Iron and heme metabolism

In hemoglobin and myoglobin, the heme iron remains in the ferrous or Fe²⁺ state throughout the cycle of oxygen binding and release. In redox-active enzymes and in the respiratory chain, heme regularly goes back and forth between the ferrous and the ferric or Fe³⁺ states, and sometimes also the ferryl or Fe⁴⁺ state.

In cytochrome P450 enzymes, the function of heme is to facilitate the formation of reactive oxygen (see slide 19.2). Similarly, formation of reactive oxygen species (ROS) may also occur as a side reaction wherever heme functions in transport of oxygen or of electrons, and protective mechanisms are required to contain damage by ROS. In hemoglobin, molecular oxygen (O₂) may abscond with an extra electron and thereby turn itself into superoxide (O₂^− •), while the heme iron is left one electron short. Hemoglobin that has lost one electron is called methemoglobin; it is unable to carry oxygen. Its iron is reduced to the ferrous form again by an NADH-dependent enzyme, methemoglobin reductase.102

Protective mechanisms that scavenge reactive oxygen species include the enzymes superoxide dismutase and catalase, as well as antioxidants such as tocopherol (vitamin E; see slide 18.7.11) and ascorbic acid (vitamin C). Organisms that lack such protective mechanisms are anaerobic, that is, they survive only in the absence of oxygen. Anaerobic organisms occur among both prokaryotes and eukaryotes.

Heme contained in the food is taken up from the intestines, but this is only relevant because of the iron it contains. The organic porphyrin ring is mostly synthesized from scratch. The synthetic pathway is split across two compartments: the initial and final steps occur in the mitochondria, while the intervening ones occur in the cytosol. Heme exerts feedback inhibition on the first step in the pathway.

The first reaction in the porphyrin synthesis pathway occurs in the mitochondria. Glycine forms a Schiff base with pyridoxal phosphate (PLP) in the aminolevulinate synthase enzyme. Pyridoxal phosphate acts as a transient “electron sink”, as it usually does and as is illustrated for serine hydroxymethyltransferase in slide 15.2.4. Decarboxylation of the enzyme-bound glycine produces a carbanion intermediate, which reacts with the carbonyl carbon in succinyl-CoA.

The aminolevulinate synthase reaction is the committed step of heme synthesis and is subject to feedback inhibition by heme, the final product of the pathway. The reaction product, δ-aminolevulinic acid (δ-ALA), is transported to the cytosol, where the next four reactions occur.

In the second reaction of heme synthesis, porphobilinogen synthase condenses two molecules of δ-ALA into one molecule of porphobilinogen, with the elimination of two molecules of water. At the outset, both substrate molecules form Schiff bases with two separate lysine residues in the active site. The water is already lost at this stage; however, the more interesting reaction steps that lead to the formation of the pyrrole ring still lie ahead. It should be noted that the reaction mechanism depicted here is still somewhat tentative [133].

This pretty picture (rendered from 1pv8.pdb) shows a model compound (structure on the right) inside the active site of porphobilinogen synthase, next to a zinc ion. It sheds little light on the reaction mechanism, or on the specific role of zinc. The only point I want to make here is that zinc is important for this enzyme. In lead intoxication, the zinc in porphobilinogen synthase is dislodged by lead, which inhibits the enzyme and therefore the biosynthesis of heme and hemoglobin.

Porphobilinogen deaminase (or hydroxymethylbilane synthase) turns four molecules of porphobilinogen into one linear tetramer. The process involves the priming of the enzyme with two porphobilinogen subunits (shown in blue and green) which remain with the enzyme molecule throughout and serve as a prosthetic group. Another unusual feature of this enzyme is that it uses ammonia as a leaving group. (In this scheme, Ac and P denote the acetate and propionate groups that are part of the porphobilinogen molecule; compare slide 17.2.3.)

Hydroxymethylbilane, when left alone, cyclizes spontaneously, giving rise to a ring with fourfold rotational symmetry, with acetyl groups and propionyl groups alternating regularly around the perimeter of the molecule. The product of this spontaneous reaction is named uroporphyrinogen I.

