SEARCH

SEARCH BY CITATION

Keywords:

  • Flavobacterium columnare;
  • chemotaxis;
  • lectin-like capsular substances;
  • gliding motility;
  • quantitative PCR;
  • fish skin mucus

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

To gain an insight into the chemotactic factors involved in chemotaxis, we exposed a virulent strain of Flavobacterium columnare to various treatments, followed by analysis of its chemotactic activity. The chemotactic activity of F. columnare was significantly (P<0.05) inhibited when cells were pretreated by sodium metaperiodate, and a major portion of the capsular layer surrounding the cells was removed. Pretreatment of F. columnare with d-mannose, d-glucose and N-acteyl-d-glucosamine significantly (P<0.05) inhibited its chemotaxis activity, whereas pretreatment of cells with d-fructose, l-fucose, d-glucosamine, d-galactosamine, d-sucrose and N-acetyl-d-galactosamine failed to inhibit its chemotactic activity. These results indicate that at least three carbohydrate-binding receptors (d-mannose, d-glucose and N-acteyl-d-glucosamine) associated with the capsule of F. columnare might be involved in the chemotactic responses. The relative transcriptional levels of three gliding motility genes (gldB, gldC, gldH) of F. columnare compared with 16S rRNA gene following the exposure of F. columnare to catfish skin mucus were evaluated by quantitative PCR (qPCR). qPCR results revealed that the transcriptional level of gldH was significantly (P<0.001) upregulated in normal F. columnare at 5 min postexposure to the catfish mucus. However, when F. columnare were pretreated with d-mannose, there was no upregulation of gliding motility genes. Taken together, our results suggest that carbohydrate-binding receptors play important roles in the chemotactic response to catfish mucus.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Flavobacterium columnare, the causative agent of columnaris disease, is responsible for significant economic losses in freshwater fish aquaculture worldwide. Many species of wild, cultured and ornamental fish are susceptible to columnaris disease (Austin & Austin, 1999). Channel catfish are especially susceptible to columnaris, with high mortality rates (Wagner et al., 2002). Columnaris disease is characterized by necrotic skin, fin and gill lesions containing yellow-pigmented bacteria aggregated in hay stack-shaped films (Austin & Austin, 1999). Flavobacterium columnare is a motile bacterium that moves by gliding motility over surfaces (McBride, 2001). It is considered to be a rapid glider (Youderian, 1998). Flavobacterium johnsoniae, a closely related species, is reported to glide at speeds up to 10 μm s−1 (Pate & Chang, 1979; Lapidus & Berg, 1982), and its gliding motion appears to require the recognition of extracellular components of the host by components of the bacterial cells to send signals to trigger the movement. Gliding motility of F. johnsoniae requires the expression of six genes: gldA, gldB, gldD, gldF, gldG and gldH (McBride et al., 2003), and it has been suggested that the mechanisms of gliding motility in F. columnare are similar (Olivares-Fuster & Arias, 2008). To our knowledge, the expression of genes involved in the gliding motility of F. columnare has not been described previously. Mucus from the skin and gills of catfish has been demonstrated to promote the chemotaxis of F. columnare (Klesius et al., 2008; LaFrentz & Klesius, 2009). The mechanisms involved in the chemotactic response of F. columnare to mucus are largely unknown. In this study, the effects of sodium metaperiodate and different carbohydrate treatments on F. columnare chemotactic activities to catfish skin mucus were examined. Furthermore, the effect of catfish skin mucus treatment on the transcriptional levels of three gliding motility genes (gldB, gldC and gldH) in F. columnare was evaluated.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Flavobacterium columnare strain and culture conditions

