• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of jcmPermissionsJournals.ASM.orgJournalJCM ArticleJournal InfoAuthorsReviewers
J Clin Microbiol. Dec 2002; 40(12): 4607–4611.
PMCID: PMC154653

Use of Subtractive Hybridization To Identify a Diagnostic Probe for a Cystic Fibrosis Epidemic Strain of Pseudomonas aeruginosa


A multiresistant strain of Pseudomonas aeruginosa is widespread among cystic fibrosis (CF) patients attending clinics in Liverpool, United Kingdom. Suppression subtractive hybridization was used to identify sequences present in the Liverpool CF epidemic strain but absent from strain PAO1. Using dot blot and PCR amplification assays, the prevalence of such sequences among a panel of CF isolates was determined. Several sequences were found only in the Liverpool epidemic strain. Some sequences were present in the Liverpool epidemic strain and in a minority of other isolates, including sequences with homology to genes implicated in O6 serotype and siderophore production. The Liverpool epidemic strain and 81% of nonepidemic isolates contained a sequence identified as part of the PAGI-1 genomic island. Other strains implicated in epidemic spread, which were from Manchester, United Kingdom, and Melbourne, Australia, were also screened. None of the sequences identified was present in the Manchester strain. However, one of two Melbourne strains contained some of the sequences found in the Liverpool epidemic strain. All isolates implicated in epidemic spread and 76% of sporadic isolates contained the exoS gene. A sequence present in all isolates of the Liverpool epidemic strain was used to develop a diagnostic PCR test for identification of the strain from colonies or directly from sputum samples.

Chronic infection with Pseudomonas aeruginosa occurs in up to 80% of adult cystic fibrosis (CF) patients and is associated with increased morbidity and mortality. In 1996, Cheng et al. used molecular typing methods to demonstrate that many CF patients in our local pediatric clinic were colonized by the same multiresistant strain of P. aeruginosa and proposed that this was due either to cross infection or to acquisition from a common environmental source (3). Despite extensive sampling, no evidence supporting the latter explanation has been found, suggesting that the strain is transmissible. Recent evidence also suggests that this epidemic strain can cause superinfection of already colonized adult CF patients (11). Again, no evidence of hospital environmental contamination was found and all patients became superinfected following close contact with colonized individuals during inpatient treatment. In addition, there is evidence that the same epidemic strain of P. aeruginosa has been transmitted from an adult CF patient to her healthy parents, with subsequent significant morbidity (12). Thus, it seems that this multiresistant, transmissible epidemic strain possesses a number of characteristics that are not shared by other P. aeruginosa CF isolates.

Suppression subtractive hybridization (SSH) is a PCR-based technique that facilitates the identification of DNA sequences present in one strain (the Tester strain) but absent from another (the Driver strain) (5). SSH has been successfully applied to a number of different bacterial pathogens (22). Other versions of subtractive hybridization have been applied to P. aeruginosa (17), and differential hybridization of arrayed libraries was used to identify a genomic island present in most pathogenic isolates of P. aeruginosa (9). However, there have been no previous reports of the use of SSH for the identification of sequence variations between P. aeruginosa strains. Here, we report the use of SSH to identify genomic sequences present in a CF epidemic strain but absent from strain PAO1, the entire genome of which has been sequenced (20). Using these data, we report the development of a simple PCR amplification assay for identification and detection of the Liverpool CF epidemic strain from colonies and directly from sputum.


Bacterial strains.

The strains of the CF P. aeruginosa isolates used in this study fall into three categories: CF nonepidemic strains from Merseyside, England, and surrounding regions, the Liverpool CF epidemic strain, and other strains implicated in epidemic spread. All strains were confirmed as P. aeruginosa by using a PCR assay for detection of the exotoxin A gene (8). Nonepidemic strains were selected on the basis of genetic distinctiveness according to macrorestriction analysis and pulsed-field gel electrophoresis (PFGE) (11). In accordance with previous studies, isolates with PFGE patterns differing by three or fewer bands were considered to harbor the same strain (21). Several separate isolates (identified by PFGE as harboring the Liverpool CF epidemic strain) from different patients were included, as were other isolates harboring strains implicated in epidemic spread, which were from Manchester, United Kingdom (7), and Melbourne, Australia (D. S. Armstrong, G. Nixon, J. Carlin, R. Carzino, and K. Grimwood, abstract from the 14th Annu. N. Am. Cystic Fibrosis Conf., Baltimore, Md. Pediatr. Pulmonol. 30(Suppl.):285, abstract 393, 2000).

