Metabonomic Identification of Two Distinct Phenotypes in Sprague-Dawley (Crl:CD(SD)) Rats
http://www.100md.com
《毒物学科学杂志》
Pfizer Global Research and Development, Metabonomics Evaluation Group, Ann Arbor, Michigan 48105
Manpower, Inc., 231 Little Lake Drive, Ann Arbor, Michigan 48103
ABSTRACT
Genetic drift in animal populations has been a recognized concern for many years. Less understood is the potential for phenotypic "drift" or variation that is not related to any genetic change. Recently, stock Sprague-Dawley (Crl:CD(SD)) rats obtained from the Charles River Raleigh facility demonstrated a distinct endogenous urinary metabonomic profile that differed from historical control SD urine spectral profiles obtained over the past several years in our laboratory. In follow-up studies, the origin of the variant phenotype was narrowed down to animals of both sexes that were housed in one specific room (Room 9) in the Raleigh facility. It is likely that the two phenotypes are related to distinct populations of gut flora that particularly impact the metabolism of aromatic molecules. The most pronounced difference between the two phenotypes is the relative amounts of hippuric acid versus other aromatic acid metabolites of chlorogenic acid. Though both molecular species are present in either phenotype, the marked variation in levels of these molecules between the two phenotypes has led to the designation of high hippuric acid (HIP) and high chlorogenic acid metabolites (CA) phenotypes. Specific urinary components that distinguish the phenotypes have been thoroughly characterized by NMR spectroscopy with additional, limited characterization by LC-MS (high performance liquid chromatography coupled with mass spectrometry). Co-habitation of rats from the two phenotypes rapidly facilitated a switch of the CA phenotype to the historical Sprague-Dawley phenotype (HIP). The impact of these variant phenotypes on drug metabolism and long-term safety assessment studies (e.g., carcinogenicity bioassays) is unknown.
Key Words: metabonomics; phenotype; Sprague-Dawley (Crl:CD(SD)); hippuric acid; chlorogenic acid.
INTRODUCTION
Metabonomics is a powerful tool that assesses the biological response to stress, such as that induced by toxins or disease states. However, any physiological or environmental change, such as age, estrus, diet, and housing can impact an animal's metabonomic profile (Beckwith-Hall et al., 2002a,b; Bollard et al., 2001; Robertson et al., 2000). Current methodology relies on detecting pattern differences in NMR or MS spectra of biofluids that correlate with physiological or pathophysiological changes in the organism. The pattern is synonymous with the endogenous metabolite profile. In a preclinical setting, the ability to detect subtle treatment-related changes with metabonomics is heavily dependent on a stable, uniform animal population in which the endogenous metabolite profile in biofluids (mainly urine and plasma/serum) is reasonably constant. It would be expected that this stability is also critical for understanding and evaluating data from proteomic and toxicogenomic studies as well. Other researchers (Gavaghan et al., 2001; Phipps et al., 1998) have noted instances where this has not been the case in urines collected from several strains. In these instances, unusually low or undetectable hippuric acid levels were observed in urine from these rat strains with concomitant increases in other urinary aromatic species including 3-(3-hydroxyphenyl)propionic acid (3-HPPA) and 3-hydroxycinnamic acid (3-HCA). In some cases, the dietary intake was implicated in the observed change in hippuric acid excretion. In other cases, the cause was more difficult to ascertain. Since gut flora are known to be involved in the metabolism of plant phenolics, it has been postulated that changes in the population of gut flora are also involved in instances where low levels of hippuric acid are excreted in the urine (Gavaghan et al., 2001).
In the course of five years of metabonomics studies conducted in this laboratory, employing mainly Wistar and Sprague-Dawley rats, the urinary metabolite profile within both strains has remained remarkably stable with high levels of hippuric acid as a prominent hallmark metabolite. In this report, we characterize a recent change in that stability with two distinct phenotypes coming from different rooms (colonies) in the same commercial facility.
MATERIALS AND METHODS
Animals.
Sprague-Dawley (Crl:CD(SD)) rats were obtained from Charles River Laboratories. Animals from historical studies (Table 1) were obtained from the facilities listed. Rats for phenotype follow-up experiments were obtained exclusively from the Charles River Raleigh facility. Room 9 at that facility was found to house animals with the variant phenotype and Room 10 housed animals with the historical phenotype observed in previous studies with SD rats from various Charles River facilities. Animals in historical studies ranged from 6 to 12 weeks of age, and all rats used for the phenotype assessment studies were between 7 and 8 weeks of age at initiation.
Routine animal husbandry.
Animals were housed in an AAALAC accredited facility and all in-vivo studies were reviewed and approved by the IACUC. When not in metabolism cages, animals were housed in individual cages in temperature (70–78°F) and humidity (30–70% RH) controlled rooms with a 12-h light cycle. Animals were fed (Lab Diet Certified Rodent Chow–5002) and watered ad libitum.
Urine collection.
When required, animals were placed into individual plastic metabolism cages (Harvard Apparatus, Holliston, MA) where they remained for the duration of urine collection. Food and water were available ad libitum. For phenotype monitoring, animals were
Cohabitation experiment.
To assess a possible change in phenotype, a group of 10 female rats from Rooms 9 and 10 were obtained and baseline phenotype identified as described above. The rats were then pair housed (one from each room) in plastic shoebox cages containing Bed O'Cob absorbent bedding. Food and water were available ad libitum. At weekly intervals, individual animals were placed in metabolism cages for 24 h and a single urine sample was subsequently collected, after which each animal was then returned to its shoebox cage. Following three weeks of cohabitation, animals were separated into individual shoebox caging. Weekly 24-h urine collections were continued for an additional two weeks to explore the possibility of reverting back to the original phenotype.
One-dimensional 1H NMR spectroscopy.
Samples for NMR analysis were prepared by mixing 500 μl of urine with 250 μl of buffer in 96-well plates. The buffer was added to provide some normalization of the urinary pH. After mixing, plates were centrifuged to sediment insolubles. The buffer was 0.2 M sodium phosphate buffer at pH 7.4 (80:20 H2O:D2O), containing 1 mM TSP (sodium 2,2',3,3'deutero-3-trimethylsilylpropionate, an internal NMR reference standard) and 3 mM sodium azide. 1H NMR data were acquired using a Varian Inova NMR spectrometer operating at 600.36 MHz for 1H and equipped with a 1H-{15N, 13C} flow probe (120 μl active volume) and a Varian automated sample transport accessory (VAST). Four-hundred-fifty microliters of sample was injected into the probe with no push solvent. Two cell rinses were completed with isotonic phosphate buffer between samples. The push buffer was prepared by mixing one part 0.2 M sodium phosphate buffer at pH 7.4 with two parts water. One-dimensional 1H NMR spectra were acquired at 27°C using a one-dimensional NOESY pulse sequence including water presaturation and a mixing time of 100 milliseconds. A total of 64 scans were collected with 64k data points, an acquisition time of 2.73 s, an inter-pulse delay of 1 s, and a sweep width of 12 kHz.
NMR data analysis.
Spectra were processed and analyzed using in-house software, Metabonomi (International Patent Application, WO2004038602A1). Spectra were normalized to the total integrated spectral area minus the region containing water and urea from 6.0 to 4.5 ppm.
Mass spectrometry.