The F. columnare ALG-00-530 strain was used in this study. This strain was isolated from channel catfish with columnaris disease in Alabama. The ALG-00-530 is a genomovar II strain that is highly virulent to channel catfish (Arias et al., 2004; Shoemaker et al., 2007). This strain was demonstrated to be chemotactic to mucus from the skin of channel catfish (Klesius et al., 2008). The bacteria were cultured in modified Shieh broth (0.5% tryptone, 0.2% yeast extract, 45.6 μM CaCl2·2H2O, 1.1 mM KH2PO4, 1.2 mM MgSO4·7H2O, 3.6 μM FeSO4·7H2O, pH 7.2) for 24 h at 28 °C on an orbital shaker set at 90 rotations min−1 (Klesius et al., 2008; LaFrentz & Klesius, 2009). The bacteria were harvested by centrifugation at 2800 g for 15 min, washed twice with sterile phosphate-buffered saline (PBS), pH 7.2 and resuspended in Hanks' balanced salt solution (HBSS, pH 7.2, Sigma, St. Louis, MO) to an OD540 nm of 1.0 (1 × 109 CFU mL−1 (LaFrentz & Klesius, 2009).

Mucus samples

Healthy channel catfish NWAC-103 strain (50–100 g) were cultured in 57-L glass aquaria with aeration and flowthrough water. Fish were anesthetized with 100 mg−1 tricane methanesulfonate (Argent Chemicals, Redmond, CA). The anesthetized fish were held vertically and mucus was collected from the skin by gently stroking with a soft rubber spatula into Petri dishes. Special care was taken to prevent damage to the skin and avoid contamination with blood or other extraneous products. The mucus from individual fish were pooled together and centrifuged at 6000 g for 15 min and the pellet (epithelium cells and cellular debris) was discarded. The mucus protein concentration was determined using the Micro BCA Protein assay (Pierce, Rockford, IL) and adjusted to 0.1–0.2 μg μL−1 with HBSS. Pooled mucus samples (10 μL) were streaked for bacterial isolation onto tryptic soy agar and modified Shieh agar plates and incubated at 28 °C for 72 h to check for contamination. The pooled mucus samples were stored at −80 °C before use.

Chemotaxis assay

Chemotaxis assays with F. columnare were performed using blind-well chambers (Corning Costar, Cambridge, MA) as described previously (LaFrentz & Klesius, 2009). Briefly, chemotaxis assays were performed in duplicate using a triad of chemotaxis chambers and mucus samples that were culture negative. The wells of the bottom chambers were filled with 200 μL of mucus (mucus test) or HBSS (negative control). Polycarbonate membranes (Nucleopore, Pleasontan, CA) with a diameter of 13 mm and a pore size of 0.8 μm were carefully placed on the top of the bottom chambers with the shiny side up. Following assembly of the chambers, 200 μL of an F. columnare cell preparation was placed in the wells of the top chambers. Triplicate chambers were used for each assay. Following incubation at room temperature for 1 h, the chambers were disassembled and the membranes were removed carefully using a PenVacuum with a 3/8″ probe (Ted Pella, Redding, CA). The contents of the bottom wells were mixed and 100-μL samples were removed and placed in flat-bottom microtiter 96-well plates (Thermo-Scientific, Milfort, MA). Each mucus test or HBSS alone was also added to the 96-well plate (100 μL) to determine the background absorbance due to the sample alone. Positive controls consisting of 100 μL of the adjusted F. columnare culture diluted 1 : 5 in HBSS were also added to the 96-well plates. To each test well that contained either mucus, positive or negative controls, 20 μL of the combined MTS/PMS [Celltiter 96 Aqueous Non-Radioactive Cell Proliferation Assay (Promega, Madison, WI) was added and mixed. The plate was covered by an aluminum foil to protect from light and incubated for 4 h at 28 °C. The A490 nm was recorded using a Model 680 microplate reader (Bio-Rad, Hercules, CA). The absorbance values of the mucus samples or HBSS alone were subtracted from mucus test samples and HBSS control to correct the absorbance values of mucus sample or HBSS control alone. Three independent assays were carried out using the pooled mucus sample. To quantify the F. columnare chemotactic response in CFU mL−1, the corrected absorbance values for the cell concentrations were plotted against the corresponding numbers of viable F. columnare CFU mL−1. Linear regression was performed using graphpad prism (version 2.01, GraphPad Software, San Diego, CA) to determine the correlation between the corrected A490 nm and the number viable CFU mL−1.

