OUP user menu

Influence of xylem fluid chemistry on planktonic growth, biofilm formation and aggregation of Xylella fastidiosa

Peter C. Andersen, Brent V. Brodbeck, Steve Oden, Anthony Shriner, Breno Leite
DOI: http://dx.doi.org/10.1111/j.1574-6968.2007.00827.x 210-217 First published online: 1 September 2007


Xylella fastidiosa is the causal agent of Pierce's disease in grapevines. The mechanisms of pathogenicity are largely due to occlusion of xylem vessels by aggregation of X. fastidiosa and biofilm formation. Xylella fastidiosa was subjected to xylem fluids with varying chemistries to examine the effects of nutritional components on bacterial growth in vitro. The exposure of X. fastidiosa to xylem fluids collected from different Vitis genotypes resulted in highly significant differences in both planktonic growth and biofilm formation. Planktonic growth of X. fastidiosa in Vitis xylem fluid was correlated to the concentration of citric acid, amino acids (glutamic acid, glutamine, histidine, valine, methionine, isoleucine and phenylalanine) and inorganic ions (copper, magnesium, phosphorus and zinc). Biofilm formation was correlated to many amino acids at 1 h of incubation. Xylem fluid from Vitis rotundifolia cv. Noble (fluid that supported low planktonic growth) was supplemented with the compounds that were correlated above to levels found in Vitis champinii cv. Ramsey (fluid that supported high planktonic growth) to determine the direct impact of xylem constituents on the growth characteristics of X. fastidiosa. Augmentation of fluid from Noble with the amino acids listed above, citric acid, calcium and magnesium resulted in increased planktonic growth and aggregation.

  • aggregation
  • amino acids
  • biofilm
  • Pierce's disease
  • Xylella fastidiosa
  • xylem fluid


Pierce's disease (PD) of grapevine is caused by the xylem-limited bacterium, Xylella fastidiosa (Wells et al., 1987). PD precludes a grape industry in the southeastern United States based on Vitis vinifera L., and recently PD has had a negative impact on grapes grown in southern California with the introduction of a new leafhopper vector, Homalodisca vitripennis (Germar) (Redak et al., 2004; Takiya et al., 2006). Other diseases caused by X. fastidiosa strains include phony peach disease, citrus variegated chlorosis and numerous scorch diseases of crop, forestry and landscape species. Xylella fastidiosa exists in a benign presence in many plants that are endemic to its native range (Hopkins & Purcell, 2002).

Cell multiplication and the formation of biofilm and aggregates are early processes that precede visual PD symptoms. PD symptomatology is mainly due to water stress as a result of occlusion of xylem vessels by X. fastidiosa and extracellular polysaccharides. In addition, xylem vessels may be plugged with pectins, tyloses and gums as a host plant response to invasion by X. fastidiosa (Fry & Milholland, 1990; Purcell & Hopkins, 1996; Hopkins & Purcell, 2002).

A functional relationship has been reported between xylem chemistry and X. fastidiosa planktonic growth, aggregation and biofilm formation within Vitis germplasms (Leite et al., 2004b). The PD strains of X. fastidiosa spread and multiply faster in xylem vessels of PD-susceptible than PD-resistant grapevine species (Hopkins, 1984; Fry & Milholland, 1990), although the underlying mechanism(s) of PD resistance is not known. For many diseases caused by X. fastidiosa, there is no cure. The initial aspects of pathogenesis associated with colonization and movement of X. fastidiosa have not been determined for PD or any other diseases caused by X. fastidiosa. The stimuli for cell aggregation and biofilm formation may involve specific plant/bacterium interactions and the nutrient status of xylem fluid. Cell aggregation was promoted by the addition of calcium chloride (CaCl2) to xylem fluid in vitro (Andersen et al., 2004b; Leite et al., 2004b). Calcium and magnesium have been implicated in the adhesion and aggregation of X. fastidiosa in planta (Leite et al., 2002). Xylem fluid from PD-resistant Vitis rotundifolia maintained X. fastidiosa in a planktonic state, whereas X. fastidiosa was more likely to form aggregates when incubated in PD-susceptible V. vinifera cultivars (Leite et al., 2004b).