Nature, however, prefers another isomer, namely, uroporphyrinogen III. To forestall its spontaneous cyclization into uroporphyrinogen I, hydroxymethylbilane is captured and processed directly by uroporphyrinogen III cosynthase, which forms a complex with porphyrinogen deaminase. The cosynthase reaction proceeds via the improbable-looking spiro intermediate shown here within brackets, which leads to the inverted orientation of the acetate and propionate side chains on the fourth ring in uroporphyrinogen III.

The next enzyme, uroporphyrinogen III decarboxylase, decarboxylates all four acetate groups to methyl groups. This gives coproporphyrinogen III, which again enters the mitochondria.

Oxidative decarboxylation converts two of the four propionate residues to vinyl groups, which produces protoporphyrinogen IX. Dehydrogenation to protoporphyrin IX is the final ring modification. Ferrochelatase then inserts the iron, which completes heme synthesis.

While the ferrochelatase reaction looks rather trivial and indeed can proceed with appreciable rate spontaneously, the enzyme is important: when it is absent, protoporphyrin IX accumulates, which gives rise to a clinically manifest form of porphyria [134].

Heme retains the vinyl groups that were introduced at the stage of protoporphyrinogen IX. Addition of the sulfhydryl groups of cysteine side chains across these vinyl groups can produce covalent bonds between heme and some of its apoproteins. Such covalent attachment occurs for example in cytochrome C oxidase, but it does not happen in hemoglobin or myoglobin.

Regardless of the underlying cause, the inhibition of heme synthesis will result in red blood cells that are smaller and contain less hemoglobin than normal ones. This condition is named microcytic, hypochromic anemia. The term “microcytic” refers to the reduced cell size, whereas “hypochromic” denotes the reduced intracellular hemoglobin concentration.

The accumulation of porphyrin precursors in various enzyme defects in porphyrin synthesis can cause photosensitization of the skin. The clinical picture can vary a bit, depending on the specific intermediate. As an example, we will consider the disease porphyria cutanea tarda (translated: chronic porphyria of the skin), which is the most common form of porphyria. Here, the deficient enzyme is uroporphyrinogen III decarboxylase (see slide 17.2.6). This leads to the accumulation of an aberrant metabolite, uroporphyrinogen III. Its likewise aberrant precursor uroporphomethene III is an inhibitor of the already deficient decarboxylase, which will tend to make things worse.

The accumulating uroporphyrinogen distributes throughout the body and becomes oxidized non-enzymatically to the non-physiological product uroporphyrin III. In the skin, uroporphyrin III can absorb photons and then react with molecular oxygen to produce reactive oxygen species; the latter inflict the skin tissue damage that is illustrated in the slide. An important aspect of treatment is the protection of skin from direct sunlight.

The absorbed wavelength range (or absorption spectrum) differs between the various porphyrins. Uroporphyrin III has an absorption peak at 405 nm, which is at the blue end of the visible spectrum; this peak is readily detectable in blood serum samples (left). The sun light is more intense in the visible range than in the UV range. Sun screen lotion, which is designed to absorb UV light but not visible light, will not prevent photosensitization by uroporphyrin III.

Sporadic PCT is often associated with disturbances of iron homeostasis. Approximately two thirds of sporadic PCT patients carry a deficient allele for the iron regulator HFE. In homozygous form, this gene defect causes hemochromatosis, a disease that is characterized by severe iron overload. HFE knockout mice show increased intestinal expression and activity of iron uptake transporters. Excess iron may facilitate the non-enzymatic oxidation of uroporphyrinogen to uroporphyrin. An intermediate of this oxidation, uroporphomethene (see slide 17.3.2), inhibits uroporphyrin decarboxylase, which would account for the role of iron overload in pathogenesis [135].

Iron overload of the liver can also occur in chronic infections and in other chronic inflammatory diseases (see section 17.5). Blood letting—which depletes iron—is reportedly beneficial in PCT, regardless of the cause of the iron overload.