Effect of pretreatment of F. columnare cells with sodium metaperiodate and carbohydrates on chemotaxis

To assess the effect of sodium metaperiodate (Sigma) on chemotaxis, bacteria were prepared in HBSS as described above and treated at concentrations of 0.5, 1.0, 1.5, 2.0 and 2.5 mM for 1 h in the dark at 28 °C. The treatments were stopped by adding three to five drops of 10% ethylene glycol. The bacteria were then washed once in HBSS, resuspended in HBSS and assayed for their chemotaxis capacity. To evaluate the effect of 50 mM of carbohydrates (Sigma) on chemotaxis, bacteria were prepared as described above and incubated with 50 mM of d-galactosamine, d-glucosamine, d-sucrose, d-fructose, l-fucose, N-actyl-d-glucosamine, N-acetyl-d-galactosamine, d-glucose or d-mannose for 1 h in the dark at 28 °C. The effect of 50 mM d-mannose alone on the chemotactic response of F. columnare to mucus samples from 24 individual catfish was also determined.

Effect of sodium metaperiodate treatment on the capsule of F. columnare

To examine the effect of 2.5 mM sodium metaperiodate treatment on the capsule, bacteria were prepared in HBSS as described above and treated for 1 h in the dark at 28 °C. Ten microliters of treated cells and three drops of 5% skim milk were mixed together on clean microscope slides and then streaked across the slides using a second slide in a swift motion. After air drying, the slides were stained with Gram's crystal violet solution for 1–2 min. The excess stain was washed off with 20% copper sulfate solution and the slides were air dried in a vertical position. The capsule was observed using a × 100 oil immersion lens with a Olympus BX41 microscope (Tokyo, Japan) at a total magnification of × 1000. The bacterial cells and skim milk are expected to appear a dark color while the capsule will remain colorless.

Effect of catfish skin mucus treatment on gliding motility gene expression in F. columnare

The F. columnare ALG-00-530 cells were cultured in modified Shieh broth. Cells were then treated with or without 50 mM d-mannose for 1 h. Cells were then harvested by centrifugation at 2800 g for 15 min. The cell pellets were then washed twice with sterile (pH 7.2) and resuspended in HBSS (pH 7.2, Sigma) with mucus proteins at a concentration of 0.2 μg μL−1 for 0, 5, 10 and 15 min, respectively. As negative controls, F. columnare cells were incubated in an HBSS solution without mucus protein for 0, 5, 10 and 15 min, respectively. Cells were then harvested by centrifugation at 3000 g for 15 min and stored at −80 °C before RNA extraction. Mucus exposure experiments were replicated three times. Total RNA was isolated from F. columnare bacterial cells using the RNeasy Mini Kit (Qiagen, Valencia, CA) following the manufacturer's protocol. Total RNAs were then treated with DNA-free (Ambion, Austin, TX). All total RNAs were quantified on a Nanodrop ND-1000 spectrophotometer (Nanodrop Technologies, Rockland, DE). Total RNAs were resuspended in distilled water and cDNA synthesis was immediately performed using an oligo-dT20 primer and AMV reverse transcriptase (Invitrogen, Carlsbad, CA). For each cDNA sample, F. columnare 16S rRNA gene (GenBank accession no. AY842899) primers were included as an internal control to normalize the variation of the cDNA amount. The primers used for the amplification of 16S rRNA gene, gldB, gldC, gldH and hsp90 are listed in Table 1. Hsp90 primers were included in all quantitative PCR (qPCR) because heat shock proteins have been widely used as internal controls in different experiments due to their ‘housekeeping’ functions (Greer et al., 2010). All qPCR was performed on an Applied Biosystems 7500 Real-Time PCR System (ABI, Foster City, CA) using Platinum® SYBR® Green qPCR SuperMix-UDG with ROX (Invitrogen) in a total volume of 12.5 μL. The qPCR mixture consisted of 1 μL of cDNA, 0.5 μL of 5 μM gene-specific forward primer, 0.5 μL of 5 μM gene-specific reverse primer and 10.5 μL of 1 × SYBR Green SuperMix. The qPCR thermal cycling parameters were 50 °C for 2 min, 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 1 min. All qPCR was run in duplicate for each cDNA sample and three F. columnare cDNA samples were analyzed by qPCR. The relative transcriptional levels of different genes were determined by subtracting the cycle threshold (Ct) of the sample by that of the 16S rRNA gene, the calibrator or internal control, as per the formula: ΔCt=Ct (sample)−Ct (calibrator). The relative transcriptional level of a specific gene in F. columnare after mucus treatment compared with that in the untreated F. columnare was then calculated using the formula 2ΔΔCt where ΔΔCtCt (with mucus)−ΔCt (without mucus) as described previously (Pridgeon et al., 2009).