Xylem fluid is extremely dilute (typically 5–20 mM) and consists mainly of amino acids, organic acids and inorganic ions (Andersen & Brodbeck, 1989a, b, 1991; Andersen et al., 1995). However, the nitrogen to carbon ratio of xylem fluid is higher than most proteins (Andersen et al., 1995). Xylem fluid chemistry is dynamic and may vary with plant genotype, time of year, time of day, soil fertilization status and temperature (Andersen & Brodbeck, 1989a, b, 1991; Andersen et al., 1995). The amide glutamine is the major nitrogen source in xylem fluid of Vitis spp. and is also the predominant nitrogen source in most chemical growth media for X. fastidiosa. It is difficult to collect sufficient quantities of xylem fluid for chemical analyses and for use as growth media, except before or during budbreak in early spring when vines bleed copious quantities of xylem fluid from cut spurs.

It has been shown that both X. fastidiosa proliferation and biofilm formation may be impacted by a variety of constituents including inorganic ions, O2, antioxidants, amino and organic acids and sulfhydryl groups (Leite et al., 2004b). Newly developed chemically defined media result in variable patterns of X. fastidiosa planktonic growth and biofilm formation. Periwinkle wilt (PW+) media (Davis et al., 1981) are a nutrient-rich growth media that result in rapid planktonic growth with comparatively little biofilm formation, whereas the newly defined media CHARD2 provide slower planktonic growth, but high biofilm formation (Leite et al., 2004a).

The objective of this study was to quantify the relationships between naturally occurring xylem constituents (inorganic ions, amino acids, organic acids and inorganic ions) and X. fastidiosa planktonic growth, biofilm formation and aggregation utilizing xylem fluid from different Vitis genotypes. Further delineation of the role of individual compounds on parameters of X. fastidiosa growth was accomplished by augmentation of xylem fluid with chemical constituents found in xylem fluid.

Materials and methods

Xylem fluid from V. rotundifolia cvs. Carlos and Noble, Vitis rupestris cv. St George, Vitis simpsoni cv. Pixiala, Vitis champinii cvs. Dogridge and Ramsey and V. vinifera cvs. Chardonnay, Chenin blanc and Exotic was collected from cut bleeding spurs during March 2005 (Andersen & Brodbeck, 1991) at the North Florida Research and Education Center Quincy Florida. Xylem fluid was also collected from V. vinifera cv. Chardonnay and V. rotundifolia cv. Noble at the University of California Davis, California, during March and April 2006.

The concentration of amino acids and organic acids in xylem fluid of nine genotypes was determined by HPLC. In brief, samples were centrifuged at 1200 g through a 10 000 MW filter. For amino acids, samples were dried and then derivitized with 2 : 2 : 1 ethanol : triethanolamine (TEA) : H2O (Heinricksen & Meredith, 1984). 7 : 1 : 1 : 1 ethanol : TEA : H2O : phenylisothiocyanate was added for a 20-min reaction time under a N2 atmosphere. The eluent for chromatography was 5 mM sodium phosphate buffer plus 6% acetonitrile. HPLC quantification was accomplished using a Pico Tag column and a UV detector (Waters Division, Millipore Corp.). Organic acid analysis was performed using a polymeric cation exchange column (Ion-300 Interaction Corp., San Jose, CA) using a 0.01 M H2SO4 buffer. Quantification was via Waters HPLC equipped with a UV detector. Phosphorus, potassium, calcium, magnesium, manganese, copper, zinc and boron in xylem fluid were quantified using an Inductively Coupled Argon Plasma Spectrophotometer. Iron and sodium were analyzed with an Atomic Absorption Spectrophotometer. pH was determined using an Orion 280a pH meter.

