5-Lipoxygenase and Leukotrienes
Transgenic Mouse and Nuclear Targeting Studies

COLIN D. FUNK and XIN-SHENG CHEN

Department of Pharmacology and Center for Experimental Therapeutics, University of Pennsylvania, Philadelphia, Pennsylvania

	INTRODUCTION

TOP INTRODUCTION A STRATEGY TO GENERATE... INITIAL CHARACTERIZATION OF... LEUKOTRIENE DEFICIENT MICE IN... 5-LIPOXYGENASE: A NUCLEAR... CONCLUDING REMARKS REFERENCES

A decade has passed since the molecular cloning of the 5-lipoxygenase gene and nearly twenty years since the discovery of leukotrienes. In our laboratory, we have strived to make advances in the realm of basic biology and understanding of the 5-lipoxygenase/leukotriene pathway by the use of knockout mice and other cell biology techniques. The aim of this article is to survey the studies conducted with these mice and to present novel information about the nuclear localization of 5-lipoxygenase.

	A STRATEGY TO GENERATE LEUKOTRIENE-DEFICIENT MICE

TOP INTRODUCTION A STRATEGY TO GENERATE... INITIAL CHARACTERIZATION OF... LEUKOTRIENE DEFICIENT MICE IN... 5-LIPOXYGENASE: A NUCLEAR... CONCLUDING REMARKS REFERENCES

In early 1991, as significant advances had been achieved in the area of in vivo manipulation of the mammalian genome through gene targeting using embryonic stem cell technology (1), we decided to adopt this approach for the study of lipoxygenase enzymes. The generation of gene "knockout" mice was in its infancy and appeared to hold great promise for unraveling biological functions of known and newly discovered genes (1, 2). Since 5-lipoxygenase was the most widely studied member of the lipoxygenase family, and perhaps the most well characterized at the time with respect to gene structure and biological roles, we initiated our experiments by generating a leukotriene deficient mouse before development of other lipoxygenase gene-disrupted animals.

The human 5-lipoxygenase gene was discovered to be a large gene of 14 exons, spread out over more than 80 kb on human chromosome 10 in a CpG-rich DNA segment (3, 4). The promoter region contained tandem repeats of CCGCCC and the regulation of 5-lipoxygenase expression was poorly understood (3, 5, 6). In fact, even at this time, little is known about 5-lipoxygenase gene regulatory mechanisms. A segment of the mouse 5-lipoxygenase gene that contained 3 of the 14 exons orthologous to the human gene was cloned (7). The exon/intron junctions for exons 4-6 were situated in analogous positions to the human gene boundaries but intron sizes differed significantly (7, 8).

A targeting vector was constructed from the mouse 5-lipoxygenase gene fragment containing exons 4 to 6 (7, 8). Exon 6 was disrupted by insertion of a neomycin resistance cassette. This portion of the gene encodes amino acids 220-277 of the 673-residue protein and lies upstream of the critical histidine iron ligand-encoding region of the protein.

	INITIAL CHARACTERIZATION OF LEUKOTRIENE-DEFICIENT MICE

TOP INTRODUCTION A STRATEGY TO GENERATE... INITIAL CHARACTERIZATION OF... LEUKOTRIENE DEFICIENT MICE IN... 5-LIPOXYGENASE: A NUCLEAR... CONCLUDING REMARKS REFERENCES

5-Lipoxygenase-deficient mice were generated with targeted embryonic stem cells (8). Mice were derived on a mixed C57BL/6 × 129 Sv genetic background and also on a "pure" 129 Sv strain. Southern blot analysis confirmed the disruption of the gene in the offspring. Western blot analysis, combined with enzyme activity assays, revealed the absence of 5-lipoxygenase protein and functional enzyme in bone marrow-derived mast cells (BMMCs) and peritoneal and alveolar macrophages. Thus, the first known mammal to lack completely the capability to synthesize leukotrienes had been artificially created in a laboratory setting.

Leukotriene-deficient mice developed without apparent abnormalities in utero, were outwardly normal in appearance, and reproduced without problem.