Table 1.   Primers used in qPCR
Gene nameAccession no.*Primer namePrimer sequence (5′–3′)
  • *

    Accession number from NCBI GenBank.

16S rRNAAY84289916S-forwardGGCTTAAATGGGAAACGAC
16S-reverseACTTCAGGTACCCCCAGC
gldBDQ911519gldB-forwardTGGTGTGAGGAAAATGATG
gldB-reverseCAACCAATCCACGCTCC
gldCDQ911520gldC-forwardGATGGAGGAGTGGAAAAGC
gldC-reverseCGGTTGCACGTTCAAAAG
gldHAY781295gldH-forwardCTTTGAAAACGGATGGCC
gldH-reverseCTTGCCCCATAAGACTTCC
Hsp90EI702724hsp90-forwardCGACGCATAAACTCTGGTTG
Hsp90-reverseCACGTGTAGACGCTGACCAT

Statistical analysis

The chemotaxis results were statistically analyzed by anova, followed by Duncan's multiple range test to determine significant differences between means of CFU mL−1 (sas, version 9.1, Cary, NC). Transcriptional-level data were analyzed by anova using sigmastat statistical analysis software (Systat Software, San Jose, CA). A 95% confidence interval was considered to be significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Chemotaxis assay results

To quantify the F. columnare chemotactic response in CFU mL−1, the corrected absorbance values for the cell concentrations were plotted against the corresponding numbers of viable F. columnare CFU mL−1. A positive linear correlation was obtained between corrected absorbance values and CFU mL−1 (Fig. 1). The coefficient of determination (r2) was 0.9831. The chemotactic response was determined from the following equation of the line [X=(Y−0.3051)/0.0000007327], where X is the number of viable F. columnare CFU mL−1 and Y is the OD490 nm or A490 nm values.

image

Figure 1.  Correlation between viable Flavobacterium columnare cells (CFU mL−1) and corrected optimal density or absorbance values using the CellTiter 96® Aqueous Assay. Each data point represents the means of triplicate assays containing different viable F. columnae cell concentrations.

Download figure to PowerPoint

Effect of sodium metaperiodate and carbohydrate treatments on F. columnare chemotaxis

The results in Table 2 show that sodium metaperiodate treatment significantly (P<0.05) inhibited the chemotactic response at all the concentrations tested. A concentration of 0.5 mM was the lowest concentration that significantly (P<0.05) inhibited chemotaxis. The effect of carbohydrate treatment on the chemotaxis of F. columnare is presented in Table 3. Pretreatment of cells with d-mannose resulted in the strongest inhibition of chemotaxis. Significant (P<0.05) inhibition was also observed following treatment with either d-glucose or N-acetyl-d-glucosamine. Other mono- or disaccharides tested failed to significantly inhibit chemotaxis. Treatment with d-mannose treatment consistently caused a significant (P<0.05) 65.9% inhibition in the chemotactic response of F. columnare to mucus samples from 24 individual healthy catfish (data not shown).

Table 2.   Effect of pretreatment of Flavobacterium columnare cells with sodium metaperiodate on their chemotaxis activity to catfish skin mucus
Sodium metaperiodate concentration (mM)Mean OD ± SE of MTS assay*
  • *

    Each value represents the mean OD ± SE of three experiments. Different letters represent statistically significant mean OD differences as determined by anova as described in Materials and methods.

Untreated positive control2.58 ± 0.18A
Untreated negative control0.05 ± 0.01E
0.50.82 ± 0.11B
1.00.62 ± 0.16B,C
1.50.24 ± 0.09C,D
2.00.28 ± 0.09C,D
2.50.21 ± 0.06F
Table 3.   Effect of pretreatment of Flavobacterium columnare cells with carbohydrates on chemotaxtic activity to catfish skin mucus
PretreatmentMean A490 nm± SE MTS assay*Mean (CFU mL−1)
  • *

    Each value represents the means ± SE of three independent assays. Different letters indicate significant differences between pretreatments and mucus-positive control as determined by one-way anova as described in Materials and methods.

