- Email: [email protected]
A bioluminescent Escherichia coli auxotroph for use in an in vitro lysine availability assay
Journal of Microbiological Methods 40 (2000) 207–212
Journal of Microbiological Methods www.elsevier.com / locate / jmicmeth
A bioluminescent Escherichia coli auxotroph for use in an in vitro lysine availability assay ´ a , Y.M. Kwon a , S.C. Ricke a , * A.M. Erickson a ,b , I.B. Zabala Dıaz a
Poultry Science Department, Texas A and M University, College Station, TX 77843 -2472, USA b Quik to Fix Foods, 113 Range Drive, Garland, TX 75243, USA
Received 22 August 1999; received in revised form 23 November 1999; accepted 4 December 1999
Abstract Microbiological methods have been used to determine the amino acid availability of a variety of animal feed and human food protein sources. Growth of Escherichia coli auxotrophs have been shown to yield a consistent linear response to lysine concentration when compared to chemical measures. Extent of total growth of E. coli lysine mutant (American Type Culture Collection [23812) when measured as optical density (OD) displays a lysine-dependent growth response that can be used to estimate lysine in feed proteins. However, typical OD-based growth studies for amino acid quantitation using the mutant may require anywhere from 12 to over 40 h. To develop an improved rapid method for lysine quantitation in protein sources, the plasmid pJHD500 carrying genes that encode for expression of bioluminescence and ampicillin resistance was transformed into the E. coli mutant by electroporation (set at 1.80 kV). The luminescence measured during early exponential growth allowed detectable differentiation of lysine concentration in the media in 4 h. When the luminescence method was compared with the conventional optical density lysine growth assay, the correlation coefficient was 0.989. Lysine availability valued for enzymatically hydrolyzed protein sources were comparable with availability measures using animal methods for lysine availability. This research shows potential applications for more rapid quantitative measurement of bioavailable lysine. 2000 Elsevier Science B.V. All rights reserved. Keywords: Lysine availability assay; Auxotroph; Bioluminescence
1. Introduction Assaying the available amino acids in protein sources for animal nutrition is critically important due to the variable quality of the protein sources. Animal by-products are becoming a primary source for protein in feeds as the cost of higher quality protein sources increases. These sources are incon*Corresponding author. Tel.: 1 1-979-862-1528; fax: 1 1-979845-1921. E-mail address: [email protected] (S.C. Ricke)
sistent in both composition and processing treatment which leads to variable nutrient digestibility and availability (Hardy, 1991). The essential amino acid lysine is of importance due to it being the first or second limiting amino acid in poultry and swine nutrition and its sensitivity to heat and processing treatments (Nordheim and Coon, 1984). Chemical methods such as high-performance liquid chromatography (HPLC) can be used to determine the lysine content of a protein source yet they cannot provide a measure of the available lysine to the animal. Animal bioassays utilizing chicks, roosters or rats
0167-7012 / 00 / $ – see front matter 2000 Elsevier Science B.V. All rights reserved. PII: S0167-7012( 00 )00121-4
208
A.M. Erickson et al. / Journal of Microbiological Methods 40 (2000) 207 – 212
serve as the standard to provide a nutritional evaluation of the protein source. Growth assays provide information such as protein efficiency ratio, gain to feed ratio and amino acid availability. These assays can take from 2 to 4 weeks and require special facilities and large amounts of raw materials. Precision fed rooster assays (Sibbald, 1979) provide information concerning the digestibility of a protein source, which is used as an indicator of availability but still require 2 days to complete and may require specialized surgery. Microbiological assays for amino acid availability are an attractive alternative to the labor intensive, time-consuming and variable nature of animal bioassays. Historically they have been promoted as a more rapid and cost effective alternative assay for assessing vitamin and amino acid contents of test materials (Shockman, 1963). A variety of microorganisms have been examined, but auxotrophic mutants of Escherichia coli have proven to be a reliable assay for quantitative measurement of amino acid availability (Payne et al., 1977; Anantharaman et al., 1983; Tuffnell and Payne, 1985). E. coli assays for lysine have shown correlation values greater than 0.9 with chemical methods for available lysine (Anantharaman et al., 1983) and have been shown to be good predictors of animal values for lysine availability for a variety of protein sources (Tuffnell and Payne, 1985). These assays take between 4 to over 40 h for the quantification of lysine depending upon the method used to measure growth response. Additionally, animal feed protein sources support a bacterial population ranging from 10 2 –10 6 colony forming units (CFU) / g and a fungal population ranging from 10 1 –10 4 cfu / g (Tabib et al., 1981; Ha et al., 1995). This microbial background contribution cannot only alter microbial growth assay results (Erickson et al., 1999a,b) but any attempt to heat sterilize feeds to prevent background microbial growth can lower lysine bioavailability (Evans and Butts, 1949; Boctor and Harper, 1968; Bjarnason and Carpenter, 1970; Carpenter and Booth, 1973; Angkanaporn et al., 1997; Johnson et al., 1998). Bioluminescence-based measurements of bacterial responses is an emerging technology that allows for rapid, non-invasive detection of microbial populations and can be used as a measure of cellular
viability (Burlage, 1997; Hill and Stewart, 1994). Naturally bioluminescent microorganisms include the marine-dwelling bacteria Vibrio fischeri and V. harveyi. Genes responsible for expression of bioluminescence can be genetically transferred from these marine bacteria to species not naturally bioluminescent. E. coli, Salmonella typhimurium, Staphylococcus aureus, Bacillus cereus and B. subtilis have been transformed into a bioluminescent phenotype for applications in food safety, pollution detection, bioavailability measurements of metals and for the study of bioluminescence and its expression (Baker et al., 1992; Selifonova et al., 1993; Duffy et al., 1995; Tauriainen et al., 1997). Advantages of bioluminescent applications include ease and rapidity of light measurement, increased sensitivity (as few as 500 cells can be detected), and in most environments bioluminescence is sufficiently rare among indigenous microbial populations, such that background microbial contribution is no longer a problem (Baker et al., 1992; Burlage, 1997). The objective of this research was to genetically transform a lysine auxotrophic strain of E. coli with bioluminescent genes for a lysine assay application. The modified E. coli was used to construct a lysinedependent standard curve that was compared to traditional microbiological assays using optical density (OD). The availability of lysine in enzymatically hydrolyzed protein sources was estimated and compared to reference values for lysine availability in vivo.