Samples were centrifuged and injected directly onto the HPLC column undiluted for metabolite profiling and characterization. For exact mass determination, urine samples were diluted 1:1000 in water prior to analysis. Metabolite profiling and characterization was performed using a Surveyor LC system (ThermoFinnigan, San Jose, CA) coupled via an electrospray interface to a LCQ Deca XP Plus ion trap mass spectrometer (ThermoFinnigan, San Jose, CA). Data was collected in both positive and negative ionization modes. Operating parameters included a source voltage of 4.5 kV, capillary temperature of 250°C, capillary voltage of 21 V, and a tube lens offset of 25V. Helium was used as the sheath gas and auxiliary gas at flow settings of 80 (arb.) and 20 (arb.), respectively. Data acquired in full scan mode was collected over a scan range of 80–1200 amu. Collision induced fragmentation was performed for selected ions of interest, using helium as a collision gas with an activation amplitude of 20–30% and an isolation width of 1.5–1.8 amu. Product ion spectra were acquired for selected ions of interest. Chromatographic separation of metabolites was achieved using a Symmetry C-18 column 150 x 2.1 mm, 5 μm (Waters, Milford, CT) under gradient conditions starting at 100% 0.1% formic acid (v/v) and 0% acetonitrile, increasing linearly to 20% acetonitrile at 4 min, then to 95% acetonitrile at 8 min and holding for 1 min before re-equilibration at initial conditions. Flow rate was held constant at 300 μl/min.
Exact mass determination was conducted at PE Sciex, Concord, Canada, using an LC-MS system consisting of an Agilent 1100 high performance liquid chromatography system (Agilent Technologies, Torrence, CA) coupled to a QSTAR XL hybrid triple quadrupole orthogonal time of flight mass spectrometer (PE Sciex, Foster City, CA) via a Turbospray interface.
LC-NMR/MS spectroscopic analysis.
LC-NMR/MS experiments were conducted using an integrated LC-NMR/MS system from Varian (Palo Alto, CA). The LC system consisted of a Star 9012 solvent delivery system and Star 9050 UV/Vis detector set at 254 nM. MS data were acquired using a 1200L triple quadrupole mass spectrometer, and NMR spectra were acquired using a Varian Inova NMR spectrometer operating at 600.36 MHz for 1H. Data acquisition was controlled with VnmrJ software, version 1.1C, operating on the NMR spectrometer's host computer. Urine samples of interest were analyzed by injecting a 0.5 ml aliquot onto a Synergi Polar RP semi-preparative column 5 μm, 250 x 10 mm (Phenomenex, Torrance, CA) and eluted under gradient conditions beginning with 80% deuterium oxide containing 0.1% formic acid: 20% acetonitrile, increasing linearly to 95% acetonitrile over 40 min and holding for 5 min before re-equilibration at initial conditions. A flow rate of 2 ml/min was used. After passing through the UV detector, the flow was split so that 5% routed to the mass spectrometer and 95% routed to the NMR flow probe. Instrument parameters were set to acquire MS and NMR data continuously during the chromatography run. The mass spectrometer was coupled to the LC system via an electrospray interface, operating in the positive ion mode. Ionization parameters for the mass spectrometer included applied needle and capillary voltages of 5000 V and 100 V, respectively with nitrogen used as the drying gas at 17 psi with a source temperature of 300°C and the nebulizing gas at 57 psi. Data were acquired in the full scan mode over a scan range of 100–1000 amu. 1H NMR spectra were acquired using the lc1d pulse sequence supplied by Varian. The spectral parameters included a sweep width of 10,000 Hz, an acquisition time of 1.1 s, 8k data points, and 8 transients per slice. The effluent from the NMR probe was collected in fractions during the run. In the specific case of identifying the metabolite with a MS peak at m/z 340 [M+H]+, mass spectrometry was used to confirm the presence of the metabolite in isolated fractions and guided the pooling of fractions for further purification. Final purification was achieved by injecting the concentrated fraction onto a Symmetry C-18 column, 5 μm, 150 x 3.9 mm (Waters, Milford, CT) and eluting the component with gradient conditions beginning with 80% deuterium oxide containing 0.1% formic acid: 20% acetonitrile, increasing linearly to 50% acetonitrile over 25 min. A flow rate of 1 ml/min was used. The effluent from the NMR probe was collected in fractions during the run. As before, mass spectrometry was used to confirm the presence of the metabolite in fractions and guided the pooling of fractions for the final sample. In the final sample, the LC solvents were removed by lyophilization and the isolated metabolite was dissolved in 550 μl of water (90:10 H2O:D2O) for analysis by 1H NMR spectroscopy.
RESULTS
Phenotype Identification and Characterization:
Prominent differences between the phenotypes were noted in 1H NMR spectra. Based on the clear and consistent predominance of aromatic components in the urine, hippuric acid, or chlorogenic acid metabolites, these two phenotypes are referred to as HIP (high hippuric acid) or CA (high chlorogenic acid metabolites). A selected region of 1H NMR spectra of urine from SD rats of HIP and CA phenotypes is shown in Figure 1. After reviewing records of this laboratory's metabonomic studies conducted using Charles River-supplied rats, the unique source of the CA phenotype was identified as the Raleigh facility, Room 9. A summary of metabonomics experiments is shown in Table 1 including the Charles River facility and room of origin and date of shipment receipt. In additional studies with Charles River-supplied rats from the Raleigh facility, the HIP phenotype was observed in male and female SD rats from Room 10, while the CA phenotype was observed in male and female SD rats from Room 9. Major metabolites contributing to each phenotype were identified and levels quantitated using NMR, LC-MS, and combined LC-NMR/MS methods. The HIP and CA metabolite profiles are detailed in Table 2. NMR spectral assignments were completed using a combination of literature (Gavaghan et al., 2001; Nicholls et al., 2003) and commercially available authentic standards in all cases except the dihydro-quinolinone glucuronide associated with the HIP phenotype. Profiling experiments by LC-MS revealed that a metabolite at [M+H]+ m/z 340 was abundant in the Room 10 sample, while present only at trace levels in the Room 9 sample. A representative subset of samples from male and female rats, originating from both rooms, was analyzed to confirm this finding. Both source-induced and collision-induced fragmentation yielded a fragment ion at m/z 164 amu consistent with a neutral loss of 176 amu, indicative of a glucuronide conjugate. Accurate mass values were determined to be m/z 340.1016 and m/z 164.0679 amu. Based on these values, the molecular formula for the unknown at m/z 340 is C15H18NO8 and the fragment at m/z 164 is consistent with C9H9NO2 (both < 5 ppm mass deviation). Further experiments using an ion trap mass spectrometer, on-line LC-NMR/MS experiments, as well as NMR experiments on the isolated material led to the proposed dihydro-quinolinone structure (Fig. 3), a cyclic derivative of cinnamic acid. The MS fragmentation pattern and 1H NMR spectrum are shown in Figures 4 and 5, respectively. Authentic 3,4-dihydro-2(1H)-quinolinone (98% pure, Sigma-Aldrich Company) fragmented in a manner similar to the aglycone, with the predominant product ion forming from loss of 42 amu. The proposed assignment was further validated through comparison of NMR spectral data with authentic 3,4-dihydro-2(1H)-quinolinone and reference to 1H chemical shift reference tables that quantitate the effect of substituents on chemical shifts (Pretsch et al., 2000).
Stability of Phenotype
The metabolite profile of 10 rats from each phenotype was monitored in SD male rats for a period of four weeks. During this time, the rats were individually housed. After four weeks, the metabolite profile in the urine remained constant for 19 out of the 20 animals. One animal, initially having the CA phenotype, spontaneously began excreting urine having a metabolite profile consistent with the HIP phenotype after two weeks. For the remaining two weeks of monitoring, the urine metabolite profile of that animal remained consistent with the HIP phenotype. In a separate experiment, five pairs (one from each phenotype) of female SD rats were co-housed. Prior to cohabitation, the phenotype was confirmed by NMR and after cohabitation of one week, urine was again collected and analyzed. Results indicated that after one week of cohabitation, the rats initially having the CA phenotype had converted to excreting urine with a metabolite profile consistent with the HIP phenotype, Figure 2. After three weeks of cohabitation, the rats were then housed individually and monitored for an additional two weeks. The urine metabolite profile remained consistent with the HIP phenotype with no reversion to the original phenotype in CA animals.
