Biosynthesis of Cholesterol, Steroids, and Isoprenoids

Cholesterol is doubtless the most publicized lipid in nature, because of the strong correlation between high levels of cholesterol in the blood and the incidence of diseases of the cardiovascular system in humans. Less well-advertised is the critical role of cholesterol in the structure of many membranes and as a precursor of steroid hormones and bile acids. Cholesterol is an essential molecule in many animals, including humans. It is not required in the mammalian diet because the liver can synthesize it from simple precursors.
Although the structure of this 27-carbon compound suggests complexity in its biosynthesis, all of its carbon atoms are provided by a single precursor-acetate (Fig. 20-30). The biosynthetic pathway to cholesterol is instructive in several respects. The study of this pathway has led to an understanding of the transport of cholesterol and other lipids between organs, of the process by which cholesterol enters cells (receptor-mediated endocytosis), of the means by which intracellular cholesterol production is influenced by dietary cholesterol, and of how failure to regulate cholesterol production affects health. Finally, the isoprene units that are key intermediates in the pathway from acetate to cholesterol are precursors to many other natural lipids, and the mechanisms by which isoprene units are polymerized are similar in all of these pathways.
We begin with an account of the major steps in the biosynthesis of cholesterol from acetate, then discuss the transport of cholesterol in the blood, its uptake by cells, and the regulation of cholesterol synthesis in normal individuals and in those with defects in cholesterol uptake or transport. We also consider other cellular components derived from cholesterol, such as bile acids and steroid hormones. Finally, the biosynthetic pathways to some of the many compounds derived from isoprene units, which share early steps with the pathway to cholesterol, are outlined to illustrate the extraordinary versatility of isoprenoid condensations in biosynthesis.

Cholesterol Is Made from Acetyl-CoA in Four Stages

Cholesterol, like long-chain fatty acids, is made from acetyl-CoA, but the assembly plan is quite different in the two cases. In early experiments animals were fed acetate labeled with 14C in either the methyl carbon or the carboxyl carbon. The pattern of labeling in the cholesterol isolated from the two groups of animals (Fig. 20-30) provided the blueprint for working out the enzymatic steps in cholesterol biosynthesis. The process occurs in four stages (Fig. 20-31). In stage 1 the three acetate units condense to form a six-carbon intermediate, mevalonate. Stage 2 involves the conversion of mevalonate into activated isoprene units, and stage 3 the polymerization of six 5-carbon isoprene units ta form the 30-carbon linear structure of squalene. Finally (stage 4, the cyclization of squalene forms the four rings of the steroid nucleus, and a further series of changes (oxidations, removal or migration of methyl groups) leads to the final product, cholesterol. 1. Synthesis of Mevalonate from Acetate The first stage in cholesterol biosynthesis leads to the intermediate mevalonate (Fig. 20-32). Two molecules of acetyl-CoA condense, forming acetoacetyl-CoA, which condenses with a third molecule of acetyl-CoA to yield the six-carbon compound β-hydroxy-β-methylglutaryl-CoA (HMG-CoA). These first two reactions, catalyzed by thiolase and HMG-CoA synthase, respectively, are reversible and do not commit the cell to the synthesis of cholesterol or other isoprenoid compounds. The third reaction is the committed step: the reduction of HMGCoA to mevalonate, for which two molecules of NADPH each donate two electrons. HMG-CoA reductase, an integral membrane protein of the smooth endoplasmic reticulum, is the major point of regulation on the pathway to cholesterol, as we shall see. 2. ConUersion of MeUalonate to Two Activated Isoprenes In the next stage of cholesterol synthesis, three phosphate groups are transferred from three ATP molecules to mevalonate (Fig. 20-33). The phosphate attached to the C-3 hydroxyl group of mevalonate in the intermediate 3-phospho-5-pyrophosphomevalonate is a good leaving group; in the next step this phosphate and the nearby carboxyl group both leave, producing a double bond in the five-carbon product, Δ3-isopentenyl pyrophosphate. This is the first of the two activated isoprenes central to cholesterol formation. Isomerization of Δ3-isopentenyl pyrophosphate yields the second activated isoprene, dimethylallyl pyrophosphate (Fig. 20-33) . Condensation of Six ActiUated Isoprene Units to Form Squalene Isopentenyl pyrophosphate and dimethylallyl pyrophosphate now undergo a "head-to-tail" condensation in which one pyrophosphate group is displaced and a 10-carbon chain, geranyl pyrophosphate, is formed (Fig. 20-34). (The "head" is the end to which pyrophosphate is joined.) Geranyl pyrophosphate undergoes another head-to-tail condensation with isopentenyl pyrophosphate, yielding the 15-carbon intermediate farnesyl pyrophosphate. Finally, two molecules of farnesyl pyrophosphate join head to head, with the elimination of both pyrophosphate groups, forming squalene (Fig. 20-34). The common names of these compounds derive from the sources from which they were first isolated. Geraniol, a component of rose oil, has the smell of geraniums, and farnesol is a scent found in the flowers of a tree, Farnese acacia. Many natural scents of plant origin are synthesized from isoprene units. Squalene, first isolated from the liver of sharks (genus Squalus), has 30 carbons, 24 in the main chain and 6 in the form of methyl group branches. 4.Conversion of Squalene to the Four-Iling Steroid Nucleus When the squalene molecule is represented as in Figure 20-35, the relationship of its linear structure to the cyclic structure of the sterols is apparent. All of the sterols have four fused rings (the steroid nucleus) and all are alcohols, with a hydroxyl group at C-3; thus the name "sterol." The action of squalene monooxygenase adds one oxygen atom from O2 to the end of the squalene chain, forming an epoxide. This enzyme is another mixed-function oxidase (Box 20-1); NADPH reduces the other oxygen atom of O2 to H2O. The double bonds of the product, squalene2,3-epoxide, are positioned so that a remarkable concerted reaction can convert the linear squalene epoxide into a cyclic structure. In animal cells, this cyclization results in the formation of lanosterol, which contains the four rings characteristic of the steroid nucleus. Lanosterol is finally converted into cholesterol in a series of about 20 reactions, including the migration of some methyl groups and the removal of others. Elucidation of this extraordinary biosynthetic pathway, one of the most complex known, was accomplished by Konrad Bloch, Feodor Lynen, John Cornforth, and George Popjak in the late 1950s. Cholesterol is the sterol characteristic of animal cells, but plants, fungi, and protists make other, closely related sterols instead of cholesterol, using the same synthetic pathway as far as squalene-2,3-epoxide. At this point the synthetic pathways diverge slightly, yielding other sterols: stigmasterol in many plants and ergosterol in fungi, for example (Fig. 20-35).