2. Materials and methods
2.1. Bacterial strain and plasmid An Escherichia coli lysine auxotroph was obtained from American Type Culture Collection (ATCC [23812 Rockville, MD, USA). Bacterial plasmid pJHD500 (Baldwin et al., 1989) with ampicillin and carbenicillin resistance and bioluminescent genes was transformed into the auxotroph by electroporation, set at 1.8 kV (Erickson et al., 1999b). Successful plasmid transfer was confirmed through plating on Luria Bertani (LB) plates with carbenicillin (100 mg / ml) and through gel electrophoresis.
A.M. Erickson et al. / Journal of Microbiological Methods 40 (2000) 207 – 212
Cultures were maintained on LB plates supplemented with carbenicillin (100 mg / ml) and stored at 48C.
2.2. Cultivation of bacteria Single colonies of bacteria not more than 24 h old were selected from LB plates containing carbenicillin (100 mg / ml). Bacteria were cultivated in 50-ml flasks of LB broth supplemented with ampicillin (100 mg / ml) at 378C in a shaking water bath. One ml of the 8-h culture (OD 5 0.05) was transferred to 50 ml of a modified M9 medium (Erickson et al., 1999b) without lysine to deplete the cells of endogenous lysine stores. This method has shown to successfully deplete the cells of lysine as indicated by a second transfer into M9 medium without lysine not supporting growth of the auxotroph (Erickson et al., 1999b).
2.3. Test media Following the depletion, 5 ml of test M9 media supplemented with L-lysine (Sigma Chemical, St. Louis, MO, USA) or protein hydrolysates was inoculated with the culture (150 ml) to reach an initial OD of 0.02. For measurement of the available lysine in the sample protein sources, M9 medium was supplemented with enzymatic hydrolysates of casein and poultry by-product meal. The proteins were enzymatically digested with protease (1.0 mg / ml, pH 6.8, Sigma Chemical, St. Louis, MO, USA) for 1 h for casein and 10 h for poultry by-product meal. Previous studies in our laboratory indicated that the 1-h time increment for casein and 10-h time increment for poultry by-product meal were sufficient incubation times for hydrolysis (Erickson et al., 1999a). Under these incubation conditions, the lysine availability of the enzyme hydrolysate was measured through the growth response of the E. coli auxotroph and compared to previously published animal assay values for availability. The lysine content of the protein sources was determined using the standard curve generated with the L-lysine-supplemented test M9 medium. The availability of each protein was calculated as the standard curve generated content divided by an HPLC or reference value for the lysine concentration of that protein source.
209
2.4. Luminescence measurements The test M9 medium was vortexed and 125 ml of the culture was pipetted to wells of microtiter plates containing 100 nM of autoinducer (n-3-oxohexanoylhomoserine lactone, Sigma). Luminescence was measured with a Dynex MLX Luminometer (Chantilly, VA, USA) after 4 h of incubation with shaking at 288C (Burlage, 1997; Thomas et al., 1997). Prior to reading, 125 ml of aldehyde was added to each well, and the wells were read for an 8-s interval with a 4-s delay. Readings were taken on duplicate wells and light measurements were reported in relative light units (RLU).
3. Results and discussion Successful plasmid transfer was confirmed through plating on LB plates with carbenicillin and through gel electrophoresis. In order to obtain consistent luminescent measurements only colonies less than 24-h old were selected from plates. The temperature was maintained at 288C due to the luciferase enzyme being heat labile above 308C (Burlage, 1997; Thomas et al., 1997). The culture was cultivated in microtiter plates with agitation to provide the aeration necessary for the biochemical reaction. Cultures were grown for 4 h to allow luciferase that may have accumulated from the overnight culture to dissipate in order to obtain a distinct response from the control. Initially the lysine concentration in the M9 medium ranged from 0.05 to 2.5 mg / ml. The lower portion of the concentration range produced a linear response to lysine with the RLU response to lysine reaching a plateau at approximately 0.5 mg / ml (Fig. 1). Due to higher lysine concentrations being above the sensitivity range of the assay, focus was centered on the 0.05 to 0.5 mg / ml lysine range using 0.05–0.1 mg / ml increments. This range produced a linear RLU response to the lysine concentration in the media with a standard curve equation of y 5 1137.2x 1 132.24, R 2 5 0.989 ( y 5 RLU, x 5 lysine concentration, Fig. 1). Low levels of light were produced in the medium containing no lysine which indicates that some cells may survive the depletion
210
A.M. Erickson et al. / Journal of Microbiological Methods 40 (2000) 207 – 212
Fig. 1. Luminescence (RLU) response of Escherichia coli lysine auxotroph to lysine-supplemented M9 medium. The standard error of each data point is based upon 34, 6, 28, 28, 8, 30, 28, 6 and 34 observations for the lysine concentrations 0.0, 0.05, 0.1, 0.2, 0.25, 0.3,0.4, 0.45 and 0.5 mg / ml respectively.