DISCUSSION
This report documents a consistent phenotypic difference within Sprague-Dawley (Crl:CD(SD)) rats supplied by the Charles River Raleigh facility that has remained stable for at least 12 months. Though a phenotype change as described here could be attributed to genetic drift, that is probably not the case here—at least in the usual sense. The changes observed in this report are most likely due to differences in the gut flora populating the GI tract. This is based primarily on two lines of evidence. Firstly, the urinary biomolecular differences observed in this study are consistent with previous reports of microflora-induced changes in urinary metabolites, particularly hippuric acid and chlorogenic acid metabolites (Gavaghan et al., 2001; Phipps et al., 1998). The second line of evidence is that when rats were housed in pairs, one from each phenotype, there was a relatively rapid switch of the phenotype of CA animals. This was most likely due to coprophagic inoculation of HIP-derived bacteria into the CA rats. The rapidity of the change ruled out an endogenous genetic component.
The GI tract is home to a vast host of microbiota with hundreds of species of bacteria present in the normal mammalian GI tract (Guarner and Malagelada, 2003). Bacterial density varies with the topography of the GI tract with < 1000 bacterial colony forming units (CFU) per ml present in the stomach to several hundred billion CFU/ml present in the colon (Hart et al., 2002). Comprehensive bacterial species identification of this complex ecology is difficult, if not impossible, particularly as speciation of the gut flora varies with species, age, and diet (Brennan-Craddock et al., 1992; Hebuterne, 2003; Rowland et al., 2000). However, an overview of the variety of gut flora populations in normal and disease states in humans has been published (Salminen et al., 1995). The role of gut flora in health is well recognized (Hart et al., 2002; Hooper et al., 1998, 2002; Hooper and Gordon, 2001; Simon and Gorbach, 1984; Stagg et al., 2004) with the fields of prebiotics and probiotics based on the concept of alteration of gut flora to treat disease or otherwise improve health (Gibson, 1998; Karimi and Pena, 2003). Therefore, it should not be terribly surprising that alterations in gut flora may cause fairly profound systemic effects. Nicholson and colleagues (Nicholson et al., 2005; Nicholson and Wilson, 2003), make a strong case for the integration of the gut floral biology into any assessment of systems biology or systems toxicology. Metabonomics as a technology excels at just such assessments, so it is not surprising that many of the anecdotal reports identifying varying phenotypes within a rat strain, have this analytical approach as their basis. Phipps et al. (1998) reported that Alp:ApfSD (Wistar derived) rats maintained using standard husbandry condition did not excrete hippuric acid in the urine with 3-HPPA constituting the primary aromatic component. These rats would be somewhat similar to the CA rats described here. Interestingly, they found that simple addition of benzoic acid to the diet caused the animals to initiate hippuric acid excretion, which did not reverse after returning to the original diet. The authors implicated microfloral changes as an explanation for this lack of reversal. While we report similar results, they vary in some important respects. Firstly, the CA rats observed in these studies had measurable, though low, levels of hippuric acid. More importantly, the diet was not the source of the variation, in our studies since animals were obtained from the same facility with identical husbandry practices between rooms. Gavaghan et al. (2001) noted phenotypic variation in Sprague-Dawley rats consistent with that observed here in that they also noted phenotypes defined by high and low levels of hippuric acid instead of its presence and absence as reported by Phipps. The authors indicate that the phenotypes represent subpopulations but are unclear as to the source of their animals or the stability of these subpopulations.
As animals are born, raised, and distributed from a single room within the facility it is possible that after an initiating (and currently unexplained) event, gut floral changes occurred within the Room 9 population and have been passed from dam to weanling propagating the changes ever since. Another alternative explanation involves the sourcing of rats for Raleigh Room 9. The colony of rats maintained in Room 9 was originally started from foundation colony rats that had only eight species of bacteria, the Charles River Altered Schaedler Flora (Charles Clifford, Charles River Laboratory, personal communication). A detailed description of the Schaedler flora can be found in the article by Sarma-Rupavtarm and colleagues (Sarma-Rupavtarm et al., 2004). These "isolators" are maintained separately and used as a source for new colony initiation. Even after several years, these isolators have very limited number of gut bacteria species. Of the facilities we evaluated, Room 9 represented the most recent colony initiation in December 2002. The colony in P09 initiated in March 2000, while the colonies in K93 and K97 initiated in May 1997. As colonies are initiated and moved to barrier rooms for propagation, a repopulation of a "normal" gut flora may occur over time. In this scenario, the gut floral variation isn't really a species shift, but is due to a lack of the "normal" speciation that occurs as rats are exposed to various bacteria, which evidently may be quite protracted. This scenario is supported by studies conducted by Nicholls et al. (2003) who monitored urine metabolic profile change in axenic Fisher-344 rats exposed to a normal laboratory environment to allow gut floral repopulation. They reported that initially low levels of urinary hippuric acid increased dramatically after 21 days of exposure to the normal laboratory environment. The primary argument against this scenario is the timing. If the colony in Room 9 had been initiated weeks or even months ago, this scenario would seem more likely. However, why would there be such a striking difference in a colony initiated three years ago versus one started five years ago Nevertheless, this scenario deserves further investigation.
Of greatest concern with regard to these findings is what, if any, significance these phenotypic changes engender. These observations raise some interesting questions from a toxicologic perspective. The fact that these changes are not simply a reflection of diet-induced modulation of the microflora, but can persist for some time is highly significant. The gut flora are known to influence metabolism of numerous compounds which can profoundly affect toxification and detoxification pathways (Boxenbaum et al., 1979; Goodwin et al., 1994; Mikov, 1994; Rowland, 1981, 1988). The role of gut flora has been directly linked to the toxicity (or lack of toxicity) of numerous compounds including nitrobenzene, amygdalin, and digoxin to name a few (Mathan et al., 1989; Reddy et al., 1976; Rowland, 1988). Activation (or lack of detoxification) by gut flora can play a role in tumor induction by exogenous compounds (Rowland and Walker, 1983; Yoshida et al., 2003). If indeed the phenotypic changes are stable for even longer than the four weeks evaluated in this report, perhaps of even greater interest is what effect microflora changes may have on exposure to natural (i.e., dietary derived) metabolites (e.g., chlorogenic acid metabolites) and what effect these changes will have on background tumor incidence. If such differences in gut flora are persistent, they represent a potential lifetime exposure to relatively high concentrations of these varying metabolites. Could altered microflora be one explanation for what otherwise seems inexplicably fast drift in background tumor incidences More generally, are these epigenetic variations hidden sources of variation in preclinical animal studies These questions remain to be answered.
It is possible that these effects simply reflect a transient phenomenon that comes and goes within populations. Alternatively, rigorous efforts to maintain "clean" animal facilities may in some cases cause problems rather than avoid them. In either case the findings provide an opportunity to more fully understand the role of microflora in rat systems biology. As it is likely these changes have occurred from time to time in the past and are almost certainly characteristic of all laboratory animal suppliers, one could argue that they have had little effect on our ability to conduct and interpret toxicology studies. However, the truth of the matter is that we don't know what we don't know. It is plausible that anomalous findings that occasionally pop up in toxicology studies are not really anomalous but attributable to unrecognized microfloral changes. Likewise, the interstudy "variability" we have grown accustomed to in toxicology studies may not really be a biologic constant but rather may be at least partially attributable to definable causes, such as microfloral differences. What is undeniable is that the newer "omic" technologies are very sensitive to off target (systemic) changes. Therefore, it is imperative that studies using these technologies understand the nature of their animal models, including the role gut flora play. The complexity of life is a beautiful thing, though it certainly can be confusing sometimes. As new technologies delve ever deeper into systems biology, it can be anticipated that the more we learn, the more we will need to critically reassess our most basic assumptions.
In conclusion, we have identified a phenotypic difference in commercially supplied Sprague-Dawley (Crl:CD(SD)) rats that has been stable within a single room of the Charles River Raleigh facility for a period of at least 12 months. This phenotypic difference is most likely due to altered gut flora and is typified by diminished urinary hippuric acid and increased chlorogenic acid metabolites.
