Most probable number method to enumerate a bioluminescent Xanthomonas campestris pv. campestris in soil

Soil

Bid.

Biochern.

Vol. 28, No. I?, pp. 1725-1728, 1996 c 1997 Elsevier Science Ltd All rights reserved. Printed in Great Britain 00%0717/96 $15.00+0.00

Pergamon PII: SOO38-0717(96)00055-7

MOST PROBABLE NUMBER METHOD TO ENUMERATE A BIOLUMINESCENT XANTHOMONAS CAMPESTRZS PV. CAMPESTRZS IN SOIL R. S. ARIAS*, Department

G. MOCHIZUKI,

of Plant Pathology,

R. FUKUI and A. M. ALVAREZ

University

of Hawaii, HI 96822, U.S.A.

(Accepted 26 Januar?; 1996) Summary-An adaptation of the most probable number (MPN) method and use of microfluor plates for monitoring a bioluminescent strain (I71LIIH-7) of Xanthomonas campestris pv. campestris (Pammel) Dowson in debris-inoculated soil, is described. Microfluor plates with liquid culture medium were inoculated with a dilution series of soil extract previously infested with cabbage debris containing 17 1LIIH-7. Light produced by the bacteria was detected on X-ray film, and absorbance readings of the images were converted to a binary set of values for estimation of MPN by the computer program MPNES (Woomer et al., 1990). The MPN-microfluor plate system was compared with direct dilution plating on FS medium and MPN-miniplate enrichment ELISA with a pathovar-specific monoclonal antibody. The detection limit of X. c. pv. cnmpest’s with FS medium was about lo4 cfu g - ’soil. The detection threshold of MPN-microfluor plates (10’ bacteria g - ’non-sterile soil) was equivalent to the MPN-miniplate enrichment ELISA and 100 times more sensitive than direct dilution plating. The MPN-microfluor plate system is an easy, accurate, sensitive and inexpensive method to detect and enumerate bioluminescent X. c. pv. campestris in soil, and permits ecological studies in the soil environment that will provide the rationale for changes in crop rotations or other cultural practices. (0 1997 Elsevier Science Ltd

INTRODUCTION

Xanthomonas

campestris pv. campestris (X. c. pv. the causal agent of the black rot of crucifers, can survive in host debris and infeet plants in subsequent cropping cycles. Because of the lack of a sensitive system to trace the bacteria in soil (Schaad and White, 1974), the ecology of the soil-borne phase of this pathogen has not been clearly established. X. c. pv. campestris is readily isolated from soil by using various semi-selective media (Fukui et al., 1994). When debris-infested soils are moist, however, the target organism is difficult to identify because of overgrowth by several groups of soil-inhabiting bacteria. The development of bioengineered microorganisms to contain a luciferase gene has greatly increased the ability to detect low numbers of marked cells in different environments (Cebolla et al., 1993). In our system, a bioluminescent strain of X. c. pv. campestris (17 I LIIH-7) was developed (R. McElhaney, unpubl. Ph.D. thesis, University of Hawaii, 1991), thereby making it possible to monitor populations of this bacterium in soil and from plant surfaces even when numerous other bacteria, including xanthomonads, are present. We have adapted the most probable number (MPN) method to microfluor plates for monitoring a bioluminescent X. c. pv. cumpestris strain in debris-infested campestris),

*Author for correspondence

soil. Although a similar approach was taken to enumerate Pseudomonas aeruginosa in soil using microtiet al., 1994). we found that ter plates (Fleming transmission of light among wells limits its use. The MPN-microfluor plate system we developed overcomes the limitations of the microtiter plates and increases the accuracy and flexibility of the assay. MATERIALS

AND METHODS

The bacterial strain we used was X. c. pv campestris This strain is chloramphenicol resistant and has the plasmid pUCD607 containing a 11*xcassette integrated in the bacterial chromosome (R. McElhaney, lot. cit.). The lux operon within the lux cassette operates under the control of the tetracycline resistance gene. pUCD607 also confers resistance to ampicilin, kanamycin and spectinomycin (Shaw and Kado, 1986). 17 1LIIH-7.