stage, yet not proliferate to the extent of the cells in the lysine-supplemented medium. Luminescent methods are a determinant of cellular metabolism as light production is dependent upon the metabolic functioning of the cells (Hill and Stewart, 1994). A higher light output (RLU value) corresponds to a larger number of cells. Substances in the medium either promoting or impairing growth will induce a concomitant change in bioluminescence. Lysine stimulates the extent of growth of the auxotroph thus variations in lysine concentration will result in a variable and quantitative growth response (Payne et al., 1977; Tuffnell and Payne, 1985; Hitchins et al., 1989). Consequently, the luminescence response to the varying concentrations of lysine can be quantified, and the response is linear due to growth of the assay organism being directly responsive to incremental increases in the lysine concentration of the medium. In conventional E. coli assays for lysine, the growth response to lysine is measured as a function of increases in optical density measurement (absorbance, A 600 ) on a spectrophotometer. When the RLU response is directly compared with the conventional method, the correlation coefficient of the methods is 0.989, P , 0.01 (Fig. 2). The conventional assay for measuring optical density requires that cell cultures be cultivated until they reach peak growth. This typically takes anywhere from 12 to 26 h. However, the luminescent assay is more sensitive and can detect changes in cellular density through light output in 4 h. By using the luminescent assay there is a ten-fold increase in the sensitivity of the
Fig. 2. Comparison of MOD response with luminescence response of an Escherichia coli lysine auxotroph to lysine-supplemented M9 medium. MOD: maximum optical density measured at 600 nm (absorbance 600 ). RLU: relative light units measured with a Dynex luminometer.
assay. The lower detection limit for optical density assays is 0.5 mg / ml while the lysine concentration of 0.05 mg / ml is detected using the luminescent method. E. coli is non-proteolytic therefore proteins must be enzymatically predigested before they can be absorbed by E. coli. Enzyme hydrolysis of proteins reflects digestion of a protein in vivo (Szmelcman and Guggenheim, 1967). Differences in the enzyme activity, specificity, and total time of digestion will cause a variety of peptides to be produced from a protein source (Longland, 1991). Protease was selected as the digestion enzyme as it has been utilized in studies to hydrolyze protein sources for utilization by E. coli (Carpenter and Booth, 1973; Payne et al., 1977; Hitchins et al., 1989). Compared to mammalian proteases such as pepsin, trypsin and chymotrypsin, protease is a bacterial enzyme that has the ability to hydrolyze the proteins into units small enough to be utilized by the bacteria (Payne et al., 1977). By using the lysine standard curve equation the lysine content of the digests was estimated using the RLU response of the luminescent E. coli to the digests. When compared to previously published animal reference values using precision fed rooster assays, the availability determined by luminescence of casein was higher, and the availability of poultry by-product meal was slightly lower than the reference values (Fig. 3; Izquierdo et al., 1988; Johnson et al., 1998). The high availability value of the casein could be a result of error in the HPLC generated
A.M. Erickson et al. / Journal of Microbiological Methods 40 (2000) 207 – 212
211
range being ten times lower than with traditional assays. Future studies should include a wider range of protein sources as well as improved digestion techniques. This may encompass using a multienzyme digest and alterations in the digestion duration in order to produce a digest that better correlates with animal digestion in vivo. This would allow the assay to serve as a potential alternative to animal methods for quality evaluation of a wide variety of animal feed protein sources. Fig. 3. Comparison of luminescence and animal assay methods for determination of lysine availability. Availability values determined by luminescence readings were derived using the lysinedependent luminescence standard curve. Standard error bars represent 23 determinations for casein and 22 determinations for poultry by-product meal. Availability values determined by animal assay was obtained from reference values of precision fed roosters (Izquierdo et al., 1988; Johnson et al., 1998).
value or possibly a growth-promoting factor in the casein digest resulting in over-estimation of availability values. The lower availability of the poultry by-product meal could be a result of batch to batch differences in poultry by-product meal or incomplete enzymatic digestion by protease. The similarity in animal values and luminescence values for availability suggests that a luminescent approach is a feasible alternative to animal assays. In theory, protein sources could be ranked according to the lysine availability determined by the luminescent assay and be supplemented with purified lysine or protein sources high in lysine if needed to produce a diet that is nutritionally complete. The luminescent method presented provides an alternative to traditional microbiological methods for lysine assays. The method highly correlates with and offers advantages over traditional methods of bacterial analytical growth assays. By using a luminescent method, the problem of background microflora in the protein sources can be eliminated. This assay measures the light output of the E. coli assay organism. Indigenous microflora will likely not possess luminescence abilities and not interfere with the luminescence-based lysine quantitation (Burlage, 1997). Assay time can be reduced using this method, as incubation time is reduced from 12 to over 40 h to a standard 4-h duration. Sensitivity is increased with the luminescent method due to the concentration
Acknowledgements This work was supported with funds from the IAMS Company, Lewisberg, Ohio and by Hatch grant H8311 administered by the Texas Agricultural Experiment Station, Texas A and M University, College Station, Texas, USA. A.M. Erickson was supported by a Regent’s Fellowship provided by ´ Texas A and M University. I.B. Zabala-Dıaz is supported by a LUZ-CONICIT Graduate Fellowship (Maracaibo, Venezuela) and Texas Good Neighbor Scholarship (Texas Higher Education Board, Austin, TX, USA). The authors wish to thank Dr. T.O. Baldwin (Department of Biochemistry and Biophysics, Texas A and M University) for the plasmid and helpful discussion, Dynex Corporation (Chantilly, VA, USA) for use of the luminometer, and A.K. Reinert for preparing the manuscript.