While the final reaction in heme synthesis—namely, the emplacement of Fe⁺⁺ into the completed protoporphyrin ring—looks like a facile reaction, and indeed proceeds with appreciable speed without ferrochelatase, the enzyme is still needed to make the reaction go fast enough. If the enzyme is deficient, protoporphyrin accumulates and causes skin manifestations similar to those observed in porphyria cutanea tarda.

The genetic defect in acute intermittent porphyria concerns the enzyme porphobilinogen deaminase. Both porphobilinogen and δ-ALA accumulate. Porphobilinogen is excreted with the urine and, through spontaneous oxidation, forms a characteristic red pigment. The more severe clinical symptoms, however, are due to accumulation of δ-aminolevulinate. This metabolite competitively inhibits the binding of γ-aminobutyrate (GABA), an inhibitory neurotransmitter, to its receptors [136], which likely causes both the psychiatric symptoms (agitation, confusion) and the neurological ones (nausea, abdominal pain) in AIP.

The dysregulation mainly affects heme synthesis in the liver. As the name of the disease indicates, it is not always manifest but only intermittently. Heme is the prosthetic group of cytochrome P450 enzymes, which are important in drug metabolism and are induced in the liver by various drugs (see slide 19.2.2). It appears that ALA synthase is induced along with the cytochrome P450 enzymes, and AIP attacks are often triggered or aggravated by the application of such drugs.

Specific drugs that induce cytochrome P450 and ALA synthase include barbituric acid derivatives and carbamazepine, which were, and occasionally still are, used in the treatment of psychiatric symptoms. Fatal outcomes have occurred when AIP patients were misdiagnosed and treated with barbituric acid derivatives.103 One element of the correct therapy consists in the application of heme, which inhibits ALA synthase.

Some fairly simple clues can narrow down the cause of jaundice in a given patient. If excretion of bilirubin is blocked, the pigments derived from it will be absent, and the stool will have a grayish color.

Hemolysis consists in the accelerated decay of red blood cells; it may result from biochemical causes such as glucose-6-phosphate dehydrogenase deficiency (see slide 9.4), or from immunological ones such as Rhesus incompatibility. In hemolysis, the serum level of unconjugated bilirubin will be more strongly increased than that of the diglucuronide. On the other hand, when the flow of the bile is backed up, the conjugated bilirubin will spill back into the serum and will be increased.

Liver diseases such as hepatitis can affect synthesis, conjugation and biliary secretion of bilirubin to various degrees, and either form of bilirubin can be more strongly increased than the other.

Neonatal jaundice is a normal event that is caused by a transiently low level of UDP-glucuronosyltransferase 1A1.104 In most cases, it does not cause any problems and simply goes away within two weeks after birth, as the level of enzyme activity increases. If the serum level of bilirubin gets too high, however, it may accumulate in the brain and cause neurological problems (see next slide). To prevent this, newborns can be treated with phototherapy (see slide 17.4.4).

Crigler-Najjar syndrome is due to the genetic deficiency of UDP-glucuronosyltransferase 1A1. It initially becomes manifest as neonatal jaundice; however, since the gene defect is permanent, the situation doesn’t improve with time. As in many other gene defects, there are variants with total or partial disruption of enzyme activity. Milder cases are classified as Gilbert’s syndrome; the boundary between Gilbert’s and Crigler-Najjar syndrome will be somewhat arbitrary. When residual enzyme activity is present, it may be increased with drugs such as phenobarbital that transcriptionally induce it.

As in neonatal jaundice, phototherapy is also used in Crigler-Najjar syndrome, but its efficiency decreases with time, since growth reduces the body’s surface-to-volume ratio, and therefore a diminishing fraction of the bilirubin in the body can be reached by the illumination. The disease is best treated with liver transplants, as the transplanted liver will not be affected by the underlying gene defect and be able to conjugate and excrete bilirubin.