Light detection

limits

To test the sensitivity of the MPN-Microfluor plate system to monitor small populations of the bioluminescent strain in soil, we studied the light detection limits on X-ray film in 3 densely pigmented 96-well Microfluor plates (Dynatech,Virginia, USA): white with flat-bottom (‘W’ FB); white with U-bottom (‘W’ UB); and black with U-bottom (‘B’ UB) and compared them to the microtiter plates used by Fleming et al. (1994) (Falcon microtest III; Becton Dickinson,

1725

I726

K.

s. Arias

Toronto, Canada). Plates were prepared with 180 111 per well of 523 medium (Kado et nl.. 1972) containing SO ltg chloramphenicol ml ‘. SO ~18spectinomycin ml-‘, I50 pg cycloheximide ml ’ and 30 pg trimethoprim ml ‘. Culture tubes containing 9 ml of phosphate buffer (pH = 7.4, ISOmM) were prepared with a IO-fold dilution series of a pure culture of the bioluminescent strain 171LIIH-7. As a positive control and to determine the lowest limits for light detection, 20 pl of each dilution were plated on the first row (I I wells) of a ‘W’ FB. ‘W’ UB and ‘B’ UB Microfluor plate. One gram of non-sterile soil was added to each dilution of the same set of tubes. and 20 ul well ’ were plated (7 wells per dilution) into the same Microfluor plates. Inoculated plates were kept in humid chambers for 4 days at 28 C without lids to facilitate aeration. Dilutions of the pure culture were plated directly on Fieldhouse-Sasser (FS) medium (Yuen rt d.. 1987) to determine the amount of inoculum in each well. Bioluminescence from the Microfluor plates was recorded every 24 h on X-ray tilm (DuPont) with 2 h exposure. The developed films were attached to a blank pattern template (Molecular device NV 62000001 ), and absorbance at 590 nm was measured with a MicroStation plate reader (Biolog. Inc.). Absorbance readings were converted to a binary set of values (positive: I .0.59 + 0.270 and negative: 0.406 k 0.004) for estimation of MPN by the computer program. MPNES (Woomer er cl/.. 1990). Analysis of variance (ANOVA) and mean comparisons of the absorbance of film for positive and negative readings were performed using the STATISTIX computer program (Siegel, 1992). Microtiter plates (Falcon Microtest III) (Fleming et nl.. 1994) were prepared with 523 broth medium and plated with a dilution series of a pure culture of 17lLIIH-7 (4 wells per dilution) and a dilution series of debris-infested soil (4 wells per dilution) to compare the quality of image on X-ray him obtained from the microtiter plate and from Microfluor plates.

Three methods were compared for their utility in estimating population densities of the strain I7 I LIIH-7 in artificially infested soil: a) direct dilution plating on FS medium (Yuen et al.. 1987) in Petri plates. b) Miniplate enrichment- ELISA (Norman and Alvarez. 1994) adapted to MPN method and c) MPNMicrofluor plates system using S23 broth medium as described above. Air-dried debris of cabbage leaves infested with the strain 17lLIIH-7 was ground (I mm’), mixed thoroughly with a non-sterile X. c. pv campestris-free soil (Typic Eutrandept from Kula, Maui, Hawaii) at ratios of I, 0.1 and 0.01 g kg - ’ soil. One gram of each infested soil sample was suspended in 9 ml of phosphate buffer (pH 7.4) in duplicate samples, and eight IO-fold dilutions were made from each soil suspension and from a pure culture of I7lLIIH-7. For di-

<‘I

L//.

lution plate counting. 100 ul of appropriate dilutions of the 4 suspensions were plated in duplicate onto FS plates, 3 plates per dilution. For the miniplate enrichment-ELISA, the X. c. pv. campesrris-specific monoclonal antibody (MAb) X21 was used as indicated Alvarez et al. (1994). Absorbance readings at 450 nm were measured with a MicroStation plate reader and converted to a binary set of values for estimation of MPN by the computer program, MPNES (Woomer er ~1.. 1990). For the Microfluor-plate assay. 180 ul well - ’ of 523 broth supplemented with antibiotics as indicated for light detection limit were dispensed in ‘W’ FB Microfluor plates. Both miniplate enrichmentELISA and microfluor plates were plated with 20 111

Fig. I. Comparison of three different Microfluor plates, (a) ‘B‘ UB, (b) ‘W’ UB and (c) ‘W’ FB to detect of a bioluminescent X. C. pv. c~tmrqwsrrisin soil. Wells Al to Al I, dilutton series of a pure culture of 17lLIIH-7; B I to HI I, seven wells per dilution of the same set of tubes, supplemented with I g of non-sterile soil on each dilution.