Eukaryotic Pathways to Phosphatidylserine, Phosphatidylethanolamine, and Phosphatidylcholine Are Interrelated

In yeast as in bacteria, phosphatidylserine can be produced by condensation of CDP-diacylglycerol and serine, and phosphatidylethanolamine can be synthesized from phosphatidylserine in the reaction catalyzed by phosphatidylserine decarboxylase (Fig. 20-25). An alternative route to phosphatidylserine is a head group exchange reaction,in which free serine displaces ethanolamine. Phosphatidylethanolamine may also be converted to phosphatidylcholine (lecithin) by the addition of three methyl groups to its amino group. All three methylation reactions are catalyzed by a single enzyme (a methyltransferase) with S-adenosylmethionine as the methyl group donor (see Fig. 17-20).

In mammals, phosphatidylserine is not synthesized from CDPdiacylglycerol; instead, it is derived from phosphatidylethanolamine via the head group exchange reaction shown in Figure 20-25. In mammals, synthesis of all nitrogen-containing phospholipids occurs by strategy 2 of Figure 20-22: phosphorylation and activation of the head group followed by condensation with diacylglycerol. For example, choline is reused ("salvaged") by being phosphorylated then converted into CDP-choline by condensation with CTP. A diacylglycerol displaces CMP from CDP-choline, producing phosphatidylcholine (Fig. 20-26).

An analogous salvage pathway converts ethanolamine obtained in the diet into phosphatidylethanolamine. In the liver, phosphatidylcholine is also produced by methylation of phosphatidylethanolamine using S-adenosylmethionine, as described above. In all other tissues, however, phosphatidylcholine is produced only by condensation of diacylglycerol and CDP-choline. The pathways to phosphatidylcholine and phosphatidylethanolamine in various organisms are summarized in Figure 20-27.

Plasmalogen Synthesis ftequires Formation of an Ether-Linked Fatty Alcohol

The biosynthetic pathway to ether lipids, including plasmalogens and the platelet-activating factor (see Fig. 9-8), involves the displacement of an esterified fatty acyl group by a long-chain alcohol to form the ether linkage (Fig. 20-28). Head group attachment follows, by mechanisms essentially like those for the common ester-linked phospholipids. Finally, the characteristic double bond of plasmalogens is introduced by the action of a mixed-function oxidase similar to that responsible for desaturation of fatty acids (Fig. 20-14).

Sphingolipid and Glycerophospholipid Synthesis Share Precursors and Some Mechanisms

The biosynthesis of sphingolipids occurs in four stages: (1) synthesis of the 18-carbon amine sphinganine from palmitoyl-CoA and serine; (2) attachment of a fatty acid in amide linkage to form ceramide; (3) desaturation of the sphinganine moiety to form sphingosine; and (4) attachment of a head group to produce a sphingolipid such as a cerebroside or sphingomyelin (Fig. 20-29). The pathway shares several features with the pathways leading to glycerophospholipids: NADPH provides reducing power and fatty acids enter as their activated CoA derivatives. In cerebroside formation, sugars enter as their activated nucleotide derivatives. Head group attachment in sphingolipid synthesis has several novel aspects. Phosphatidylcholine, rather than CDP-choline, serves as the donor of phosphocholine in the synthesis of sphingomyelin from the ceramide (Fig. 20-29). In glycolipids, the cerebrosides and gangliosides (see Fig. 9-9), the head group is a sugar, attached directly to the C-1 hydroxyl of sphingosine in glycosidic linkage, rather than through a phosphodiester bond; the sugar donor is a UDP-sugar (UDP-glucose or UDP-galactose).

Polar Lipids Are Targeted to Specific Cell Membranes

After their synthesis on the smooth endoplasmic reticulum, the polar lipids, including the glycerophospholipids, sphingolipids, and glycolipids, are inserted into different cell membranes in different proportions. The mechanism by which specific lipids are targeted for insertion into specific intracellular membranes is not yet understood. Because membrane lipids are insoluble in water, they cannot simply diffuse from their point of synthesis (the endoplasmic reticulum) to their point of insertion. Instead, they are delivered in membrane vesicles that bud from the Golgi complex then move to and fuse with the target membrane (see Figs. 2-10, 10-14). There are also cytosolic proteins that bind phospholipids and sterols and carry them from one cell membrane to another and from one face of a lipid bilayer to the other. The combined action of transport vesicles and these proteins (and perhaps other proteins yet to be discovered) produces the characteristic lipid composition of each organelle membrane

Biosynthesis of Triacylglycerols

Most of the fatty acids synthesized or ingested by an organism have one of two fates: incorporation into triacylglycerols for the storage of metabolic energy or incorporation into the phospholipid components of membranes. The partitioning between these alternative fates depends on the requirements of the organism. During rapid growth, the synthesis of new membranes requires membrane phospholipid synthesis; organisms that have a plentiful supply of food but are not actively growing shunt most of their fatty acids into storage fats. The pathways to storage fats and several classes of membrane phospholipids begin at the same point: the formation of fatty acyl esters of glycerol. First we discuss the route to triacylglycerols and its regulation.

