Genetic Variations in the 5-Lipoxygenase Core Promoter
Description and Functional Implications

ERIC S. SILVERMAN and JEFFREY M. DRAZEN

Division of Pulmonary and Critical Care Medicine, Departments of Medicine, Brigham and Women's Hospital and Harvard Medical School, Boston, Massachusetts

	INTRODUCTION

TOP INTRODUCTION THE 5-LO GENE PROMOTER A BRIEF REVIEW OF... MUTATIONS IN THE 5-LO... FUNCTIONAL CONSEQUENCES OF... FUNCTIONAL CONSEQUENCES OF... BINDING OF Sp1 AND... PROPOSED MODEL OF 5-LO... IMPLICATIONS OF THESE IN... REFERENCES

The regulation of 5-lipoxygenase (5-LO) activity occurs at multiple steps and is quite complex; each of the regulatory steps has been the subject of intense investigation. Regulation of 5-LO product output is known to occur by regulation of availability and catalytic action of 5-LO. With respect to 5-LO protein availability there are data documenting regulation at the level of gene transcription (1), gene translation (5), and gene product translocation within the cell (6, 7). With respect to catalytic action, regulation of 5-LO product production is known to be influenced by the availability of its major substrate arachidonic acid (8). Another mechanism regulating 5-LO catalytic activity is termed suicide inactivation and results from the oxidative inactivation of the functional enzyme by its primary catalytic product, leukotrienne A₄ (LTA₄). Suicide inactivation is irreversible and occurs only when 5-LO is catalytically active (9). Thus, if a cell with catalytically active 5-LO is to sustain product output, there must be a means by which to replace the inactivated enzyme. Since 5-LO output is known to be sustained, de novo production of the enzyme must take place; therefore, it is critical to examine the factors regulating 5-LO gene transcription and translation. Indeed, among these points of regulation of 5-LO gene transcription and translation, we have identified a novel mechanism potentially accounting for variability of 5-LO transcriptional regulation among individuals, namely genetic variation due to mutations in the 5-LO core promoter. We review effects of these mutations in the context of regulation of 5-LO gene transcription.

	THE 5-LO GENE PROMOTER

TOP INTRODUCTION THE 5-LO GENE PROMOTER A BRIEF REVIEW OF... MUTATIONS IN THE 5-LO... FUNCTIONAL CONSEQUENCES OF... FUNCTIONAL CONSEQUENCES OF... BINDING OF Sp1 AND... PROPOSED MODEL OF 5-LO... IMPLICATIONS OF THESE IN... REFERENCES

Several lines of evidence from other investigators suggest that alterations in the transcriptional regulation of the 5-LO gene are important for functional expression of 5-LO and thus may have clinical relevance. In response to differentiation and activation by dimethyl sulfoxide (DMSO), phorbol 12-myristate 13-acetate (PMA), transforming growth factor  (TGF-), granulocyte-macrophage colony-stimulating factor (GM-CSF), interleukin 3 (IL-3) low-density lipoprotein (LDL), or Ca²⁺ ionophore, several leukocyte cell lines exhibit increased steady state 5-LO mRNA levels (2, 3, 10). In some studies, at least part of this increase has been attributable to augmented transcription, as demonstrated by nuclear runoff analysis and alterations in chromatin conformation (2, 4). The mechanisms by which an inflammatory microenvironment augments 5-LO gene transcription have not been established. However, since the 5-LO gene promoter contains numerous consensus-binding sites for many known transcription factors, including Sp1, Sp3, Egr-1, Egr-2, NF-B, GATA, Myb, and AP family members, it is not unreasonable to suggest that one or more of these factors is involved.