Biochemical and genetic screening of isolates.

A number of other properties were assessed to ensure that important phenotypes and genotypes of P. aeruginosa were represented within the panel. We used frequency of spontaneous resistance to rifampin (300 μg ml−1), measured using the method described by Oliver et al. (15), to identify hypermutable strains among P. aeruginosa isolates. Unlike the Liverpool and Manchester epidemic strains, 4 of 15 nonepidemic strains screened were hypermutable. In addition, we used a PCR amplification assay for detection of the presence of the exoS gene (10), which is associated with invasive ability (6). The exoS gene was present in the Liverpool epidemic strain, each of the Manchester and Melbourne isolates, and 16 of the 21 sporadic isolates. Nonepidemic strains also differed with respect to auxotrophy (growth on glucose M9 salt minimal medium), motility, siderophore production (detected using chrome azurol S agar [18]), and protease activity (determined using a plate assay for the detection of casein hydrolysis [19]).

Construction and screening of subtraction libraries.

Genomic DNA was isolated as described previously (23) from strain PAO1 and an isolate identified as the Liverpool CF epidemic strain. SSH was carried out using the CLONTECH PCR-Select bacterial genome subtraction kit (Clontech) as recommended by the supplier, except a hybridization temperature of 73°C was used to take account of the higher G+C content of the organism. In the hybridization, DNA from the CF epidemic strain was used as the tester and PAO1 DNA was used as the driver. PCR products obtained following SSH were cloned into pGEM-T (Invitrogen) to produce a subtracted DNA library. The subtraction library was screened by sequencing of PCR amplicons from individual clones. Amplification and sequencing were carried out by using M13 forward and reverse vector primers. Clones for sequencing were selected on the basis of variations in amplicon size and restriction site. Sequences obtained were used in BLAST searches at the Pseudomonas genome project web site (http://www.pseudomonas.com) to determine for each sequence its presence in or absence from the PAO1 genome. Tester-specific sequences were further analyzed by using BLASTN and BLASTX searches of the database at the website http://www.ncbi.nlm.nih.gov.

PCR amplification and dot blot screening of strains.

The oligonucleotide primers (Sigma-Genosys) used in PCR assays and for the labeling of probes are listed in Table Table1,1, along with the annealing temperatures used. Typically, crude DNA in boiled suspensions (1 μl) was used directly in 50-μl volumes containing 2.5 U of Taq DNA polymerase (Helena Biosciences, Sunderland, Tyne & Wear, United Kingdom), 1× TaqMaster (Helena Biosciences), 200 nM (each) primers, 1× Taq buffer, 1.5 mM MgCl2, and 100 μM nucleotides (dATP, dCTP, dGTP, and dTTP). Amplifications were carried out in an Eppendorf MasterCycler thermal cycler for 30 cycles consisting of 95°C (1 min), the annealing temperature (1 min), and 72°C (2 min), with an additional extension time at 72°C (10 min) following completion of the 30 cycles. At the end of the amplification, 5-μl samples were subjected to electrophoresis on a standard 1.0% (wt/vol) agarose gel to confirm the presence of an amplified product. Dot blot hybridization of genomic DNA was carried out as described previously (23). Digoxigenin-11-2′-dUTP (DIG) (Roche)-labeled probes were made by carrying out PCR amplification in the presence of 60 μM DIG by using vector or internal primers. Hybridization and subsequent detection of DIG was carried out following the manufacturer's instructions (Roche).

Oligonucleotide primers used for PCR amplification

PCR assays of sputum.

Sputum samples were emulsified with mixing at room temperature for 20 min in an equal volume of Sputasol (Oxoid). Subsequently, DNA was isolated from samples by using a QIAamp DNA Mini kit (Qiagen) and by employing the manufacturer's protocol for tissue. The presence of P. aeruginosa DNA was determined by PCR amplification of the exotoxin A gene (8). As a control for bacterial DNA, amplification of 16S ribosomal DNA conserved sequences by using the oligonucleotide primers PSR and PSL (2) was carried out.