ACKNOWLEDGMENTS
The authors gratefully acknowledge Pauline Vollmerhaus, Ron Bonner, and Lyle Burton at MDS Sciex for their assistance in obtaining accurate masses that helped lead to the identification of the dihydro-quinolinone glucuronide. The authors further gratefully acknowledge the co-operation and valuable insights into Charles River Breeding facility practices provided by Dr. Charles Clifford. Conflict of interest: none declared.
REFERENCES
Beckwith-Hall, B. M., Brindle, J. T., Barton, R. H., Coen, M., Holmes, E., Nicholson, J. K., and Antti, H. (2002a). Application of orthogonal signal correction to minimise the effects of physical and biological variation in high resolution 1H NMR spectra of biofluids. Analyst 127, 1283–1288.
Beckwith-Hall, B. M., Holmes, E., Lindon, J. C., Gounarides, J., Vickers, A., Shapiro, M., and Nicholson, J. K. (2002b). NMR-based metabonomic studies on the biochemical effects of commonly used drug carrier vehicles in the rat. Chem. Res. Toxicol. 15, 1136–1141.
Bollard, M. E., Holmes, E., Lindon, J. C., Mitchell, S. C., Branstetter, D., Zhang, W., and Nicholson, J. K. (2001). Investigations into biochemical changes due to diurnal variation and estrus cycle in female rats using high-resolution (1)H NMR spectroscopy of urine and pattern recognition. Anal. Biochem. 295, 194–202.
Boxenbaum, H. G., Bekersky, I., Jack, M. L., and Kaplan, S. A. (1979). Influence of gut microflora on bioavailability. Drug Metab. Rev. 9, 259–279.
Brennan-Craddock, W. E., Mallett, A. K., Rowland, I. R., and Neale, S. (1992). Development changes to gut microflora metabolism in mice. J. Appl. Bacteriol. 73, 163–167.
Gavaghan, C. L., Nicholson, J. K., Connor, S. C., Wilson, I. D., Wright, B., and Holmes, E. (2001). Directly coupled high-performance liquid chromatography and nuclear magnetic resonance spectroscopic with chemometric studies on metabolic variation in Sprague-Dawley rats. Anal. Biochem. 291, 245–252.
Gibson, G. R. (1998). Dietary modulation of the human gut microflora using prebiotics. Br. J. Nutr. 80, S209–S212.
Goodwin, B. L., Ruthven, C. R., and Sandler, M. (1994). Gut flora and the origin of some urinary aromatic phenolic compounds. Biochem. Pharmacol. 47, 2294–2297.
Guarner, F., and Malagelada, J. R. (2003). Gut flora in health and disease. Lancet 361, 512–519.
Hart, A. L., Stagg, A. J., Frame, M., Graffner, H., Glise, H., Falk, P., and Kamm, M. A. (2002). The role of the gut flora in health and disease, and its modification as therapy. Aliment. Pharmacol. Ther. 16, 1383–1393.
Hebuterne, X. (2003). Gut changes attributed to ageing: Effects on intestinal microflora. Curr. Opin. Clin. Nutr. Metab. Care 6, 49–54.
Hooper, L. V., Bry, L., Falk, P. G., and Gordon, J. I. (1998). Host-microbial symbiosis in the mammalian intestine: exploring an internal ecosystem. Bioessays 20, 336–343.
Hooper, L. V., and Gordon, J. I. (2001). Commensal host-bacterial relationships in the gut. Science 292, 1115–1118.
Hooper, L. V., Midtvedt, T., and Gordon, J. I. (2002). How host-microbial interactions shape the nutrient environment of the mammalian intestine. Annu. Rev. Nutr. 22, 283–307.
Karimi, O., and Pena, A. S. (2003). Probiotics: Isolated bacteria strain or mixtures of different strains Two different approaches in the use of probiotics as therapeutics. Drugs Today (Barc.) 39, 565–597.
Mathan, V. I., Wiederman, J., Dobkin, J. F., and Lindenbaum, J. (1989). Geographic differences in digoxin inactivation, a metabolic activity of the human anaerobic gut flora. Gut 30, 971–977.
Mikov, M. (1994). The metabolism of drugs by the gut flora. Eur. J. Drug Metab. Pharmacokinet. 19, 201–207.
Nicholls, A. W., Mortishire-Smith, R. J., and Nicholson, J. K. (2003). NMR spectroscopic-based metabonomic studies of urinary metabolite variation in acclimatizing germ-free rats. Chem. Res. Toxicol. 16, 1395–1404.
Nicholson, J. K., Holmes, E., and Wilson, I. D. (2005). Gut microorganisms, mammalian metabolism and personalized health care. Nat. Rev. Microbiol. 3, 431–438.
Nicholson, J. K., and Wilson, I. D. (2003). Opinion: understanding ‘global’ systems biology: Metabonomics and the continuum of metabolism. Nat. Rev. Drug Discov. 2, 668–676.
Phipps, A. N., Stewart, J., Wright, B., and Wilson, I. D. (1998). Effect of diet on the urinary excretion of hippuric acid and other dietary-derived aromatics in rat. A complex interaction between diet, gut microflora and substrate specificity. Xenobiotica 28, 527–537.
Pretsch, E., Buhlmann, P., and Affolter, C. (2000). Structure Determination of Organic Compounds. Tables of Spectral Data. Springer-Verlag, Berlin.
Reddy, B. G., Pohl, L. R., and Krishna, G. (1976). The requirement of the gut flora in nitrobenzene-induced methemoglobinemia in rats. Biochem. Pharmacol. 25, 1119–1122.
Robertson, D. G., Reily, M. D., Sigler, R. E., Wells, D. F., Paterson, D. A., and Braden, T. K. (2000). Metabonomics: Evaluation of nuclear magnetic resonance (NMR) and pattern recognition technology for rapid in vivo screening of liver and kidney toxicants. Toxicol. Sci. 57, 326–337.
Rowland, I. (1981). The influence of the gut microflora on food toxicity. Proc. Nutr. Soc. 40, 67–74.
Rowland, I. R. (1988). Interactions of the gut microflora and the host in toxicology. Toxicol. Pathol. 16, 147–153.
Rowland, I. R., and Walker, R. (1983). The gastrointestinal tract in food toxicology. In Toxic Hazards in Food (D. M. Conning and A. B. G. Lansdown, Eds.), pp. 183–274. Croom Helm, London.
Rowland, I. R., Wiseman, H., Sanders, T. A., Adlercreutz, H., and Bowey, E. A. (2000). Interindividual variation in metabolism of soy isoflavones and lignans: Influence of habitual diet on equol production by the gut microflora. Nutr. Cancer 36, 27–32.
Salminen, S., Isolauri, E., and Onnela, T. (1995). Gut flora in normal and disordered states. Chemotherapy 41(Suppl. 1), 5–15.
Sarma-Rupavtarm, R. B., Ge, Z., Schauer, D. B., Fox, J. G., and Polz, M. F. (2004). Spatial distribution and stability of the eight microbial species of the altered schaedler flora in the mouse gastrointestinal tract. Appl. Environ. Microbiol. 70, 2791–2800.
Simon, G. L., and Gorbach, S. L. (1984). Intestinal flora in health and disease. Gastroenterology 86, 174–193.
Stagg, A. J., Hart, A. L., Knight, S. C., and Kamm, M. A. (2004). Microbial-gut interactions in health and disease. Interactions between dendritic cells and bacteria in the regulation of intestinal immunity. Best Pract. Res. Clin. Gastroenterol. 18, 255–270.