The human 5-LO gene promoter was first characterized by Hoshiko and associates (1). To identify important cis-regulatory elements, these investigators created a panel of chloramphenicol acetyltransferase (CAT) promoter-reporter constructs consisting of 5' deletions of the genomic DNA upstream from the translational start site for the 5-LO gene. Transient transfection of these constructs into HeLa and HL-60 cells indicated that the 5' flanking region of the 5-LO gene contained DNA sequences with both positive and negative regulatory influences on reporter gene transcription. One region, located 212 to 88 base pairs (bp) upstream from the translation start site, was found to be particularly important for normal promoter activity. This region is highly G+C rich and was designated by Hoshiko and coworkers as the transcription factor-bindng region because it contains a series of five binding motifs for the transcription factors Sp1 and Egr-1. As is discussed in detail below, we have shown that this region of the promoter can bind both Sp1 and Egr-1. Furthermore, as is also detailed, we have identified a series of relatively common and naturally occurring mutations in this region of the 5-LO gene that modify transcription factor binding and reporter gene transcription.

	A BRIEF REVIEW OF Sp1 AND Egr-1 AS DNA-BINDING PROTEINS INVOLVED IN 5-LO TRANSCRIPTION

TOP INTRODUCTION THE 5-LO GENE PROMOTER A BRIEF REVIEW OF... MUTATIONS IN THE 5-LO... FUNCTIONAL CONSEQUENCES OF... FUNCTIONAL CONSEQUENCES OF... BINDING OF Sp1 AND... PROPOSED MODEL OF 5-LO... IMPLICATIONS OF THESE IN... REFERENCES

Sp1 is a ubiquitous zinc finger protein expressed in nearly all cell types and is required for the expression of many essential genes. Sp1 binds to G+C-rich sites containing the consensus sequence GGGCGG and similar variants. In general, levels of Sp1 expression are highest in cells undergoing differentiation, and these high levels of Sp1 may be required for the subsequent induction of tissue-specific genes (13). Egr-1 and related family members Egr-2 and Egr-3 are also zinc finger transcription factors that bind to similar G+C-rich sequences containing the consensus sequence GCG(T/G)GGGCG. Egr family members are examples of "immediate-early response" proteins and are rapidly and transiently induced by a large number of growth factors, cytokines, and injurious stimuli. Because their consensus-binding sequences are similar [i.e., -GGGCGG- for Sp1 and GCG(T/G)GGGCG for Egr-1], when there are tandem arrays of Sp1-binding motifs (with a few flanking G+Cs) there are also tandem arrays of Egr-1-binding motifs. These binding motif similarities are consistent with interactions of these two trans-activating proteins.

Indeed, in other systems Egr-1 has been shown to displace Sp1 bound to promoter regions of several genes and to increase transcription above basal levels (14). This mechanism of transcription activation has been linked with the induction of platelet-derived growth factor (PDGF) in several models of vascular injury (15). Egr-1 and Egr-2 levels increase during granulocyte differentiation and after cytokine (M-CSF) activation concurrent with 5-LO induction. On the basis of these data, we speculate that 5-LO induction during cytokine activation of certain leukocytes is mediated by Sp1 and/or Egr-1. For example, an Sp1 Egr-1 displacement mechanism may be involved in the transition from the basal to the cytokine-activated state; such activated (or transition) states are likely to occur in an asthmatic microenvironment. If so, the naturally occurring 5-LO promoter mutations that we have identified may alter 5-LO functional expression by interfering with this trans-activation process.

	MUTATIONS IN THE 5-LO GENE

TOP INTRODUCTION THE 5-LO GENE PROMOTER A BRIEF REVIEW OF... MUTATIONS IN THE 5-LO... FUNCTIONAL CONSEQUENCES OF... FUNCTIONAL CONSEQUENCES OF... BINDING OF Sp1 AND... PROPOSED MODEL OF 5-LO... IMPLICATIONS OF THESE IN... REFERENCES

The potential importance of differences in 5-LO gene expression as a mechanism regulating 5-LO activity led us to look for mutations within the 5-LO gene that would cause alterations in function or structure. We reasoned that such mutations would likely be present in both nonasthmatic subjects and patients with asthma but that they would probably be functionally silent in nonasthmatic individuals because the 5-LO pathway would not be activated in this group. We examined genomic DNA from 25 nonasthmatic subjects and 31 patients with asthma, using single-stranded conformational polymorphism (SSCP) analysis, for mutations in the 5-LO gene that would cause altered amino acid sequence, mRNA splicing, or transcriptional regulation. We identified three mutations within the protein-encoding region (16) that would not modify amino acid sequence or mRNA splicing and that therefore were not pursued further.