Analysis of the subtraction library.

Using SSH, we obtained a total of 55 subtractive clones. These clones were prescreened by using PCR amplification coupled with restriction enzyme digestion to ensure that only clones with different inserts were subjected to sequencing. Sequence data was obtained from 22 of the clones, 8 of whose sequences matched strongly (>85% nucleotide identity) with the PAO1 sequence. A further three clones contained sequences that were partial matches with that of PAO1 but also contained novel sequences. Therefore, 64% of clones contained partial or total tester-specific sequences. These sequences were used as probes in BLASTN and BLASTX searches against the general database (Table (Table2).2). Three of the sequences (PS40, PS43, and PS48) were 100% identical in nucleotide sequence to previously reported non-PAO1 P. aeruginosa sequences: the sequences of wbpO and wzx, which are different parts of a gene cluster encoding the biosynthesis of B-band O-antigen of P. aeruginosa serotype O6 (1), and that of the genomic island reported by Liang et al. (9). Other homologies identified using BLASTX searches included one match with a PAO1 protein of unknown function (PS9) and a sequence that matched with a TonB-dependent receptor from PAO1 but gave a better match with an iron-siderophore receptor from a strain of P. putida (PS5). This has subsequently been identified as a type III pyoverdine receptor (P. Cornelis, personal communication). One sequence (PSA) was homologous to the C-terminal sequence of Neisseria meningitidis FhaB, a putative adhesin, and is the subject of further study.

Analysis of subtracted library sequences

Screening of P. aeruginosa isolates.

Thirteen of the subtractive clones were used in PCR and dot blot assays to determine the prevalence of subtracted library sequences among members of the panel of P. aeruginosa CF isolates. Dot blot assays were carried out for all thirteen sequences. Subsequently, simple PCR assays were developed for PS1, PS3, PS5, PS9, PS12, PS15, PS21, PS43, PS54, and PSA. For the subtracted sequences that included partial overlap with the PAO1 genome, PCR and dot blot assays were used to target only the sequences absent from PAO1. The distribution of subtraction library sequences among the CF isolates is summarized in Table Table3.3. PS40 was present in most of the isolates tested, indicating the widespread distribution of the previously reported genomic island (9) among CF isolates. Interestingly, PS40 was absent from the Manchester and Melbourne strains. PS43 was present in eight nonepidemic isolates and in the Melbourne strain C3789, identifying these strains as serotype O6. The siderophore receptor sequence (PS5) was found in only three nonepidemic isolates and was present in one of the Melbourne strains implicated in epidemic spread (C3789). PS2 was also present in several strains, including C3789. None of the subtracted sequences were present in the Manchester epidemic strain. Several of the sequences were present only in isolates identified as harboring the Liverpool CF epidemic strain or were found infrequently among nonepidemic strains. Notably, PS1, PS3, PS12, PS21, and PS54 were found only in the isolates identified (by using PFGE) as the CF epidemic strain. Of these, only PS21 could be identified in all of the epidemic strain isolates (Table (Table3).3). The other sequences were lacking in isolate 109 but present in all other isolates representing the Liverpool epidemic strain. Isolate 109 differs by at least one band from the other epidemic strain isolates, as determined by PFGE. Because our primary aim was to identify a sequence diagnostic for the epidemic strain while including all isolates that are presently designated (on the basis of PFGE results) as harboring the epidemic strain, we chose the sequence of PS21 as the preferred marker for the epidemic strain. PS21 contains no complete reading frames, and when predicted protein or nucleotide sequences were used as probes, no significant homology of the sequence of PS21 with other database sequences was found.

Distribution of subtracted sequences among P. aeruginosa isolates

PCR assays from sputum samples.

Typical results obtained for PCR assays of sputum samples by using the PS21 primer set are presented in Fig. Fig.11 for sputum samples (i) from a patient colonized with the Liverpool epidemic strain, (ii) from a patient colonized by a different strain of P. aeruginosa, or (iii) from a patient not colonized by P. aeruginosa.

FIG. 1.
PCR assays from sputum samples. The figure is a composite of three agarose gels, showing PCR assay results for PS21 (specific for the epidemic strain) (A), the exotoxin A gene (specific for P. aeruginosa) (B), and the 16S ribosomal DNA conserved sequence ...


