The Role of Metals in Gene Expression
Raymond K. Blanchard and Robert J. Cousings1
The concept that metals are able to influence gene expression has been understood in general terms for decades. Growth responses associated with the addition of trace elements to the nutrient supply of both plants and animals supported such a role. Subsequent to this, the demonstration that enzymes associated with nucleic acid biochemistry were metalloenzymes more firmly established the metal-gene expression relationship.
All living organisms require metals to sustain various cellular processes. These metals include the macrominerals, for example calcium, which is not believed to directly influence gene expression, but may do so indirectly through secondary messenger roles. Trace metals (elements) required in the diet include copper, iron, manganese, nickel, selenium, and zinc. In humans, these trace elements are required in microgram to milligram amounts per day. Cellular mechanisms control the transport of trace metals into and out of cells so that the
influences of natural abundance and thermodynamic properties do not determine the cellular content and function of a specific metal (da Silva and Williams, 1991). Nevertheless, these mechanisms can be overwhelmed by very high intakes of a trace metal.
INVOLVEMENT OF METALS IN GENE EXPRESSION
There are three ways trace metals can be involved in gene expression. One is structural, where metals facilitate interaction among various binding groups to provide the altered conformation necessary for interactions such as between specific proteins and DNA (Cousins, 1995). The second type of involvement is catalytic, where the metal is required for the activity of an enzyme associated with gene expression. The third class involves specific regulation, where metal occupancy of a transacting protein modulates transcription of a specific gene. This type of involvement is different from the first in that it is much more specific, being more interactive than structural in function. Since the catalytic role appears to be relatively unalterable in humans except, perhaps, in extreme deficiency situations during development, this chapter will concentrate on the structural and regulatory aspects of metals in gene expression.
The best examples of the regulation of gene expression by metals are iron and zinc. In the case of iron, metal occupancy decreases the binding of a metal-regulatory binding protein to ferritin mRNA, allowing the translation of ferritin mRNA to increase while simultaneously increasing binding to transferrin receptor mRNA, which increases the degradation of mRNA (O'Halloran, 1993). Because iron exhibits oxidation-reduction (redox) chemistry, rapid control of ferritin synthesis at the level of translation is necessary to provide rapid control of free iron levels within cells.
Far more is known about the involvement of zinc in gene expression than that of other elements. The intracellular binding affinity is greater for zinc than for virtually all other metals found in cells, with the exception of copper. However, unlike iron, zinc does not exhibit redox chemistry but has the properties of a Lewis acid and exhibits fast ligand exchange, which is important for its catalytic role (da Silva and Williams, 1991). A principal example of this catalytic role in gene expression is exhibited by the family of RNA nucleotidyl transferases (RNA polymerases I, II, and III). Zinc also plays a structural role in the zinc-finger motif of proteins that are involved in DNA binding, as is discussed below. Finally, as an activator of trans-acting factors,2 zinc is responsible for regulating the expression of specific genes. The latter is discussed below.
Zinc ions serve an important structural function by tetrahedrally coordinating to cysteine or histidine residues of certain proteins to stabilize the structure of a small functional domain. The most prominent role of proteins with zinc-finger motifs is sequence-specific binding to DNA during transcription, and this is one of the most common eukaryotic DNA-binding motifs (Klug and Schwabe, 1995). Transcription factors that use zinc fingers as DNA-binding domains range from basal transcription components such as Sp1, to tissue type-specific factors such as GATA-1, to inducible factors such as glucocorticoid receptors (Lewin, 1994). In addition, some zinc fingers have been shown to mediate protein-protein interactions. The degree of influence of the level of dietary zinc on zinc interaction with finger proteins is not known.
Four major but distinct families of zinc-finger motifs have been identified. The first type is the original or classic zinc finger, as exemplified in transcription factor TFIIIA (Klug and Schwabe, 1995). A single, independent motif contains 1 zinc ion coordinated to 2 cysteine and 2 histidine residues, with a loop or finger of 12 amino acids between the 2 pairs of coordinating residues. This DNA-binding structure often functions as a multimer in which the finger motif is repeated in tandem one to nine times, with the spacing between the coordinating residues being highly conserved. This primary structure has served as the starting point for identifying other zinc-finger motifs based on DNA-and amino acid-sequence homology.
The second family of zinc-finger proteins contains a "zinc twist" (Vallee et al., 1991). This motif consists of two tandem zinc fingers with only cysteines as the zinc coordinating residues. Unlike the classic zinc finger, these two fingers twist around each other to form a single functional unit with two faces. The zinc-twist motif is characteristic of the DNA-binding domain of the steroid hormone receptor super family. These receptors frequently function as dimers (Lewin, 1994).