Of greater significance was a family of mutations found within the functionally important G+C-rich core promoter region. These mutations consisted of the addition of one or deletion of one or two Sp1 consensus-binding motifs. Because of the close similarity between the Sp1 and Egr-1 domains, the changes also resulted in the addition or deletion of Egr-1-binding domains. These mutations are fairly common; about 20% of the alleles in the normal and asthmatic U.S. populations are mutant alleles. We next asked whether these mutations are associated with a functional difference in the activity of the 5-LO gene promoter.

	FUNCTIONAL CONSEQUENCES OF PROMOTER MUTATIONS IN HeLa CELLS

TOP INTRODUCTION THE 5-LO GENE PROMOTER A BRIEF REVIEW OF... MUTATIONS IN THE 5-LO... FUNCTIONAL CONSEQUENCES OF... FUNCTIONAL CONSEQUENCES OF... BINDING OF Sp1 AND... PROPOSED MODEL OF 5-LO... IMPLICATIONS OF THESE IN... REFERENCES

Functional studies with CAT reporter constructs containing approximately 300 bp of flanking sequence revealed significant differences among the wild-type and mutant promoters in the ability to drive transcription in HeLa cells (Figure 1). In these cells, all constructs containing mutant alleles were less effective at directing CAT reporter transcription than were constructs containing the wild-type sequence. The magnitude of the decrease in CAT reporter activity was small (on the order of 20 to 35%); however; clinical data for zileuton obtained in asthma treatment trials (17) indicate the potential for a clinically significant benefit if this magnitude of effect was translated into a 20-35% decrease in functional enzyme.

View larger version (17K): [in this window] [in a new window]   Figure 1.   Relative CAT activity, corrected for transfection efficiency, of HeLa cells transfected with pCAT inserts containing either wild-type 5-LO upstream DNA (assigned a value of 100) or a mutant transcription factor-binding region. The capacity of the mutant DNA to promote transcription of the CAT reporter constructs was significantly weaker than that of the wild-type DNA. Results are the mean of five experiments, each performed in triplicate. This result is slightly different from that obtained in cotransfection studies of Schneider cells (see Figure 2). In Schneider cells the 6-bp addition alleles were more effective in directing the transcription of a reporter construct than the wild-type alleles. [Reprinted, with permission, from the Journal of Clinical Investigation.]

View larger version (30K): [in this window] [in a new window]   View larger version (26K): [in this window] [in a new window]   Figure 2.   CAT reporter activity of Schneider cells cotransfected with pCAT basic constructs containing either the wild-type (WT), 6-bp addition (+6), 6-bp deletion (6) or 12-bp deletion (12) form of the 5-LO promoter region. (Left panel  ) Cells were cotransfected with a plasmid containing a cDNA encoding Sp1 (pPAC Sp1). (Right panel  ) Cells were cotransfected with a plasmid encoding Egr-1 (pPAC Egr-1). In both cases the plasmids containing the addition alleles were more effective and the plasmid containing the deletional alleles less effective than the wild-type allele in driving transcription of the reporter constructs [Reprinted, with permission, from the American Journal of Respiratory Cell and Molecular Biology.]

	FUNCTIONAL CONSEQUENCES OF PROMOTER MUTATIONS IN SCHNEIDER CELLS

TOP INTRODUCTION THE 5-LO GENE PROMOTER A BRIEF REVIEW OF... MUTATIONS IN THE 5-LO... FUNCTIONAL CONSEQUENCES OF... FUNCTIONAL CONSEQUENCES OF... BINDING OF Sp1 AND... PROPOSED MODEL OF 5-LO... IMPLICATIONS OF THESE IN... REFERENCES

HeLa cells are known to produce a wide array of transcription factors; thus we performed additional cotransfection studies in Drosophila SL2 cells (Schneider cells). Schneider cells, unlike most mature mammalian cells, do not naturally express Sp1 or Egr-1 and provide a low background of basal transcription. In this system, overexpression of either Egr-1 or Sp1 by cotransfection with the respective expression construct increased CAT activity of all promoter-reporter constructs (Figure 2). However, not all promoter mutants were equally effective at driving transcription. Mutant Add(+6) was the most active promoter, with wild-type, Del(6), and Del(12) constructs showing progressively less responsiveness to Egr-1. Sp1 overexpression resulted in a similar pattern of reporter gene transcription among the mutant forms of the promoter; however, the relative changes were less impressive than those observed with Egr-1.