In our study, we found that the levels of G+C content of subtractive clone inserts were significantly lower than the overall level for the genome (66.6% [20]). This phenomenon has been observed with other G+C-rich bacterial pathogens (16) and suggests that the method of SSH is particularly effective at detecting regions of DNA that might have been acquired from other, less G+C-rich organisms. Atypical G+C content is a common feature of pathogenicity islands, and it is likely that some of the sequences identified in this study form part of the genomic regions implicated in the greater transmissibility of the Liverpool CF epidemic strain. These possibilities are being investigated currently.

The effectiveness of SSH as a technique was demonstrated by the identification of a number of sequences associated with strain differences in P. aeruginosa. In a previous study, the incidence of the genomic island PAGI-1 among clinical isolates of P. aeruginosa was reported as 85%. These isolates included blood and urinary tract infection isolates as well as CF isolates (9). Our study suggests that PAGI-1 was present in the Liverpool epidemic strain and in 81% of the CF nonepidemic strains, which is a finding consistent with the reported figure. P. aeruginosa produces pyoverdine, a fluorescent yellow-green siderophore. Three structurally different pyoverdines have been identified (named type I, type II, and type III) (13). In a previous study of pyoverdine production by CF isolates of P. aeruginosa, only type I and type II pyoverdine producers were identified (4). Our study suggests that some CF isolates, including the Liverpool epidemic strain, are type III pyoverdine producers.

A number of isolates, including those harboring the Liverpool epidemic strain and one of the Melbourne strains, carried at least one gene specific to the P. aeruginosa O6 serotype, which is not an unusual serotype for CF isolates (14). PS2, which has some homology with an O-antigen modification protein, was found only in strains that were also positive for the O6 serotype wbpO gene (PS43). However, three strains identified as positive for PS43 were negative for PS2.

Up to 2/3 of the adults attending the Cardiothoracic Centre in Liverpool are colonized by the Liverpool transmissible strain. In addition, four adults previously colonized with unique strains of P. aeruginosa were subsequently superinfected by the epidemic strain (11) and it has been shown to cause pneumonia in parents of a CF patient infected with the strain (12). In our previous study, we found that 92 (76.7%) of 120 CF-infected children were colonized with the Liverpool epidemic strain (3). It is interesting that none of the sequences identified in our study were found in the Manchester strain reported as epidemic (7). This strain was present in 22 (14%) of 154 CF patients colonized with P. aeruginosa (7). The Liverpool and Manchester strains are genetically distinct, as demonstrated by PFGE results. These observations suggest that the Liverpool epidemic strain is fundamentally different from the Manchester strain. Although none of the sequences identified as specific to the Liverpool epidemic strain were found in the Melbourne strains, one of the strains did contain some of the other sequences found in the Liverpool epidemic strain and a minority of nonepidemic strains (PS2, PS5, and PS43).

We believe that the panel of strains used to test for the incidence of subtracted sequences was representative of P. aeruginosa CF isolates. Strains differing in PFGE type, as well as in characteristics such as auxotrophy, hypermutability, motility, and siderophore production, were included. We found that the exoS gene, the presence of which is indicative of an invasive phenotype (6), was present in all of the isolates implicated in epidemic spread and in 76% of the sporadic isolates. In a previous study of P. aeruginosa isolates associated with various infections, a majority (65% of 145 strains) possessed exoS. Indeed, 80 to 100% of strains from urine, lungs, wounds, and feces were positive for exoS (10). Our study indicates that there is a similarly high prevalence of the exoS gene among CF isolates.

In summary, this study indicates that the multiresistant Liverpool CF epidemic strain is serotype O6, carries the exoS gene, is not hypermutable, carries a genomic island associated with pathogenic strains, and contains a type III pyoverdine receptor. Since the results of PCR and dot blot assays for the PS21 sequence correlate, we suggest that a simple PCR assay, using oligonucleotide primers PS21F and PS21R, can be used for effective identification of the Liverpool CF epidemic strain without the need for screening by PFGE, which is time consuming and requires greater levels of technical expertise. Other PCR assays are also available for PS1, PS3, PS12, PS21, and PS54, which are sequences found only in the Liverpool epidemic strain although they were absent from one isolate (isolate 109). Such assays should enable researchers and physicians in other CF units throughout the world to determine whether strains related to the Liverpool CF epidemic strain are distributed widely. However, since our strain collection cannot be assumed to be representative of all P. aeruginosa strains throughout the world, positive PCR tests should be followed up by molecular typing. We have demonstrated that the PCR assay can be used directly to determine the presence of the epidemic strain in sputum. Previous analysis has relied upon culture followed by molecular typing of colony types. This approach is time consuming and cannot guarantee detection of the epidemic strain when other P. aeruginosa strain types are present in higher numbers. We are now conducting a wider-ranging study to assess the prevalence of the Liverpool epidemic strain in our CF centers by using the diagnostic PCR assay directly on sputum samples.