The third zinc-finger type motif to be recognized is referred to as a "zinc cluster." In this structure, just six conserved cysteines3 coordinate two zinc ions in a single cluster, and the first and third cysteines are shared between both zinc ions. This motif is responsible for sequencing specific DNA binding of proteins, such as in the transcription factor GAL4.
The LIM domain is the fourth type of zinc-finger motif. It is composed of two tandem finger-like zinc-binding sites (Dawid et al., 1995). The first zinc is coordinated by three cysteines and a histidine, while the second zinc is coordinated by four cysteines. The LIM domain occurring in some proteins is followed by a homeodomain,4 which may mediate DNA binding. In contrast, for
LIM-only proteins, there is mounting evidence that the LIM motif is involved in protein-protein interactions (Schmeichel and Beckerle, 1994).
With such a wide distribution among proteins involved in different cellular processes, zinc fingers are an important focus of research for therapeutic applications. For example, with their high sequence specificity, they are being evaluated as alternatives to anti-sense DNA approaches to modulating gene expression, and new DNA specificities for zinc fingers are being developed by mutagenesis and phage display library screening (Rebar and Pabo, 1994; Wu et al., 1995). In this way, engineered zinc-finger proteins could be selected based on their ability to bind a specific sequence of interest in order to target specific genes or points of regulation.
METAL-RESPONSIVE GENE REGULATION
The mechanisms of transcriptional regulation operate through protein interactions with specific sequences of DNA known as response elements. These response elements provide the specificity for protein factors to interact with each other on the promoter and ultimately result in the unique regulation of different genes. One widely studied eukaryotic response element that confers increased transcription by metals was first identified in the promoter of the metallothionein I gene (Stuart et al., 1984). This metal response element, or MRE, is a 15-base pair (bp) consensus sequence found in multiple copies that potentiates a large increase in gene expression when zinc or cadmium is present (Cousins, 1994). MRE consensus sequences are generally found in the first several hundred base pairs of the promoter and are often near or overlapping with response elements for other transcriptional factors, such as AP1 (Lewin, 1994). In addition, MREs are orientation independent and confer metal responsiveness when placed in heterologous promoters.
Recently a protein factor that binds to MREs, and is essential for metal responsiveness, has been cloned and characterized from the mouse and human sources. This MRE-binding transcription factor (MTF-1) has a DNA-binding domain consisting of six classical zinc fingers and a separate transcriptional activation domain (Brugnera et al., 1994). It is not yet clear, however, whether MTF-1 binds or interacts directly with zinc or cadmium to activate transcription or whether another metalloregulatory protein binds zinc or cadmium and then interacts with MTF-1.
The fact that MRE elements function independently of other regulatory elements makes them valuable in the construction of chimeric genes for transgenic animals (Palmiter et al., 1982). One or several MREs can be incorporated into the promoter of a chimeric gene and allow the expression of the gene to be controlled in vivo. In this way, dietary zinc acting through MREs
might eventually be coupled to gene therapy to provide some degree of control for therapeutic gene expression, as discussed below.
mRNA DIFFERENTIAL DISPLAY
As international genome mapping projects progress, it has become an increasing priority in biology to identify the genes contained in these vast sequences in order to characterize the function of each gene product. Consequently, detection of genes regulated by nutritional status or altered physiological situations has become increasingly important. Most of the techniques for analyzing regulation by nutrition and other factors, however, require information about the gene as a prerequisite, and it is estimated that current international databases have only identified approximately 30 percent of the total genes in the human genome (Orr et al., 1994). A recently developed polymerase chain reaction (PCR) technique, mRNA differential display, can detect genes that are regulated under different physiological states with no prior information about the gene (Liang et al., 1993). This method is currently being used to detect genes regulated by dietary micronutrients such as zinc and selenium (Blanchard and Cousins, 1996; Kendall and Christensen, 1995).