Bioluminescent

based MPN method

1234567891Ollli

1721

to detect a bioluminescent Pseudomonas in soil. However, light contamination observed among wells and even on the margin of the transparent plates (Falcon Microtest III) after 1 h exposure to X-ray film (Fig. 2) constitutes a limitation, preventing their use in bioluminescence detection. Samples that were calculated to contain a single cell of 171LIIH-7 in the 20 l,d of inoculum of a soil suspension (1 g 9 ml - ’ phosphate buffer) were visible as dark spots when white microfluor plates were exposed to X-ray film [Table 1, Fig. l(b-c)]. The comparison of 3 microfluor plates showed that ‘B’ UB plates absorbed much light, producing a weak image on the X-ray film and therefore lower MPN estimation, [Fig. l(a), Table 11. Both white-microfluor plates showed conspicuous images on X-ray film; additionally, the border present on the ‘W’ FB plates provides a clearly delimited image and prevented light contamination during long exposures [Fig. l(b-c)]. Differences in absorbance values between positiveand negative-reacting wells obtained on the X-ray film from ‘W’ FB and ‘W’ UB Microfluor plates were highly significant (P
aeruginosa

Fig. 2. X-ray film obtained from a microtiter plate (Falcon Microtest III), following 1 h exposure arranged in 4 wells per dilution of a pure culture of 171LIIH-7 (A to D) and 4 well per dilution of soil infested with 17lLIIH-7 (E to H).

from each tube of a lo-fold dilution series (10 - ’ to 10 ‘) in 3 wells per row. Plates were incubated and bioluminescence was detected as indicated above. Two replicates were used for each method and data analyzed as a completely randomized design. ANOVA and differences among means were determined using Tukey’s Studentized range test and the computer program STATISTIX (Siegel, 1992). The time and cost to process each sample by these three methods were recorded.

RESULTS AND DISCUSSION

The maximum number of positive readings determined by the X-ray film was observed between 48 and 72 h of incubation. Exposures at two different times of incubation assured maximum detection. The strain 171LIIH-7 incubated in liquid medium can bioluminesce for up to 10 days in pure culture, and up to 5 days in liquid medium containing non-sterile soil. The maximum light emission is produced at 25°C. Since the amount of light produced by this bioluminescent strain is a function of environmental factors such as temperature, aeration and carbohydrate source (R. McElhaney, lot. cit.), exposures of up to 3 h may be required to detect low light emission. No light contamination (transmission of light among wells) was detected on any densely-pigmented Microfluor plate after 2 h exposure to X-ray film [Fig. l(a-c)]. ‘W’ FB Microfluor plates exposed for up to 6 h shown no light contamination. Fleming et al. (1994) proposed the use of the transparent microtiter plates Falcon Microtest III on a simi-

Table

I, Bacterial

numbers resulting from direct (FS agar plates) or indirect (MPN-Microfluor

2.43 x IO’ ( + ;I.70 x IO’)

‘B’ UB 3.89 x IO6

‘MPN estimated using MPNES program (Woomer er al., 1990).

Standard error is in parentheses.

plate) counts on debris-infested

non-sterile soil

MPN-microfluor plate Y

FS agar plates

‘W’ UB 1.48 x IO8

‘W’ FB I .94 x IO8

R. S. Anas e/ t/l.

172X

Most Probable Number FSagar

ELISA from ET microplates

plats

Y (mean X-ray

of duplicates) film

from

Microfluor

lP

Y1S x loFh*

1.10 x 10hh

1.25 x 10hh

IO0 mg

2 IS x IO”

7 00 x 102d

2.00 x IO-”

ND

IO rng

3 IO _.

pure culture ‘MPN

determined

’ Same superscript ND:

011 IO-fold letter<

dilutwm

indicate

plated

no significant

x

IO’.’

on 3 well\ difference\

I .ROx I O’e

4.60 x IO”

I .25 x IO’.’

4.00 x 107.’

plates

per dilutmn. wthln

row

at fhr

0.05%

level,

as determined

by Scheffe’h

method

(Siegel.