The Temecula strain of X. fastidiosa was cultivated in PW+ medium until the mid log phase of growth (OD600 nm=0.3). One milliliter aliquots were dispensed into 15 mL polypropylene tubes and each tube was centrifuged at 2380 g to pellet the bacterial populations. The supernatant was discarded and the pellets were washed in a phosphate-buffered saline solution (0.1 M potassium phosphate buffer and 0.8% saline solution at pH 6.8). This solution was centrifuged at 2380 g and the supernatant was discarded and the pellet was washed in the appropriate experimental medium (xylem fluid). This was once again centrifuged at 2380 g and the bacteria were resuspended in xylem fluid. The bacteria were incubated at 28°C in xylem fluid from the varying cultivars for 1 h, 5 and 12 days while rotating at 150 r.p.m. on a mechanical shaker (New Brunswick Scientific Co., model-25R). OD was measured using a Genosys 8 spectrophotometer (Spectronic Unicam Corp., Rochester, NY) at a wavelength of 600 nm. The formation of biofilm on the surface of polypropylene tubes was assayed by staining with crystal violet and destaining using 70% isopropyl alcohol (Espinosa-Urgel et al., 2000). The amount of biofilm was quantified at a wavelength of 600 nm. Measurements of planktonic growth and biofilm were replicated three times.

The concentrations of individual amino acids, organic acids and inorganic ions were established for the xylem fluid from each genotype. Tests were run in March 2006 and again in June 2006 to ensure repeatability. Linear regression was used to determine the relationship between planktonic growth (OD600 nm) and biofilm to xylem constituents using the mean values for each genotype (SAS Institute, 2003).

Xylem chemistry was manipulated by the addition of chemical constituents to xylem fluid that promoted low rates of growth (Noble) to that which promoted the most growth (Ramsey). The treatments were: (1) amino acids (glutamine, histidine, valine, methionine, isoleucine and phenylalanine), (2) citric acid, (3) calcium chloride and magnesium sulfate, (4) a combination of (1–3); and (5) unaltered Noble xylem fluid. The concentrations of each treatment were the amount of each compound required to increase the levels from that found in Noble to that found in Ramsey. Methodology to quantify planktonic growth and biofilm formation was as before; however, in this experiment % aggregation was also quantified. Percentage aggregation was measured using an improved Neubauer apparatus (Clay Adams, New York) according to a bioassay designed for X. fastidiosa (Leite et al., 2004b). Approximately 20 µL of Xf in suspension at the bottom of a 1.5 mL Eppendorf vial was pipetted onto a Neubauer chamber. The Neubauer chamber was placed under a Nikon 80i light microscope and the cell suspension was photographed. Percentage aggregation (the fraction of the field occupied by X. fastidiosa aggregates) was quantified using National Institute of Health, image j software. Each treatment was replicated four times.

Results and discussion

Xylella fastidiosa survived in xylem fluid from all genotypes, and growth occurred in PW+ media at the end of the experiment, indicating that X. fastidiosa cells were alive. In addition, biofilm formation increased over time for all genotypes. It is not possible to draw definitive conclusions about the resistance of individual Vitis germplasms based on xylem fluid collected during late winter. For example, in this study Ramsey (tolerant but shows symptoms of PD, A. Walker, pers. commun.) manifested the highest bacterial growth but was low in biofilm formation. The only way to collect sufficient quantities of xylem fluid for this experiment was to collect bleeding xylem fluid from cut Vitis spurs, which is only available in late winter/early spring. Xylem fluid from dormant vines may or may not be representative of xylem fluid from specific genotypes in summer when X. fastidiosa is more actively multiplying (Andersen et al., 2004a). Thus, this methodology was developed to provide a range of xylem profiles to assess the effects of individual xylem components on X. fastidiosa growth rather than confirming the relative resistance of Vitis genotypes.