The technique of mRNA differential display begins with the isolation of RNA from animals, tissues, or cells exposed to different physiological conditions. The RNA is reverse transcribed using an oligo d(T) primer that has an additional two non-T bases at its 3' end in order to anchor the start of the cDNA synthesis to the junction of the 3' untranslated region and the poly A tail. A total of four anchor primers are needed. The resulting cDNAs are then subjected to the polymerase chain reaction using the oligo d(T) primer and a decanucleotide primer of arbitrary sequence that will only amplify cDNAs representing a small subset of mRNAs from the original sample and incorporate a radioactive label. Gel electrophoresis and autoradiography are used to display the resulting PCR products. Any mRNA that contains the arbitrary decanucleotide sequence within approximately 400 bp of poly A tail will have that portion of its 3' end amplified by the primers in the PCR reaction, and this will produce a band of a specific size on the autoradiograph. In order to screen the entire population of mRNA in a given sample effectively, it is necessary to use a battery of at least 26 different arbitrary decamers, as well as the 4 oligo d(T) primers for all possible combinations at the 3' end. The intensity of the autoradiographic image is proportional to the amount of a particular mRNA in the samples. Therefore, cDNA bands of the same size that vary in intensity between experimental conditions represent an mRNA that is differentially expressed under those conditions. In addition, since the samples are displayed adjacent to each other on a gel with many lanes, more than two physiological conditions can be evaluated at the same time, which increases the versatility of this technique.
The cDNA for each differentially expressed mRNA is recovered from the electrophoresis gel and cloned into a plasmid to maintain a stable copy of cDNA for further analysis. The cDNA is first used as a probe for a Northern blot analysis of the original RNA samples to confirm the differential expression in the actual RNA and to quantify the levels of expression. Northern blot confirmation of mRNA differential display is important to eliminate false positives from further evaluation.
DNA sequencing is the next step in analyzing the cDNA clone. The sequence generally contains the 3' untranslated region of the mRNA and a small portion of the carboxyl terminal of the protein coding region. This incomplete sequence of an expressed mRNA is referred to as a 3' expressed sequence tag (EST) (Okubo et al., 1992). The sequence of the EST is used to search DNA sequence databases to determine if the mRNA has been previously identified and characterized. The novel aspect of 3' ESTs from mRNA differential display is that some information is already known about the expression of the gene. This information will greatly assist in selecting which newly identified genes should receive priority for further characterization in different metabolic processes.
USES OF METALS IN GENE REGULATION
There are two general areas in which metals introduced in the diet could regulate genes of interest to the support of military troops in the field. The first would relate to regulation of chimeric genes introduced through genetic manipulation. The field of molecular medicine has shown that it is possible to target specific genes, introduce them into human subjects, and observe expression (Morrow and Kucherlapati, 1995). This area is clearly at the forefront of molecular biology, but practical applications and resolution of ethical ramifications of these applications are at least decades away.
The regulation of transgenes by zinc is feasible because: metal response elements have been characterized, it is possible to produce multiple MREs in a single promoter, low basal-level expression of MRE-regulated genes usually occurs, there is a potential for a high level of expression, and duration of the elevated expression is directly correlated to the duration of increased zinc intake. The approach envisioned would be similar to that of the technology demonstrated earlier through use of a metal-responsive promoter (metallothionein promoter) linked to the growth hormone genes (Palmiter et al., 1982). Specifically, in mice and farm animals, transgenes were activated by increasing the level of zinc in the diet. The metal binds to the MRE-binding protein, which in turn initiates increased transcription of the growth hormone gene. This technology could be applied to virtually any gene of interest to military or public health situations.
The second area where metals could contribute to applications of interest to the military is that of gene expression associated with stress. It has been demonstrated repeatedly that zinc in high levels provides protection against
cellular damage caused by ionizing radiation and various chemical toxins (Willson, 1989). The mechanism of this protection is believed to be through prevention of free radical formation, although this mechanism has yet to be firmly established. Since many genes are activated by zinc, it is likely that a spectrum of zinc-responsive genes are activated during a distinct supplementation period that provides the observed cellular protection. Providing supplemental zinc via the diet at high levels during periods of stress may well be advantageous to allow military troops the ability to better withstand stressful situations. Zinc is a relatively nontoxic metal and could easily be introduced through dietary means that would provide a programmed duration of response.
AUTHORS' CONCLUSIONS AND RECOMMENDATIONS
This brief review has tried to demonstrate how some aspects of metal-regulated gene expression could be put to use in applications related to military activities. Molecular biology has clearly evolved to the point where it is providing tools, such as differential display, that will allow identification of genes that are beneficial to control during stressful situations. It should then be possible through newly evolving techniques to use that information in a way that will allow altering the diet through various combinations of dietary components that activate genes to select specific types of responses. The key task for the immediate future is to identify the genes most likely to be of benefit when regulated by nutritional means. As technology advances, the application of this information will unfold.
Blanchard, R.K., and R.J. Cousins 1996 Differential display of intestinal mRNAs regulated by dietary zinc. Proc. Natl. Acad. Sci. USA 93:6863–6868.
Brugnera, E., O. Georgiev, F. Radtke, R. Heuchel, E. Baker, G.R. Sutherland, and W. Schaffner 1994 Cloning, chromosomal mapping and characterization of the human metal-regulatory transcription factor MTF-1. Nucleic Acids Res. 22:3167–3173.
