Brief Critical Reviews September 1996: 285-292

Discovery of the Hemochromatosis Gene Will Require Rethinking the Regulation of Iron Metabolism

The identity of the protein responsible for hemo- chromatosis, the iron overload disease, has elud- ed scientists for years. However, a recent report identifies the gene where the hemochromatosis defect lies. It is a gene that encodes a major his- tocompatibility complex (MHC) class-I-like protein called HLA-H. The mechanism by which an HLA-H defect alters iron metabolism is still un- identified. However, this new discovery will cer- tainly ignite a new wave of study into the physi- ology of iron metabolism and its regulation.

Hemochromatosis is an autosomal recessive disor- der of iron metabolism that is one of the most prev- alent genetic diseases in people of Northern Euro- pean descent-1 in 300 individuals are afflicted and 1 in 10 people carry the gene for the disease.' The primary defect of iron metabolism in hemochro- matosis appears to be an inability to down-regulate intestinal iron absorption when iron stores are high. Since the ability to excrete iron is limited to the loss of blood and the sloughing off of intestinal cells loaded with iron, inappropriately high iron absorp- tion leads directly to excess iron accumulation.

Despite its high prevalence, hemochromatosis is often diagnosed later in life, after organ failure oc- curs. Untreated patients may suffer from a number of serious diseases, such as cirrhosis, hepatomas, diabetes, cardiomyopathy, and arthritis. If diagnosed early, these conditions can be avoided simply by elimination of excess iron through regular blood do- nation.

The identity of the hemochromatosis gene de- fect has eluded investigators despite the fact that they have mapped the gene to a spot near the major histocompatibility complex (MHC) on chromosome 6p. However, a recent paper by Feder and his as- sociates2 has now identified a gene that accounts for 83% of the cases of hemochromatosis.

In order to identify the hemochromatosis gene, Feder et a1.2 had to first characterize the entire he- mochromatosis gene region on chromosome 6. This

This review was prepared by James C. Fleet, Ph.D. at the Jean Mayer Human Nutrition Research Center on Aging at Tufts University, 71 1 Washington Street, Boston MA 021 11, USA.

was done by identifying the region of maximum linkage disequilibrium using both pointwise analy- sis and haplotype analysis to identify historic cross- over events. This approach allowed them to narrow the search area to a 250-kb region on chromosome 6. At this point, they identified potential genes with- in the 250-kb region by cDNA selection, exon trap- ping, and genomic DNA sequencing. The combined analysis of the 250-kb region showed that it con- tained 15 genes.

indicate the influence of a founder effect in hemo- chromatosis, confirming a likelihood that a majority of affected chromosomes have the same mutation. According to Feder, 80% or more of patients have the same genetic mutation. Conversely, in control subjects, the mutation should be rare (5% or less). With this in mind, Feder et aL2 analyzed all 15 genes in the 250-kb hemochromatosis region for se- quence differences using DNA from two patients with hemochromatosis and two control subjects. They found 18 differences in the DNA between the control and hemochromatosis patients, of which 15 were silent changes that did not alter the amino acid coding of the genes. Two of the 15 genes contained changes that would alter amino acid composition of their gene products. Of the three remaining muta- tions, two changes were in a histone H1 gene. How- ever, because of their frequency in control subjects (61% and 32% of 28 control subjects had the first or the second change), neither alteration was con- sistent with the calculated frequency of the hemo- chromatosis mutation.

The only mutation consistent with the observed frequency of the hemochromatosis haplotype was noted in a gene encoding an MHC class-I-like pro- tein, designated by the authors as HLA-H. The mu- tation is a single-base mutation of G to A resulting in a cysteine to tyrosine substitution at amino acid 282 in the 352-residue protein. It was found to oc- cur in 85% of the chromosomes from 178 hemo- chromatosis patients studied, but in only 3.2% of control chromosomes (n = 310). The fact that only 83% of the 178 hemochromatosis patients studied were homozygous for the G- to A-mutation sug- gests that additional mutations within HLA-H can also lead to hemochromatosis. No additional mu- tations in HLA-H were identified that might account

Linkage analyses by Feder et al.2 and

Nutrition Reviews, Vol. 54, No. 9 285

for the 17% of affected patients who do not have the G- to A-mutation.