Yoshida, K., Kuroda, K., Zhou, X., Inoue, Y., Date, Y., Wanibuchi, H., Fukushima, S., and Endo, G. (2003). Urinary sulfur-containing metabolite produced by intestinal bacteria following oral administration of dimethylarsinic acid to rats. Chem. Res. Toxicol. 16, 1124–1129.(Lora C. Robosky, Dale F. )
Manpower, Inc., 231 Little Lake Drive, Ann Arbor, Michigan 48103
ABSTRACT
Genetic drift in animal populations has been a recognized concern for many years. Less understood is the potential for phenotypic "drift" or variation that is not related to any genetic change. Recently, stock Sprague-Dawley (Crl:CD(SD)) rats obtained from the Charles River Raleigh facility demonstrated a distinct endogenous urinary metabonomic profile that differed from historical control SD urine spectral profiles obtained over the past several years in our laboratory. In follow-up studies, the origin of the variant phenotype was narrowed down to animals of both sexes that were housed in one specific room (Room 9) in the Raleigh facility. It is likely that the two phenotypes are related to distinct populations of gut flora that particularly impact the metabolism of aromatic molecules. The most pronounced difference between the two phenotypes is the relative amounts of hippuric acid versus other aromatic acid metabolites of chlorogenic acid. Though both molecular species are present in either phenotype, the marked variation in levels of these molecules between the two phenotypes has led to the designation of high hippuric acid (HIP) and high chlorogenic acid metabolites (CA) phenotypes. Specific urinary components that distinguish the phenotypes have been thoroughly characterized by NMR spectroscopy with additional, limited characterization by LC-MS (high performance liquid chromatography coupled with mass spectrometry). Co-habitation of rats from the two phenotypes rapidly facilitated a switch of the CA phenotype to the historical Sprague-Dawley phenotype (HIP). The impact of these variant phenotypes on drug metabolism and long-term safety assessment studies (e.g., carcinogenicity bioassays) is unknown.
Key Words: metabonomics; phenotype; Sprague-Dawley (Crl:CD(SD)); hippuric acid; chlorogenic acid.
INTRODUCTION
Metabonomics is a powerful tool that assesses the biological response to stress, such as that induced by toxins or disease states. However, any physiological or environmental change, such as age, estrus, diet, and housing can impact an animal's metabonomic profile (Beckwith-Hall et al., 2002a,b; Bollard et al., 2001; Robertson et al., 2000). Current methodology relies on detecting pattern differences in NMR or MS spectra of biofluids that correlate with physiological or pathophysiological changes in the organism. The pattern is synonymous with the endogenous metabolite profile. In a preclinical setting, the ability to detect subtle treatment-related changes with metabonomics is heavily dependent on a stable, uniform animal population in which the endogenous metabolite profile in biofluids (mainly urine and plasma/serum) is reasonably constant. It would be expected that this stability is also critical for understanding and evaluating data from proteomic and toxicogenomic studies as well. Other researchers (Gavaghan et al., 2001; Phipps et al., 1998) have noted instances where this has not been the case in urines collected from several strains. In these instances, unusually low or undetectable hippuric acid levels were observed in urine from these rat strains with concomitant increases in other urinary aromatic species including 3-(3-hydroxyphenyl)propionic acid (3-HPPA) and 3-hydroxycinnamic acid (3-HCA). In some cases, the dietary intake was implicated in the observed change in hippuric acid excretion. In other cases, the cause was more difficult to ascertain. Since gut flora are known to be involved in the metabolism of plant phenolics, it has been postulated that changes in the population of gut flora are also involved in instances where low levels of hippuric acid are excreted in the urine (Gavaghan et al., 2001).
In the course of five years of metabonomics studies conducted in this laboratory, employing mainly Wistar and Sprague-Dawley rats, the urinary metabolite profile within both strains has remained remarkably stable with high levels of hippuric acid as a prominent hallmark metabolite. In this report, we characterize a recent change in that stability with two distinct phenotypes coming from different rooms (colonies) in the same commercial facility.
MATERIALS AND METHODS
Animals.
Sprague-Dawley (Crl:CD(SD)) rats were obtained from Charles River Laboratories. Animals from historical studies (Table 1) were obtained from the facilities listed. Rats for phenotype follow-up experiments were obtained exclusively from the Charles River Raleigh facility. Room 9 at that facility was found to house animals with the variant phenotype and Room 10 housed animals with the historical phenotype observed in previous studies with SD rats from various Charles River facilities. Animals in historical studies ranged from 6 to 12 weeks of age, and all rats used for the phenotype assessment studies were between 7 and 8 weeks of age at initiation.
Routine animal husbandry.
Animals were housed in an AAALAC accredited facility and all in-vivo studies were reviewed and approved by the IACUC. When not in metabolism cages, animals were housed in individual cages in temperature (70–78°F) and humidity (30–70% RH) controlled rooms with a 12-h light cycle. Animals were fed (Lab Diet Certified Rodent Chow–5002) and watered ad libitum.
Urine collection.
When required, animals were placed into individual plastic metabolism cages (Harvard Apparatus, Holliston, MA) where they remained for the duration of urine collection. Food and water were available ad libitum. For phenotype monitoring, animals were
Cohabitation experiment.
To assess a possible change in phenotype, a group of 10 female rats from Rooms 9 and 10 were obtained and baseline phenotype identified as described above. The rats were then pair housed (one from each room) in plastic shoebox cages containing Bed O'Cob absorbent bedding. Food and water were available ad libitum. At weekly intervals, individual animals were placed in metabolism cages for 24 h and a single urine sample was subsequently collected, after which each animal was then returned to its shoebox cage. Following three weeks of cohabitation, animals were separated into individual shoebox caging. Weekly 24-h urine collections were continued for an additional two weeks to explore the possibility of reverting back to the original phenotype.
One-dimensional 1H NMR spectroscopy.
Samples for NMR analysis were prepared by mixing 500 μl of urine with 250 μl of buffer in 96-well plates. The buffer was added to provide some normalization of the urinary pH. After mixing, plates were centrifuged to sediment insolubles. The buffer was 0.2 M sodium phosphate buffer at pH 7.4 (80:20 H2O:D2O), containing 1 mM TSP (sodium 2,2',3,3'deutero-3-trimethylsilylpropionate, an internal NMR reference standard) and 3 mM sodium azide. 1H NMR data were acquired using a Varian Inova NMR spectrometer operating at 600.36 MHz for 1H and equipped with a 1H-{15N, 13C} flow probe (120 μl active volume) and a Varian automated sample transport accessory (VAST). Four-hundred-fifty microliters of sample was injected into the probe with no push solvent. Two cell rinses were completed with isotonic phosphate buffer between samples. The push buffer was prepared by mixing one part 0.2 M sodium phosphate buffer at pH 7.4 with two parts water. One-dimensional 1H NMR spectra were acquired at 27°C using a one-dimensional NOESY pulse sequence including water presaturation and a mixing time of 100 milliseconds. A total of 64 scans were collected with 64k data points, an acquisition time of 2.73 s, an inter-pulse delay of 1 s, and a sweep width of 12 kHz.
NMR data analysis.
Spectra were processed and analyzed using in-house software, Metabonomi (International Patent Application, WO2004038602A1). Spectra were normalized to the total integrated spectral area minus the region containing water and urea from 6.0 to 4.5 ppm.
Mass spectrometry.
Samples were centrifuged and injected directly onto the HPLC column undiluted for metabolite profiling and characterization. For exact mass determination, urine samples were diluted 1:1000 in water prior to analysis. Metabolite profiling and characterization was performed using a Surveyor LC system (ThermoFinnigan, San Jose, CA) coupled via an electrospray interface to a LCQ Deca XP Plus ion trap mass spectrometer (ThermoFinnigan, San Jose, CA). Data was collected in both positive and negative ionization modes. Operating parameters included a source voltage of 4.5 kV, capillary temperature of 250°C, capillary voltage of 21 V, and a tube lens offset of 25V. Helium was used as the sheath gas and auxiliary gas at flow settings of 80 (arb.) and 20 (arb.), respectively. Data acquired in full scan mode was collected over a scan range of 80–1200 amu. Collision induced fragmentation was performed for selected ions of interest, using helium as a collision gas with an activation amplitude of 20–30% and an isolation width of 1.5–1.8 amu. Product ion spectra were acquired for selected ions of interest. Chromatographic separation of metabolites was achieved using a Symmetry C-18 column 150 x 2.1 mm, 5 μm (Waters, Milford, CT) under gradient conditions starting at 100% 0.1% formic acid (v/v) and 0% acetonitrile, increasing linearly to 20% acetonitrile at 4 min, then to 95% acetonitrile at 8 min and holding for 1 min before re-equilibration at initial conditions. Flow rate was held constant at 300 μl/min.