There was no significant difference in planktonic (OD600 nm) growth 1 h after the initiation of the experiment (Table 1). Planktonic growth was significantly (two- to four-fold) greater for X. fastidiosa in Ramsey xylem fluid than in the other genotypes at 5 and 12 days. Rapid and significant effects on biofilm also occurred 1 h, 5 and 12 days after the initiation of the experiment. Noble had the most and Ramsey had the least biofilm and the amount of biofilm varied as much as five-fold between genotypes. Planktonic growth and biofilm formation were often inversely related. For example, planktonic growth for X. fastidiosa incubated in Ramsey fluid was consistently higher than in other fluids at each time period, but was also consistently lower in biofilm formation.

View this table:
Table 1

Planktonic growth and biofilm formation (OD600 nm) of Xylella fastidiosa Temecula in xylem fluid from nine different cultivars of Vitis for periods of 1 h, 5 days and 12 days

1 h5 days12 days
Chenin Blanc0.1230.060bc0.128b0.474b0.140b0.636bcd
St George0.1290.057bc0.140b0.549b0.166b0.794bc
Statistics (P<)NS0.00010.00010.00010.00010.0001
  • Numbers within columns with different letters are significantly different by Duncan's multiple-range test (P<0.05).

  • Data represent the mean of two experiments.

Analyses of xylem fluids showed that many xylem constituents were highly correlated to planktonic growth and biofilm formation. The relationships between planktonic growth and biofilm formation to chemical constituents in xylem fluid were consistent in the two separate experiments (pooled results are presented in Tables 2 and 3). Xylella fastidiosa planktonic growth became more highly correlated with these constituents over time, with only two significant correlations at 1 h (Table 2). Glutamine, the predominant amino acid in Vitis xylem fluid, was weakly, but consistently, correlated with bacterial growth after 5 and 12 days. Some of the minor amino acids (histidine, valine, methionine, isoleucine and phenylalanine) were much more strongly correlated to planktonic growth. Citric acid was also very highly correlated to planktonic growth after 5 and 12 days. Phosphorus, copper and zinc were inorganic ions well correlated with planktonic growth. For all of the constituents mentioned above, the results appeared to be consistent over time as significant relationships apparent after 5 days also persisted through 12 days. The importance of calcium, magnesium, phosphorus and citric acid to X. fastidiosa growth has been hypothesized previously (Leite et al., 2002, 2004b; Andersen et al., 2004b). In the current study, the equation [(citric acid × P)/(Ca+Mg)] yielded higher correlations by regression analyses than did any single chemical constituent (P<0.0001; R2=0.90). The strength of these relationships suggests that the original hypothesis merits further investigation.

View this table:
Table 2

Relationship between planktonic growth (OD600 nm) of Xylella fastidiosa Temecula strain and constituents in xylem fluid (amino acids and organic acids, µM; P, Mg and Zn, mg L−1; Cu, µg L−1) from nine different cultivars of Vitis for periods of 1 h, 5 days and 12 days

1 h5 days12 days
Amino acids
Organic acids
Inorganic ions
    (Citric acid × P)/(Ca+Mg)NSy=0.101+0.00142x0.00010.90y=0.0927+0.0021x0.00010.90
  • Data from two experiments.

View this table:
Table 3

Relationship between biofilm formation (OD600 nm) of Xylella fastidiosa Temecula strain and constituents in xylem fluid (amino acids and tartaric acid, µM; Cu, µg L−1) from nine different cultivars of Vitis for periods of 1 h, 5 and 12 days

1 h5 days12 days
Amino acids
Organic acids
Inorganic ions
  • Data represent the means of two experiments.

Biofilm formation was also correlated to xylem constituents (Table 3). Both the specific compounds correlated to biofilm formation and the timing of possible effects varied drastically from those found with planktonic growth. Six amino acids (glycine, alanine, threonine, arginine, leucine and lysine) and one organic acid (tartaric acid) were related to biofilm formation, but only at 1 h. None of these relationships persisted for 5 and 12 days. The only correlations between xylem constituents and biofilm formation that persisted were for copper (5 days) and xylem pH (5 and 12 days).