C.W. and C.A.H. acknowledge funding from the United Kingdom Cystic Fibrosis Trust.

We also thank John Corkill for assistance with PFGE typing and John Govan (Department of Medical Microbiology, University Medical School, Edinburgh, Scotland) for providing some of the strains.


1. Belanger, M., L. L. Burrows, and J. S. Lam. 1999. Functional analysis of genes responsible for the synthesis of the B-band O antigen of Pseudomonas aeruginosa serotype O6 lipopolysaccharide. Microbiology 145:3505-3521. [PubMed]
2. Campbell, P. W., III, J. A. Phillips III, G. J. Heidecker, M. R. Krishnamani, R. Zahorchak, and T. L. Stull. 1995. Detection of Pseudomonas (Burkholderia) cepacia using PCR. Pediatr. Pulmonol. 20:44-49. [PubMed]
3. Cheng, K., R. L. Smyth, J. R. Govan, C. Doherty, C. Winstanley, N. Denning, D. P. Heaf, H. van Saene, and C. A. Hart. 1996. Spread of beta-lactam-resistant Pseudomonas aeruginosa in a cystic fibrosis clinic. Lancet 348:639-642. [PubMed]
4. De Vos, D., M. De Chial, C. Cochez, S. Jansen, B. Tummler, J. M. Meyer, and P. Cornelis. 2001. Study of pyoverdine type and production by Pseudomonas aeruginosa isolated from cystic fibrosis patients: prevalence of type II pyoverdine isolates and accumulation of pyoverdine-negative mutations. Arch. Microbiol. 175:384-388. [PubMed]
5. Diatchenko, L., S. Lukyanov, Y. F. Lau, and P. D. Siebert. 1999. Suppression subtractive hybridization: a versatile method for identifying differentially expressed genes. Methods Enzymol. 303:349-380. [PubMed]
6. Fleiszig, S. M., J. P. Wiener-Kronish, H. Miyazaki, V. Vallas, K. E. Mostov, D. Kanada, T. Sawa, T. S. Yen, and D. W. Frank. 1997. Pseudomonas aeruginosa-mediated cytotoxicity and invasion correlate with distinct genotypes at the loci encoding exoenzyme S. Infect. Immun. 65:579-586. [PMC free article] [PubMed]
7. Jones, A. M., J. R. Govan, C. J. Doherty, M. E. Dodd, B. J. Isalska, T. N. Stanbridge, and A. K. Webb. 2001. Spread of a multiresistant strain of Pseudomonas aeruginosa in an adult cystic fibrosis clinic. Lancet 358:557-558. [PubMed]
8. Khan, A. A., and C. E. Cerniglia. 1994. Detection of Pseudomonas aeruginosa from clinical and environmental samples by amplification of the exotoxin A gene using PCR. Appl. Environ. Microbiol. 60:3739-3745. [PMC free article] [PubMed]
9. Liang, X., X. Q. Pham, M. V. Olson, and S. Lory. 2001. Identification of a genomic island present in the majority of pathogenic isolates of Pseudomonas aeruginosa. J. Bacteriol. 183:843-853. [PMC free article] [PubMed]
10. Lomholt, J. A., K. Poulsen, and M. Kilian. 2001. Epidemic population structure of Pseudomonas aeruginosa: evidence for a clone that is pathogenic to the eye and that has a distinct combination of virulence factors. Infect. Immun. 69:6284-6295. [PMC free article] [PubMed]
11. McCallum, S. J., J. Corkill, M. Gallagher, M. J. Ledson, C. A. Hart, and M. J. Walshaw. 2001. Superinfection with a transmissible strain of Pseudomonas aeruginosa in adults with cystic fibrosis chronically colonised by P. aeruginosa. Lancet 358:558-560. [PubMed]
12. McCallum, S. J., M. J. Gallagher, J. E. Corkill, C. A. Hart, M. J. Ledson, and M. J. Walshaw. 2002. Spread of an epidemic Pseudomonas aeruginosa strain from a patient with cystic fibrosis (CF) to non-CF relatives. Thorax 57:559-560. [PMC free article] [PubMed]
