Approaches to Bacterial RNA Isolation and Purification for Microarray Analysis of Escherichia coli K1 Interaction with Human Brain Microvasc
http://www.100md.com
微生物临床杂志 2005年第8期
Division of Pediatric Infectious Diseases, Johns Hopkins University School of Medicine, 600 North Wolfe Street, Park 256, Baltimore, MD 21287
ABSTRACT
We established a protocol for isolation of microarray-grade bacterial RNA from Escherichia coli K1 interacting with human brain microvascular endothelial cells. The extracted RNA was free of human RNA contamination. More importantly, microarray analysis demonstrated that no bias was introduced in the gene expression pattern during the RNA isolation procedure.
TEXT
DNA microarrays offer new opportunities for exploring the molecular pathogenesis of infectious diseases. It is possible to analyze the whole gene expression of a bacterial pathogen during its interaction with the host. Such information can lead to the identification of virulence factors, the elements to which they respond, and the mechanisms by which they are regulated (1, 2, 6, 14, 15, 17). The feasibility of this approach, however, depends upon the ability to recover biologically relevant bacterial RNA, and serious consideration should be given to prevent gene expression changes associated with preparative procedures (6, 10). This is a difficult task, since bacteria are capable of rapid transcriptional responses to the environment, a fact extensively mentioned in the past but hardly addressed in full. Moreover, RNA must be obtained from biologically relevant models of infection that reflect the complexity of pathogen-host interaction. Such models are amalgams of pathogens and host cells, which make DNA microarray analysis even more challenging.
We developed a protocol for RNA extraction from Escherichia coli K1 interacting with human brain microvascular endothelial cells (HBMEC). HBMEC represent the in vitro model of the human blood-brain barrier, and we have previously demonstrated that E. coli K1 interaction with HBMEC is a biologically relevant model pertaining to the pathogenesis of E. coli meningitis in vivo (7-9, 19, 20). We initially attempted to coextract bacterial and human RNA from a mixed sample of E. coli K1 and HBMEC. This approach did not require time-consuming steps that can affect bacterial gene expression but creates other technical issues for microarray analysis, as human RNA can compete with bacterial RNA during cDNA synthesis and labeling, limit the amount of bacterial RNA that can be used, and impede accurate quantitation of bacterial RNA. Moreover, the ratio between host and pathogen RNA may vary among samples, making normalization of microarray data difficult. Thus, it was necessary to eliminate the human RNA, which could be done by hybridization capture using MICROBEnrich (Ambion). Total RNA was extracted from HBMEC and associated bacteria a using RiboPure-Bacteria kit (Ambion) that includes zirconia-silica beads for bead beating with a special vortex adapter (Ambion). Bead beating was necessary to lyse E. coli K1. Human RNA extracted using this method was found to be degraded, whether extracted from the mixed HBMEC-bacteria sample (28S/18S = 0.33; Fig. 1A) or a pure HBMEC culture (28S/18S = 1.2; data not shown). RNA extracted from a pure culture of E. coli K1 (data not shown) produced high quality RNA as expected (23S/16S = 1.8). Degraded human RNA cannot be removed by hybridization capture and could affect microarray analysis.