Exact mass determination was conducted at PE Sciex, Concord, Canada, using an LC-MS system consisting of an Agilent 1100 high performance liquid chromatography system (Agilent Technologies, Torrence, CA) coupled to a QSTAR XL hybrid triple quadrupole orthogonal time of flight mass spectrometer (PE Sciex, Foster City, CA) via a Turbospray interface.
LC-NMR/MS spectroscopic analysis.
LC-NMR/MS experiments were conducted using an integrated LC-NMR/MS system from Varian (Palo Alto, CA). The LC system consisted of a Star 9012 solvent delivery system and Star 9050 UV/Vis detector set at 254 nM. MS data were acquired using a 1200L triple quadrupole mass spectrometer, and NMR spectra were acquired using a Varian Inova NMR spectrometer operating at 600.36 MHz for 1H. Data acquisition was controlled with VnmrJ software, version 1.1C, operating on the NMR spectrometer's host computer. Urine samples of interest were analyzed by injecting a 0.5 ml aliquot onto a Synergi Polar RP semi-preparative column 5 μm, 250 x 10 mm (Phenomenex, Torrance, CA) and eluted under gradient conditions beginning with 80% deuterium oxide containing 0.1% formic acid: 20% acetonitrile, increasing linearly to 95% acetonitrile over 40 min and holding for 5 min before re-equilibration at initial conditions. A flow rate of 2 ml/min was used. After passing through the UV detector, the flow was split so that 5% routed to the mass spectrometer and 95% routed to the NMR flow probe. Instrument parameters were set to acquire MS and NMR data continuously during the chromatography run. The mass spectrometer was coupled to the LC system via an electrospray interface, operating in the positive ion mode. Ionization parameters for the mass spectrometer included applied needle and capillary voltages of 5000 V and 100 V, respectively with nitrogen used as the drying gas at 17 psi with a source temperature of 300°C and the nebulizing gas at 57 psi. Data were acquired in the full scan mode over a scan range of 100–1000 amu. 1H NMR spectra were acquired using the lc1d pulse sequence supplied by Varian. The spectral parameters included a sweep width of 10,000 Hz, an acquisition time of 1.1 s, 8k data points, and 8 transients per slice. The effluent from the NMR probe was collected in fractions during the run. In the specific case of identifying the metabolite with a MS peak at m/z 340 [M+H]+, mass spectrometry was used to confirm the presence of the metabolite in isolated fractions and guided the pooling of fractions for further purification. Final purification was achieved by injecting the concentrated fraction onto a Symmetry C-18 column, 5 μm, 150 x 3.9 mm (Waters, Milford, CT) and eluting the component with gradient conditions beginning with 80% deuterium oxide containing 0.1% formic acid: 20% acetonitrile, increasing linearly to 50% acetonitrile over 25 min. A flow rate of 1 ml/min was used. The effluent from the NMR probe was collected in fractions during the run. As before, mass spectrometry was used to confirm the presence of the metabolite in fractions and guided the pooling of fractions for the final sample. In the final sample, the LC solvents were removed by lyophilization and the isolated metabolite was dissolved in 550 μl of water (90:10 H2O:D2O) for analysis by 1H NMR spectroscopy.
RESULTS
Phenotype Identification and Characterization:
Prominent differences between the phenotypes were noted in 1H NMR spectra. Based on the clear and consistent predominance of aromatic components in the urine, hippuric acid, or chlorogenic acid metabolites, these two phenotypes are referred to as HIP (high hippuric acid) or CA (high chlorogenic acid metabolites). A selected region of 1H NMR spectra of urine from SD rats of HIP and CA phenotypes is shown in Figure 1. After reviewing records of this laboratory's metabonomic studies conducted using Charles River-supplied rats, the unique source of the CA phenotype was identified as the Raleigh facility, Room 9. A summary of metabonomics experiments is shown in Table 1 including the Charles River facility and room of origin and date of shipment receipt. In additional studies with Charles River-supplied rats from the Raleigh facility, the HIP phenotype was observed in male and female SD rats from Room 10, while the CA phenotype was observed in male and female SD rats from Room 9. Major metabolites contributing to each phenotype were identified and levels quantitated using NMR, LC-MS, and combined LC-NMR/MS methods. The HIP and CA metabolite profiles are detailed in Table 2. NMR spectral assignments were completed using a combination of literature (Gavaghan et al., 2001; Nicholls et al., 2003) and commercially available authentic standards in all cases except the dihydro-quinolinone glucuronide associated with the HIP phenotype. Profiling experiments by LC-MS revealed that a metabolite at [M+H]+ m/z 340 was abundant in the Room 10 sample, while present only at trace levels in the Room 9 sample. A representative subset of samples from male and female rats, originating from both rooms, was analyzed to confirm this finding. Both source-induced and collision-induced fragmentation yielded a fragment ion at m/z 164 amu consistent with a neutral loss of 176 amu, indicative of a glucuronide conjugate. Accurate mass values were determined to be m/z 340.1016 and m/z 164.0679 amu. Based on these values, the molecular formula for the unknown at m/z 340 is C15H18NO8 and the fragment at m/z 164 is consistent with C9H9NO2 (both < 5 ppm mass deviation). Further experiments using an ion trap mass spectrometer, on-line LC-NMR/MS experiments, as well as NMR experiments on the isolated material led to the proposed dihydro-quinolinone structure (Fig. 3), a cyclic derivative of cinnamic acid. The MS fragmentation pattern and 1H NMR spectrum are shown in Figures 4 and 5, respectively. Authentic 3,4-dihydro-2(1H)-quinolinone (98% pure, Sigma-Aldrich Company) fragmented in a manner similar to the aglycone, with the predominant product ion forming from loss of 42 amu. The proposed assignment was further validated through comparison of NMR spectral data with authentic 3,4-dihydro-2(1H)-quinolinone and reference to 1H chemical shift reference tables that quantitate the effect of substituents on chemical shifts (Pretsch et al., 2000).
Stability of Phenotype
The metabolite profile of 10 rats from each phenotype was monitored in SD male rats for a period of four weeks. During this time, the rats were individually housed. After four weeks, the metabolite profile in the urine remained constant for 19 out of the 20 animals. One animal, initially having the CA phenotype, spontaneously began excreting urine having a metabolite profile consistent with the HIP phenotype after two weeks. For the remaining two weeks of monitoring, the urine metabolite profile of that animal remained consistent with the HIP phenotype. In a separate experiment, five pairs (one from each phenotype) of female SD rats were co-housed. Prior to cohabitation, the phenotype was confirmed by NMR and after cohabitation of one week, urine was again collected and analyzed. Results indicated that after one week of cohabitation, the rats initially having the CA phenotype had converted to excreting urine with a metabolite profile consistent with the HIP phenotype, Figure 2. After three weeks of cohabitation, the rats were then housed individually and monitored for an additional two weeks. The urine metabolite profile remained consistent with the HIP phenotype with no reversion to the original phenotype in CA animals.