References Anantharaman, K., Parard, C., Decarli, B., Reinhardt, P., Graf, E., Finot, P.A., 1983. The microbiological estimation of available lysine in industrial milk products using a lysine Escherichia coli. m. 26–26. In: McLoughlin, J.V., McKenna), B.M. (Eds.), Basic Studies in Food Science, Research in Food Science and Nutrition, Vol. 2, Boole Press Ltd, Dublin, Ireland, pp. 81–82. Angkanaporn, K., Ravindan, V., Bryden, W.L., 1997. Influence of caecectomy and dietary protein concentration on apparent excreta amino acid digestibility in adult cockerels. Br. Poult. Sci. 38, 270–276. Baker, J.M., Griffiths, M.W., Collins-Thompson, D.L., 1992. Bacterial bioluminescence: applications in food microbiology. J. Food Prot. 55, 62–70. Baldwin, T.O., Devine, J.H., Heckel, R.C., Lin, J.W., Shadel, G.S., 1989. The complete nucleotide sequence of the lux regulon of Vibrio fischeri and the luxABN region of Photobacterium
212
A.M. Erickson et al. / Journal of Microbiological Methods 40 (2000) 207 – 212
leiognathi and the mechanism of control of bacterial bioluminescence. J. Biolum. Chemilum. 4, 326–341. Bjarnason, J., Carpenter, K.J., 1970. Mechanism of heat damage in proteins. 2. Chemical changes in pure proteins. Br. J. Nutr. 24, 313–329. Boctor, A.M., Harper, A.E., 1968. Measurement of available lysine in heated and unheated foodstuffs by chemical and biological methods. J. Nutr. 94, 289–296. Burlage, R.S., 1997. Emerging technologies: bioreporters, biosensors and microprobes. In: Hurst, C.J., Knudsen, G.R., McInerney, M.J., Stetzenbach, L.D., Walter, M.V. (Eds.), Manual of Environmental Microbiology, American Society for Microbiology, Washington, DC, pp. 115–123. Carpenter, K.J., Booth, V.H., 1973. Damage to lysine in foods processing: its measurement and significance. Nutr. Abstr. Rev. 43, 423–451. Duffy, G., Ellison, A., Anderson, W., Cole, M.B., Stewart, G.S.A.B., 1995. Use of bioluminescence to model the thermal inactivation of Salmonella typhimurium in the presence of a competitive microflora. Appl. Env. Microbiol. 61, 3463–3465. Erickson, A.M., Li, X., Woodward, C.L., Ricke, S.C., 1999a. Optimisation of enzyme treatment for the degradation of feed proteins for an Escherichia coli auxotroph lysine availability assay. J. Sci. Food Agric. 79, 1929–1935. Erickson, A.M., Zabala Diaz, I.B., Ricke, S.C., 1999b. Antibiotic amendment for suppression of indigenous microflora in feed sources for an Escherichia coli auxotroph lysine assay. J. Appl. Microbiol. 87, 125–130. Evans, R.J., Butts, H.A., 1949. Inactivation of amino acids by autoclaving. Science 109, 569–571. Ha, S.D., Pillai, S.D., Maciorowski, K.G., Ricke, S.C., 1995. Cycloheximide as a media amendment for enumerating bacterial populations in animal feeds. J. Rapid Meth. Automation Microbiol. 4, 95–105. Hardy, B., 1991. The future requirements of the feed industry. In: Fuller, M.F. (Ed.), In Vitro Digestion for Pigs and Poultry, CAB International, Wallingford, UK, pp. 181–192. Hill, P.J., Stewart, G.S.A.B., 1994. Use of lux genes in applied biochemistry. J. Biolumin. Chemilumin. 9, 211–215. Hitchins, A.D., McDonough, F.E., Wells, P.A., 1989. The use of Escherichia coli mutants to measure the bioavailability of essential amino acids in foods. Plant Food Hum. Nutr. 39, 109–120.
Izquierdo, O.A., Parsons, C.M., Baker, D.H., 1988. Bioavailability of L-lysine in L-lysine HCl. J. Anim. Sci. 66, 2590–2597. Johnson, M.L., Parsons, C.M., Fahey, Jr. G.C., Merchen, N.R., Aldrich, C.G., 1998. Effect of species raw material source, ash content, and processing temperature on amino acid digestibility of animal by-product meals by cecectomized roosters and ileally cannulated dogs. J. Anim. Sci. 76, 1112–1122. Longland, A.C., 1991. Digestive enzyme activities in pigs and poultry. In: Fuller, M.F. (Ed.), In Vitro Digestion for Pigs and Poultry, CAB International, Wallingford, UK, pp. 3–18. Nordheim, J.P., Coon, C.N., 1984. A comparison of four methods for determining available lysine in animal protein meals. Poult. Sci. 63, 1040–1051. Payne, J.W., Bell, G., Higgins, C.F., 1977. The use of an Escherichia coli lys auxotroph to assay nutritionally available lysine in biological materials. J. Appl. Bact. 42, 165–177. Selifonova, O., Burlage, R., Barkay, T., 1993. Bioluminescent sensors for detection of bioavailable Hg(II) in the environment. Appl. Env. Microbiol. 59, 3083–3090. Shockman, G.D., 1963. Amino acids. In: Kavanagh), F. (Ed.), Analytical Microbiology, Academic Press Inc, New York, pp. 567–673. Sibbald, I.P., 1979. A bioassay for available amino acids and true metabolizable energy in feeding stuffs. Poult. Sci. 58, 668– 673. Szmelcman, S., Guggenheim, K., 1967. Availability of amino acids in processed plant protein foodstuffs. J. Sci. Food Agric. 18, 347–350. Tabib, Z., Jones, F.T., Hamilton, P.B., 1981. Microbiological quality of protein feed and ingredients. Poult. Sci. 60, 1392– 1397. Tauriainen, S., Karp, M., Chang, W., Virta, M., 1997. Recombinant luminescent bacteria for measuring bioavailable arsenite and antimonite. Appl. Env. Microbiol. 63, 4456–4461. Thomas, M.D., Schineller, J.B., Baldwin, T.O., 1997. lux autoinducer response in E. coli and Vibrio fischeri. In: Hastings, J.W., Kricka, L.J., Stanley, P.E. (Eds.), Bioluminescence and Chemiluminescence: Molecular Reporting with Photons, John Wiley and Sons, New York, pp. 109–112. Tuffnell, J.M., Payne, J.W., 1985. A colorimetric enzyme assay using Escherichia coli to determine nutritionally available lysine in biological materials. J. Appl. Bact. 58, 333–341.