Triacylglycerols and Glycerophospholipids Are Synthesized from Common Precursors

Animals can synthesize and store large quantities of triacylglycerols, to be used later as fuel In humans only a few hundred grams of glycogen can be stored in the liver and muscles, barely enough to supply the body's energy needs for 12 hours. In contrast, the total amount of stored triacylglycerol in a 70 kg man of average build is about 15 kg, enough to supply his basal energy needs for as long as 12 weeks Whenever carbohydrate is ingested in excess of the capacity to store glycogen, it is converted into triacylglycerols and stored in adipose tissue. Plants also manufacture triacylglycerols as an energy-rich fuel, stored especially in fruits, nuts, and seeds.
Triacylglycerols and glycerophospholipids such as phosphatidylethanolamine share two precursors (fatty acyl-CoAs and glycerol-3-phosphate) and several enzymatic steps in their biosynthesis in animal tissues. Glycerol-3-phosphate can be formed in two ways (Fig. 20-18). It can arise from dihydroxyacetone phosphate generated during glycolysis by the action of the cytosolic NAD-linked glycerol-3phosphate dehydrogenase, and in liver and kidney it is also formed from glycerol by the action of glycerol kinase. The other precursors of triacylglycerols are fatty acyl-CoAs, formed from fatty acids by acylCoA synthetases (Fig. 20-18), the same enzymes responsible for the activation of fatty acids for β oxidation

The first stage in the biosynthesis of triacylglycerols is the acylation of the two free hydroxyl groups of glycerol-3-phosphate by two molecules of fatty acyl-CoA to yield diacylglycerol-3-phosphate, more commonly called phosphatidate (Fig. 20-18). Phosphatidate occurs in only trace amounts in cells, but is a central intermediate in lipid biosynthesis; it can be converted either to a triacylglycerol or to a glycerophospholipid. In the pathway to triacylglycerols, phosphatidate is hydrolyzed by phosphatidate phosphatase to form a 1,2-diacylglycerol (Fig. 20-19). Diacylglycerols are then converted into triacylglycerols by transesterification with a third fatty acyl-CoA.

Triacylglycerol Biosynthesis in Animals Is Regulated by Hormones In humans, the amount of body fat stays relatively constant over long periods, although there may be minor short-term changes as the caloric intake fluctuates. However, if carbohydrate, fat, or protein is consumed in amounts exceeding energy needs, the excess is stored in the form of triacylglycerols. The fat stored in this way can be drawn upon for energy and enables the body to withstand periods of fasting. The biosynthesis and degradation of triacylglycerols are regulated reciprocally, with the favored path depending upon the metabolic resources and requirements of the moment. The rate of triacylglycerol biosynthesis is profoundly altered by the action of several hormones. Insulin, for example, promotes the conversion of carbohydrate into triacylglycerols (Fig. 20-20). People with severe diabetes mellitus, due to failure of insulin secretion or action, not only are unable to use glucose properly but also fail to synthesize fatty acids from carbohydrates or amino acids. They show increased rates of fat oxidation and ketone body formation (Chapter 16). As a consequence they lose weight. Triacylglycerol metabolism is also influenced by glucagon (Chapter 22), and by pituitary growth hormone and adrenal cortical hormones.

Biosynthesis of Membrane Phospholipids

In Chapter 9 we introduced two major classes of membrane phospholipids: glycerophospholipids and sphingolipids. We noted that many different phospholipid species can be constructed by combining various fatty acids and polar head groups with the glycerol or sphingosine backbones (see Figs. 9-7, 9-9). Although the number of different end products of phospholipid biosynthesis is very large, all of these diverse products are synthesized according to a few basic patterns. We will describe the biosynthesis of selected membrane lipids to illustrate these patterns. In general, the assembly of phospholipids from simple precursors requires (1) synthesis of the backbone molecule (glycerol or sphingosine); (2) attachment of fatty acid(s) to the backbone, in ester or amide linkage; (3) addition of a hydrophilic head group, joined to the backbone through a phosphodiester linkage; and in some cases, (4) alteration or exchange of the head group to yield the fmal phospholipid product.
In eukaryotic cells, phospholipid synthesis occurs primarily at the surface of the smooth endoplasmic reticulum. Some newly synthesized phospholipids remain in that membrane, but most are destined for other cellular locations. The process by which water-insoluble phospholipids move from the site of their synthesis to the point of their eventual function is not fully understood, but we will conclude by discussing some mechanisms that have emerged in recent years.

There Are Two Strategies for Attaching Head Groups

The first steps of glycerophospholipid synthesis are shared with the pathway to triacylglycerols (Fig. 20-19): two fatty acyl groups are esterified to C-1 and C-2 of L-glycerol-3-phosphate to form phosphatidate. Commonly but not invariably, the fatty acid at C-1 is saturated and that at C-2 is unsaturated. A second route to phosphatidate is the phosphorylation of a diacylglycerol by a specific kinase. The polar head group of glycerophospholipids is attached through a phosphodiester bond, in which each of two alcoholic hydroxyls (one on the polar head group and one on C-3 of glycerol) forms an ester with phosphoric acid (Fig. 20-21). In the biosynthetic process, one of the hydroxyls is first activated by attachment of a nucleotide, cytidine diphosphate (CDP). Cytidine monophosphate (CMP) is then displaced in a nucleophilic attack by the other hydroxyl (Fig. 20-22, p. 662). The CDP is attached either to the diacylglycerol, forming in effect an activated phosphatidate, CDP-diacylglycerol (strategy 1), or to the hydroxyl of the head group (strategy 2). The central importance of cytidine nucleotides in lipid biosynthesis was discovered by Eugene P. Kennedy in the early 1960s. Phospholipid Synthesis in E. coli Employs CDP-Diacylglycerol The first strategy for head group attachment is illustrated by the synthesis of phosphatidylserine, phosphatidylethanolamine, and phosphatidylglycerol in E. coli. The diacylglycerol is activated by condensation of phosphatidate with CTP to form CDP-diacylglycerol, with the elimination of pyrophosphate (Fig. 20-23). Displacement of CMP through nucleophilic attack by the hydroxyl group of serine or by the C-1 hydroxyl of glycerol-3-phosphate yields phosphatidylserine or phosphatidylglycerol-3-phosphate, respectively. The latter is processed further by cleavage of the phosphate monoester (with release of Pi) to yield phosphatidylglycerol. Phosphatidylserine and phosphatidylglycerol can both serve as precursors of other membrane lipids in bacteria (Fig. 20-23). Decarboxylation of the serine moiety in phosphatidylserine by phosphatidylserine decarboxylase yields phosphatidylethanolamine. In E. coli, condensation of two molecules of phosphatidylglycerol, with the elimination of one glycerol, yields cardiolipin, in which two diacylglycerols are joined through a common head group.