There is an additional line of evidence that sup- ports an important role for an MHC class-I-like pro- tein in hemochromatosis. Two groups of researchers have demonstrated hemochromatosis-like iron over- load in mice from which the p2 microglobulin gene has been eliminated (a gene ‘‘kno~kout’’) .~~~ p2 mi- croglobulin is a small-molecular-weight protein that forms a complex with MHC class I proteins on the surface of cells. Found on almost all nucleated cells, p2 microglobulin-MHC class I protein complexes present fragments of foreign proteins (frequently of viral origin) to cytotoxic T cells in a process essen- tial for the elimination of infected cells from the body. Without p2 microglobulin, the MHC class I proteins appear to be destabilized; the cell surface levels of HLAs are reduced in p2 microglobulin knockout mice. Thus, while the observation of iron overload in these mice did not point to a specific gene, it caused Rothenberg and Voland3 to hypoth- esize that the hemochromatosis gene was an HLA protein. It is noteworthy that the cysteine-to-tyro- sine mutation in HLA-H is very close to the ex- pected site of p2 microglobulin association. It is therefore possible that the association of HLA-H with pz microglobulin is disrupted in hemochro- matosis. Alternatively, a similar mutation in a dif- ferent HLA prevents the movement of the protein from the endoplasmic reticulum to the cell surface. In either event, the end result is the same-less cell surface expression of HLA-H.

It has been the expectation of most scientists that the discovery of the hemochromatosis gene would clarify our understanding of how the body regulates iron metabolism. However, since neither an MHC class I-like protein in iron metabolism nor intestinal iron absorption conveniently fits our cur- rent understanding of these areas, the finding of Feder et al.2 appears to do just the opposite. The models currently developed to explain the role of HLA-H in iron metabolism are purely speculative. Rothenberg and Voland3 suggest that HLA-H is found on the brush border membrane of enterocytes and serves as a receptor for an iron chelate. While this hypothesis is supported by the observation that other nonclassical MHC class I proteins are found on the brush border surface of enterocytes, it ap- pears to ignore the fact that heme and nonheme iron in the diet are absorbed by separate mechanisms and that absorption of both forms is elevated in hemo- chromat~sis.~ Thus, if the hemochromatosis defect was in a brush border receptor, leading to elevated iron uptake, two receptors would have to be affect- ed. And since HLA-H levels are likely to be lower in hemochromatosis, models in which HLA-H pro- motes iron absorption are not feasible. HLA-H must

therefore be involved in the down-regulation of the normal transport process.

McLaren et a1.6 have conducted kinetic studies of iron absorption in patients with hemochromato- sis. These studies suggest that the primary defect in the disease results from the excessive and unregu- lated transfer of iron from intestinal epithelial cells to the blood, i.e., a basolateral membrane defect. Feder et al.2 used these data to suggest that HLA-H is a basolateral protein with several potential func- tions in enterocytes. First, it could serve as an iron- binding ligand that is then internalized to signal to the enterocyte that iron stores are high. Alternative- ly, HLA-H could be a receptor linked to a signal transduction pathway with activation of the receptor leading to the regulation of genes that control iron absorption. In either case, if HLA-H was not present or could not bind iron from the plasma, the enter- ocyte would not recognize that iron stores had be- come elevated and iron absorption would continue to be high. Finally, Feder et aL2 suggest that the protein may act indirectly through association with components of the immune system that influence iron metabolism. What these components are re- mains to be determined. However, the fact that HLA-H mRNA can be found in a wide variety of tissues (as is the case with most other HLAs) sug- gests that the defect is not limited to the intestine.

While it was reasonable to expect that the dis- covery of the hemochromatosis gene would clarify the mechanism of intestinal iron absorption and iron metabolism in humans, we now realize that the task of fitting HLA-H into a model will not be so simple. The identification of HLA-H as the primary target affected by hemochromatosis raises far more ques- tions than it answers. However, this finding is likely to stimulate renewed interest in the regulation of iron metabolism. In addition, the fact that HLA-H is related to proteins classically involved in immune function suggests a more intimate relationship be- tween iron metabolism and immunology than ever before realized.