Journal of Microbiological Methods www.elsevier.com / locate / jmicmeth
A bioluminescent Escherichia coli auxotroph for use in an in vitro lysine availability assay ´ a , Y.M. Kwon a , S.C. Ricke a , * A.M. Erickson a ,b , I.B. Zabala Dıaz a
Poultry Science Department, Texas A and M University, College Station, TX 77843 -2472, USA b Quik to Fix Foods, 113 Range Drive, Garland, TX 75243, USA
Received 22 August 1999; received in revised form 23 November 1999; accepted 4 December 1999
Abstract Microbiological methods have been used to determine the amino acid availability of a variety of animal feed and human food protein sources. Growth of Escherichia coli auxotrophs have been shown to yield a consistent linear response to lysine concentration when compared to chemical measures. Extent of total growth of E. coli lysine mutant (American Type Culture Collection [23812) when measured as optical density (OD) displays a lysine-dependent growth response that can be used to estimate lysine in feed proteins. However, typical OD-based growth studies for amino acid quantitation using the mutant may require anywhere from 12 to over 40 h. To develop an improved rapid method for lysine quantitation in protein sources, the plasmid pJHD500 carrying genes that encode for expression of bioluminescence and ampicillin resistance was transformed into the E. coli mutant by electroporation (set at 1.80 kV). The luminescence measured during early exponential growth allowed detectable differentiation of lysine concentration in the media in 4 h. When the luminescence method was compared with the conventional optical density lysine growth assay, the correlation coefficient was 0.989. Lysine availability valued for enzymatically hydrolyzed protein sources were comparable with availability measures using animal methods for lysine availability. This research shows potential applications for more rapid quantitative measurement of bioavailable lysine. 2000 Elsevier Science B.V. All rights reserved. Keywords: Lysine availability assay; Auxotroph; Bioluminescence
1. Introduction Assaying the available amino acids in protein sources for animal nutrition is critically important due to the variable quality of the protein sources. Animal by-products are becoming a primary source for protein in feeds as the cost of higher quality protein sources increases. These sources are incon*Corresponding author. Tel.: 1 1-979-862-1528; fax: 1 1-979845-1921. E-mail address: [email protected] (S.C. Ricke)
sistent in both composition and processing treatment which leads to variable nutrient digestibility and availability (Hardy, 1991). The essential amino acid lysine is of importance due to it being the first or second limiting amino acid in poultry and swine nutrition and its sensitivity to heat and processing treatments (Nordheim and Coon, 1984). Chemical methods such as high-performance liquid chromatography (HPLC) can be used to determine the lysine content of a protein source yet they cannot provide a measure of the available lysine to the animal. Animal bioassays utilizing chicks, roosters or rats
0167-7012 / 00 / $ – see front matter 2000 Elsevier Science B.V. All rights reserved. PII: S0167-7012( 00 )00121-4
208
A.M. Erickson et al. / Journal of Microbiological Methods 40 (2000) 207 – 212
serve as the standard to provide a nutritional evaluation of the protein source. Growth assays provide information such as protein efficiency ratio, gain to feed ratio and amino acid availability. These assays can take from 2 to 4 weeks and require special facilities and large amounts of raw materials. Precision fed rooster assays (Sibbald, 1979) provide information concerning the digestibility of a protein source, which is used as an indicator of availability but still require 2 days to complete and may require specialized surgery. Microbiological assays for amino acid availability are an attractive alternative to the labor intensive, time-consuming and variable nature of animal bioassays. Historically they have been promoted as a more rapid and cost effective alternative assay for assessing vitamin and amino acid contents of test materials (Shockman, 1963). A variety of microorganisms have been examined, but auxotrophic mutants of Escherichia coli have proven to be a reliable assay for quantitative measurement of amino acid availability (Payne et al., 1977; Anantharaman et al., 1983; Tuffnell and Payne, 1985). E. coli assays for lysine have shown correlation values greater than 0.9 with chemical methods for available lysine (Anantharaman et al., 1983) and have been shown to be good predictors of animal values for lysine availability for a variety of protein sources (Tuffnell and Payne, 1985). These assays take between 4 to over 40 h for the quantification of lysine depending upon the method used to measure growth response. Additionally, animal feed protein sources support a bacterial population ranging from 10 2 –10 6 colony forming units (CFU) / g and a fungal population ranging from 10 1 –10 4 cfu / g (Tabib et al., 1981; Ha et al., 1995). This microbial background contribution cannot only alter microbial growth assay results (Erickson et al., 1999a,b) but any attempt to heat sterilize feeds to prevent background microbial growth can lower lysine bioavailability (Evans and Butts, 1949; Boctor and Harper, 1968; Bjarnason and Carpenter, 1970; Carpenter and Booth, 1973; Angkanaporn et al., 1997; Johnson et al., 1998). Bioluminescence-based measurements of bacterial responses is an emerging technology that allows for rapid, non-invasive detection of microbial populations and can be used as a measure of cellular
viability (Burlage, 1997; Hill and Stewart, 1994). Naturally bioluminescent microorganisms include the marine-dwelling bacteria Vibrio fischeri and V. harveyi. Genes responsible for expression of bioluminescence can be genetically transferred from these marine bacteria to species not naturally bioluminescent. E. coli, Salmonella typhimurium, Staphylococcus aureus, Bacillus cereus and B. subtilis have been transformed into a bioluminescent phenotype for applications in food safety, pollution detection, bioavailability measurements of metals and for the study of bioluminescence and its expression (Baker et al., 1992; Selifonova et al., 1993; Duffy et al., 1995; Tauriainen et al., 1997). Advantages of bioluminescent applications include ease and rapidity of light measurement, increased sensitivity (as few as 500 cells can be detected), and in most environments bioluminescence is sufficiently rare among indigenous microbial populations, such that background microbial contribution is no longer a problem (Baker et al., 1992; Burlage, 1997). The objective of this research was to genetically transform a lysine auxotrophic strain of E. coli with bioluminescent genes for a lysine assay application. The modified E. coli was used to construct a lysinedependent standard curve that was compared to traditional microbiological assays using optical density (OD). The availability of lysine in enzymatically hydrolyzed protein sources was estimated and compared to reference values for lysine availability in vivo.