Some Fatty Acids Are Desaturated

Palmitate and stearate serve as precursors of the two most common monounsaturated fatty acids of animal tissues: palmitoleate, 16:1(Δ⁹), and oleate, 18:1(Δ⁹) (Fig. 20-13). Each of these fatty acids has a single cis double bond in the Δ⁹ position (between C-9 and C-10). The double bond is introduced into the fatty acid chain by an oxidative reaction catalyzed by fatty acyl-CoA desaturase (Fig. 20-14). This enzymeis an example of a mixed-function oxidase (Box 20-1). Two different substrates, the fatty acid and NADPH, simultaneously undergo twoelectron oxidations. The path of electron flow includes a cytochrome (cytochrome b5) and a flavoprotein (cytochrome b5 reductase), both of which, like fatty acyl-CoA desaturase itself, are present in the smooth endoplasmic reticulum.
Mammalian hepatocytes can readily introduce double bonds at the Δ⁹ position of fatty acids but cannot introduce additional double bonds in the fatty acid chain between C-10 and the methyl-terminal end.

Linoleate, 18:2(Δ^9,12), and α-linolenate, 18:3(Δ^9,12,15, cannot be synthesized by mammals, but plants can synthesize both. The plant desaturases that introduce double bonds at Δ¹² and Δ¹⁵ positions are located in the endoplasmic reticulum. These enzymes act not on free fatty acids but on a phospholipid, phosphatidylcholine, containing at least one oleate linked to glycerol (Fig. 20-15, p. 656).
Because they are necessary precursors for the synthesis of other products, linoleate and linolenate are essential fatty acids for mammals; they must be obtained from plant material in the diet. Once ingested, linoleate may be converted into certain other polyunsaturated acids, particularly y-linolenate, eicosatrienoate, and eicosatetraenoate (arachidonate), which can be made only from linoleate (Fig. 20-13). Arachidonate, 20: 4(Δ^5,8,11,14), is an essential precursor of regulatory lipids, the eicosanoids.

Eicosanoids Are Formed from Arachidonate

Arachidonate is parent to the eicosanoids, a family of very potent biological signaling molecules that act as short-range messengers, affecting tissues near the cells that produce them. In response to a hormonal or other stimulus, a specific phospholipase present in most types of mammalian cells attacks membrane phospholipids, releasing arachidonate. Enzymes of the smooth endoplasmic reticulum then convert arachidonate into prostaglandins, beginning with the formation of PGH2, the immediate precursor of many other prostaglandins and thromboxanes (Fig. 20-16a). The two reactions that lead to PGH2 involve the addition of molecular oxygen; both are catalyzed by a bifunctional enzyme, prostaglandin endoperoxide synthase. Aspirin (acetylsalicylate; Fig. 20-16b) irreversibly inactivates this enzyme (thus blocking the synthesis of prostaglandins and thromboxanes) by acetylating a Ser residue essential to catalytic activity. Ibuprofen, a widely used nonsteroidal antiinflammatory drug (Fig. 20-16c), also acts by inhibiting this enzyme. Thromboxane synthase present in blood platelets (thrombocytes) converts PGH2 into thromboxane A2, from which other thromboxanes are derived (Fig. 20-16a). Thromboxanes induce blood vessel constriction and platelet aggregation, early steps in blood clotting. Low doses of aspirin, taken regularly, are believed to reduce the probability of heart attacks and strokes by reducing thromboxane production. Thromboxanes, like prostaglandins, contain a ring of five or six atoms, and the pathway that leads from arachidonate to these two classes of compounds is sometimes called the "cyclic" pathway, to distinguish it from the "linear" pathway that leads from arachidonate to the leukotrienes, which are linear (Fig. 20-17, p. 658). Leukotriene synthesis begins with the action of several lipoxygenases that catalyze the incorporation of molecular oxygen into arachidonate. These enzymes, found in leukocytes and in heart, brain, lung, and spleen, are mixed-function oxidases that use cytochrome P-450  The various leukotrienes differ in the position of the peroxide that is introduced by these lipoxygenases. This linear pathway from arachidonate, unlike the cyclic pathway, is not inhibited by aspirin or the other nonsteroidal antiinflammatory drugs.

The Fatty Acid Synthase Reactions Are Repeated to Form Palmitate

The production of the four-carbon, saturated fatty acyl-ACP completes one pass through the fatty acid synthase complex. The butyryl group is now transferred from the phosphopantetheine -SH group of ACP to the Cys -SH group of β-ketoacyl-ACP synthase, which initially bore the acetyl group (Fig. 20-5). To start the next cycle of four reactions that lengthens the chain by two more carbons, another malonyl group is linked to the now unoccupied phosphopantetheine -SH group of ACP (Fig. 20-6). Condensation occurs as the butyryl group, acting exactly as did the acetyl group in the first cycle, is linked to two carbons of the malonyl-ACP group with concurrent loss of CO₂. The product of this condensation is a six-carbon acyl group, covalently bound to the phosphopantetheine -SH group. Its β-keto group is reduced in the next three steps of the synthase cycle to yield the six-carbon saturated acyl group, exactly as in the first round of reactions.
Seven cycles of condensation and reduction produce the 16-carbon saturated palmitoyl group, still bound to ACP. For reasons not well understood, chain elongation generally stops at this point, and free palmitate is released from the ACP molecule by the action of a hydrolytic activity in the synthase complex. Small amounts of longer fatty acids such as stearate (18:0) are also formed. In certain plants (coconut and palm, for example) chain termination occurs earlier; up to 90% of the fatty acids in the oils of these plants are between 8 and 14 carbons long.