Cousins, R.J. 1994 Metal elements and gene expression. Annu. Rev. Nutr. 14:449–469.
1995 Trace element micronutrients. Pp. 898–901 in Molecular Biology and Biotechnology: A Comprehensive Desk Reference, R.A. Meyers, ed. New York: VCH Publishers.
da Silva, J.J.R., and R.J.P. Williams, eds. 1991 The Biological Chemistry of the Elements: The Inorganic Chemistry of Life. Oxford: Clarendon Press.
Dawid, I.B., R. Toyama, and M. Taira 1995 LIM domain proteins. C. R. Acad. Sci. Paris 318:295–306.
Kendall, S.D., and M.J. Christensen 1995 Modified differential display identifies genes in rat liver differentially regulated by selenium [abstract]. FASEB J. 9:A158.
Klug, A., and J.W.R. Schwabe 1995 Zinc fingers. FASEB J. 9:597–604.
Lewin, B. 1994 Genes V. Oxford: Oxford University Press.
Liang, P., L. Averboukh, and A.B. Pardee 1993 Distribution and cloning of eukaryotic mRNAs by means of differential display: Refinements and optimization. Nucleic Acids Res. 21:3269–3275.
Morrow, B.E., and R. Kucherlapati 1995 Molecular genetic medicine. Pp. 556–561 in Molecular Biology and Biotechnology: A Comprehensive Desk Reference, R.A. Meyers, ed. New York: VCH Publishers.
O'Halloran, T.V. 1993 Transition metals in control of gene expression. Science 261:712–725.
Okubo, K., N. Hori, R. Matoba, T. Niiyama, A. Fukushima, Y. Kojima, and K. Matsubara 1992 Large-scale cDNA sequencing for analysis of quantitative and qualitative aspects of gene expression. Nature Genetics 2:173–179.
Orr, S., T.P. Hughes, C.L. Sawyers, R.M. Kato, S.G. Quan, S.P. Williams, O.N. Witte, and L. Hood 1994 Isolation of unknown genes from human bone marrow by differential screening and single-pass cDNA sequence determination. Proc. Natl. Acad. Sci. USA 91:11869–11873.
Palmiter, R.D., R.L. Brinster, R.E. Hammer, M.E. Trumbauer, M.G. Rosenfeld, N.C. Birnberg, and R.M. Evans 1982 Dramatic growth of mice that develop from eggs microinjected with metallothionein-growth hormone fusion genes. Nature 300:611–615.
Rebar, E.J., and C.O. Pabo 1994 Zinc finger phage: Affinity selection of fingers with new DNA-binding specificities. Science 263:671–673.
Schmeichel, K.L., and M.C. Beckerle 1994 The LIM domain is a modular protein-binding interface. Cell 79:211–219.
Stuart, G.W., P.F. Searle, H.Y. Chen, R.L. Brinster, and R.D. Palmiter 1984 A 12-base-pair DNA motif that is repeated several times in metallothionein gene promoters confers metal regulation to a heterologous gene. Proc. Natl. Acad. Sci. USA 81:7318–7322.
Vallee, B.L., J.E. Coleman, and D.S. Auld 1991 Zinc fingers, zinc clusters, and zinc twists in DNA-binding protein domains. Proc. Natl. Acad. Sci. USA 88:999–1003.
Willson, R.L. 1989 Zinc and iron in free radical pathology and cellular control. Pp. 147–172 in Zinc in Human Biology, C.F. Mills, ed. New York: Springer-Verlag.
Wu, H., W-P. Yang, and C.F. Barbas, III 1995 Building zinc fingers by selection: Toward a therapeutic application. Proc. Natl. Acad Sci. USA 92:344–348.
HARRIS LIEBERMAN: How widely used is the technique, the differential display of RNA, and how practical is it?
ROBERT COUSINS: It was developed in late 1992. There was a very significant modification in 1993. Since then there have been a number of papers out on it. The first use with intact animals, I think, will be coming out fairly shortly. It is a technique that has some limitations, but nonetheless, it is, in my view, the most powerful screening method that one has available to look at what genes are turned on and turned off in a given physiologic state or a given nutritional state. I think the sequence information that is available in data banks, which this has to be drawn against, is the limiting factor.
G. RICHARD JANSEN: What is the physiological significance of the enhanced expression of metallothionein beyond the requirement level?