Mucus-positive control2.43 ± 0.17A2.90 × 106
HBSS-negative control0.06 ± 0.01B3.34 × 105
Carbohydrate 50 mM
 d-Galactosamine1.94 ± 0.11A2.61 × 106
 d-Glucosamine2.48 ± 0.12A2.97 × 106
 d-Sucrose2.31 ± 0.33A2.74 × 106
 d-Fructose2.01 ± 0.27A2.33 × 106
 l-Fucose2.24 ± 0.20A2.64 × 106
 N-Acetyl-d-glucosamine1.24 ± 0.25C1.28 × 106
 N-Acetyl-d-galactosamine2.53 ± 0.27A3.04 × 106
 d-Glucose1.19 ± 0.09C1.14 × 106
 d-Mannose0.96 ± 0.05C8.93 × 105

Effect of sodium metaperiodate treatment on the capsule of F. columnare

The capsule of untreated F. columnare cells is shown in Fig. 2a. The effect of sodium metaperiodate treatment on the capsule of F. columnare is shown in Fig. 2b. In Fig. 2a, the bacterial cells were surrounded by a thick capsular layer. However, sodium metaperiodate treatment considerably reduced the thickness of the capsule to a very thin layer surrounding the cells (Fig. 2b).

image

Figure 2.  Capsule staining of Flavobacterium columnare (× 100). (a) Untreated F. columnare (control) covered by a thicker capsule and (b) F. columnare treated with 2.5 mM sodium metaperiodate (NaIO4) covered by a thinner capsule.

Download figure to PowerPoint

Expression of gliding motility genes in untreated F. columnare in response to mucus treatment

The relative transcriptional levels of three gliding motility genes (gldB, gldC and gldH) of normal (untreated) F. columnare following exposure to catfish mucus were compared with that of normal F. columnare not exposed to catfish mucus. qPCR results revealed that the transcriptional level of gldH was significantly (P<0.001) upregulated at 5 min postexposure to the catfish mucus (Fig. 3). However, the transcriptional levels of gldB and gldC in mucus-treated F. columnare were not significantly different from that in F. columnare not treated by mucus. As a negative control, the expression of the gene encoding Hsp90 of F. columnare was not affected by the mucus treatment (Fig. 3).

image

Figure 3.  Effect of mucus treatment on the transcriptional levels of three gliding motility proteins (gldB, gldC and gldH) and Hsp90 in normal Flavobacterium columnare. The relative transcriptional levels of different genes were determined by subtracting the cycle threshold (Ct) of the sample by that of the 16S rRNA gene, the calibrator or internal control, as per the formula: ΔCt=Ct (sample)−Ct (calibrator). The relative transcriptional level of a specific gene in normal F. columnare after mucus treatment compared with that in the untreated F. columnare was then calculated using the formula inline image, where ΔΔCtCt (with mucus)−ΔCt (without mucus). Data are presented as mean ± SE.

Download figure to PowerPoint

Expression of gliding motility genes in d-mannose-treated F. columnare in response to mucus treatment

The relative transcriptional levels of three gliding motility genes (gldB, gldC and gldH) of d-mannose-treated F. columnare following exposure to catfish mucus were compared with that of treated F. columnare not exposed to catfish mucus. qPCR results revealed that the transcriptional level of gldB, gldC and gldH in mucus-treated F. columnare was similar to that in the PBS-treated F. columnare (Fig. 4). Similarly, the transcriptional level of the negative control Hsp90 was not affected by the mucus treatment in the d-mannose-pretreated F. columnare (Fig. 4).