As the coextraction protocol did not produce microarray-grade RNA, we developed an alternative method using differential lysis. HBMEC were eliminated using RLT lysis buffer (QIAGEN) that caused immediate and complete lysis of HBMEC but did not affect E. coli K1 (Fig. 2). Intact bacteria were separated from HBMEC quickly to avoid any alteration of bacterial gene expression. This was necessary to obtain a sample that was representative of E. coli K1 interacting with HBMEC. HBMEC infected with E. coli K1 strain RS218 (18) were prepared as previously described (13) in 150-mm dishes. Nonassociated bacteria were collected from the culture supernatant by centrifugation and resuspended in RLT buffer (QIAGEN, Valencia, CA). The infected monolayers were then washed once with M199-Ham F12 (1:1; Invitrogen), resuspended in RLT buffer (QIAGEN), and immediately collected from the plates. Both suspensions were vortexed for 1 min and centrifuged for 1 min at high speed. The supernatant (RLT buffer with or without HBMEC lysate) was discarded, and the bacterial pellet was extracted using a RiboPure-Bacteria kit (Ambion). RNA was cleaned up and concentrated using an RNeasy Mini kit (QIAGEN) with on-column DNase treatment. Samples that were processed with this method yielded 25 μg of bacterial RNA of outstanding quality, free of human RNA contamination (Fig. 1B). This bacterial RNA sample was found to be suitable for microarray analysis. We compared the expression profiles of associated and nonassociated bacteria using a DNA microarray designed in our laboratory. Our E. coli DNA microarray consists of 8,239 oligonucleotides (50-mer) arrayed onto aminosilane slides (UltraGAPS; Corning) and covers each open reading frame in E. coli K-12 strain MG1655, E. coli O157:H7 strains EDL 933 and RIMD0509952/VT2-Sakai, most open reading frames in uropathogenic E. coli strains (CFT073, 536), and meningitis-causing E. coli strains (RS218, C5). This DNA microarray analysis was carried out as previously described (21).
Differential lysis was previously applied to pathogen-host systems (3-5, 10-12, 16). However, it was unclear whether or not bacterial gene expression could be altered during differential lysis. Applying RNA stabilization reagents (RNAlater, Ambion; or RNAprotect bacteria; QIAGEN) caused a carry-over of degraded host RNA (data not shown) which was not suitable for microarray analysis. We examined, using our E. coli DNA microarray, whether our differential lysis protocol could introduce any bias in the bacterial gene expression pattern. E. coli RS218 in early-log phase was prepared and split into three aliquots. One was immediately extracted (reference) while the others were treated with RLT buffer (QIAGEN) for 5 or 15 min to simulate differential lysis. Except for the absence of HBMEC, RNA was extracted from these samples as in our differential lysis protocol. Samples were compared using our E. coli microarray as previously described (21). The expression pattern of bacteria incubated in RLT buffer for 5 min was nearly identical to that of untreated E. coli RS218 (Fig. 3), and the correlation coefficient (r2) was essentially indistinguishable between a reference-reference self hybridization (0.995) versus a 5-min-RLT-treated-reference hybridization (0.992). In contrast, longer RLT buffer treatment (15 min) resulted in higher variability, and more than 90 genes appeared differentially expressed (data not shown). Based on these findings we concluded that E. coli K1 gene expression was not significantly altered during a 5-min differential lysis treatment with RLT buffer. Our protocol easily allows the procedure to stay within this time limit. Indeed, a few seconds are sufficient to completely lyse the HBMEC. Most mammalian cell lines will be lysed equally fast, thereby our protocol can be applied to a variety of bacteria-host infection models as several bacteria can withstand a short incubation in RLT buffer. Other lysis buffers can also be applied providing that preservation of bacterial gene expression is properly verified.
ACKNOWLEDGMENTS
This work was supported by National Institutes of Health grants NS-26310 and AI-47225.
REFERENCES
Carulli, J. P., M. Artinger, P. M. Swain, C. D. Root, L. Chee, C. Tulig, J. Guerin, M. Osborne, G. Stein, J. Lian, and P. T. Lomedico. 1998. High throughput analysis of differential gene expression. J. Cell Biochem. 31(Suppl. 30):286-296.
Cummings, C. A., and D. A. Relman. 2000. Using DNA microarrays to study host-microbe interactions. Emerg. Infect. Dis. 6:513-525.
Eriksson, S., S. Lucchini, A. Thompson, M. Rhen, and J. C. Hinton. 2003. Unravelling the biology of macrophage infection by gene expression profiling of intracellular Salmonella enterica. Mol. Microbiol. 47:103-118.
Grifantini, R., E. Bartolini, A. Muzzi, M. Draghi, E. Frigimelica, J. Berger, F. Randazzo, and G. Grandi. 2002. Gene expression profile in Neisseria meningitidis and Neisseria lactamica upon host-cell contact: from basic research to vaccine development. Ann. N. Y. Acad. Sci. 975:202-216.