2. Materials and methods
2.1. Bacterial strain and plasmid An Escherichia coli lysine auxotroph was obtained from American Type Culture Collection (ATCC [23812 Rockville, MD, USA). Bacterial plasmid pJHD500 (Baldwin et al., 1989) with ampicillin and carbenicillin resistance and bioluminescent genes was transformed into the auxotroph by electroporation, set at 1.8 kV (Erickson et al., 1999b). Successful plasmid transfer was confirmed through plating on Luria Bertani (LB) plates with carbenicillin (100 mg / ml) and through gel electrophoresis.
A.M. Erickson et al. / Journal of Microbiological Methods 40 (2000) 207 – 212
Cultures were maintained on LB plates supplemented with carbenicillin (100 mg / ml) and stored at 48C.
2.2. Cultivation of bacteria Single colonies of bacteria not more than 24 h old were selected from LB plates containing carbenicillin (100 mg / ml). Bacteria were cultivated in 50-ml flasks of LB broth supplemented with ampicillin (100 mg / ml) at 378C in a shaking water bath. One ml of the 8-h culture (OD 5 0.05) was transferred to 50 ml of a modified M9 medium (Erickson et al., 1999b) without lysine to deplete the cells of endogenous lysine stores. This method has shown to successfully deplete the cells of lysine as indicated by a second transfer into M9 medium without lysine not supporting growth of the auxotroph (Erickson et al., 1999b).
2.3. Test media Following the depletion, 5 ml of test M9 media supplemented with L-lysine (Sigma Chemical, St. Louis, MO, USA) or protein hydrolysates was inoculated with the culture (150 ml) to reach an initial OD of 0.02. For measurement of the available lysine in the sample protein sources, M9 medium was supplemented with enzymatic hydrolysates of casein and poultry by-product meal. The proteins were enzymatically digested with protease (1.0 mg / ml, pH 6.8, Sigma Chemical, St. Louis, MO, USA) for 1 h for casein and 10 h for poultry by-product meal. Previous studies in our laboratory indicated that the 1-h time increment for casein and 10-h time increment for poultry by-product meal were sufficient incubation times for hydrolysis (Erickson et al., 1999a). Under these incubation conditions, the lysine availability of the enzyme hydrolysate was measured through the growth response of the E. coli auxotroph and compared to previously published animal assay values for availability. The lysine content of the protein sources was determined using the standard curve generated with the L-lysine-supplemented test M9 medium. The availability of each protein was calculated as the standard curve generated content divided by an HPLC or reference value for the lysine concentration of that protein source.
209
2.4. Luminescence measurements The test M9 medium was vortexed and 125 ml of the culture was pipetted to wells of microtiter plates containing 100 nM of autoinducer (n-3-oxohexanoylhomoserine lactone, Sigma). Luminescence was measured with a Dynex MLX Luminometer (Chantilly, VA, USA) after 4 h of incubation with shaking at 288C (Burlage, 1997; Thomas et al., 1997). Prior to reading, 125 ml of aldehyde was added to each well, and the wells were read for an 8-s interval with a 4-s delay. Readings were taken on duplicate wells and light measurements were reported in relative light units (RLU).
3. Results and discussion Successful plasmid transfer was confirmed through plating on LB plates with carbenicillin and through gel electrophoresis. In order to obtain consistent luminescent measurements only colonies less than 24-h old were selected from plates. The temperature was maintained at 288C due to the luciferase enzyme being heat labile above 308C (Burlage, 1997; Thomas et al., 1997). The culture was cultivated in microtiter plates with agitation to provide the aeration necessary for the biochemical reaction. Cultures were grown for 4 h to allow luciferase that may have accumulated from the overnight culture to dissipate in order to obtain a distinct response from the control. Initially the lysine concentration in the M9 medium ranged from 0.05 to 2.5 mg / ml. The lower portion of the concentration range produced a linear response to lysine with the RLU response to lysine reaching a plateau at approximately 0.5 mg / ml (Fig. 1). Due to higher lysine concentrations being above the sensitivity range of the assay, focus was centered on the 0.05 to 0.5 mg / ml lysine range using 0.05–0.1 mg / ml increments. This range produced a linear RLU response to the lysine concentration in the media with a standard curve equation of y 5 1137.2x 1 132.24, R 2 5 0.989 ( y 5 RLU, x 5 lysine concentration, Fig. 1). Low levels of light were produced in the medium containing no lysine which indicates that some cells may survive the depletion
210
A.M. Erickson et al. / Journal of Microbiological Methods 40 (2000) 207 – 212
Fig. 1. Luminescence (RLU) response of Escherichia coli lysine auxotroph to lysine-supplemented M9 medium. The standard error of each data point is based upon 34, 6, 28, 28, 8, 30, 28, 6 and 34 observations for the lysine concentrations 0.0, 0.05, 0.1, 0.2, 0.25, 0.3,0.4, 0.45 and 0.5 mg / ml respectively.