image

Figure 4.  Effect of mucus treatment on the transcriptional levels of three gliding motility proteins (gldB, gldC and gldH) and Hsp90 in d-mannose-treated Flavobacterium columnare. The relative transcriptional levels of different genes were determined by subtracting the cycle threshold (Ct) of the sample by that of the 16S rRNA gene, the calibrator or internal control, as per the formula: ΔCt=Ct (sample)−Ct (calibrator). The relative transcriptional level of a specific gene in d-mannose-treated F. columnare after mucus treatment compared with that in the untreated F. columnare was then calculated using the formula inline image, where ΔΔCtCt (with mucus)−ΔCt (without mucus). Data are presented as mean ± SE.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

When F. columnare cells were pretreated by sodium metaperiodate, their chemotactic response to catfish skin mucus was significantly inhibited. Sodium metaperiodate treatment also resulted in a partial loss of its capsule. A previous study demonstrated that sodium metaperiodate treatment of a F. columnare isolate resulted in significant inhibition of adherence to gill tissue and a 90% loss of capsule (Decostere et al., 1999). Decostere et al. (1999) hypothesized that sodium metaperiodate treatment removed or inactivated the lectin chemotactic receptor associated with the capsule by cleaving the C–C bond between vicinal hydroxyl groups of sugar, thus removing or loosening the capsule of F. columnare. We hypothesize that the sodium metaperiodate treatment removed or inactivated the sugar-binding receptor associated with capsule, thus inhibiting the F. columnare chemotactic response to mucus. The treatments of d-mannose, d-glucose and N-acetyl-d-galactosamine resulted in significant inhibition of the chemotactic responses of F. columanare to catfish skin mucus, suggesting that at least three carbohydrate-binding receptors of the capsule are involved in chemotactic responses. These receptors may recognize and bind to the d-mannose, d-glucose and N-acetyl-d- galactosamine structure of the chemoattractants associated with the fish mucus. d-Glucose and N-acetyl-d-galactosamine treatment of F. columnare was previously shown to significantly inhibit adherence to gill tissue (Decostere et al., 1999).

Several genes are required for F. johnsoniae gliding motility (Agarwal et al., 1997; Hunnicutt & McBride, 2000, 2001; Hunnicutt et al., 2002). The GldH protein is a lipoprotein and has been demonstrated to be required for F. johnsoniae gliding motility (McBride et al., 2003). We examined the expression of gldB, gldC and gldH following the exposure of F. columnare cells to catfish skin mucus and found that the transcriptional level of gldH was significantly upregulated in normal F. columnare at 5 min postexposure to the mucus. However, when F. columnare cells were pretreated with 50 mM d-mannose, the catfish skin mucus failed to induce the upregulation of gldH, suggesting that gldH might play an important role in the chemotactic response of F. columnare to catfish skin mucus and that pretreatment of F. columnare with d-mannose might be able to block the chemotactic response of F. columnare to catfish. Whether pretreatment of F. columnare with d-mannose will affect the virulence of F. columnare to catfish merits further study.

In summary, using a different pretreatment of F. columnare cells and an in vitro chemotaxis assay, we found that at least two major components were involved in the chemotactic responses of F. columnare to catfish skin mucus. Firstly, the capsule of F. columnare plays an important role in recognizing the extracellular chemoattractants from the catfish mucus through lectin-like receptors. Secondly, one or more gliding motility proteins are involved in the chemotactic response of F. columnare to catfish skin mucus. These components might play important roles in the cell-to-cell communication necessary for gliding the chemotaxis of F. columnare toward catfish skin mucus. However, the exact roles of F. columnare gliding motility proteins in chemotaxis and the identities of the lectin-like receptors on the capsule of F. columnare receptors and the chemoattractants of the catfish skin mucus remain to be further studied.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We thank Drs Benjamin LaFrentz (USDA-ARS) and Victor Panangala (USDA collaborator) for critical reviews of the manuscript. We thank Beth Peterman and Stacey LaFrentz (USDA-ARS) for their excellent technical support. We also thank the management team of the Aquatic Animal Health Research Unit for daily care and management of the fish. This study was supported by the USDA/ARS CRIS project #6420-32000-024-00D. The use of trade, firm or corporate names in this publication is for the information and convenience of the reader. Such use does not constitute an official endorsement or approval by the United States Department of Agriculture or the Agricultural Research Service of any product or service to the exclusion of others that may be suitable.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References