There are discrepancies between the HeLa and Schneider cell data sets; the HeLa cell experiments suggest that the wild-type promoter is more active than all mutant promoters, whereas the Schneider cell data suggest that promoter activity is proportional to the number of Egr1/Sp-1-binding motifs. The reason for this discrepancy is not known; we speculate that as yet unidentified transcription factors may be interacting with this region of the promoter.

	BINDING OF Sp1 AND Egr-1 TO THE 5-LO CORE PROMOTER

TOP INTRODUCTION THE 5-LO GENE PROMOTER A BRIEF REVIEW OF... MUTATIONS IN THE 5-LO... FUNCTIONAL CONSEQUENCES OF... FUNCTIONAL CONSEQUENCES OF... BINDING OF Sp1 AND... PROPOSED MODEL OF 5-LO... IMPLICATIONS OF THESE IN... REFERENCES

Because the mutant alleles differed in the number of Sp1- and Egr-1-binding motifs, we next asked whether this change in CAT reporter activity was reflected in an altered capacity of oligonucleotides corresponding to the mutant forms of the transcription factor-binding region to bind nuclear extracts from cells known to produce Sp1 and Egr-1 when cultured under various conditions. For this analysis, we used nuclear extracts from human umbilical vein endothelial cells (HUVECs). When HUVECs are cultured in the absence of PMA, Sp1, but not Egr-1, is a known constituent of their nuclear extract; in contrast, when HUVECs are cultured in the presence of PMA, their nuclear extracts contain both Sp1 and Egr-1 (14). Nuclear extracts from HUVECs cultured in the absence and presence of PMA were used in an electrophoretic mobility shift assay (EMSA). Synthetic oligonucleotides corresponding to the wild-type transcription factor-binding region were subjected to electrophoresis in the absence of nuclear extract, in the presence of nuclear extract from HUVECs cultured without PMA, and with nuclear extract from HUVECs cultured in the presence of PMA. In addition, supershift analyses were performed with antibodies to Sp1, Egr-1, Sp4, and ETS to confirm the identity of each band.

The EMSA showed a single band when nuclear extract from HUVECs cultured without PMA was used and two bands when extract from HUVECs cultured with PMA was used. Furthermore, the band with the lower mobility supershifted with the antibody to Sp1, and the band with the higher mobility supershifted with the antibody to Egr-1. When EMSAs were performed with oligonucleotides corresponding to the wild-type and various mutant forms of the 5-LO promoter, the same qualitative pattern emerged but the bands from the mutant forms were 47 to 54% less intense for Sp1 and 42 to 67% less intense for Egr-1 than those from the wild type (16). These data are in general agreement with the functional studies in HeLa cells and suggest that binding intensity of Sp1 and Egr-1 to the 5-LO promoter is proportional to its trans-activation potential.

	PROPOSED MODEL OF 5-LO TRANSCRIPTION: ROLE OF THE TRANSCRIPTIONAL COACTIVATORS CBP AND p300

TOP INTRODUCTION THE 5-LO GENE PROMOTER A BRIEF REVIEW OF... MUTATIONS IN THE 5-LO... FUNCTIONAL CONSEQUENCES OF... FUNCTIONAL CONSEQUENCES OF... BINDING OF Sp1 AND... PROPOSED MODEL OF 5-LO... IMPLICATIONS OF THESE IN... REFERENCES