Cells that normally express 5-lipoxygenase and release leukotrienes when activated, including macrophages and BMMCs were isolated from wild-type and knockout mice. In vitro, the 5-lipoxygenase-deficient cells appeared to behave normally in most respects (8, 9). Thus, BMMCs released normal amounts of granule components when challenged with calcium ionophore A23187 and synthesized prostaglandins at levels approximately equivalent to those of wild-type mice. Peritoneal macrophages from 5-lipoxygenase-deficient mice synthesized normal prostaglandin levels in response to zymosan and had a partly diminished capacity to undergo phagocytosis when compared with wild-type cells. Since many of the cell-type specific functions of mast cells and macrophages have not been investigated yet it is likely that future studies will unravel a more definite function of the 5-lipoxygenase pathway.

	LEUKOTRIENE DEFICIENT MICE IN MODELS OF INFLAMMATION

TOP INTRODUCTION A STRATEGY TO GENERATE... INITIAL CHARACTERIZATION OF... LEUKOTRIENE DEFICIENT MICE IN... 5-LIPOXYGENASE: A NUCLEAR... CONCLUDING REMARKS REFERENCES

The characterization of leukotriene deficient mice has been reviewed (10). The major phenotypic findings in these mice from numerous experimental models are outlined in Table 1 (8). In essence, the results can be simplified into three general categories: no evident role for leukotrienes, "detrimental" pathophysiological effects of leukotrienes, and "beneficial" roles for leukotrienes (Table 2).

In the majority of models tested to date, leukotriene-deficient mice generally fare much better or the same as their wild-type littermates. In only one case, Klebsiella pneumoniae infection, did 5-lipoxygenase-deficient mice exhibit a poor outcome in terms of survival, thus indicating a beneficial role for leukotrienes in host defense in this laboratory model of airway infection (16). It remains to be determined if leukotrienes are important in host defense in human settings of pneumonial infection.

Presumably, the 5-lipoxygenase/leukotriene pathway did not evolve to perform detrimental actions to the host. Since this pathway is expressed primarily in activated inflammatory cells, leukotriene biosynthesis is probably playing a protective mechanism against infection or invasion of the host by specific pathogens whose identity remains to be determined. The 5-lipoxygenase/leukotriene system may represent a redundant pathway in a network of inflammatory mechanisms. The 5-lipoxygenase pathway does not appear to be expressed in lower organisms but it is present in some plants such as potato tubers (18). Most plant lipoxygenases represent pathways for generation of specific signaling molecules (e.g., jasmonic acid) or protective mechanisms in response to wounding (19). 5-Lipoxygenase-derived leukotrienes are likely serving a similar role in higher animals. Future studies with leukotriene-deficient mice should resolve these "beneficial" roles of leukotrienes.

	5-LIPOXYGENASE: A NUCLEAR PROTEIN IN MAST CELLS

TOP INTRODUCTION A STRATEGY TO GENERATE... INITIAL CHARACTERIZATION OF... LEUKOTRIENE DEFICIENT MICE IN... 5-LIPOXYGENASE: A NUCLEAR... CONCLUDING REMARKS REFERENCES

5-Lipoxygenase has long been known as a soluble cytosolic protein isolated from human and porcine leukocytes (20, 21). Around 1993, Peters-Golden and colleagues made observations that were to change our thinking about 5-lipoxygenase and leukotrienes (22). We were intrigued by their reports that 5-lipoxygenase was located at the nuclear envelope or in the nucleus of alveolar macrophages and rat basophilic leukemia (RBL-1) cells (23, 24). RBL-1 cells are phenotypically related to mast cells. As many of our studies in 5-lipoxygenase deficient mice had been carried out with BMMCs, we sought to characterize the localization of the enzyme in BMMCs. To our surprise, the enzyme was localized virtually exclusively in the nucleus of this primitive mast cell population (25). Activation of these cells with calcium ionophore or IgE/antigen led to altered intracellular expression patterns that included either punctate distribution in the nucleus and by the nuclear envelope or a perinuclear/nuclear envelope pattern, respectively (25).

