933

Journal of Food Protection, Vol. 74, No. 6, 2011, Pages 933–938 doi:10.4315/0362-028X.JFP-10-418 Copyright G, International Association for Food Protection

Detection of Alicyclobacillus in Fruit Juice Using a Random Genomic DNA Microarray Chip

JUN HYEONG JANG, SUN-JOONG KIM, BO HYUN YOON, JEE-HOON RYU, MAN BOCK GU, AND HYO-IHL CHANG*

College of Life Sciences and Biotechnology, Korea University, 5-1 Anam-dong, Seongbuk-gu, Seoul 136-701, Republic of Korea

MS 10-418: Received 27 September 2010/Accepted 9 February 2011 Downloaded from http://meridian.allenpress.com/jfp/article-pdf/74/6/933/1685326/0362-028x_jfp-10-418.pdf by guest on 28 September 2021

ABSTRACT This study describes a method using a DNA microarray chip to rapidly and simultaneously detect Alicyclobacillus species in orange juice based on the hybridization of genomic DNA with random probes. Three food spoilage were used in this study: Alicyclobacillus acidocaldarius, Alicyclobacillus acidoterrestris, and Alicyclobacillus cycloheptanicus. The three Alicyclobacillus species were adjusted to 2 | 103 CFU/ml and inoculated into pasteurized 100% pure orange juice. Cy5-dCTP labeling was used for reference signals, and Cy3-dCTP was labeled for target genomic DNA. The molar ratio of 1:1 of Cy3-dCTP and Cy5-dCTP was used. DNA microarray chips were fabricated using randomly fragmented DNA of Alicyclobacillus spp. and were hybridized with genomic DNA extracted from spp. Genomic DNA extracted from Alicyclobacillus spp. showed a significantly higher hybridization rate compared with DNA of Bacillus spp., thereby distinguishing Alicyclobacillus spp. from Bacillus spp. The results showed that the microarray DNA chip containing randomly fragmented genomic DNA was specific and clearly identified specific food spoilage bacteria. This microarray system is a good tool for rapid and specific detection of thermophilic spoilage bacteria, mainly Alicyclobacillus spp., and is useful and applicable to the fruit juice industry.

Alicyclobacillus spp. are acidophilic, thermophilic detection methods require extensive amounts of genetic sporeformers commonly found in acidic fruit juices. They information for spoilage microbes and cannot detect many cause a sour taste and a bad smell and are known to be the spoilage bacteria simultaneously (5, 7, 8, 13). Therefore, main cause of lowered quality in fruit juice (3). Since food there is a critical need for an advanced diagnostic system contains a variety of nutrients, microbial growth can lead to that can rapidly and simultaneously detect various types of adverse changes in taste, smell, and color that negatively food spoilage bacteria. impact food quality. Since information on the genome sequence for specific Spoilage bacteria that form heat-resistant spores can microorganisms that cause food decomposition and poor germinate and grow in industrially sterilized fruit juice, food quality is not always available, genomic DNA from posing a serious threat to the beverage industry. pure cultures and random genomic DNA libraries can be The representative spoilage bacteria Alicyclobacillus built. The DNA chip is made as a product of that library. spp. was first identified in 1971 (6) and was later found to With random priming, hybridization detection of a specific be the cause of decreasing quality in various foods including microorganism can be confirmed (9). As such, DNA fruit juice, ice tea, and canned tomatoes. Classical detection microarray technology can be useful for detecting and methods for spoilage bacteria in various foods are based on analyzing specific microorganisms in foods. the results from enrichment, microscopic observations, and In this study, we developed a DNA chip to detect and biochemical assays. These procedures have several limita- identify Alicyclobacillus spp. A DNA chip containing tions, such as long culture time, dependency on enrichment randomly fragmented genomic DNA of Alicyclobacillus and selective culture, and difficulties in simultaneous spp. was fabricated and assessed for its ability to distinguish detection of spoilage bacteria. Alicyclobacillus spp. from other strains and other bacteria To overcome these problems, molecular-based methods such as Bacillus spp. such as real-time PCR and immunoassay have been developed. These methods offer specificity and high MATERIALS AND METHODS affinity, depending on target genes of specific spoilage Bacterial strains and isolation of genomic DNA. Alicyclo- bacteria. Although these molecular-based methods or bacillus acidocaldarius (ATCC 43030, 27009), Alicyclobacillus immunoassays can shorten the detection time, they also acidoterrestris (ATCC 49025, 49027), and Alicyclobacillus have several drawbacks. For example, real-time PCR cycloheptanicus (ATCC 49029, 49028) were obtained from the American Type Culture Collection (ATCC; Manassas, VA) and grown in ATCC medium 573, 1655, and 1656, respectively, * Author for correspondence. Tel: 82-2-3290-3421; according to the manufacturer’s instructions. Bacillus subtilis E-mail: [email protected]. ATCC 6633 and Bacillus cereus F3812 strains were also used. 934 JANG ET AL. J. Food Prot., Vol. 74, No. 6