DISCUSSION
This report documents a consistent phenotypic difference within Sprague-Dawley (Crl:CD(SD)) rats supplied by the Charles River Raleigh facility that has remained stable for at least 12 months. Though a phenotype change as described here could be attributed to genetic drift, that is probably not the case here—at least in the usual sense. The changes observed in this report are most likely due to differences in the gut flora populating the GI tract. This is based primarily on two lines of evidence. Firstly, the urinary biomolecular differences observed in this study are consistent with previous reports of microflora-induced changes in urinary metabolites, particularly hippuric acid and chlorogenic acid metabolites (Gavaghan et al., 2001; Phipps et al., 1998). The second line of evidence is that when rats were housed in pairs, one from each phenotype, there was a relatively rapid switch of the phenotype of CA animals. This was most likely due to coprophagic inoculation of HIP-derived bacteria into the CA rats. The rapidity of the change ruled out an endogenous genetic component.
The GI tract is home to a vast host of microbiota with hundreds of species of bacteria present in the normal mammalian GI tract (Guarner and Malagelada, 2003). Bacterial density varies with the topography of the GI tract with < 1000 bacterial colony forming units (CFU) per ml present in the stomach to several hundred billion CFU/ml present in the colon (Hart et al., 2002). Comprehensive bacterial species identification of this complex ecology is difficult, if not impossible, particularly as speciation of the gut flora varies with species, age, and diet (Brennan-Craddock et al., 1992; Hebuterne, 2003; Rowland et al., 2000). However, an overview of the variety of gut flora populations in normal and disease states in humans has been published (Salminen et al., 1995). The role of gut flora in health is well recognized (Hart et al., 2002; Hooper et al., 1998, 2002; Hooper and Gordon, 2001; Simon and Gorbach, 1984; Stagg et al., 2004) with the fields of prebiotics and probiotics based on the concept of alteration of gut flora to treat disease or otherwise improve health (Gibson, 1998; Karimi and Pena, 2003). Therefore, it should not be terribly surprising that alterations in gut flora may cause fairly profound systemic effects. Nicholson and colleagues (Nicholson et al., 2005; Nicholson and Wilson, 2003), make a strong case for the integration of the gut floral biology into any assessment of systems biology or systems toxicology. Metabonomics as a technology excels at just such assessments, so it is not surprising that many of the anecdotal reports identifying varying phenotypes within a rat strain, have this analytical approach as their basis. Phipps et al. (1998) reported that Alp:ApfSD (Wistar derived) rats maintained using standard husbandry condition did not excrete hippuric acid in the urine with 3-HPPA constituting the primary aromatic component. These rats would be somewhat similar to the CA rats described here. Interestingly, they found that simple addition of benzoic acid to the diet caused the animals to initiate hippuric acid excretion, which did not reverse after returning to the original diet. The authors implicated microfloral changes as an explanation for this lack of reversal. While we report similar results, they vary in some important respects. Firstly, the CA rats observed in these studies had measurable, though low, levels of hippuric acid. More importantly, the diet was not the source of the variation, in our studies since animals were obtained from the same facility with identical husbandry practices between rooms. Gavaghan et al. (2001) noted phenotypic variation in Sprague-Dawley rats consistent with that observed here in that they also noted phenotypes defined by high and low levels of hippuric acid instead of its presence and absence as reported by Phipps. The authors indicate that the phenotypes represent subpopulations but are unclear as to the source of their animals or the stability of these subpopulations.
As animals are born, raised, and distributed from a single room within the facility it is possible that after an initiating (and currently unexplained) event, gut floral changes occurred within the Room 9 population and have been passed from dam to weanling propagating the changes ever since. Another alternative explanation involves the sourcing of rats for Raleigh Room 9. The colony of rats maintained in Room 9 was originally started from foundation colony rats that had only eight species of bacteria, the Charles River Altered Schaedler Flora (Charles Clifford, Charles River Laboratory, personal communication). A detailed description of the Schaedler flora can be found in the article by Sarma-Rupavtarm and colleagues (Sarma-Rupavtarm et al., 2004). These "isolators" are maintained separately and used as a source for new colony initiation. Even after several years, these isolators have very limited number of gut bacteria species. Of the facilities we evaluated, Room 9 represented the most recent colony initiation in December 2002. The colony in P09 initiated in March 2000, while the colonies in K93 and K97 initiated in May 1997. As colonies are initiated and moved to barrier rooms for propagation, a repopulation of a "normal" gut flora may occur over time. In this scenario, the gut floral variation isn't really a species shift, but is due to a lack of the "normal" speciation that occurs as rats are exposed to various bacteria, which evidently may be quite protracted. This scenario is supported by studies conducted by Nicholls et al. (2003) who monitored urine metabolic profile change in axenic Fisher-344 rats exposed to a normal laboratory environment to allow gut floral repopulation. They reported that initially low levels of urinary hippuric acid increased dramatically after 21 days of exposure to the normal laboratory environment. The primary argument against this scenario is the timing. If the colony in Room 9 had been initiated weeks or even months ago, this scenario would seem more likely. However, why would there be such a striking difference in a colony initiated three years ago versus one started five years ago Nevertheless, this scenario deserves further investigation.
Of greatest concern with regard to these findings is what, if any, significance these phenotypic changes engender. These observations raise some interesting questions from a toxicologic perspective. The fact that these changes are not simply a reflection of diet-induced modulation of the microflora, but can persist for some time is highly significant. The gut flora are known to influence metabolism of numerous compounds which can profoundly affect toxification and detoxification pathways (Boxenbaum et al., 1979; Goodwin et al., 1994; Mikov, 1994; Rowland, 1981, 1988). The role of gut flora has been directly linked to the toxicity (or lack of toxicity) of numerous compounds including nitrobenzene, amygdalin, and digoxin to name a few (Mathan et al., 1989; Reddy et al., 1976; Rowland, 1988). Activation (or lack of detoxification) by gut flora can play a role in tumor induction by exogenous compounds (Rowland and Walker, 1983; Yoshida et al., 2003). If indeed the phenotypic changes are stable for even longer than the four weeks evaluated in this report, perhaps of even greater interest is what effect microflora changes may have on exposure to natural (i.e., dietary derived) metabolites (e.g., chlorogenic acid metabolites) and what effect these changes will have on background tumor incidence. If such differences in gut flora are persistent, they represent a potential lifetime exposure to relatively high concentrations of these varying metabolites. Could altered microflora be one explanation for what otherwise seems inexplicably fast drift in background tumor incidences More generally, are these epigenetic variations hidden sources of variation in preclinical animal studies These questions remain to be answered.
It is possible that these effects simply reflect a transient phenomenon that comes and goes within populations. Alternatively, rigorous efforts to maintain "clean" animal facilities may in some cases cause problems rather than avoid them. In either case the findings provide an opportunity to more fully understand the role of microflora in rat systems biology. As it is likely these changes have occurred from time to time in the past and are almost certainly characteristic of all laboratory animal suppliers, one could argue that they have had little effect on our ability to conduct and interpret toxicology studies. However, the truth of the matter is that we don't know what we don't know. It is plausible that anomalous findings that occasionally pop up in toxicology studies are not really anomalous but attributable to unrecognized microfloral changes. Likewise, the interstudy "variability" we have grown accustomed to in toxicology studies may not really be a biologic constant but rather may be at least partially attributable to definable causes, such as microfloral differences. What is undeniable is that the newer "omic" technologies are very sensitive to off target (systemic) changes. Therefore, it is imperative that studies using these technologies understand the nature of their animal models, including the role gut flora play. The complexity of life is a beautiful thing, though it certainly can be confusing sometimes. As new technologies delve ever deeper into systems biology, it can be anticipated that the more we learn, the more we will need to critically reassess our most basic assumptions.
In conclusion, we have identified a phenotypic difference in commercially supplied Sprague-Dawley (Crl:CD(SD)) rats that has been stable within a single room of the Charles River Raleigh facility for a period of at least 12 months. This phenotypic difference is most likely due to altered gut flora and is typified by diminished urinary hippuric acid and increased chlorogenic acid metabolites.