The finding of 5-lipoxygenase in the nucleus has raised many questions of novel roles for the enzyme in the nucleus that may or may not be leukotriene dependent. Furthermore, the discovery that LTB₄ could act through a nuclear hormone receptor (PPAR-) has intensified this area of research (26). We carried out a study to examine determinants of 5-lipoxygenase nuclear targeting in various cell types (27). To analyze the phenomenon of nuclear 5-lipoxygenase localization in more detail we prepared a vector to produce fusion protein, with the enzyme tagged to the green fluorescent protein (GFP) from Aequoria victoria. In this way we could follow movement of the enzyme in living cells. Interestingly, when we transfected a plasmid for this fusion protein into various cell types the enzyme was invariably expressed in the nucleus. This was true for HEK 293 cells, COS-1 cells, Chinese hamster ovary cells, and NIH 3T3 fibroblasts (Figure 1). In all cases, the fusion protein colocalized with the nuclear DNA dye Hoechst 33258. We also transfected BMMCs obtained from 5-lipoxygenase knockout mice with a 5-lipoxygenase vector and again the enzyme was expressed just like in the native state (i.e., in the nucleus).

View larger version (36K): [in this window] [in a new window]   Figure 1.   Cells (NIH 3T3, Chinese hamster ovary [CHO], COS-1, and human embryonic kidney [HEK] 293) transfected with a GFP- 5-lipoxygenase fusion protein vector express in the nucleus enzyme that colocalizes with the nuclear DNA stain Hoechst 33258.

Many proteins contain classic nuclear localization signal (NLS) sequences consisting of a basic stretch of amino acids (mainly arginine and lysine) either in a cluster, or in a bipartite organization (28). 5-Lipoxygenase, in the original cDNA cloning article (5), was noted to contain several clusters of basic residues. Although none were perfect matches to a consensus classic NLS, some were potential NLS motifs. We set out by in vitro mutagenesis and transfection studies with GFP-5LO to analyze the importance of the three major basic regions (Table 3) in nuclear targeting. Basic region 1 or basic region 3 mutations (double replacements of lysine and/or arginine to glutamine) did not affect enzyme activity or the ability of the enzyme to enter the nucleus (Table 4). In contrast, basic region 2 mutation abolished enzyme activity and nuclear localization. Since mutation of this region abolished enzyme activity it was troublesome that nuclear localization was also disrupted. The altered cellular localization may have been due to a protein folding problem. Therefore, we prepared other GFP-5LO mutant proteins in which mutations in the native protein are known to abolish enzyme activity. These mutations were placed at points critical for binding iron (29, 30). 5-Lipoxygenase contains a non-heme iron atom essential for catalytic activity (31). On the basis of the crystal structures of soybean and rabbit reticulocyte lipoxygenases and mutational analyses of 5-lipoxygenase (29, 30, 32, 33), there are three important histidine residues (H367, H372, and H550) and a C-terminal isoleucine (I673) that participate in binding the iron atom and presumably that are essential for maintaining proper structural integrity for enzyme activity. Mutations known to abolish completely the presence of iron in 5-lipoxygenase (H372Q, H550Q, and C6-deletion [deletion of the C-terminal six amino acid residues, including I673]) led to blockade of nuclear targeting with most of the enzyme in the transfected cells being localized in the cytosol (Table 4). Mutations known to abolish enzyme activity yet leave a partial presence of iron in the protein (H367Q, H367N) showed a graded distribution of 5-lipoxygenase in the nucleus and cytosol depending on the iron content (27). Thus, these studies were suggestive of a proper folding requirement for nuclear localization.

To address the question of whether specific regions of 5-lipoxygenase are sufficient for directing nuclear localization, various fragments of 5-lipoxygenase were fused to GFP to see if a specific region of the protein could direct nuclear localization. An N-terminal 5-LO fragment of either 127 or 80 amino acids could direct a nuclear localization pattern (Figure 2). The pattern of nuclear localization gave a mottled appearance as previously determined for 5-lipoxygenase in alveolar macrophages (24). In contrast, neither an N-terminal fragment of platelet 12-lipoxygenase nor a 90-residue C-terminal 5-lipoxygenase fragment directed nuclear localization (Figure 2). A segment within the latter region was reported by Lepley and co-workers (34) to possibly control 5-lipoxygenase nuclear localization. This result appears to contradict our findings and may be due to different modes of detection of the enzyme (direct visualization in living cells versus membrane fractionation combined with Western blot analysis).

View larger version (41K): [in this window] [in a new window]   Figure 2.   N-terminal 5-lipoxygenase fragments direct nuclear localization of GFP. Both 80- and 127-residue N-terminal fragments direct nuclear localization of GFP that gives a mottled-like appearance (top two panels). A 90-residue C-terminal 5-lipoxygenase fragment and a platelet 12-lipoxygenase 75-residue fragment did not lead to nuclear localization (i.e., not localized to Hoechst 33258 fluorescent regions) (bottom two panels).