stage, yet not proliferate to the extent of the cells in the lysine-supplemented medium. Luminescent methods are a determinant of cellular metabolism as light production is dependent upon the metabolic functioning of the cells (Hill and Stewart, 1994). A higher light output (RLU value) corresponds to a larger number of cells. Substances in the medium either promoting or impairing growth will induce a concomitant change in bioluminescence. Lysine stimulates the extent of growth of the auxotroph thus variations in lysine concentration will result in a variable and quantitative growth response (Payne et al., 1977; Tuffnell and Payne, 1985; Hitchins et al., 1989). Consequently, the luminescence response to the varying concentrations of lysine can be quantified, and the response is linear due to growth of the assay organism being directly responsive to incremental increases in the lysine concentration of the medium. In conventional E. coli assays for lysine, the growth response to lysine is measured as a function of increases in optical density measurement (absorbance, A 600 ) on a spectrophotometer. When the RLU response is directly compared with the conventional method, the correlation coefficient of the methods is 0.989, P , 0.01 (Fig. 2). The conventional assay for measuring optical density requires that cell cultures be cultivated until they reach peak growth. This typically takes anywhere from 12 to 26 h. However, the luminescent assay is more sensitive and can detect changes in cellular density through light output in 4 h. By using the luminescent assay there is a ten-fold increase in the sensitivity of the
Fig. 2. Comparison of MOD response with luminescence response of an Escherichia coli lysine auxotroph to lysine-supplemented M9 medium. MOD: maximum optical density measured at 600 nm (absorbance 600 ). RLU: relative light units measured with a Dynex luminometer.
assay. The lower detection limit for optical density assays is 0.5 mg / ml while the lysine concentration of 0.05 mg / ml is detected using the luminescent method. E. coli is non-proteolytic therefore proteins must be enzymatically predigested before they can be absorbed by E. coli. Enzyme hydrolysis of proteins reflects digestion of a protein in vivo (Szmelcman and Guggenheim, 1967). Differences in the enzyme activity, specificity, and total time of digestion will cause a variety of peptides to be produced from a protein source (Longland, 1991). Protease was selected as the digestion enzyme as it has been utilized in studies to hydrolyze protein sources for utilization by E. coli (Carpenter and Booth, 1973; Payne et al., 1977; Hitchins et al., 1989). Compared to mammalian proteases such as pepsin, trypsin and chymotrypsin, protease is a bacterial enzyme that has the ability to hydrolyze the proteins into units small enough to be utilized by the bacteria (Payne et al., 1977). By using the lysine standard curve equation the lysine content of the digests was estimated using the RLU response of the luminescent E. coli to the digests. When compared to previously published animal reference values using precision fed rooster assays, the availability determined by luminescence of casein was higher, and the availability of poultry by-product meal was slightly lower than the reference values (Fig. 3; Izquierdo et al., 1988; Johnson et al., 1998). The high availability value of the casein could be a result of error in the HPLC generated
A.M. Erickson et al. / Journal of Microbiological Methods 40 (2000) 207 – 212
211
range being ten times lower than with traditional assays. Future studies should include a wider range of protein sources as well as improved digestion techniques. This may encompass using a multienzyme digest and alterations in the digestion duration in order to produce a digest that better correlates with animal digestion in vivo. This would allow the assay to serve as a potential alternative to animal methods for quality evaluation of a wide variety of animal feed protein sources. Fig. 3. Comparison of luminescence and animal assay methods for determination of lysine availability. Availability values determined by luminescence readings were derived using the lysinedependent luminescence standard curve. Standard error bars represent 23 determinations for casein and 22 determinations for poultry by-product meal. Availability values determined by animal assay was obtained from reference values of precision fed roosters (Izquierdo et al., 1988; Johnson et al., 1998).
value or possibly a growth-promoting factor in the casein digest resulting in over-estimation of availability values. The lower availability of the poultry by-product meal could be a result of batch to batch differences in poultry by-product meal or incomplete enzymatic digestion by protease. The similarity in animal values and luminescence values for availability suggests that a luminescent approach is a feasible alternative to animal assays. In theory, protein sources could be ranked according to the lysine availability determined by the luminescent assay and be supplemented with purified lysine or protein sources high in lysine if needed to produce a diet that is nutritionally complete. The luminescent method presented provides an alternative to traditional microbiological methods for lysine assays. The method highly correlates with and offers advantages over traditional methods of bacterial analytical growth assays. By using a luminescent method, the problem of background microflora in the protein sources can be eliminated. This assay measures the light output of the E. coli assay organism. Indigenous microflora will likely not possess luminescence abilities and not interfere with the luminescence-based lysine quantitation (Burlage, 1997). Assay time can be reduced using this method, as incubation time is reduced from 12 to over 40 h to a standard 4-h duration. Sensitivity is increased with the luminescent method due to the concentration
Acknowledgements This work was supported with funds from the IAMS Company, Lewisberg, Ohio and by Hatch grant H8311 administered by the Texas Agricultural Experiment Station, Texas A and M University, College Station, Texas, USA. A.M. Erickson was supported by a Regent’s Fellowship provided by ´ Texas A and M University. I.B. Zabala-Dıaz is supported by a LUZ-CONICIT Graduate Fellowship (Maracaibo, Venezuela) and Texas Good Neighbor Scholarship (Texas Higher Education Board, Austin, TX, USA). The authors wish to thank Dr. T.O. Baldwin (Department of Biochemistry and Biophysics, Texas A and M University) for the plasmid and helpful discussion, Dynex Corporation (Chantilly, VA, USA) for use of the luminometer, and A.K. Reinert for preparing the manuscript.