IYY?)

none detected.

ing the base dilution ratio. and increasing the number of wells per dilution. The use of the program MPNES 1990) in combination with the (Woomer et al., Microfluor plates gives flexibility to the assay. MPNES program permits evaluation of any number of wells per dilution and any base of dilution. in contrast to published MPN tables that are restricted in the dilution ratios and number of tubes per dilution (Woomer it nl.. 1990). Both the coyt and time of processing soil samples by using the MPN-Microfluor plate system using 1 wells per dilution and up to 6 dilutions per sample were about 60% less than the traditional plating method. The cost for MPN-miniplate enrichmentELISA was about the same as for the MPN-Microfluor plate system. MPN-enrichment-ELISA is accurate for the detection and enumeration of indigenous populations of X. c. pv. campestris in soil. while MPNMicrofluor plate system allows tracing a single target strain in non-sterile soil. The MPN-Microfluor plate system is an accurate. sensitive and inexpensive method to detect and enumerate bioluminescent X. c. pv. carnpstris in soil, and permits ecological studies that will provide the rationale for changes in crop rotation or other cultural practices. This method also may be used to detect and enumerate other bioluminescent bacteria in soil. A~.knoM,/edKrmrnts-Research was supported in part by a grant from the U. S. Agency for International Development. Program in Science and Technology Cooperation (PSTC).

REFERENCES

Alvarez A.M.. Benedict A.A.. Mizumoto C.Y.. Hunter J.E. Gabriel D.W. (1994) Serological. pathological. and genetic

diversity among strains of Xanthomonas campestris infecting crucifers. Phytymtholog~ 84, I449- 1457. Bashan Y., Mitiku G.. Ziv-Vecht 0. and Levanony H. (1991) Estimation of minimal numbers of Aw~pirillrtm hrc~.si/erwe using time-limited liquid enrichment combined with enzyme-linked lmmunosorbent assay. Soil Biology & /3iochrmi.\t,~ 23, 13.5-138. Cebolla A., Ruir-Berraquero F. and Palomares A.J. (1993) Stable tagging of Rhi;ohium me/i/& with the tirefly lucifcra\e gene for environmental monitoring. Applied aml Ewironmrntcrl Microhiolog~ 59, 25 I I-25 19. Fleming C.A., Lee H. and Trevors J.T. (1994) Bioluminescent Most-Probable-Number method to enumerate lux-marked Preudomomu rrerugino.w UG2Lr in Soil. .~/~~>lie~l trrztl O~~~ironmental Microhiolog~ 60, 345X-346 1. Fukui R.. Arias R.S. and Alvarez R. (1994) Efticacy of foul semi-selective media for recovery of Xarzthomontrc- ctrmpes~ tric pv. cwnpc.stri.\ from tropical soils. Journtrl of Applied Brrc~trriok>g~ 77, 534-540. Kado C.I.. Herkett M.G. and Langley R.A. (1972) Studies of A,qro/xwtrrinm rrrmqfircYens: characterization of strains ID135 and B6 and analysis of the bacterial chromosome. transfer RNA and riboaomeh for tumor inducing ability. P/~~.~io/o,~iccr/ Pkmt P&7o/o/o,qy 2. 47-57. Norman D. and Alvarez A.M. (1994) Rapid detection of pv. diyffenbrrchiur in anthurium Xtr~zrhomo~lusumprstris plant\ with a miniplate enrichment/ELISA system. Plnru fxsecl,r 78, 954-958. Schaad N.W. and White W.C. (1974) Survival of 64, IS 1% Xunrho/norms umzpestris in soil. Ph~topathology 1520. Shaw J.J. and Kado C.I. ( 1986) Development of a Vihrio bioluminescence gene-set to monitor phytopathogenic bacteria during the ongoing disease process in a non-disruptive manner. BioTrc~hnolo~.v 4, 560-564 SIegeI J. (1992) STATfSTIX (J. Siegel Ed.). Analytical Software Publisher, St Paul, MN. Woomer P., Bennet J. and Yost R. (1990) Overcoming the inRexibility of Most-Probable-Number Procedures. Agromnq .Jouma/ 82, 349-3.53. Yuen G.Y., Alvarez A.M., Benedict A.A. and Trotter K.J. ( 19X7) Use of monoclonal antibodies to monitor the dissemination of Xunthomoncts ccrmpestris pv, ccrmprstris Pk~topcuholog~ 77, 366-370.