1. Crawford DHG, Powell LW, Halliday JW, Leggett BA. Factors influencing disease expression in hemochro- matosis. Ann Rev Nutr 1996;16:139-60

2. Feder JN, Gnirke A, Thomas W, Tsuchihashi Z, et al. A novel MHC class I-like gene is mutated in patients with hereditary haemochrmatosis. Nat Genet 1996; 13: 399-408

3. Rothenberg BE, Voland JR. pz Knockout mice develop parenchymal iron overload: A putative role for class I genes of the major histocompatibility complex in iron metabolism. Proc Natl Acad Sci USA 1996;93:152%34

4. de Sousa M, Reimao R, Lacerda R, Hugo P, Kauf- mann SHE, Port0 G. Iron overload in p,-microglobulin- deficient mice. lmmunol Lett 1994;39:105-11

286 Nutrition Reviews, Vol. 54, No. 9

5. Lynch SR, Skikne BS, Cook JD. Food iron absorption in idiopathic hemochromatosis. Blood 1989;74:2187-93

6. McLaren GD, Nathanson MH, Jacobs A, Trevett D,

Thomson W. Regulation of intestinal iron absorption and mucosal iron kinetics in hereditary hemochroma- tosis. J Lab Clin Med 1991;117:390-401

Differential Display PCR: A New Age in Nutrition Investigation

Molecular biology has provided nutrition science with a powerful experimental tool for exploring the molecular basis of essential nutrient deficiencies. Differential display polymerase chain reaction has emerged as an instrument of unlimited po- tential for assessing the manner by which nutri- ents regulate cell functions.

Nutrition investigations have relied almost entirely on data from human or whole animal studies to de- termine the roles of vitamins, minerals, and essen- tial amino acids. A missing dietary component can leave an indelible mark on biologic systems, such as a failed enzyme or an impaired physiologic func- tion. Today, nutrition science has a new tool for spotting effects caused by missing nutrients. This technique is called differential display polymerase chain reaction (PCR). As the name implies, PCR is put to the task of examining which genes are and which are not being expressed.

PCR is a highly sensitive and rapid technique for detecting extremely small amounts of a specific DNA or mRNA. A reverse transcription enzyme is used to convert the mRNAs to their double-stranded complementary DNA (cDNA), which can then be amplified millions of times to detectable levels. With the use of differential display PCR, one is able to make comparisons between mRNA populations from control and experimental animals, paying special attention to mRNA species that have been augment- ed or diminished. In the past, amplifying mRNAs meant inserting all messages into plasmids and let- ting a bacteria replicate the lot, then waiting 3-5 days for an answer. In differential display PCR, the DNA polymerase enzyme works outside the cell and meets the same objective in one day or less.

Differential display PCR is a relatively new technique. Its origins go back about 4 years, when Liang and Pardee’ first determined that a combi- nation of 12 random primers with an anchored cDNA primer would successfully generate a set of

This review was prepared by Edward D. Harris at the Department of Biochemistry and Biophysics and the Faculty of Nutrition, Texas A&M University, Col- lege Station, TX 77843-2128, USA.

fragments of DNA that represented the total mRNA of a cell. Using the new technology, one was readily able to show that the pattern of fragments derived from one cell type was both reproducible and easily discernible from those of other cell types.* In time, other disciplines saw the value of using differential display, and soon the method was applied to as- sessing effects of retinoic acid3 and growth factors4 on cells, and even to studying molecular mecha- nisms in cellular differentiation, aging, and senes- ~ e n c e . ~ - ~ To date, few studies have determined mod- ifications caused by a nutrient intake or a variable nutrient environment.

What makes differential display PCR so re- markable is the ability to obtain measurable quan- tities of a low-abundance nucleic acid beginning with only traces of starting nucleic acid. A typical protocol used by Wang et a1.8 is shown in Figure 1. Starting with about 0.2 pg (2 X g) of DNA- free RNA, the RNA is reverse transcribed with a primer designated T,, MN, i.e., 12 thymidines linked to two other nucleotides, MN, at the 3’ end, where M is either an A, C, or G (any nucleotide except T) and N is A, C, G, or T. The string of T’s will bind to the poly A tail of the mRNA and the M and N will assure randomness from the point of extension. In this way, a large number of mRNAs in the total population will be reverse transcribed, each extending the so-called anchored primer.

In step 2, the newly formed cDNAs react with arbitrary primers and the anchored primers, ampli- fying the section between the two primers. A Taq polymerase enzyme is required for this step. To as- sure that the highest populations of mRNAs are af- fected, 10 different variable primers are used, re- sulting in 40 primer set combinations. The frag- ments generated during the PCR phase are present in microgram quantities, sufficient enough to visu- alize by ethidium fluorescence on acrylamide gels and to perform sequence analysis and other appli- cations. It should be noted that the genes identified by differential display PCR must be confirmed as being amplified by using either a more conventional analysis such as Northern blot analysis, or by nu- clear run-on assays. Quantitatjon by these methods can readily be achieved using fragments created by the PCR DNA as probes.

Nutrition Reviews, Vol. 54, No. 9 287