FIGURE 1. Schematic of nonsequenced random genomic DNA probes for three different Alicyclobacillus species. All probes were spotted in duplicate. Downloaded from http://meridian.allenpress.com/jfp/article-pdf/74/6/933/1685326/0362-028x_jfp-10-418.pdf by guest on 28 September 2021

Pasteurized 100% pure orange juice was purchased from a genes were prepared by RT-PCR from AGS cancer cell RNA with local grocer. One liter of orange juice was filtered through a 0.45- primers 59-cattctggcccaccaactttgg-39 and 59-tggagatgcaggctc- mm-pore-size membrane filter under vacuum, after which 5 ml each cactttg-39 (Fig. 1). of A. acidocaldarius (ATCC 43030), A. acidoterrestris (ATCC 49025), and A. cycloheptanicus (ATCC 49029) (7, 10, 11) culture Genomic DNA labeling and hybridization. Genomic DNA broth (2 | 103 CFU/ml) were inoculated into 45-ml aliquots of the (50 ng) from all strains, including the reference strains, was labeled pasteurized orange juice and incubated at 47uC for 5 days. The with Cy5-dCTP (Amersham Bioscience, Uppsala, Sweden) for the cells were then recovered by centrifugation at 5,400 | g for reference signals and with Cy3-dCTP by random priming for 10 min. genomic DNA (High Prime DNA labeling kit, Roche, Mannheim, Genomic DNA from Alicyclobacillus spp. was extracted from Germany) (9). Mixtures of genomic DNA (50 ng) from the three 50 ml of culture (Flexible Kit, NucleoGen, Seoul, Korea). DNA reference strains (1:1) used for microarray fabrication were labeled concentration was determined using NanoDrop ND-1000 (Nano- with Cy3-dCTP (Amersham Bioscience) and used as reference Drop Technologies, Inc., Wilmington, DE) by calculating the ratio DNA for signal ratio calculations (Cy3-test/Cy5-reference). of absorbance (260 nm/280 nm) and agarose gel electrophoresis. Bacteriophage MJ1 (10 ng) was included in each labeling reaction The DNA was stored at 220uC until further use. as a positive control. The microarrays were prehybridized in buffer (1.7| SSC [1| SSC is 0.15 M NaCl plus 0.015 M sodium DNA microarray chip fabrication. Genomic DNA from citrate], 0.1% sodium dodecyl sulfate [SDS], 10 mg/ml bovine three different food spoilage bacteria (A. acidocaldarius ATCC serum) for 30 min at 65uC and then hybridized with approximately 43030, A. acidoterrestris ATCC 49025, and A. cycloheptanicus 50 ng of genomic DNA with Cy3-labeled (test)/Cy5-labeled ATCC 49029) was isolated and digested with several pairs of (reference) DNA mixture (1:1) in hybridization buffer (6| SSC, restriction enzymes, i.e., EcoRI/BamHI, HindIII/XhoI, HindIII/ 0.2% SDS, 5| Denhardt’s solution, 0.1 mg/ml denatured salmon SacII, and EcoRI/XhoI. The digested fragments were separated by sperm DNA) at 65uC overnight. Then the microarrays were washed agarose gel electrophoresis, after which bands ranging from 100 bp once with primary wash buffer 1 (2| SSC, 0.1% SDS) at room to 1.5 kb were excised from the gel. QIAquick gel extraction kits temperature to remove the cover glass, followed by 10 min with (QIAGEN, Valencia, CA) were used to elute and purify DNA from buffer 2 (2| SSC, 0.2% SDS prewarmed to 65uC), and finally 1% agarose gels. The genomic DNA fragments were inserted into a twice with wash buffer 3 (0.05| SSC) for 5 min; after this, the pPCR-Script Amp vector (Stratagene, La Jolla, CA) and then chips were dried by centrifugation at 2,200 | g at 4uC for 3 min. transformed into DH5a E. coli (RBC Bioscience, Taipei, Taiwan). All experiments were replicated at least three times. Clones were with amplified with a T3 to T7 promoter primer set. Amplified genomic DNA fragments were purified with a PCR Scanning and image analysis. After hybridization, all purification kit (QIAGEN) (4, 9). After purification, these genomic microarrays were scanned with a GenePix 4000B laser scanner DNA probes were sent to GenoCheck Co. (Gyeonggi, Korea), (Axon Instruments, Foster City, CA). The intensity values of where they were denatured and printed on amine-coated glass fluorescent signals from each spot were measured with GenePix slides in duplicate. In the DNA microarray chip assay, bacterio- Pro 6.0 software (Axon) and normalized by a correction factor phage integrase (MJ 1) gene (14) and peroxisome proliferator- calculated from an internal standard (c ~ Cy5-IS/Cy3-IS). activated receptor c gene (1) were used as positive and negative The intensity values of local background material were controls, respectively. The positive control gene was amplified subtracted from the normalized nonintensive values. To determine from WFC1 DNA with primers 59-cgggatccatgaaacgtgcagcattg-39 positive binding of target genomic DNA, the threshold values were and 59-cgggaattcaccgaatgcatgttcgta-39, and the negative control set up based on average intensity of the negative control and blank J. Food Prot., Vol. 74, No. 6 DETECTION OF ALICYCLOBACILLUS SPP. USING DNA CHIP 935 Downloaded from http://meridian.allenpress.com/jfp/article-pdf/74/6/933/1685326/0362-028x_jfp-10-418.pdf by guest on 28 September 2021