Grifantini, R., E. Bartolini, A. Muzzi, M. Draghi, E. Frigimelica, J. Berger, G. Ratti, R. Petracca, G. Galli, M. Agnusdei, M. M. Giuliani, L. Santini, B. Brunelli, H. Tettelin, R. Rappuoli, F. Randazzo, and G. Grandi. 2002. Previously unrecognized vaccine candidates against group B meningococcus identified by DNA microarrays. Nat. Biotechnol. 20:914-921.
Hinton, J. C., I. Hautefort, S. Eriksson, A. Thompson, and M. Rhen. 2004. Benefits and pitfalls of using microarrays to monitor bacterial gene expression during infection. Curr. Opin. Microbiol. 7:277-282.
Huang, S. H., C. Wass, Q. Fu, N. V. Prasadarao, M. Stins, and K. S. Kim. 1995. Escherichia coli invasion of brain microvascular endothelial cells in vitro and in vivo: molecular cloning and characterization of invasion gene ibe10. Infect. Immun. 63:4470-4475.
Kim, K. S. 2001. Escherichia coli translocation at the blood-brain barrier. Infect. Immun. 69:5217-5222.
Kim, K. S. 2003. Pathogenesis of bacterial meningitis: from bacteraemia to neuronal injury. Nat. Rev. Neurosci. 4:376-385.
Mangan, J. A., I. M. Monahan, and P. D. Butcher. 2002. Gene expression during host-pathogen interactions: approaches to bacterial mRNA extraction and labeling for microarray analysis, p. 137-151. In B. W. Wren and N. Dorrel (ed.), Functional microbial genomics, vol. 33. Academic Press, London, United Kingdom.
Mangan, J. A., K. M. Sole, D. A. Mitchison, and P. D. Butcher. 1997. An effective method of RNA extraction from bacteria refractory to disruption, including mycobacteria. Nucleic Acids Res. 25:675-676.
Orihuela, C. J., J. N. Radin, J. E. Sublett, G. Gao, D. Kaushal, and E. I. Tuomanen. 2004. Microarray analysis of pneumococcal gene expression during invasive disease. Infect. Immun. 72:5582-5596.
Prasadarao, N. V., C. A. Wass, J. N. Weiser, M. F. Stins, S. H. Huang, and K. S. Kim. 1996. Outer membrane protein A of Escherichia coli contributes to invasion of brain microvascular endothelial cells. Infect. Immun. 64:146-153.
Rappuoli, R. 2000. Pushing the limits of cellular microbiology: microarrays to study bacteria-host cell intimate contacts. Proc. Natl. Acad. Sci. USA 97:13467-13469.
Relman, D. A. 2002. Genome-wide responses of a pathogenic bacterium to its host. J. Clin. Investig. 110:1071-1073.
Schnappinger, D., S. Ehrt, M. I. Voskuil, Y. Liu, J. A. Mangan, I. M. Monahan, G. Dolganov, B. Efron, P. D. Butcher, C. Nathan, and G. K. Schoolnik. 2003. Transcriptional adaptation of Mycobacterium tuberculosis within macrophages: insights into the phagosomal environment. J. Exp. Med. 198:693-704.
Schoolnik, G. K. 2002. Functional and comparative genomics of pathogenic bacteria. Curr. Opin. Microbiol. 5:20-26.
Silver, R. P., W. Aaronson, A. Sutton, and R. Schneerson. 1980. Comparative analysis of plasmids and some metabolic characteristics of Escherichia coli K1 from diseased and healthy individuals. Infect. Immun. 29:200-206.
Stins, M. F., P. V. Nemani, C. Wass, and K. S. Kim. 1999. Escherichia coli binding to and invasion of brain microvascular endothelial cells derived from humans and rats of different ages. Infect. Immun. 67:5522-5525.
Stins, M. F., N. V. Prasadarao, L. Ibric, C. A. Wass, P. Luckett, and K. S. Kim. 1994. Binding characteristics of S fimbriated Escherichia coli to isolated brain microvascular endothelial cells. Am. J. Pathol. 145:1228-1236.