The overall reaction for the synthesis of palmitate from acetyl-CoA can be broken down into two parts. First, the formation of seven malonyl-CoA molecules:

7 Acetyl-CoA + 7CO₂ + 7ATP  -------->7 malonyl-CoA + 7ADP + 7P_i (20-1)

then seven cycles of condensation and reduction:

Acetyl-CoA + 7 malonyl-CoA + l4NADPH + 14H⁺ -------------->

palmitate + 7C0₂ + 8CoA + l4NADP⁺ + 6H₂O (20-2) The overall process (the sum of Eqns 20-1 and 20-2) is

8 Acetyl-CoA + 7ATP + l4NADPH + 14H⁺ ------------> palmitate + 8CoA + 6H₂O + 7ADP + 7P_i + l4NADP⁺ (20-3)

The biosynthesis of fatty acids such as palmitate thus requires acetyl-CoA and the input of chemical energy in two forms: the group transfer potential of ATP and the reducing power of NADPH. The ATP is required to attach CO₂ to acetyl-CoA to make malonyl-CoA; the NADPH is required to reduce the double bonds. We shall return to the sources of acetyl-CoA and NADPH soon, but let us first consider the structure of the remarkable enzyme complex that catalyzes the synthesis of fatty acids.

The Fatty Acid Synthase of Some Organisms Is Composed of Multifunctional Proteins

We noted earlier that the seven active sites for fatty acid synthesis (six enzymes and ACP) reside in seven separate polypeptides in the fatty acid synthase of E. coli; the same is true of the enzyme complex from higher plants (Fig. 20-7). In these complexes each enzyme is positioned with its active site near that of the preceding and succeeding enzymes of the sequence. The flexible pantetheine arm of ACP can reach all of the active sites, and it carries the growing fatty acyl chain from one site to the next; the intermediates are not released from the enzyxne complex until the finished product is obtained. As in the cases we have encountered in earlier chapters, this channeling of intermediates from one active site to the next increases the efficiency of the overall process.