In contrast to our results with a GFP reporter protein, 5-lipoxygenase fusions to the cytosolic protein pyruvate kinase could not direct nuclear localization (27). The combined data from more than 25 different fusion protein constructs indicate that 5-lipoxygenase nuclear targeting is dependent on a set of parameters that is not easily discernible from typical targeting experiments. Thus, 5-lipoxygenase does not possess a "classic" basic region NLS and proper folding of the enzyme is critical for nuclear localization. A weak NLS of undefined sequence seems to be present in the N-terminal 80 residues of 5-lipoxygenase. This segment of 5-lipoxygenase is predicted by molecular modeling to lie within a short -barrel N-terminal domain of unknown function. Gillmor and colleagues (33) have hypothesized that the -barrel N-terminal region of 5-lipoxygenase, with homology to lipoprotein lipase, may interact with 5-lipoxygenase-activating protein (FLAP), a coaccessory protein in leukotriene biosynthesis, and participate in binding to membrane. Our data support a potential role for the N-terminal domain of 5-lipoxygenase in aiding nuclear localization. Data from our laboratory (27) suggest that the folding of this -barrel domain is important for nuclear localization. Alternative mechanisms that need to be more fully characterized are masking of the weak NLS by other proteins or the catalytic domain of 5-lipoxygenase itself, the ability of other proteins to "piggyback" 5-lipoxygenase to the nucleus (e.g., chaperone protein mechanisms) and posttranslational modification such as phosphorylation. A minor phosphorylated nuclear 5-lipoxygenase fraction has been detected in HL-60 cells but the site of phosphorylation has not been identified (35).

	CONCLUDING REMARKS

TOP INTRODUCTION A STRATEGY TO GENERATE... INITIAL CHARACTERIZATION OF... LEUKOTRIENE DEFICIENT MICE IN... 5-LIPOXYGENASE: A NUCLEAR... CONCLUDING REMARKS REFERENCES

Leukotriene-deficient mice have been generated and are awaiting testing by researchers from all disciplines to examine their function in models of disease and inflammation. Some interesting areas to explore will be neuroinflammation, reperfusion injury, inflammatory bowel disease, as well as various immune functions. The mice are freely available from the Induced Mutant Resource of the Jackson Laboratory (Bar Harbor, ME).

Research efforts will intensify to gain more insight into the nuclear functions of 5-lipoxygenase or leukotrienes. We are beginning to understand the mechanisms for regulation of nuclear import for this enzyme. Future studies should illuminate the control of nuclear targeting of 5-lipoxygenase and its importance in novel, as yet undiscovered, cellular functions.

	Footnotes

Correspondence and requests for reprints should be addressed to Colin D. Funk, Ph.D., Department of Pharmacology and Center for Experimental Therapeutics, Room 814, BRBII/III, University of Pennsylvania, Philadelphia, PA 19104.

Acknowledgments: Supported in part by NIH Grant HL58464.

	References

TOP INTRODUCTION A STRATEGY TO GENERATE... INITIAL CHARACTERIZATION OF... LEUKOTRIENE DEFICIENT MICE IN... 5-LIPOXYGENASE: A NUCLEAR... CONCLUDING REMARKS REFERENCES

1. Capecchi, M. R.. 1989. Altering the genome by homologous recombination. Science 244: 1288-1292 [Abstract/Free Full Text].

2. Koller, B. H., and O. Smithies. 1992. Altering genes in animals by gene targeting. Annu. Rev. Immunol. 10: 705-730 [Medline].

3. Funk, C. D., T. Matsumoto, S. Hoshiko, O. Rådmark, and B. Samuelsson. 1989. Characterization of the human 5-lipoxygenase gene. Proc. Natl. Acad. Sci. U.S.A. 86: 2587-2591 [Abstract/Free Full Text].

4. Funk, C. D., L. B. Funk, G. A. FitzGerald, and B. Samuelsson. 1992. Characterization of human 12-lipoxygenase genes. Proc. Natl. Acad. Sci. U.S.A. 89: 3962-3966 [Abstract/Free Full Text].