FIGURE 2. Microarray images for hybridization with genomic DNA from A. acidocaldarius ATCC 43030 (A), A. acidoterrestris ATCC 49025 (B), and A. cycloheptanicus ATCC 49029 (C). Hybridizations were performed with reference genomic DNA labeled with Cy5 and target genomic DNA labeled with Cy3. Each strain was detected only in target spots. The graph shows the intensity values based on Cy3 analysis (y axis). All experiments were replicated at least three times. regions (no probes deposited). The presence of target genomic these chips could identify other strains of A. acidocaldarius DNA in samples was determined from the average intensity of the ATCC 27009, A. acidoterrestris ATCC 49027, and A. target region. Finally, the ratios of median intensity values (test and cycloheptanicus ATCC 49028. Genomic DNA from A. reference) were log2-transformed. acidocaldarius ATCC 27009, A. acidoterrestris ATCC 49027, and A. cycloheptanicus ATCC 49028 was hybrid- RESULTS AND DISCUSSION ized with probes on chips. Detection of specific bacteria. In this study, a DNA All test strains of Alicyclobacillus spp. gave a positive microarray chip was fabricated for the identification of hybridization signal with the other strain of Alicyclobacillus specific bacterial strains using random fragmented probes spp. To determine whether the DNA chip could detect A. without any information about the sequence of the probes. To acidocaldarius ATCC 43030 and A. acidoterrestris ATCC verify that the DNA chip was properly fabricated, the genomic 49025, genomic DNA extracted from these two strains was DNA extracted from A. acidocaldarius ATCC 43030, A. hybridized with probes on the chip. Approximately 95% of acidocaldarius ATCC 27009, and A. cycloheptanicus ATCC the DNA probes on the chip gave a positive signal for 49029 was hybridized with probes on the chip, because those hybridization. However, 16S rRNA comparison analysis probes originated from the same strain. Results showed 100% found that A. acidocaldarius and A. acidoterrestris belong hybridization between genomic DNA of all three Alicycloba- to practically the same genus (98.8%) (12), which explains cillus species and the probe on the chip (Fig. 2). why cross-hybridization occurred between the two strains. This result indicated that the chip was properly To determine whether DNA probes on the chips could fabricated. Because the DNA chip was also fabricated using distinguish Alicyclobacillus spp. from other bacteria of randomly fragmented DNA originating from A. acidocal- different genera, genomic DNA of B. cereus F381284 and darius ATCC 43030, A. acidoterrestris ATCC 49025, and B. subtilis ATCC 6633 was prepared and hybridized with A. cycloheptanicus ATCC 49029, we tried to confirm that probes on the chips. Results indicated that less than 1% of 936 JANG ET AL. J. Food Prot., Vol. 74, No. 6 Downloaded from http://meridian.allenpress.com/jfp/article-pdf/74/6/933/1685326/0362-028x_jfp-10-418.pdf by guest on 28 September 2021