13. Meyer, J. M., A. Stintzi, D. De Vos, P. Cornelis, R. Tappe, K. Taraz, and H. Budzikiewicz. 1997. Use of siderophores to type pseudomonads: the three Pseudomonas aeruginosa pyoverdine systems. Microbiology 143:35-43. [PubMed]
14. Millesimo, M., G. de Intinis, M. G. Chirillo, T. Musso, and D. Savoia. 1996. Pseudomonas aeruginosa clinical isolates: serotypes, resistance phenotypes and plasmid profiles. Eur. J. Epidemiol. 12:123-129. [PubMed]
15. Oliver, A., R. Canton, P. Campo, F. Baquero, and J. Blazquez. 2000. High frequency of hypermutable Pseudomonas aeruginosa in cystic fibrosis lung infection. Science 288:1251-1254. [PubMed]
16. Reckseidler, S. L., D. DeShazer, P. A. Sokol, and D. E. Woods. 2001. Detection of bacterial virulence genes by subtractive hybridization: identification of capsular polysaccharide of Burkholderia pseudomallei as a major virulence determinant. Infect. Immun. 69:34-44. [PMC free article] [PubMed]
17. Schmidt, K. D., T. Schmidt-Rose, U. Romling, and B. Tummler. 1998. Differential genome analysis of bacteria by genomic subtractive hybridization and pulsed field gel electrophoresis. Electrophoresis 19:509-514. [PubMed]
18. Schwyn, B., and J. B. Neilands. 1987. Universal chemical assay for the detection and determination of siderophores. Anal. Biochem. 160:47-56. [PubMed]
19. Sokol, P. A., D. E. Ohman, and B. H. Iglewski. 1979. A more sensitive plate assay for detection of protease production by Pseudomonas aeruginosa. J. Clin. Microbiol. 9:538-540. [PMC free article] [PubMed]
20. Stover, C. K., X. Q. Pham, A. L. Erwin, S. D. Mizoguchi, P. Warrener, M. J. Hickey, F. S. Brinkman, W. O. Hufnagle, D. J. Kowalik, M. Lagrou, R. L. Garber, L. Goltry, E. Tolentino, S. Westbrock-Wadman, Y. Yuan, L. L. Brody, S. N. Coulter, K. R. Folger, A. Kas, K. Larbig, R. Lim, K. Smith, D. Spencer, G. K. Wong, Z. Wu, and I. T. Paulsen. 2000. Complete genome sequence of Pseudomonas aeruginosa PA01, an opportunistic pathogen. Nature 406:959-964. [PubMed]
21. Tenover, F. C., R. D. Arbeit, R. V. Goering, P. A. Mickelsen, B. E. Murray, D. H. Persing, and B. Swaminathan. 1995. Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: criteria for bacterial strain typing. J. Clin. Microbiol. 33:2233-2239. [PMC free article] [PubMed]
22. Winstanley, C. 2002. Spot the difference: applications of subtractive hybridisation to the study of bacterial pathogens. J. Med. Microbiol. 51:459-467. [PubMed]
23. Winstanley, C., and C. A. Hart. 2000. Presence of type III secretion genes in Burkholderia pseudomallei correlates with Ara phenotypes. J. Clin. Microbiol. 38:883-885. [PMC free article] [PubMed]

Articles from Journal of Clinical Microbiology are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • Gene (nucleotide)
    Gene (nucleotide)
    Records in Gene identified from shared sequence links
  • GSS
    Published GSS sequences
  • MedGen
    Related information in MedGen
  • Nucleotide
    Published Nucleotide sequences
  • Protein
    Published protein sequences
  • PubMed
    PubMed citations for these articles
  • Taxonomy
    Related taxonomy entry
  • Taxonomy Tree
    Taxonomy Tree

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...