5. Matsumoto, T., C. D. Funk, O. Rådmark, J. O. Höög, H. Jörnvall, and B. Samuelsson. 1988. Molecular cloning and amino acid sequence of human 5-lipoxygenase. Proc. Natl. Acad. Sci. U.S.A. 85: 26-30 [Abstract/Free Full Text].

6. Hoshiko, S., O. Rådmark, and B. Samuelsson. 1990. Characterization of the human 5-lipoxygenase gene promoter. Proc. Natl. Acad. Sci. U.S.A. 87: 9073-9077 [Abstract/Free Full Text].

7. Funk, C. D., U. Kurre, and G. Griffis. 1994. Targeted gene disruption by homologous recombination: toward an understanding of specific blood cell functions. Ann. N.Y. Acad. Sci. 714: 253-258 [Medline].

8. Chen, X. S., J. R. Sheller, E. N. Johnson, and C. D. Funk. 1994. Role of leukotrienes revealed by targeted disruption of the 5-lipoxygenase gene. Nature (London) 372: 179-182 [Medline].

9. Funk, C. D., X. S. Chen, U. Kurre, and G. Griffis. 1995. Leukotriene-deficient mice generated by targeted disruption of the 5-lipoxygenase gene. Adv. Prostaglandin Thromboxane Leukotriene Res. 23: 145-150 [Medline].

10. Funk, C. D.. 1996. The molecular biology of mammalian lipoxygenases and the quest for eicosanoid functions using lipoxygenase-deficient mice. Biochim. Biophys. Acta 1304: 65-84 [Medline].

11. Funk, C. D., X.-S. Chen, Y.-P. Tu, J. R. Sheller, and C. G. Irvin. 1998. The role of 5-lipoxygenase products in a mouse model of allergic airway inflammation. In A. Szczeklik, R. J. Gryglewski, and J. R. Vane, editors. Eicosanoids, Aspirin and Asthma. Marcel Dekker, New York. 187-200.

12. Funk, C. D. 1999. Lipid mediator deficient mice in mouse models of inflammation. In C. N. Serhan and P. A. Ward, editors. Molecular and Cellular Basis of Inflammation. Humana Press, Totowa, NJ.

13. Johnson, E. N., D. Sun, X.-S. Chen, and C. D. Funk. 1999. Lipoxygenase gene disruption studies: status and applications. Adv. Exp. Med. Biol. 447: 63-73 [Medline].

14. Irvin, C. G., Y. P. Tu, J. R. Sheller, and C. D. Funk. 1997. 5-Lipoxygenase products are necessary for ovalbumin-induced airway responsiveness in mice. Am. J. Physiol. 272: L1053-L1058 [Abstract/Free Full Text].

15. Secor, W. E., M. R. Powell, J. Morgan, T. A. Wynn, and C. D. Funk. 1999. Mice deficient for 5-lipoxygenase, but not 12-lipoxygenase, display altered immune responses during infection with Schistosoma mansoni. Prostaglandins Other Lipid Mediators 56: 291-304 .

16. Bailie, M. B., T. J. Standiford, L. L. Laichalk, M. J. Coffey, R. Strieter, and M. Peters-Golden. 1996. Leukotriene-deficient mice manifest enhanced lethality from Klebsiella pneumoniae in association with decreased alveolar macrophage phagocytic and bactericidal activities. J. Immunol. 157: 5221-5224 [Abstract].

17. Byrum, R. S., J. L. Goulet, R. J. Griffiths, and B. H. Koller. 1997. Role of the 5-lipoxygenase-activating protein (FLAP) in murine acute inflammatory responses. J. Exp. Med. 185: 1065-1075 [Abstract/Free Full Text].

18. Ho, H. Y., and P. Y. Wong. 1989. The formation of lipoxins and 6,7-dihydrolipoxins by 5-lipoxygenase and reductase partially purified from potato tubers. Adv. Prostaglandin Thromboxane Leukotriene Res. 19: 137-140 [Medline].

19. Bell, E., R. A. Creelman, and J. E. Mullet. 1995. A chloroplast lipoxygenase is required for wound-induced jasmonic acid accumulation in Arabidopsis. Proc. Natl. Acad. Sci. U.S.A. 92: 8675-8679 [Abstract/Free Full Text].