Teng, C. H., M. Cai, S. Shin, Y. Xie, K. J. Kim, N. A. Khan, F. Di Cello, and K. S. Kim. 2005. Escherichia coli K1 RS218 interacts with human brain microvascular endothelial cells via type 1 fimbria phase-on bacteria. Infect. Immun. 73:2923-2931.(Francescopaolo Di Cello, )
ABSTRACT
We established a protocol for isolation of microarray-grade bacterial RNA from Escherichia coli K1 interacting with human brain microvascular endothelial cells. The extracted RNA was free of human RNA contamination. More importantly, microarray analysis demonstrated that no bias was introduced in the gene expression pattern during the RNA isolation procedure.
TEXT
DNA microarrays offer new opportunities for exploring the molecular pathogenesis of infectious diseases. It is possible to analyze the whole gene expression of a bacterial pathogen during its interaction with the host. Such information can lead to the identification of virulence factors, the elements to which they respond, and the mechanisms by which they are regulated (1, 2, 6, 14, 15, 17). The feasibility of this approach, however, depends upon the ability to recover biologically relevant bacterial RNA, and serious consideration should be given to prevent gene expression changes associated with preparative procedures (6, 10). This is a difficult task, since bacteria are capable of rapid transcriptional responses to the environment, a fact extensively mentioned in the past but hardly addressed in full. Moreover, RNA must be obtained from biologically relevant models of infection that reflect the complexity of pathogen-host interaction. Such models are amalgams of pathogens and host cells, which make DNA microarray analysis even more challenging.
We developed a protocol for RNA extraction from Escherichia coli K1 interacting with human brain microvascular endothelial cells (HBMEC). HBMEC represent the in vitro model of the human blood-brain barrier, and we have previously demonstrated that E. coli K1 interaction with HBMEC is a biologically relevant model pertaining to the pathogenesis of E. coli meningitis in vivo (7-9, 19, 20). We initially attempted to coextract bacterial and human RNA from a mixed sample of E. coli K1 and HBMEC. This approach did not require time-consuming steps that can affect bacterial gene expression but creates other technical issues for microarray analysis, as human RNA can compete with bacterial RNA during cDNA synthesis and labeling, limit the amount of bacterial RNA that can be used, and impede accurate quantitation of bacterial RNA. Moreover, the ratio between host and pathogen RNA may vary among samples, making normalization of microarray data difficult. Thus, it was necessary to eliminate the human RNA, which could be done by hybridization capture using MICROBEnrich (Ambion). Total RNA was extracted from HBMEC and associated bacteria a using RiboPure-Bacteria kit (Ambion) that includes zirconia-silica beads for bead beating with a special vortex adapter (Ambion). Bead beating was necessary to lyse E. coli K1. Human RNA extracted using this method was found to be degraded, whether extracted from the mixed HBMEC-bacteria sample (28S/18S = 0.33; Fig. 1A) or a pure HBMEC culture (28S/18S = 1.2; data not shown). RNA extracted from a pure culture of E. coli K1 (data not shown) produced high quality RNA as expected (23S/16S = 1.8). Degraded human RNA cannot be removed by hybridization capture and could affect microarray analysis.
As the coextraction protocol did not produce microarray-grade RNA, we developed an alternative method using differential lysis. HBMEC were eliminated using RLT lysis buffer (QIAGEN) that caused immediate and complete lysis of HBMEC but did not affect E. coli K1 (Fig. 2). Intact bacteria were separated from HBMEC quickly to avoid any alteration of bacterial gene expression. This was necessary to obtain a sample that was representative of E. coli K1 interacting with HBMEC. HBMEC infected with E. coli K1 strain RS218 (18) were prepared as previously described (13) in 150-mm dishes. Nonassociated bacteria were collected from the culture supernatant by centrifugation and resuspended in RLT buffer (QIAGEN, Valencia, CA). The infected monolayers were then washed once with M199-Ham F12 (1:1; Invitrogen), resuspended in RLT buffer (QIAGEN), and immediately collected from the plates. Both suspensions were vortexed for 1 min and centrifuged for 1 min at high speed. The supernatant (RLT buffer with or without HBMEC lysate) was discarded, and the bacterial pellet was extracted using a RiboPure-Bacteria kit (Ambion). RNA was cleaned up and concentrated using an RNeasy Mini kit (QIAGEN) with on-column DNase treatment. Samples that were processed with this method yielded 25 μg of bacterial RNA of outstanding quality, free of human RNA contamination (Fig. 1B). This bacterial RNA sample was found to be suitable for microarray analysis. We compared the expression profiles of associated and nonassociated bacteria using a DNA microarray designed in our laboratory. Our E. coli DNA microarray consists of 8,239 oligonucleotides (50-mer) arrayed onto aminosilane slides (UltraGAPS; Corning) and covers each open reading frame in E. coli K-12 strain MG1655, E. coli O157:H7 strains EDL 933 and RIMD0509952/VT2-Sakai, most open reading frames in uropathogenic E. coli strains (CFT073, 536), and meningitis-causing E. coli strains (RS218, C5). This DNA microarray analysis was carried out as previously described (21).