FIGURE 3. Specificity of the microarray assay using genomic DNA from other strains of Alicyclobacillus species and Bacillus species. The test was done with 50 ng of genomic DNA from A. acidocaldarius ATCC 27009 (A), A. acidoterrestris ATCC 49027 (B), A. cycloheptanicus ATCC 49028 (C), B. cereus F381284 (D), and B. subtilis ATCC 6633 (E). The microarray image shows that only Alicyclobacillus species were detected (A through C) and clearly distinguishes Alicyclobacillus species from Bacillus species (D and E). The graph shows the intensity values of fluorescent signal of Alicyclobacillus species and Bacillus species based on Cy3 analysis (y axis). All experiments were replicated at least three times. the genomic DNA from Bacillus spp. hybridized with DNA preparation protocols for detection and identification of probes of Alicyclobacillus spp. (Fig. 3), suggesting that the Alicyclobacillus spp. using this chip were optimized. DNA chip did not exhibit significant cross-hybridization According to our recent study, positive hybridizations with Bacillus spp. have typically been above 88% for the spotted probes, while averages for different strains of the same species were as Detection of Alicyclobacillus spp. in orange juice by high as 70% (2). Therefore, a positive result for Alicyclo- microarray. The DNA microarray chip was applied to food bacillus should yield a hybridization rate of at least 80%. products and evaluated. Three Alicyclobacillus spp. were Using nonspecific genomic DNA fragments from tested in pasteurized 100% pure orange juice, and genomic individual organisms, this research showed the possibility DNA was extracted. Genomic DNA from Alicyclobacillus of making a DNA microarray capable of detecting specific spp. (A. acidocaldarius, A. acidoterrestris, and A. cyclo- organisms without having sequence information for probe heptanicus) was tested in the DNA microarray using dye construction. The problem of potential nonspecific hybrid- analysis with Cy3 for target genomic DNA. ization that appeared using the existing 16s rRNA was As shown in Figure 4, the specificity tests for each resolved by using a combination of many random probes. target spoilage bacteria gave clear results, with meaningful The time for detection was reduced by using the intensities of Cy3 observed in regions of the chip where the hybridization reaction directly, with no amplification of reference probes of the target spoilage bacteria were placed. the DNA. To interpret the DNA chip information, two types Therefore, the DNA microarray assay afforded reliable and of samples (control and target) were marked with different clear detection. fluorescent dyes (Cy3, Cy5). Hybridization was conducted The ability of this DNA chip to detect and identify with the DNA chip, and the molar ratio was determined. The Alicyclobacillus spp. in food samples was also investigated. basic principle of DNA microarray assays is hydrogen The DNA microarray chip showed high specificity to the bonding between bases (in DNA, the double hydrogen bond target food spoilage bacteria. Also, real food sample of A and T bases and the triple hydrogen bond of C-G; in J. Food Prot., Vol. 74, No. 6 DETECTION OF ALICYCLOBACILLUS SPP. USING DNA CHIP 937 Downloaded from http://meridian.allenpress.com/jfp/article-pdf/74/6/933/1685326/0362-028x_jfp-10-418.pdf by guest on 28 September 2021

FIGURE 4. Detection of genomic DNA extracted from Alicyclobacillus species in orange juice samples by microarray images (A through C). The microarray images with A. acidocaldarius ATCC 43030 (A), A. acidoterrestris ATCC 49025 (B), and A. cycloheptanicus ATCC 49029 (C) were detected only in target spots. The graph shows the intensity values of the fluorescent signal for Alicyclobacillus species inoculated into orange juice samples based on Cy3 analysis (y axis). All experiments were replicated at least three times.