From the previously described studies it is clear that the 5-LO promoter mutations can alter Sp1/Egr-1 binding and reporter construct activity in vitro. In some models the intensity of Sp1/ Egr-1 binding is proportional to the degree of trans-activation (HeLa cells), whereas in other models (Schneider cells) the number of Sp1/Egr-1-binding sites appears to be the major factor. Which model best reflects in vivo conditions remains to be seen and is the subject of ongoing investigation. However, both the HeLa and Schneider cell models support the notion that it is the number of Sp1/Egr-1 molecules bound to this region of the promoter that is the major determinant of trans- activation. To determine how this occurs at a molecular level, we are investigating the possibility that the nuclear coactivators CREB-binding protein (CBP) and p300 may be involved. CBP and p300 are ubiquitous transcriptional coactivators capable of physically linking sequence specific transcription factors with components of the basal transcription apparatus. They serve a key regulatory role in gene expression by integrating diverse signaling pathways at the transcriptional level. Tandem Egr-1 consensus binding sites may serve as a platform for the recruitment of CBP/p300 through two or more protein-protein contact points on the 5-LO promoter. The larger the promoter platform, the more contact points there are between CBP/p300 and Egr-1; the more contact points between CBP/p300 and Egr-1, the greater the stability of the transcriptional initiation complex. Figure 3 illustrates the proposed model of 5-LO activation. In leukocytes under quiescent conditions, basal levels of 5-LO message are produced through an Sp1-mediated transcriptional mechanism. In an inflammatory microenvironment Egr-1 levels increase, displace Sp1 from the 5-LO promoter, and further increase transcription. The efficiency of the transcription initiation complex reflects the recruitment of CBP/p300 and is therefore proportional to the number of Egr-1-binding events reflected by intensity (HeLa cell) or the number of binding sites (Schneider cell). This model is still somewhat speculative, and the role of Sp1/Egr-1 and CBP/p300 in 5-LO transcription upregulation in mast cells and eosinophils remains to be defined. Population studies correlating genotype with the expression of 5-LO should help clarify the true effects of these promoter mutations.

View larger version (26K): [in this window] [in a new window]   Figure 3.   Proposed model of 5-LO transcription induction. Under normal conditions Sp1 occupies the G+C-rich promoter region and mediates basal levels of 5-LO transcription. In response to inflammation, Egr-1 levels increase and displace Sp1 from the G+C-rich region of the 5-LO promoter. Unlike Sp1, Egr-1 is capable of recruiting the transcriptional coactivators CBP/ p300, resulting in enhanced transcription. Alleles with more Egr-1-binding sites enhance transcription more efficiently by supporting multiple protein-protein contact points that result in a more stable platform for the recruitment of CBP/p300. Other transcription factors, not yet identified, may also bind this region of the 5-LO promoter and have regulatory roles.

	IMPLICATIONS OF THESE IN VITRO FINDINGS

TOP INTRODUCTION THE 5-LO GENE PROMOTER A BRIEF REVIEW OF... MUTATIONS IN THE 5-LO... FUNCTIONAL CONSEQUENCES OF... FUNCTIONAL CONSEQUENCES OF... BINDING OF Sp1 AND... PROPOSED MODEL OF 5-LO... IMPLICATIONS OF THESE IN... REFERENCES

These data indicated that the various mutations we identified are more than a simple tandem nucleotide repeat; rather, they are positioned in such a fashion as to modify gene transcription and therefore to potentially modify 5-LO product formation. However, at this time, these are only potential effects of the mutations, and additional experiments will be required before their true functional implications can be appreciated.

	Footnotes

Correspondence and requests for reprints should be addressed to Jeffrey M. Drazen, M.D., Division of Pulmonary and Critical Care Medicine, Brigham and Women's Hospital, 75 Francis Street, Boston, MA 02115. E-mail: jdrazen{at}rics.bwh.harvard.edu

	References

TOP INTRODUCTION THE 5-LO GENE PROMOTER A BRIEF REVIEW OF... MUTATIONS IN THE 5-LO... FUNCTIONAL CONSEQUENCES OF... FUNCTIONAL CONSEQUENCES OF... BINDING OF Sp1 AND... PROPOSED MODEL OF 5-LO... IMPLICATIONS OF THESE IN... REFERENCES

1. 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].

2. Stankova, J., M. Rolapleszczynski, and C. M. Dubois. 1995. Granulocyte-macrophage colony-stimulating factor increases 5-lipoxygenase gene transcription and protein expression in human neutrophils. Blood 85: 3719-3726 [Abstract/Free Full Text].

