PO4, K2HPO4	KH2PO4, K2HPO4
Inorganic salts	MgSO4·7H2O	MgSO4·7H2O	MgSO4·7H2O	MgSO4·7H2O	MgSO4·7H2O	MgSO4·7H2O	MgSO4·7H2O	MgSO4·7H2O
		NaCl, CaCl2
Iron source	hemin chloride			ferric pyrophosphate	ferric pyrophosphate	ferric pyrophosphate	ferric pyrophosphate	ferric pyrophosphate
Organic acids		sodium lactate	Tris sodium citrate, disodium succinate	glucose	glucose	glucose		glucose
Chelator agent		nitrilo triacetic acid
Micronutrients		FeSO4·7H2, Na2MoO4·2H2O, MnSO4·H2O, H3BO3, CuSO4·5H2O, ZnSO4·7H2O
Vitamin		biotin		biotin, thiamine, pyridoxine, nicotinic acid, B12, myo-inositol	biotin, thiamine, pyridoxine, nicotinic acid, B12, myo-inositol	biotin, thiamine, pyridoxine, nicotinic acid, B12, myo-inositol	biotin	biotin
Fungal growth inhibitor	cycloheximide
Amino acids	Gln, His		Ala, Arg, Asp, Cys, Gly, Gln, His, Iso, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Val	Ser, Met, Gln, Asp	Gln	Ser, Met, Gln, Asp	Ala, Arg, Asp, Cys, Gln	Gln
Amino acids source (undefined)	soytone, tryptone
Detoxifier	BSA, starch		starch				starch only for solid	starch only for solid
Antioxidant								glutathione (reduced)
pH indicator	phenol red 0.02%			phenol red 0.1%		phenol red 0.02%	phenol red 0.02%	phenol red 0.02%

d-biotin was the only vitamin in the formulation of 3G10-R and CHARD2, contrasting with a mixture of six vitamins used in the XDM media [9]. Other undefined media which contained the vitamin-rich yeast extract have also been proposed [23]. Another major difference in media was the concentration of phenol red (PR), which was 5-fold higher in XDM media compared to all other media (Table 1). Glutathione is a tripeptide composed of glutamate, l-cysteine and l-glycine. This antioxidant was added to the formulation of 3G10-R (Table 1), since glutathione is a component of xylem fluid [24,25]. Other reducing agents such as l-cysteine and l-methionine may also be found in the xylem fluid [18,19]. l-cysteine is contained in medium CHARD2. The antioxidant concentration is 7.5-fold higher in 3G10-R (glutathione) and approximately 32-fold higher in CHARD2 (l-cysteine), compared to XDM₂ (l-methionine). Our results support the contention that glutathione is acting as an antioxidant rather than a nutrient supplement since we observed better growth in media with reduced compared to the oxidized form of glutathione (data not shown). Glutathione and l-cysteine, serving as reducing agents, may be important to the growth of X. fastidiosa in planta and, as a consequence, was added to the nutrient media.

The capacity of 3G10-R and CHARD2 to alter colony size was tested for the UCLA-PD strain. After 14 days of incubation, 3G10-R and CHARD2 clearly promoted cell aggregation to form bigger colonies (Fig. 1). The same phenomenon was verified with medium XF-26 but with less intensity. By contrast, PW⁺ incubated cells exhibited a pattern of a homogeneous distribution of small colonies. The results suggest that the medium with the highest nutritional value (PW⁺) apparently stimulated X. fastidiosa cells to be planktonic. More medium-sized colonies were found in 3G10-R and CHARD2 than in PW⁺ (Fig. 2). Large colonies were also monitored. However, the differences were as not clear as observed for small- and medium-sized colonies (data not shown).

1

Colony distribution of X. fastidiosa after 10 days of incubation on PW⁺, XF-26, CHARD2 and 3G10-R media. Plating was performed in triplicate for each medium. Colonies were maintained at 28°C. Digital images of the colonies were taken with a digital camera (Sony).

2

Colony size distribution of X. fastidiosa after 10 days of incubation on PW⁺, XF-26, CHARD2 and 3G10-R media. Colonies were maintained for 10 days at 28°C.

Bacterial growth and protein were monitored in the media: XF-26, XDM₂, XDM₄, XDM₂*, CHARD2 and 3G10-R (Fig. 3). XDM₂* had the highest bacterial growth (as measured by OD_600nm and protein concentration) and CHARD2 the lowest (Fig. 3A and B). Better growth was observed for XDM₂* than for XDM₂ (Fig. 3A and B). The only difference between the two media is the concentration of PR (XDM₂ (PR=0.1%) and XDM₂* (PR=0.02%)). The advised concentration of PR as pH indicator is 0.02%[26].