RNA, A-U and G-C). Based on the hydrogen bond between REFERENCES bases, when an unknown gene was hybridized with probes 1. Andersen, O., V. G. H. Eijsink, and M. Thomassen. 2000. Multiple on the chip, the gene could be determined by the degree of variants of the peroxisome proliferator-activated receptor (PPAR) c hybridization. Moreover, the analysis of thousands of genes are expressed in the liver of Atlantic salmon (Salmo salar). Gene 255: rapidly and simultaneously in large quantities was achiev- 411–418. able. This technique, which could be used to discriminate 2. Bang, J., L. R. Beuchat, M. B. Gu, H. I. Chang, and J. H. Ryu. 2010. Identification of Yersinia enterocolitica using a random genomic within the genus level of bacteria, was applied to several DNA microarray chip. Lett. Appl. Microbiol. 51: 665–670. strains of spoilage bacteria that are known to cause a 3. Chang, S. S., and D. H. Kang. 2004. Alicyclobacillus spp. in the fruit decrease in food quality. juice industry: history, characteristics and current isolation/detection In conclusion, this study supports the use of genomic procedures. Crit. Rev. Microbiol. 30:55–74. 4. Cho, J. C., and J. M. Tiedje. 2001. Bacterial species determination DNA hybridization with DNA microarrays for the identi- from DNA-DNA hybridization by using genome fragments and DNA fication of bacteria, as tested on three reference strains of microarray. Appl. Environ. Microbiol. 67:3677–3682. Alicyclobacillus spp. The DNA chip developed in this study 5. Connor, C. J., H. Luo, B. B. Gardener, and H. H. Wang. 2005. can be used to determine the existence of specific food Development of a real-time PCR-based system targeting the decomposition bacteria prior to distribution, by predicting 16SrRNA gene sequence for rapid detection of Alicyclobacillus spp. in juice products. Int. J. Food Microbiol. 99:229–235. the period of circulation and detecting spoilage bacteria in 6. Darland, G., and T. D. Brock. 1971. Bacillus acidocaldarius sp. nov., an advance. If applied during food processing, this DNA chip acidophilic thermophilic spore-forming bacterium. J. Gen. Microbiol. could reduce product losses caused by spoilage bacteria. 67:9–15. 7. Gouws, P. A., L. Gie, A. Pretorius, and N. Dhansay. 2005. Isolation and identification of Alicyclobacillus acidocaldarius by 16S rDNA ACKNOWLEDGMENT from mango juice and concentrate drinks. Int. J. Food Sci. Technol. 40:789–792. This work was supported by a grant (code # 20080401034020) from 8. Groenewald, W. H., P. A. Gouws, and R. C. Witthuhn. 2009. Isolation, the BioGreen 21 Program. identification and typification of Alicyclobacillus acidoterrestris and 938 JANG ET AL. J. Food Prot., Vol. 74, No. 6

Alicyclobacillus acidocaldarius strains from orchard soil and the fruit 12. Wisotzkey, J. D., P. Jurtshuk, G. E. Fox, G. Deinhard, and K. Poralla. processing environment in South Africa. Food Microbiol. 26:71–76. 1992. Comparative sequence analyses on the 16S rRNA (rDNA) of 9. Kim, B. C., J. H. Park, and M. B. Gu. 2004. Development of a DNA Bacillus acidocaldarius, Bacillus acidoterrestris, and Bacillus microarray chip for the identification of sludge bacteria using an cycloheptanicus and proposal for creation of a new genus, unsequenced random genomic DNA hybridization method. Environ. Alicyclobacillus gen. nov. Int. J. Syst. Bacteriol. 42:263–269. Sci. Technol. 38:6767–6774. 13. Yamazaki, K., H. Teduka, N. Inoue, and H. Shinano. 1996. Specific 10. Pettipher, G. L., M. E. Osmundson, and J. M. Murphy. 1997. Methods primers for detection of Alicyclobacillus acidoterrestris by RT-PCR. for the detection and enumeration of Alicyclobacillus acidoterrestris Lett. Appl. Microbiol. 23:350–354. and investigation of growth and production of taint in fruit juice and 14. Yang, H. Y., Y. W. Kim, and H. I. Chang. 2002. Construction of an fruit juice-containing. Lett. Appl. Microbiol. 24:185–189. integration-proficient vector based on the site-specific recombination 11. Walls, I., and R. Chuyate. 2000. Isolation of Alicyclobacillus mechanism of enterococcal temperate phage FC1. J. Bacteriol. 184: acidoterrestris from fruit juices. J. AOAC Int. 83:1115–1120. 1859–1864. Downloaded from http://meridian.allenpress.com/jfp/article-pdf/74/6/933/1685326/0362-028x_jfp-10-418.pdf by guest on 28 September 2021