The fatty acid synthases of yeast and of vertebrates are also multienzyme complexes, but their integration is even more complete than in E. coli and plants. In yeast, the seven distinct active sites reside in only two large, multifunctional polypeptides, and in vertebrates, a single large polypeptide (Mr 240,000) contains all seven enzymatic activities as well as a hydrolytic activity that cleaves the fatty acid from the ACP-like part of the enzyme complex. The active form of this multifunctional protein is a dimer (Mr 480,000). Fatty Acid Synthesis Occurs in the Cytosol of Many Organisms but in the Chloroplasts of Plants In mammals, the fatty acid synthase complex is found exclusively in the cytosol (Fig. 20-8), as are the biosynthetic enzymes for nucleotides, amino acids, and glucose. This location segregates synthetic processes from degradative reactions, many of which take place in the mitochondrial matrix. There is a corresponding segregation of electron-carrying cofactors for anabolism (generally a reductive process) and those for catabolism (generally oxidative). Usually, NADPH is the electron carrier for anabolic reactions, and NAD+ serves in catabolic reactions. In hepatocytes, the ratio [NADPH]/[NADP+] is very high (about 75) in the cytosol, furnishing a strongly reducing environment for the reductive synthesis of fatty acids and other biomolecules. Because the cytosolic [NADH]/[NAD+] ratio is much smaller (only about 8 ×10-4), the NAD+-dependent oxidative catabolism of glucose can occur in the same compartment, at the same time, as fatty acid synthesis. The [NADH]/ [NAD+] ratio within the mitochondrion is much higher than in the cytosol because of the flow of electrons into NAD+ from the oxidation of fatty acids, amino acids, pyruvate, and acetyl-CoA. This high [NADH]/ [NAD+] ratio favors the reduction of oxygen via the respiratory chain. In adipocytes cytosolic NADPH is largely generated by malic enzyme (Fig. 20-9). (We encountered an NAD-linked malic enzyme in the carbon fixation pathway of C4 plants (see Fig. 19-37); this enzyme is unrelated in function.) The pyruvate produced in this reaction reenters the mitochondrion. In hepatocytes and in the mammary gland of lactating animals, the NADPH required for fatty acid biosynthesis is supplied primarily by the reactions of the pentose phosphate pathway In the photosynthetic cells of plants, fatty acid synthesis occurs not in the cytosol, but in the chloroplast stroma (Fig. 20-8). This location makes sense when we recall that NADPH is produced in chloroplasts by the light reactions of photosynthesis (Fig. 20-10). Again, the resulting high [NADPH I/[NADP+] ratio provides the reducing environment that favors reductive anabolic processes such as fatty acid synthesis. light H2O + NADP+ > z0~ + NADPH + H~ Figure 20-10 Production of NADPH by photosynthesis. Acetate Is Shuttled out of Mitochondria as Citrate In nonphotosynthetic eukaryotes, nearly all the acetyl-CoA used in fatty acid synthesis is formed in mitochondria from pyruvate oxidation and from the catabolism of the carbon skeletons of amino acids. AcetylCoA arising from the oxidation of fatty acids does not represent a significant source of acetyl-CoA for fatty acid biosynthesis in animals because the two pathways are regulated reciprocally, as described below. Because the mitochondrial inner membrane is impermeable to acetylCoA, an indirect shuttle transfers acetyl group equivalents across the inner membrane (Fig. 20-11). Intramitochondrial acetyl-CoA first reacts with oxaloacetate to form citrate, in the citric acid cycle reaction catalyzed by citrate synthase (see Fig. 15-7). Citrate then passes into the cytosol through the mitochondrial inner membrane on the tricarboxylate transporter. In the cytosol, citrate cleavage by citrate lyase regenerates acetyl-CoA; this reaction is driven by the investment of energy from ATP. Oxaloacetate cannot return to the matrix directly; there is no transporter for it. Instead, oxaloacetate is reduced by cytosolic malate dehydrogenase to malate, which returns to the mitochondrial matrix on the malate-a-ketoglutarate transporter in exchange for citrate, and is reoxidized to oxaloacetate to complete the shuttle. Alternatively, the malate produced in the cytosol is used to generate cytosolic NADPH through the activity of malic enzyme, as described above. Plants have another means of acquiring acetyl-CoA for fatty acid synthesis. They produce acetyl-CoA from pyruvate using a stromal isozyme of pyruvate dehydrogenase (see Fig. 15-2). Fatty Acid Biosynthesis Is Tightly Regulated When a cell or organism has more than enough metabolic fuel available to meet its energetic needs, the excess is generally converted to fatty acids and stored as lipids such as triacylglycerols. The reaction catalyzed by acetyl-CoA carboxylase is the rate-limiting step in the biosynthesis of fatty acids, and this enzyme is an important site of regulation. In vertebrates, palmitoyl-CoA, the principal product of fatty acid synthesis, acts as a feedback inhibitor of the enzyme, and citrate is an allosteric activator (Fig. 20-12a). When there is an increase in the concentrations of mitochondrial acetyl-CoA and of ATP, citrate is transported out of the mitochondria and becomes both the precursor of cytosolic acetyl-CoA and an allosteric signal for the activation of acetyl-CoA carboxylase. Acetyl-CoA carboxylase is also regulated by covalent alteration. Phosphorylation triggered by the hormones glucagon and epinephrine inactivates it, thereby slowing fatty acid synthesis. In its active (dephosphorylated) form, acetyl-CoA carboxylase polymerizes into long filaments (Fig. 20-12b); phosphorylation is accompanied by dissociation into monomeric subunits and loss of activity. The acetyl-CoA carboxylase from plants and bacteria is not regulated by citrate or by a phosphorylation-dephosphorylation cycle. The plant enzyme is activated by an increase in stromal pH and Mg2+ concentration, both of which occur upon illumination of the plant (p. 630). Bacteria do not use triacylglycerols as energy stores. The primary role of fatty acid synthesis in E. coli is to provide precursors for membrane lipids, and the regulation of this process is complex, involving certain guanine nucleotides that coordinate cell growth with membrane formation. Other enzymes in the pathway of fatty acid synthesis are also regulated. The pyruvate dehydrogenase complex and citrate lyase, both of which supply acetyl-CoA, are activated by insulin (Fig. 20-12a) through a cascade of protein phosphorylation. Insulin and glucagon are released in response to blood glucose concentrations that are too high or too low, respectively. Their broad metabolic effects and molecular mechanisms will be discussed further in If fatty acid synthesis and β oxidation were to occur simultaneously, the two processes would constitute a futile cycle, wasting energy. We noted earlier (p. 496) that β oxidation is blocked by malonylCoA, which inhibits carnitine acyltransferase I. Thus during fatty acid synthesis, the production of the first intermediate, malonyl-CoA, shuts down β oxidation at the level of a transport system in the mitochondrial inner membrane. This control mechanism illustrates another advantage to a cell of segregating synthetic and degradative pathways in different cellular compartments. Long-Chain Fatty Acids Are Synthesized from Palmitate Palmitate, the principal product of the fatty acid synthase system in animal cells, is the precursor of other long-chain fatty acids (Fig. 2013). It may be lengthened to form stearate (18:0) or even longer saturated fatty acids by further additions of acetyl groups, through the action of fatty acid elongation systems present in the smooth endoplasmic reticulum and the mitochondria. The more active elongation system of the endoplasmic reticulum extends the 16-carbon chain of palmitoyl-CoA by two carbons, forming stearoyl-CoA. Although different enzyme systems are involved, and coenzyme A rather than ACP is the acyl carrier directly involved in the reaction, the mechanism of elongation is otherwise identical with that employed in palmitate synthesis: donation of two carbons by malonyl-ACP, followed by reduction, dehydration, and reduction to the saturated 18-carbon product, stearoyl-CoA.

Lipid Biosynthesis

Lipids play a variety of cellular roles, including some only recently recognized. They are the principal form of stored energy in most organisms, as well as major constituents of cell membranes. Specialized lipids serve as pigments (retinal), cofactors (vitamin K), detergents (bile salts), transporters (dolichols), hormones (vitamin D derivatives, sex hormones), extracellular and intracellular messengers (eicosanoids and derivatives of phosphatidylinositol), and anchors for membrane proteins (covalently attached fatty acids, prenyl groups, and phosphatidylinositol). The ability to synthesize a variety of lipids is therefore essential to all organisms. This chapter describes biosynthetic pathways for some of the principal lipids present in most cells, illustrating the strategies employed in assembling these water-insoluble products from simple, water-soluble precursors such as acetate. Like other biosynthetic pathways, these reaction sequences are endergonic and reductive. They use ATP as a source of metabolic energy and a reduced electron carrier (usually NADPH) as a reductant.
We first describe the biosynthesis of fatty acids, the major components of both triacylglycerols and phospholipids. Then we examine the assembly of fatty acids into triacylglycerols and into the simpler types of membrane phospholipids. Finally, we consider the synthesis of cholesterol, a component of some membranes and the precursor of such steroid products as the bile acids, sex hormones, and adrenal cortical hormones.