3. Steinhilber, D., S. Hoshiko, J. Grunewald, O. Rådmark, and B. Samuelsson. 1993. Serum factors regulate 5-lipoxygenase activity in maturating HL60 cells. Biochim. Biophys. Acta 1178: 1-8 [Medline].

4. Ponton, A., J. P. Thirion, and P. Sirois. 1997. Repression of the 5-lipoxygenase gene by c-myb overexpression in differentiated HL-60 cells. Prostaglandins 53: 49-58 [Medline].

5. Pouliot, M., P. P. McDonald, L. Khamzina, P. Borgeat, and S. R. McColl. 1994. Granulocyte-macrophage colony-stimulating factor enhances 5-lipoxygenase levels in human polymorphonuclear leukocytes. J. Immunol. 152: 851-858 [Abstract].

6. 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].

7. 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].

8. Bergstrom, S.. 1967. Prostaglandins: members of a hormonal system. Science 157: 382-391 [Free Full Text].

9. Lepley, R. A., and F. A. Fitzpatrick. 1994. Irreversible inactivation of 5-lipoxygenase by leukotriene A4: characterization of product inactivation with purified enzyme and intact leukocytes. J. Biol. Chem. 269: 2627-2631 [Abstract/Free Full Text].

10. Jakobsson, P. J., D. Steinhilber, B. Odlander, O. Rådmark, H. E. Claesson, and B. Samuelsson. 1992. On the expression and regulation of 5-lipoxygenase in human lymphocytes. Proc. Natl. Acad. Sci. U.S.A. 89: 3521-3525 [Abstract/Free Full Text].

11. Piechele, G., S. Colli, E. Tremoli, and A. Maggi. 1993. 5-Lipoxygenase gene expression in HL60 cells during differentiation with DMSO. Pharmacol. Res. 27: 53-60 [Medline].

12. Steinhilber, D., O. Rådmark, and B. Samuelsson. 1993. Transforming growth factor-beta upregulates 5-lipoxygenase activity during myeloid cell maturation. Proc. Natl. Acad. Sci. U.S.A. 90: 5984-5988 [Abstract/Free Full Text].

13. Courey, A. J., and R. Tjian. 1992. Mechanisms of transcriptional control as revealed by studies of human transcription factor Sp1. In S. L. McKnight and K. R. Yamamoto, editors. Transcriptional Regulation. Cold Spring Harbor Press, Cold Spring Harbor, NY. 743-769.

14. Khachigian, L. M., V. Linder, A. J. Williams, and T. Collins. 1996. Egr-1-induced endothelial gene expression: a common theme in vascular injury. Science 271: 1427-1431 [Abstract].

15. Silverman, E. S., L. M. Khachigian, V. Lindner, A. J. Williams, and T. Collins. 1997. Inducible PDGF-A chain transcription in smooth muscle cells is mediated by Egr-1 displacement of Sp1 and Sp3. Am. J. Physiol. 273: H1415-H1426 [Abstract/Free Full Text].

16. In, K. H., K. Asano, D. Beier, J. Grobholz, P. W. Finn, E. K. Silverman, E. S. Silverman, T. Collins, A. R. Fischer, T. P. Keith, K. Serino, S. W. Kim, G. T. Desanctis, C. Yandava, A. Pillari, P. Rubin, J. Kemp, E. Israel, W. Busse, D. Ledford, J. J. Murray, A. Segal, D. Tinkleman, and J. M. Drazen. 1997. Naturally occurring mutations in the human 5-lipoxygenase gene promoter that modify transcription factor binding and reporter gene transcription. J. Clin. Invest. 99: 1130-1137 [Medline].

17. Israel, E., P. Rubin, J. P. Kemp, J. Grossman, W. E. Pierson, S. C. Siegel, D. Tinkelman, J. J. Murray, W. Busse, A. T. Segal, J. Fish, H. B. Kaiser, D. Ledford, S. Wenzel, R. Rosenthal, J. Cohn, C. Lanni, H. Pearlman, P. Karahalios, and J. M. Drazen. 1993. The effect of inhibition of 5-lipoxygenase by zileuton in mild to moderate asthma. Ann. Intern. Med. 119: 1059-1066 [Abstract/Free Full Text].