References Anantharaman, K., Parard, C., Decarli, B., Reinhardt, P., Graf, E., Finot, P.A., 1983. The microbiological estimation of available lysine in industrial milk products using a lysine Escherichia coli. m. 26–26. In: McLoughlin, J.V., McKenna), B.M. (Eds.), Basic Studies in Food Science, Research in Food Science and Nutrition, Vol. 2, Boole Press Ltd, Dublin, Ireland, pp. 81–82. Angkanaporn, K., Ravindan, V., Bryden, W.L., 1997. Influence of caecectomy and dietary protein concentration on apparent excreta amino acid digestibility in adult cockerels. Br. Poult. Sci. 38, 270–276. Baker, J.M., Griffiths, M.W., Collins-Thompson, D.L., 1992. Bacterial bioluminescence: applications in food microbiology. J. Food Prot. 55, 62–70. Baldwin, T.O., Devine, J.H., Heckel, R.C., Lin, J.W., Shadel, G.S., 1989. The complete nucleotide sequence of the lux regulon of Vibrio fischeri and the luxABN region of Photobacterium
212
A.M. Erickson et al. / Journal of Microbiological Methods 40 (2000) 207 – 212
leiognathi and the mechanism of control of bacterial bioluminescence. J. Biolum. Chemilum. 4, 326–341. Bjarnason, J., Carpenter, K.J., 1970. Mechanism of heat damage in proteins. 2. Chemical changes in pure proteins. Br. J. Nutr. 24, 313–329. Boctor, A.M., Harper, A.E., 1968. Measurement of available lysine in heated and unheated foodstuffs by chemical and biological methods. J. Nutr. 94, 289–296. Burlage, R.S., 1997. Emerging technologies: bioreporters, biosensors and microprobes. In: Hurst, C.J., Knudsen, G.R., McInerney, M.J., Stetzenbach, L.D., Walter, M.V. (Eds.), Manual of Environmental Microbiology, American Society for Microbiology, Washington, DC, pp. 115–123. Carpenter, K.J., Booth, V.H., 1973. Damage to lysine in foods processing: its measurement and significance. Nutr. Abstr. Rev. 43, 423–451. Duffy, G., Ellison, A., Anderson, W., Cole, M.B., Stewart, G.S.A.B., 1995. Use of bioluminescence to model the thermal inactivation of Salmonella typhimurium in the presence of a competitive microflora. Appl. Env. Microbiol. 61, 3463–3465. Erickson, A.M., Li, X., Woodward, C.L., Ricke, S.C., 1999a. Optimisation of enzyme treatment for the degradation of feed proteins for an Escherichia coli auxotroph lysine availability assay. J. Sci. Food Agric. 79, 1929–1935. Erickson, A.M., Zabala Diaz, I.B., Ricke, S.C., 1999b. Antibiotic amendment for suppression of indigenous microflora in feed sources for an Escherichia coli auxotroph lysine assay. J. Appl. Microbiol. 87, 125–130. Evans, R.J., Butts, H.A., 1949. Inactivation of amino acids by autoclaving. Science 109, 569–571. Ha, S.D., Pillai, S.D., Maciorowski, K.G., Ricke, S.C., 1995. Cycloheximide as a media amendment for enumerating bacterial populations in animal feeds. J. Rapid Meth. Automation Microbiol. 4, 95–105. Hardy, B., 1991. The future requirements of the feed industry. In: Fuller, M.F. (Ed.), In Vitro Digestion for Pigs and Poultry, CAB International, Wallingford, UK, pp. 181–192. Hill, P.J., Stewart, G.S.A.B., 1994. Use of lux genes in applied biochemistry. J. Biolumin. Chemilumin. 9, 211–215. Hitchins, A.D., McDonough, F.E., Wells, P.A., 1989. The use of Escherichia coli mutants to measure the bioavailability of essential amino acids in foods. Plant Food Hum. Nutr. 39, 109–120.
Izquierdo, O.A., Parsons, C.M., Baker, D.H., 1988. Bioavailability of L-lysine in L-lysine HCl. J. Anim. Sci. 66, 2590–2597. Johnson, M.L., Parsons, C.M., Fahey, Jr. G.C., Merchen, N.R., Aldrich, C.G., 1998. Effect of species raw material source, ash content, and processing temperature on amino acid digestibility of animal by-product meals by cecectomized roosters and ileally cannulated dogs. J. Anim. Sci. 76, 1112–1122. Longland, A.C., 1991. Digestive enzyme activities in pigs and poultry. In: Fuller, M.F. (Ed.), In Vitro Digestion for Pigs and Poultry, CAB International, Wallingford, UK, pp. 3–18. Nordheim, J.P., Coon, C.N., 1984. A comparison of four methods for determining available lysine in animal protein meals. Poult. Sci. 63, 1040–1051. Payne, J.W., Bell, G., Higgins, C.F., 1977. The use of an Escherichia coli lys auxotroph to assay nutritionally available lysine in biological materials. J. Appl. Bact. 42, 165–177. Selifonova, O., Burlage, R., Barkay, T., 1993. Bioluminescent sensors for detection of bioavailable Hg(II) in the environment. Appl. Env. Microbiol. 59, 3083–3090. Shockman, G.D., 1963. Amino acids. In: Kavanagh), F. (Ed.), Analytical Microbiology, Academic Press Inc, New York, pp. 567–673. Sibbald, I.P., 1979. A bioassay for available amino acids and true metabolizable energy in feeding stuffs. Poult. Sci. 58, 668– 673. Szmelcman, S., Guggenheim, K., 1967. Availability of amino acids in processed plant protein foodstuffs. J. Sci. Food Agric. 18, 347–350. Tabib, Z., Jones, F.T., Hamilton, P.B., 1981. Microbiological quality of protein feed and ingredients. Poult. Sci. 60, 1392– 1397. Tauriainen, S., Karp, M., Chang, W., Virta, M., 1997. Recombinant luminescent bacteria for measuring bioavailable arsenite and antimonite. Appl. Env. Microbiol. 63, 4456–4461. Thomas, M.D., Schineller, J.B., Baldwin, T.O., 1997. lux autoinducer response in E. coli and Vibrio fischeri. In: Hastings, J.W., Kricka, L.J., Stanley, P.E. (Eds.), Bioluminescence and Chemiluminescence: Molecular Reporting with Photons, John Wiley and Sons, New York, pp. 109–112. Tuffnell, J.M., Payne, J.W., 1985. A colorimetric enzyme assay using Escherichia coli to determine nutritionally available lysine in biological materials. J. Appl. Bact. 58, 333–341.