Biosynthesis of Fatty Acids and Eicosanoids

When fatty acid oxidation was found to occur by oxidative removal of successive two-carbon (acetyl-CoA) units (see Fig. 16-8), biochemists thought that the biosynthesis of fatty acids might proceed by simple reversal of the same enzymatic steps used in their oxidation. However, fatty acid biosynthesis and breakdown occur by different pathways, are catalyzed by different sets of enzymes, and take place in different parts of the cell. Moreover, a three-carbon intermediate, malonylCoA, participates in the biosynthesis of fatty acids but not in their breakdown.
We focus first on the pathway of fatty acid synthesis, then turn our attention to regulation of the pathway and to the biosynthesis of longchain fatty acids, unsaturated fatty acids, and their eicosanoid derivatives .

Malonyl-CoA Is Formed from Acetyl-CoA and Bicarbonate

The irreversible formation of malonyl-CoA from acetyl-CoA is catalyzed by acetyl-CoA carboxylase (Fig. 20-1). Acetyl-CoA carboxylase contains biotin as its prosthetic group, covalently bound in amide linkage to the ε-amino group of a Lys residue on one of the three subunits of the enzyme molecule. The two-step reaction is very similar to other biotin-dependent carboxylation reactions, such as those catalyzed by pyruvate carboxylase (see Fig. 15-13) and propionyl-CoA carboxylase (see Fig. 16-12). The carboxyl group, derived from bicarbonate (HCO₃ ), is first transferred to biotin in an ATP-dependent reaction. The biotinyl group serves as a temporary carrier of C0₂, transferring it to acetyl-CoA in the second step to yield malonyl-CoA.
Acetyl-CoA carboxylase from bacteria has three separate polypeptide subunits, as shown in Figure 20-1. In higher plants and animals, all three activities are part of a single multifunctional polypeptide.

The Biosynthesis of Fatty Acids Proceeds by a Distinctive Pathway

The fundamental reaction sequence by which the long chains of carbon atoms in fatty acids are assembled consists of four steps (Fig. 20-2). The saturated acyl group produced during this set of reactions is recycled to become the substrate in another condensation with an activated malonyl group. With each passage through the cycle, the fatty acyl chain is extended by two carbons. When the chain length reaches 16, the product (palmitate, 16:0; see Table 9-1) leaves the cycle. The methyl and carboxyl carbon atoms of the acetyl group become C-16 and C-15, respectively, of the palmitate (Fig. 20-3); the rest of the carbon atoms are derived from malonyl-CoA.
Both the electron carrier cofactor and the activating groups in the reductive anabolic sequence are different from those that act in the oxidative catabolic process. Recall that in β oxidation, NAD+ and FAD serve as electron acceptors, and the activating group is the thiol (-SH) group of coenzyme A (see Fig. 16-8). By contrast, the reducing agent in the synthetic sequence is NADPH, and the activating groups are two different enzyme-bound -SH groups, to be described below.
All of the reactions in the synthetic process are catalyzed by a multienzyme complex, the fatty acid synthase. The detailed structure of this multienzyme complex and its location in the cell differ from one species to another, but the reaction sequence is identical in all organisms.

The Fatty Acid Synthase Complex Has Seven Different Active Sites

The fatty acid synthase system from E. coli consists of seven separate polypeptides that are tightly associated in a single, organized complex (Table 20-1). The proteins act together to catalyze the formation of fatty acids from acetyl-CoA and malonyl-CoA. Throughout the process, the intermediates remain covalently attached to one of two thiol groups of the complex. One point of attachment is the -SH group of a Cys residue in one of the seven proteins (β-ketoacyl-ACP synthase, described below); the other is the -SH group of acyl carrier protein, with which the acyl intermediates of fatty acid synthesis form a thioester.
Acyl carrier protein (ACP) of E. coli (Table 20-1) is a small protein (M,. 8,860) containing the prosthetic group 4'-phosphopantetheine (Fig. 20-4), an intermediate in the synthesis of coenzyme A (see Fig. 12-41). The thioester that links ACP to the fatty acyl group has a high free energy of hydrolysis, and the energy released when this bond is broken helps to make the first reaction in fatty acid synthesis, condensation, thermodynamically favorable. The 4'-phosphopantetheine prosthetic group of ACP is believed to serve as a flexible arm, tethering the growing fatty acyl chain to the surface of the fatty acid synthase complex and carrying the reaction intermediates from one enzyme active site to the next.
Although the details of enzyme structure differ in prokaryotes such as E. coli and in eukaryotes, the four-step process of fatty acid synthesis is the same in all organisms. We first describe the process as it occurs in E. coli, then consider the differences in enzyme structure among other organisms.

Fatty Acid Synthase Receives the Acetyl and Malonyl Groups Before the condensation