This slide shows a brain section from a newborn with severe bilirubin encephalopathy. The yellow color in the deeper structures of the forebrain, the so-called basal ganglia, is due to bilirubin accumulation. Through an as yet unknown biochemical mechanism [137–139], bilirubin causes damage to the basal ganglia, which results in motor dysfunction and other neurological symptoms.

In phototherapy, bilirubin absorbs photons and subsequently undergoes cis-trans isomerization across the two remaining double bonds between the pyrrole rings of the bilirubin molecule, as well as ring formation [140]. This slide shows some of the photochemical reaction products. The 4Z,15Z isomer of bilirubin (top left) is the one that is produced directly by biliverdin reductase, and which is eliminated very slowly in the unconjugated form. The other isomers are eliminated more rapidly; the fastest rate of elimination is observed with lumirubin (cyclobilirubin).

While the absorption maximum of bilirubin is in the blue wavelength band, green light reportedly produces lumirubin more efficiently, and it also induces less cytotoxic byproducts in cell culture models [141]. It seems, however, that blue lamps are still more widely used in practice, and the literature makes no mention of significant side effects of blue light, even in the long-term treatment of Crigler-Najjar patients [142].

The inhibition of heme oxygenase with Sn-mesoporphyrin has been used successfully in clinical studies to treat neonatal jaundice. The study summarized in this slide examined the effectiveness of Sn-mesoporphyrin in the treatment of newborns with glucose-6-phosphate dehydrogenase deficiency (see slide 9.4). In this condition, the lifespan of red blood cells is diminished, which increases the rate of heme degradation; newborns therefore are at an increased risk of severe jaundice.

Remarkably, a single injection of the drug was sufficient to reduce the peak levels of bilirubin to a greater extent than the reference treatment (phototherapy). It should be noted that both forms of treatment sufficed to keep the peak bilirubin levels below 25 mg/dL, which is considered sufficient to avoid damage to the brain. Phototherapy currently remains the standard treatment in clinical practice. Figure prepared from original data in [143].

Nitric oxide is an important signaling molecule. As a small molecule, it can easily diffuse out of one cell and into another. Inside the target cell, it binds to soluble guanylate cyclase (sGC). Interestingly, the NO-binding site on sGC is a heme molecule. Binding of NO to one face of the heme releases a histidine side chain on the other, which causes a conformational change and activation of the sGC molecule.

In vitro experiments show that CO can also bind and activate sGC. It has therefore been proposed that heme oxygenase, which produces CO, has a regulatory role in addition to its metabolic one. However, while this idea has been around for awhile, I have not come across solid evidence that supports a significant signaling role of CO in vivo.

Ferritin is a hollow protein particle that consists of 24 identical subunits. Up to ~4,500 iron ions can be stored inside the particle, in the form of FeOOH.

On the left, two neighboring subunits have been removed from the front of the sphere, and we are peeking into the interior of the shell, which in vivo would be packed with iron. On the right, some more subunits have been removed, and the iron ions actually contained in the crystal structure—just one per subunit—are shown as spheres. Can you figure out why the crystallographers did not try to obtain a structure of the fully loaded particle? Rendered from 1fha.pdb.

Much of the iron regulation is accomplished right in the small intestine. Iron that is taken up is initially stored inside the epithelial cells; surplus iron that is not requested from this store is simply lost when the epithelial cells are replaced, which happens about once a week.

In diseases that cause iron overload, the storage capacity of ferritin will be exceeded. Excess iron will then precipitate around the ferritin particles and cause them to aggregate. These iron-rich aggregates are called hemosiderin.

Iron overload can occur as a result of multiple blood transfusions, or of genetic defects that interfere with iron transport. An example is a defect of ceruloplasmin, a copper-containing serum protein that oxidizes Fe²⁺ to Fe³⁺.

In the figure, the brown stipples represent the hemosiderin. The dark blue blotches are out of focus and are just precipitated dye particles, that is, artifacts; this tends to happen if dye solutions stand around on the shelf for too long.