These consistent correlations between planktonic growth and xylem constituents are important. The chemical profiles of amino acids, organic acids and inorganic ions are presented for Noble (low values of planktonic growth but the highest biofilm producer) and Ramsey (high planktonic growth but a low biofilm producer) (Table 4). Nineteen amino acids were detected in xylem fluid. The concentrations of organic constituents were c. two-fold higher in Ramsey compared with Noble (Table 4). All Vitis genotypes had a very unbalanced chemical profile, with glutamine accounting for 80% or more of the total amino acids (only data for Noble and Ramsey are presented). Only three organic acids were detected (citric, tartaric and malic acids). The concentrations of phosphorus, potassium, calcium, magnesium and manganese were generally higher in Ramsey than Noble; iron, boron and sodium were below detectable concentrations in Noble, and iron was below detectable concentrations in Ramsey.

View this table:
Table 4

Concentration of amino acids, organic acids and inorganic ions in xylem fluid of Vitis rotundifolia cv. Noble and Vitis champinii cv. Ramsey

Amino acid (µM)
Organic acids (µM)
Inorganic ions (mg L−1)
Inorganic ions (µg L−1)
  • ND, not detected.

Augmentation of compounds raising xylem concentrations from that of a resistant cultivar to those of a susceptible cultivar showed a functional relationship between these compounds and Xylella growth. Planktonic growth (OD600 nm), biofilm and % aggregation were measured after incubation of X. fastidiosa for 1 h, 5 and 12 days in fluids augmented with amino acids (glutamine, histidine, valine, methionine, isoleucine and phenylalanine) citric acid, inorganic ions (calcium chloride and magnesium sulfate), a combination treatment and unaltered Noble fluid (Table 5). Planktonic growth was highly variable at 1 h. No explanation can be given, except that it was likely due to lack of consistency during the pelleting and rinsing stages. Biofilm formation was significantly influenced by treatment and was the highest for citric acid and the lowest for the amino acid treatment at 1 h and 5 days. It was not possible to measure % aggregation using image j software at 1 h because there was no adequate contrast between the bacteria and the background; however, visual observations indicated that % aggregation was <2% for all treatments. After 5 days of incubation, the combination treatment had the highest planktonic growth, and after 12 days both the calcium inorganic ion treatment and the combination treatment resulted in the highest planktonic growth. After 5 days, the inorganic ion treatment produced the most biofilm, but significant differences no longer occurred after 12 days. The fraction of the Neubauer chamber covered in bacterial aggregates (% aggregation) was the highest for the combination treatment at 5 and 12 days. Representative presentations of % aggregation of the five treatments after 12 days are presented in Fig. 1.

View this table:
Table 5

Planktonic growth (OD600 nm), biofilm formation and the percentage of the field occupied by aggregates (% aggregation) of Xylella fastidiosa, Temecula strain

1 h5 days12 days
TreatmentODBiofilm% AggregationODBiofilm% AggregationODBiofilm% Aggregation
1 Amino acids0.113b0.055cND0.082c0.598b4.7b0.114b0.888a18.2ab
2 Citric acid0.157ab0.100aND0.117b0.664b7.0b0.112b0.969a16.1b
3 CaCl2 & MgSO40.181ab0.064bcND0.113b1.200a8.8b0.216a0.943a23.9ab
4 Combination 1-30.182ab0.080bND0.206a0.704b16.6a0.226a0.949a31.9a
5 Noble xylem fluid0.213a0.068bcND0.103bc0.740b8.2b0.119b1.101a14.7b
  • Treatments represent the addition of compounds (based on the chemistry of Vitis champinii cv. Ramsey) to xylem fluid of Vitis rotundifolia cv. Noble

  • Amino acids supplemented to Noble xylem fluid include GLN, HIS, VAL, MET, ILE and PHE.

  • Numbers within a column with different letters are significantly different by Duncan's multiple-range test (P<0.05).