Differential lysis was previously applied to pathogen-host systems (3-5, 10-12, 16). However, it was unclear whether or not bacterial gene expression could be altered during differential lysis. Applying RNA stabilization reagents (RNAlater, Ambion; or RNAprotect bacteria; QIAGEN) caused a carry-over of degraded host RNA (data not shown) which was not suitable for microarray analysis. We examined, using our E. coli DNA microarray, whether our differential lysis protocol could introduce any bias in the bacterial gene expression pattern. E. coli RS218 in early-log phase was prepared and split into three aliquots. One was immediately extracted (reference) while the others were treated with RLT buffer (QIAGEN) for 5 or 15 min to simulate differential lysis. Except for the absence of HBMEC, RNA was extracted from these samples as in our differential lysis protocol. Samples were compared using our E. coli microarray as previously described (21). The expression pattern of bacteria incubated in RLT buffer for 5 min was nearly identical to that of untreated E. coli RS218 (Fig. 3), and the correlation coefficient (r2) was essentially indistinguishable between a reference-reference self hybridization (0.995) versus a 5-min-RLT-treated-reference hybridization (0.992). In contrast, longer RLT buffer treatment (15 min) resulted in higher variability, and more than 90 genes appeared differentially expressed (data not shown). Based on these findings we concluded that E. coli K1 gene expression was not significantly altered during a 5-min differential lysis treatment with RLT buffer. Our protocol easily allows the procedure to stay within this time limit. Indeed, a few seconds are sufficient to completely lyse the HBMEC. Most mammalian cell lines will be lysed equally fast, thereby our protocol can be applied to a variety of bacteria-host infection models as several bacteria can withstand a short incubation in RLT buffer. Other lysis buffers can also be applied providing that preservation of bacterial gene expression is properly verified.
ACKNOWLEDGMENTS
This work was supported by National Institutes of Health grants NS-26310 and AI-47225.
REFERENCES
Carulli, J. P., M. Artinger, P. M. Swain, C. D. Root, L. Chee, C. Tulig, J. Guerin, M. Osborne, G. Stein, J. Lian, and P. T. Lomedico. 1998. High throughput analysis of differential gene expression. J. Cell Biochem. 31(Suppl. 30):286-296.
Cummings, C. A., and D. A. Relman. 2000. Using DNA microarrays to study host-microbe interactions. Emerg. Infect. Dis. 6:513-525.
Eriksson, S., S. Lucchini, A. Thompson, M. Rhen, and J. C. Hinton. 2003. Unravelling the biology of macrophage infection by gene expression profiling of intracellular Salmonella enterica. Mol. Microbiol. 47:103-118.
Grifantini, R., E. Bartolini, A. Muzzi, M. Draghi, E. Frigimelica, J. Berger, F. Randazzo, and G. Grandi. 2002. Gene expression profile in Neisseria meningitidis and Neisseria lactamica upon host-cell contact: from basic research to vaccine development. Ann. N. Y. Acad. Sci. 975:202-216.
Grifantini, R., E. Bartolini, A. Muzzi, M. Draghi, E. Frigimelica, J. Berger, G. Ratti, R. Petracca, G. Galli, M. Agnusdei, M. M. Giuliani, L. Santini, B. Brunelli, H. Tettelin, R. Rappuoli, F. Randazzo, and G. Grandi. 2002. Previously unrecognized vaccine candidates against group B meningococcus identified by DNA microarrays. Nat. Biotechnol. 20:914-921.