ACKNOWLEDGMENTS
The authors gratefully acknowledge Pauline Vollmerhaus, Ron Bonner, and Lyle Burton at MDS Sciex for their assistance in obtaining accurate masses that helped lead to the identification of the dihydro-quinolinone glucuronide. The authors further gratefully acknowledge the co-operation and valuable insights into Charles River Breeding facility practices provided by Dr. Charles Clifford. Conflict of interest: none declared.
REFERENCES
Beckwith-Hall, B. M., Brindle, J. T., Barton, R. H., Coen, M., Holmes, E., Nicholson, J. K., and Antti, H. (2002a). Application of orthogonal signal correction to minimise the effects of physical and biological variation in high resolution 1H NMR spectra of biofluids. Analyst 127, 1283–1288.
Beckwith-Hall, B. M., Holmes, E., Lindon, J. C., Gounarides, J., Vickers, A., Shapiro, M., and Nicholson, J. K. (2002b). NMR-based metabonomic studies on the biochemical effects of commonly used drug carrier vehicles in the rat. Chem. Res. Toxicol. 15, 1136–1141.
Bollard, M. E., Holmes, E., Lindon, J. C., Mitchell, S. C., Branstetter, D., Zhang, W., and Nicholson, J. K. (2001). Investigations into biochemical changes due to diurnal variation and estrus cycle in female rats using high-resolution (1)H NMR spectroscopy of urine and pattern recognition. Anal. Biochem. 295, 194–202.
Boxenbaum, H. G., Bekersky, I., Jack, M. L., and Kaplan, S. A. (1979). Influence of gut microflora on bioavailability. Drug Metab. Rev. 9, 259–279.
Brennan-Craddock, W. E., Mallett, A. K., Rowland, I. R., and Neale, S. (1992). Development changes to gut microflora metabolism in mice. J. Appl. Bacteriol. 73, 163–167.
Gavaghan, C. L., Nicholson, J. K., Connor, S. C., Wilson, I. D., Wright, B., and Holmes, E. (2001). Directly coupled high-performance liquid chromatography and nuclear magnetic resonance spectroscopic with chemometric studies on metabolic variation in Sprague-Dawley rats. Anal. Biochem. 291, 245–252.
Gibson, G. R. (1998). Dietary modulation of the human gut microflora using prebiotics. Br. J. Nutr. 80, S209–S212.
Goodwin, B. L., Ruthven, C. R., and Sandler, M. (1994). Gut flora and the origin of some urinary aromatic phenolic compounds. Biochem. Pharmacol. 47, 2294–2297.
Guarner, F., and Malagelada, J. R. (2003). Gut flora in health and disease. Lancet 361, 512–519.
Hart, A. L., Stagg, A. J., Frame, M., Graffner, H., Glise, H., Falk, P., and Kamm, M. A. (2002). The role of the gut flora in health and disease, and its modification as therapy. Aliment. Pharmacol. Ther. 16, 1383–1393.
Hebuterne, X. (2003). Gut changes attributed to ageing: Effects on intestinal microflora. Curr. Opin. Clin. Nutr. Metab. Care 6, 49–54.
Hooper, L. V., Bry, L., Falk, P. G., and Gordon, J. I. (1998). Host-microbial symbiosis in the mammalian intestine: exploring an internal ecosystem. Bioessays 20, 336–343.
Hooper, L. V., and Gordon, J. I. (2001). Commensal host-bacterial relationships in the gut. Science 292, 1115–1118.
Hooper, L. V., Midtvedt, T., and Gordon, J. I. (2002). How host-microbial interactions shape the nutrient environment of the mammalian intestine. Annu. Rev. Nutr. 22, 283–307.
Karimi, O., and Pena, A. S. (2003). Probiotics: Isolated bacteria strain or mixtures of different strains Two different approaches in the use of probiotics as therapeutics. Drugs Today (Barc.) 39, 565–597.
Mathan, V. I., Wiederman, J., Dobkin, J. F., and Lindenbaum, J. (1989). Geographic differences in digoxin inactivation, a metabolic activity of the human anaerobic gut flora. Gut 30, 971–977.
Mikov, M. (1994). The metabolism of drugs by the gut flora. Eur. J. Drug Metab. Pharmacokinet. 19, 201–207.
Nicholls, A. W., Mortishire-Smith, R. J., and Nicholson, J. K. (2003). NMR spectroscopic-based metabonomic studies of urinary metabolite variation in acclimatizing germ-free rats. Chem. Res. Toxicol. 16, 1395–1404.
Nicholson, J. K., Holmes, E., and Wilson, I. D. (2005). Gut microorganisms, mammalian metabolism and personalized health care. Nat. Rev. Microbiol. 3, 431–438.
Nicholson, J. K., and Wilson, I. D. (2003). Opinion: understanding ‘global’ systems biology: Metabonomics and the continuum of metabolism. Nat. Rev. Drug Discov. 2, 668–676.
Phipps, A. N., Stewart, J., Wright, B., and Wilson, I. D. (1998). Effect of diet on the urinary excretion of hippuric acid and other dietary-derived aromatics in rat. A complex interaction between diet, gut microflora and substrate specificity. Xenobiotica 28, 527–537.
Pretsch, E., Buhlmann, P., and Affolter, C. (2000). Structure Determination of Organic Compounds. Tables of Spectral Data. Springer-Verlag, Berlin.
Reddy, B. G., Pohl, L. R., and Krishna, G. (1976). The requirement of the gut flora in nitrobenzene-induced methemoglobinemia in rats. Biochem. Pharmacol. 25, 1119–1122.
Robertson, D. G., Reily, M. D., Sigler, R. E., Wells, D. F., Paterson, D. A., and Braden, T. K. (2000). Metabonomics: Evaluation of nuclear magnetic resonance (NMR) and pattern recognition technology for rapid in vivo screening of liver and kidney toxicants. Toxicol. Sci. 57, 326–337.
Rowland, I. (1981). The influence of the gut microflora on food toxicity. Proc. Nutr. Soc. 40, 67–74.
Rowland, I. R. (1988). Interactions of the gut microflora and the host in toxicology. Toxicol. Pathol. 16, 147–153.
Rowland, I. R., and Walker, R. (1983). The gastrointestinal tract in food toxicology. In Toxic Hazards in Food (D. M. Conning and A. B. G. Lansdown, Eds.), pp. 183–274. Croom Helm, London.
Rowland, I. R., Wiseman, H., Sanders, T. A., Adlercreutz, H., and Bowey, E. A. (2000). Interindividual variation in metabolism of soy isoflavones and lignans: Influence of habitual diet on equol production by the gut microflora. Nutr. Cancer 36, 27–32.
Salminen, S., Isolauri, E., and Onnela, T. (1995). Gut flora in normal and disordered states. Chemotherapy 41(Suppl. 1), 5–15.
Sarma-Rupavtarm, R. B., Ge, Z., Schauer, D. B., Fox, J. G., and Polz, M. F. (2004). Spatial distribution and stability of the eight microbial species of the altered schaedler flora in the mouse gastrointestinal tract. Appl. Environ. Microbiol. 70, 2791–2800.
Simon, G. L., and Gorbach, S. L. (1984). Intestinal flora in health and disease. Gastroenterology 86, 174–193.
Stagg, A. J., Hart, A. L., Knight, S. C., and Kamm, M. A. (2004). Microbial-gut interactions in health and disease. Interactions between dendritic cells and bacteria in the regulation of intestinal immunity. Best Pract. Res. Clin. Gastroenterol. 18, 255–270.
Yoshida, K., Kuroda, K., Zhou, X., Inoue, Y., Date, Y., Wanibuchi, H., Fukushima, S., and Endo, G. (2003). Urinary sulfur-containing metabolite produced by intestinal bacteria following oral administration of dimethylarsinic acid to rats. Chem. Res. Toxicol. 16, 1124–1129.(Lora C. Robosky, Dale F. )