20. Rouzer, C. A., and B. Samuelsson. 1985. On the nature of the 5-lipoxygenase reaction in human leukocytes: enzyme purification and requirement for multiple stimulatory factors. Proc. Natl. Acad. Sci. U.S.A. 82: 6040-6044 [Abstract/Free Full Text].

21. Ueda, N., S. Kaneko, T. Yoshimoto, and S. Yamamoto. 1986. Purification of arachidonate 5-lipoxygenase from porcine leukocytes and its reactivity with hydroperoxyeicosatetraenoic acids. J. Biol. Chem. 261: 7982-7988 [Abstract/Free Full Text].

22. Peters-Golden, M., and R. W. McNish. 1993. Redistribution of 5-lipoxygenase and cytosolic phospholipase A2 to the nuclear fraction upon macrophage activation. Biochem. Biophys. Res. Commun. 196: 147-153 [Medline].

23. Brock, T. G., R. Paine, and M. Peters-Golden. 1994. Localization of 5-lipoxygenase to the nucleus of unstimulated rat basophilic leukemia cells. J. Biol. Chem. 269: 22059-22066 [Abstract/Free Full Text].

24. Woods, J. W., M. J. Coffey, T. G. Brock, I. I. Singer, and M. Peters-Golden. 1995. 5-Lipoxygenase is located in the euchromatin of the nucleus in resting human alveolar macrophages and translocates to the nuclear envelope upon cell activation. J. Clin. Invest. 95: 2035-2046 .

25. Chen, X. S., T. A. Naumann, U. Kurre, N. A. Jenkins, N. G. Copeland, and C. D. Funk. 1995. cDNA cloning, expression, mutagenesis, intracellular localization, and gene chromosomal assignment of mouse 5-lipoxygenase. J. Biol. Chem. 270: 17993-17999 [Abstract/Free Full Text].

26. Devchand, P. R., H. Keller, J. M. Peters, M. Vazquez, F. J. Gonzalez, and W. Wahli. 1996. The PPARalpha-leukotriene B4 pathway to inflammation control. Nature (London) 384: 39-43 [Medline].

27. Chen, X.-S., Y. Y. Zhang, and C. D. Funk. 1998. Determinants of 5-lipoxygenase nuclear localization using green fluorescent protein/5-lipoxygenase fusion proteins. J. Biol. Chem. 273: 31237-31244 [Abstract/Free Full Text].

28. Corbett, A. H., and P. A. Silver. 1997. Nucleocytoplasmic transport of macromolecules. Microbiol. Mol. Biol. Rev. 61: 193-211 . [Abstract]

29. Zhang, Y. Y., B. Lind, O. Rådmark, and B. Samuelsson. 1993. Iron content of human 5-lipoxygenase, effects of mutations regarding conserved histidine residues. J. Biol. Chem. 268: 2535-2541 [Abstract/Free Full Text].

30. Hammarberg, T., Y. Y. Zhang, B. Lind, O. Rådmark, and B. Samuelsson. 1995. Mutations at the C-terminal isoleucine and other potential iron ligands of 5-lipoxygenase. Eur. J. Biochem. 230: 401-407 [Medline].

31. Percival, M. D.. 1991. Human 5-lipoxygenase contains an essential iron. J. Biol. Chem. 266: 10058-10061 [Abstract/Free Full Text].

32. Boyington, J. C., B. J. Gaffney, and L. M. Amzel. 1993. The three- dimensional structure of an arachidonic acid 15-lipoxygenase. Science 260: 1482-1486 [Abstract/Free Full Text].

33. Gillmor, S. A., A. Villasenor, R. Fletterick, E. Sigal, and M. F. Browner. 1997. The structure of mammalian 15-lipoxygenase reveals similarity to the lipases and the determinants of substrate specificity. Nature Struct. Biol. 4: 1003-1009 [Medline].

34. Lepley, R. A., and F. A. Fitzpatrick. 1998. 5-Lipoxygenase compartmentalization in granulocytic cells is modulated by an internal bipartite nuclear localizing sequence and nuclear factor kappa B complex formation. Arch. Biochem. Biophys. 356: 71-76 [Medline].

35. Lepley, R. A., D. T. Muskardin, and F. A. Fitzpatrick. 1996. Tyrosine kinase activity modulates catalysis and translocation of cellular 5-lipoxygenase J. Biol. Chem. 271: 6179-6184 [Abstract/Free Full Text].