  • ND, not determined.

Figure 1

Percentage aggregation of Xylella fastidiosa subjected to the treatments in Table 5 for 12 days. Treatment 1, amino acid supplementation; treatment 2, citric acid supplementation; treatment 3, calcium chloride and magnesium sulfate supplementation; treatment 4, combination of treatments 1–3; treatment 5, Noble xylem fluid.

Planktonic growth was consistently the highest and biofilm formation was the lowest for V. champinii cv. Ramsey and the converse was true for V. rotundifolia cv. Noble. Ramsey was, by far, the most nutritional food source of all the xylem fluids (Table 3), and it promoted the most vigorous planktonic growth, but produced very little biofilm (Table 1). Analogously, nutrient-rich X. fastidiosa growth media such as PW+ (Davis et al., 1981) usually support high planktonic growth, but little biofilm formation or aggregates compared with less rich chemically defined media such as CHARD2 (Leite et al., 2004a). Planktonic growth (1 h, 5 and 12 days), biofilm (1 h and 5 days) and % aggregation (5 and 12 days) were the highest for the combination treatment indicating that aspects of bacterial growth had been successfully manipulated by the addition of chemical constituents. On day 12, biofilm was similar for all treatments, although colony growth and aggregation were higher for the calcium chloride and magnesium sulfate treatment and the combination treatment, but not the amino acid or citric acid treatments, indicating that the addition of inorganic ions was crucial to achieving maximum growth and aggregation. This is consistent with the hypothesis that xylem constituents which precipitate calcium (citric acid and phosphorus) promote aggregation and biofilm formation (Leite et al., 2004b). Measurements of aggregation likely represent clumps of bacterial cells and microbial byproducts such as extracellular polysaccharides (De Souza et al., 2004). Aggregation of oral streptococcal cells has been correlated to the concentration of free calcium (Rose, 2000). It is hypothesized that citric acid and phosphorus remove calcium and magnesium from solution to form complexes and insoluble salts (Van Der Houwen & Valsami-Jones, 2001).

Many microbial species in nature grow predominantly in biofilms rather than in planktonic form. Biofilms may be defined as structured communities of microbial aggregates that are enclosed in a polymeric matrix (Costerton et al., 1999). Bacteria in a biofilm or aggregate may be 1000-fold more resistant to immune responses, antibiotics and biocides than planktonic cells (Schembri et al., 2002). Marques (2002) suggested that X. fastidiosa biofilm formation plays a decisive role in vessel occlusion and is a key factor in the virulence to host plants. Xylella fastidiosa strains differ in their ability to form biofilms and aggregates (Marques et al., 2002). Subculturing of X. fastidiosa can also result in a loss of virulence and can reduce the tendency to form aggregates (Hopkins, 1984).

It has been shown that planktonic growth, biofilm formation and aggregation are dependent on the chemistry of xylem fluid and can be manipulated by altering xylem chemistry. Leite (2002) presented a model describing a mechanism whereby Ca+2 and Mg2+ could mediate adhesion between X. fastidiosa and xylem vessels before the development of fastidium gum. De Souza (2004) used analysis of microarray to examine X. fastidiosa genes that were activated in response to biofilm formation. They found that overexpressed genes in biofilm formation promote the synthesis of proteins that are associated with attachment to surfaces, and other genes conferred advantages in the colonization environment. The dependence of aggregation and biofilm formation on the nutrient content of xylem fluid (or growth media) suggests that xylem chemistry is important in the resistance/susceptibility of PD. Future work will entail the delineation of the role of specific chemical factors involved in planktonic growth, aggregation and biofilm formation in vitro and in planta.


The authors are grateful for the review provided by Dr James Marois and Dr Tim Momol. The authors would like to acknowledge the financial support provided by the American Vineyard Foundation and the California Department of Food and Agriculture, and the United States Department of Agriculture.


  • Editor: Yaacov Okon


View Abstract