Hinton, J. C., I. Hautefort, S. Eriksson, A. Thompson, and M. Rhen. 2004. Benefits and pitfalls of using microarrays to monitor bacterial gene expression during infection. Curr. Opin. Microbiol. 7:277-282.
Huang, S. H., C. Wass, Q. Fu, N. V. Prasadarao, M. Stins, and K. S. Kim. 1995. Escherichia coli invasion of brain microvascular endothelial cells in vitro and in vivo: molecular cloning and characterization of invasion gene ibe10. Infect. Immun. 63:4470-4475.
Kim, K. S. 2001. Escherichia coli translocation at the blood-brain barrier. Infect. Immun. 69:5217-5222.
Kim, K. S. 2003. Pathogenesis of bacterial meningitis: from bacteraemia to neuronal injury. Nat. Rev. Neurosci. 4:376-385.
Mangan, J. A., I. M. Monahan, and P. D. Butcher. 2002. Gene expression during host-pathogen interactions: approaches to bacterial mRNA extraction and labeling for microarray analysis, p. 137-151. In B. W. Wren and N. Dorrel (ed.), Functional microbial genomics, vol. 33. Academic Press, London, United Kingdom.
Mangan, J. A., K. M. Sole, D. A. Mitchison, and P. D. Butcher. 1997. An effective method of RNA extraction from bacteria refractory to disruption, including mycobacteria. Nucleic Acids Res. 25:675-676.
Orihuela, C. J., J. N. Radin, J. E. Sublett, G. Gao, D. Kaushal, and E. I. Tuomanen. 2004. Microarray analysis of pneumococcal gene expression during invasive disease. Infect. Immun. 72:5582-5596.
Prasadarao, N. V., C. A. Wass, J. N. Weiser, M. F. Stins, S. H. Huang, and K. S. Kim. 1996. Outer membrane protein A of Escherichia coli contributes to invasion of brain microvascular endothelial cells. Infect. Immun. 64:146-153.
Rappuoli, R. 2000. Pushing the limits of cellular microbiology: microarrays to study bacteria-host cell intimate contacts. Proc. Natl. Acad. Sci. USA 97:13467-13469.
Relman, D. A. 2002. Genome-wide responses of a pathogenic bacterium to its host. J. Clin. Investig. 110:1071-1073.
Schnappinger, D., S. Ehrt, M. I. Voskuil, Y. Liu, J. A. Mangan, I. M. Monahan, G. Dolganov, B. Efron, P. D. Butcher, C. Nathan, and G. K. Schoolnik. 2003. Transcriptional adaptation of Mycobacterium tuberculosis within macrophages: insights into the phagosomal environment. J. Exp. Med. 198:693-704.
Schoolnik, G. K. 2002. Functional and comparative genomics of pathogenic bacteria. Curr. Opin. Microbiol. 5:20-26.
Silver, R. P., W. Aaronson, A. Sutton, and R. Schneerson. 1980. Comparative analysis of plasmids and some metabolic characteristics of Escherichia coli K1 from diseased and healthy individuals. Infect. Immun. 29:200-206.
Stins, M. F., P. V. Nemani, C. Wass, and K. S. Kim. 1999. Escherichia coli binding to and invasion of brain microvascular endothelial cells derived from humans and rats of different ages. Infect. Immun. 67:5522-5525.
Stins, M. F., N. V. Prasadarao, L. Ibric, C. A. Wass, P. Luckett, and K. S. Kim. 1994. Binding characteristics of S fimbriated Escherichia coli to isolated brain microvascular endothelial cells. Am. J. Pathol. 145:1228-1236.
Teng, C. H., M. Cai, S. Shin, Y. Xie, K. J. Kim, N. A. Khan, F. Di Cello, and K. S. Kim. 2005. Escherichia coli K1 RS218 interacts with human brain microvascular endothelial cells via type 1 fimbria phase-on bacteria. Infect. Immun. 73:2923-2931.(Francescopaolo Di Cello, )