3

Growth curves of X. fastidiosa in chemically defined media over a period of 336 h at 28°C. A: OD_600nm. B: Protein concentration. C: Biofilm formation. D: Percentage of aggregation. Bars above each data point correspond to least significant difference 0.05. Significance was as follows: Cells in suspension (96 h, P<0.0001; 192 h, P<0.0002; 336 h, P<0.0001); protein concentration (96 h, P<0.0001; 192 h, P<0.1678; 336 h, P<0.0001); biofilm formation (96 h, P<0.0006; 192 h, P<0.363; 336 h, P<0.0001); % aggregation (96 h, P<0.0266: 192 h, P<0.0001; 336 h, P<0.0118).

The assumption that genomics-based media (XDM series) and xylem fluid chemistry-based media (3G10-R and CHARD2) play a decisive role in determining the behavior of X. fastidiosa cells in vitro was reinforced during the comparisons of percentage of aggregation and biofilm formation (Fig. 3C and D). CHARD2 and 3G10-R promoted more aggregation and biofilm formation compared to the other media. The actual media components responsible for the tendency of X. fastidiosa form biofilm and aggregates in CHARD2 and 3G10-R media may relate to the fact that these are chemically-defined media and do not contain high concentration of sugars or organic acids in their formulations. Another difference in our media is the presence of reducing agents (Fig. 3C and D and Table 1).

The redox environment of media can play a critical role in determining the tendency for aggregation and biofilm formation [27,28]. The survival of planktonic cells is likely to be affected by changes in redox conditions; ostensibly, aggregated cells would be more resistant to oxidative environment than free planktonic cells. For example, aggregation patterns of the bacterium Marichromatium gracile were interpreted as effective protection strategies against high oxygen concentrations and represented the first stages of biofilm formation [28]. Reduced surface to volume ratio and limited diffusion of oxygen into the interior of aggregates are proposed as a protective mechanism for A streptococci [27]. Thus, aggregation in X. fastidiosa may represent an alternative colony organization to cope with oxidative stress levels that could be a threat to cell survival.

A hypothetical model proposed by Leite and collaborators [29] to explain how X. fastidiosa cells form aggregates and how they adhere to the xylem vessel may help explain our results. In summary, the thiol groups of the X. fastidiosa cell membrane could stimulate an interaction with divalent cations, which results in the attraction of cells to hydrophobic surfaces. Free cells could form aggregates of X. fastidiosa based on surface charges, followed by the production of gum [30], culminating in a stable community of cells (the biofilm). The enhanced biofilm formation observed with CHARD2 and 3G10-R may be due to the mechanism cited above, i.e., the presence of reducing agents, glutathione in 3G10-R and l-cysteine in CHARD2. Hightower [31], working with membrane SH radicals, demonstrated that a concentration of 5 mM or less was enough to maintain critical external SH groups in the reduced state. Our best results were obtained with media in the concentration of 20 μM of glutathione reduced form (GSH), as found 3G10-R, when compared to similar media formulations, but containing 20 mM of GSH (data not shown). The importance of a reducing environment on the surface of a bacterial Gram-negative pathogen is pointed out as being important in the maintenance of adhesins [32]. Methionine sulfoxide reductase (MsrA) from Erwinia chrysanthemi has been considered to repair oxidative stress damage generated by host defense mechanisms [33].

In conclusion, we established that simplified media rich in amino acids (xylem fluid-based media), rather than carbohydrates (genomics-based media), may stimulate aggregation and biofilm formation of X. fastidiosa. Since the chemistry of xylem fluid is likely to be involved in bacterial growth, aggregation and biofilm formation, we hypothesized that the carbohydrate-rich media preferentially maintained cells in a planktonic state. It is possible that our media, CHARD2 and 3G10-R, could offer conditions to simulate the first stages of the interaction between X. fastidiosa and the cell walls of xylem vessels. The magnitude of these initial responses may ultimately determine disease progression. Since media composition had a profound effect on the behavior of X. fastidiosa, it is possible that the chemical composition of xylem fluid may be critical in forming aggregates and biofilm in planta. We feel that this study may provide the groundwork for future research concerning surface characteristics of X. fastidiosa and pathogenicity.

Acknowledgments

We are grateful to all valuable suggestions of Brent Brodbeck, James Marois, Hidevaldo B. Machado and Fabio O. Pedrosa. We also would like to recognize the support provided by the American Vineyard Foundation and the California Department of Food and Agriculture. This is the University of Florida Agricultural Experiment Station Journal No. R-09403.