ROBERT COUSINS: That is another very active area of investigation, but it appears now that with cells that are transfected with this gene, the cells are protected from various types of radiation damage. Transgenic animals that are now commercially available are being studied. Animals that have overexpression of the gene are being studied to determine what physiologic changes they have in response to having more copies of that gene and more of the protein produced. I think it is a host defense type of mechanism that we are looking at.
WILLIAM BEISEL: This is a fascinating presentation, and it surely brings things up to date. Originally we had discovered the sequestration of zinc, the movement of zinc into the cells during the acute-phase response, and we were always wondering what it meant, why did the body do this? We were thinking years ago that it was probably to get more zinc into enzymes, because it was present in so many enzymes, but these studies of the gene expression certainly …
ROBERT COUSINS: You are right. Your work showed that the plasma zinc decreases right after exposure to endotoxin and other things. The plasma zinc goes into the liver and other tissues that produce metallothionein, as it turns out, because, kinetically, you can actually show where the metal goes. That has some function, but whether the function is related to controlling zinc fingers or signaling mechanisms has yet to be looked at.
ROBERT NESHEIM: Orville, I am sure you have some comment, particularly with your recent article in Nature discussing the role of selenium in this area.
ORVILLE LEVANDER: There is quite a bit of work going on with selenium right now in differential display as well. What I was wondering about, Bob, is, on your differential display analysis, that was rat intestine and you had a lot of sequences up there, but is anything happening with metallothionein under those conditions? I did not see it on the slides.
ROBERT COUSINS: No. In the one slide I showed where we wanted to prove that the technique was working, we actually used kidney RNA as the source material because in a dietary study, where zinc is provided over a long period of time, the kidney metallothionein RNA is a very nice titrator of intake. So we just wanted to show that the technique under nutritional conditions, feeding rats for two weeks, would show a difference in MRE [metal-responsive element]-regulated genes, so that is why we used the kidney for that purpose.
But in the intestine, where metallothionein also is produced, you should see that response, and we have not as yet picked it up. It is kind of interesting. I do not know what the reason is—we do not select for any given size—it just probably has not come up yet. We have gone through roughly half the combinations that are possible.
EDWARD HIRSCH: What was the time course of the response to the dietary supplement?
ROBERT COUSINS: Very quickly. We see changes within hours.
GABRIEL VIRELLA: Have any of the components of the immune system, for example cytokines or complement, been tested in this system?
ROBERT COUSINS: Yes, again, using a metallothionein system as the prototype, there has been a lot of research done on that.
GABRIEL VIRELLA: I mean in working with cytokines.
ROBERT COUSINS: Oh, working with the cytokines in differential displays and so on? Not to my knowledge, no. I think it is a technology that up until now has been limited primarily to differences in transformed versus nontransformed cells, and looking for what genes are differentially expressed. I do not think the physiologists and immunologists have caught on to it quite yet, but they will, certainly.
JOHN FERNSTROM: I am ignorant in this area. Is zinc something for which there is a regulated storage pool, such that one can think about the responsiveness from dietary changes as needing to be short responses or long responses?
ROBERT COUSINS: Yes. I did not have time to go into that aspect of things. There are at least two pools. One is the vast majority of zinc that is tied up in zinc-finger proteins and in enzymes and so on in cells, which are turning over, of course, at varying rates, and then there is a shorter response, a very small pool, that is yet to be defined. Some of that may well be bound to metallothionein.
There is something in the literature called a rapidly exchanging pool. The nature of that, outside of the fact that it occurs and can be shown kinetically to occur, is not known. But metallothionein could well serve part of that function because it expands and contracts, depending upon the amount of zinc available.
JOHN FERNSTROM: For a known zinc function, how fast does the deficiency become expressed in terms of loss of a particular function?
ROBERT COUSINS: It is believed to occur very quickly. Good examples of that are really hard to come by, but it is a Type II deficiency [tissue levels are maintained], where, as soon as you get to a certain point, things crash in a hurry. You will see reductions in growth and changes in various other things. In the case of zinc you see skin problems and so on. Immune problems occur very quickly.
WILLIAM BEISEL: I was just going to comment on the still very unexplored aspect of the cytokines because the proinflammatory cytokines do turn on this sequestration of zinc. Cytokines are the trigger. Within 15 minutes you see the increase in metallothionein generation, so it is a fantastically rapid response system and needs a lot more exploration.
ROBERT COUSINS: That is why I mentioned that zinc is in some ways an ideal entity if someone wanted to use a technology such as introduction of transgenes into human subjects by various means. You could then regulate those genes very quickly and with somewhat nontoxic conditions.
ROBERT NESHEIM: Thank you very much, Bob. In addition to the discussion of zinc, I learned that there is another MRE to explain.