reactions that build up the fatty acid chain can begin, the two thiol groups on the enzyme complex must be charged with the correct acyl groups (Fig. 20-5). First, the acetyl group of acetyl-CoA is transferred to the Cys -SH group of the β-ketoacylACP synthase. This reaction is catalyzed by acetyl-CoA-ACP transacetylase. The second reaction, transfer of the malonyl group from malonyl-CoA to the -SH group of ACP, is catalyzed by malonyl-CoA-ACP transferase, also part of the complex. In the charged synthase complex, the acetyl and malonyl groups are very close to each other and are activated for the chain-lengthening process, which consists of the four steps outlined earlier. These steps are now considered in some detail.
1. Condensation The first step in the formation of a fatty acid chain is condensation of the activated acetyl and malonyl groups to form an acetoacetyl group bound to ACP through the phosphopantetheine -SH group: acetoacetyl-ACP; simultaneously, a molecule of COZ is produced (Fig. 20-5). In this reaction, catalyzed by β-ketoacyl-ACP synthase, the acetyl group is transferred from the Cys -SH group of this enzyme to the malonyl group on the -SH of ACP, becoming the methyl-terminal two-carbon unit of the new acetoacetyl group.
The carbon atom in the CO₂ formed in this reaction is the same carbon atom that was originally introduced into malonyl-CoA from HCO3 by the acetyl-CoA carboxylase reaction (Fig. 20-1). Thus CO₂ is only transiently in covalent linkage during fatty acid biosynthesis; it is removed as each two-carbon unit is inserted.
Why do cells go to the trouble of adding CO₂ to make a malonyl group from an acetyl group, only to lose CO₂ again during the formation of acetoacetate? Remember that in the β oxidation of fatty acids, cleavage of the bond between two acyl groups (the cleavage of an acetyl unit from the acyl chain) is highly exergonic. Therefore, the simple condensation of two acyl groups (of two acetyl-CoA molecules, for example) is endergonic. The condensation reaction is made thermodynamically favorable by the involvement of activated malonyl, rather than acetyl, groups. The methylene carbon (C-2) of the malonyl group, sandwiched between carbonyl and carboxyl carbons, is an especially good nucleophile. In the condensation step (Fig. 20-5), decarboxylation of the malonyl group facilitates the nucleophilic attack of this methylene carbon on the thioester linking the acetyl group to β-ketoacyl-ACP synthase, displacing the enzyme's -SH group. Coupling the condensation to the decarboxylation of the malonyl group renders the overall process highly exergonic. Recall that a similar carboxylationdecarboxylation sequence facilitates the formation of phosphoenolpyruvate from pyruvate in gluconeogenesis (see Fig. 19-3).
By using activated malonyl groups in the synthesis of fatty acids and activated acetate in their degradation, the cell manages to make both processes favorable, although one is effectively the reversal of the other. The extra energy required to make fatty acid synthesis favorable is provided by the ATP used to synthesize malonyl-CoA from acetylCoA and HCO₃^-- (Fig. 20-1).
2. Reduction of the Carbonyl Group The acetoacetyl-ACP formed in the condensation step next undergoes reduction of the carbonyl group at C-3 to form D-β-hydroxybutyryl-ACP (Fig. 20-5). This reaction is catalyzed by /3-ketoacyl-ACP/ reductase, and the electron donor is NADPH. Notice that the D-β-hydroxybutyryl group does not have the same stereoisomeric form as the L-β-hydroxyacyl intermediate in fatty acid oxidation (see Fig. 16-8).

3. Dehydration In the third step, the elements of water are removed from C-2 and C-3 of D-β-hydroxybutyryl-ACP to yield a double bond in the product, trans-Δ2-butenoyl-ACP (Fig. 20-5). The enzyme that catalyzes this dehydration is /i-hydroxyacyl-ACP dehydratase.
4.Reduction of the Double Bond Finally, the double bond of trans-Δ²-butenoyl-ACP is reduced (saturated) to form butyryl-ACP by the action of enoyl-ACP reductase (Fig. 20-5); again, NADPH is the electron donor.

Regulation of Carbohydrate Metabolism in Plants

Carbohydrate metabolism in plant cells is more complex than in animal cells or nonphotosynthetic microorganisms. In addition to the universal pathways of glycolysis and gluconeogenesis, plants have the unique reaction sequences for CO₂ reduction to triose phosphates and the associated reductive pentose phosphate pathway-all of which must be coordinately regulated to avoid wasteful futile cycling and to ensure proper allocation of carbon to energy production and synthesis of starch and sucrose. One level of coordination is achieved by light activation of certain enzymes associated with the carbon-fixing reactions of photosynthesis. Some enzymes are activated by changes in pH that result from light-induced proton movements; some enzymes are activated by reduction of disulfide bonds involved in their catalytic activity, using electrons flowing from photosystem I; other enzymes are subject to more conventional allosteric regulation by one or more metabolic intermediates. The compartmentation of metabolic sequences in organelles also contributes to the regulation of carbohydrate metabolism.
We will end with a discussion of another reaction catalyzed by rubisco, the condensation of O₂ with ribulose-1,5-bisphosphate. This is the starting point for a process called photorespiration, which greatly affects the efficiency of carbon fixation in plants.

Rubisco Is Subject to Both Positive and Negative Regulation

As the site where photosynthetic CO₂ fixation is initiated, rubisco is a prime target for regulation. One type of regulation involves the carbamylation of a Lys residue (Fig. 19-30a, b). At high CO₂ levels this occurs nonenzymatically. However, the substrate for this enzyme, ribulose-1,5-bisphosphate, inhibits carbamylation, and this effect is almost complete at physiological CO₂ concentrations. An enzyme called rubisco activase overcomes this inhibition and promotes an ATPdependent activation of rubisco that results in carbamylation. The mechanism of activation is unknown, and it is not clear why ATP is required.
Rubisco is also inhibited by 2-carboxyarabinitol-1-phosphate, a naturally occurring transition-state analog (see Box 8-3) with a structure similar to that of the β-keto acid intermediate of the rubisco reaction (Figs. 19-20, 19-30c). This compound is synthesized in the dark by some plants to depress rubisco activity, and it is sometimes called the "nocturnal inhibitor." It is broken down when light returns, permitting a reactivation of rubisco.

Certain Enzymes of the Calvin Cycle Are Indirectly Activated by Light

The reductive fixation of CO₂ requires ATP and NADPH, and their stromal concentrations increase when chloroplasts are illuminated (Fig. 19-31). The light-induced transport of protons across the thylakoid membrane (Chapter 18) also makes the stromal compartment alkaline and is accompanied by a flow of Mg²⁺ out of the thylakoid compartment into the stroma. Several stromal enzymes have evolved to take advantage of these light-dependent conditions that signal the availability of ATP and NADPH; they have pH or Mg²⁺ optima that are better suited to alkaline conditions and high [Mg²⁺]. Activation of rubisco by formation of the lysyl carbamate is faster at alkaline pH, and high stromal [Mg²⁺] favors formation of the active Mg²⁺ complex. Fructose-1,6-bisphosphatase requires Mg²⁺ and is very dependent upon pH (Fig. 19-32). Its activity increases by a factor of more than 100 when the pH and [Mg²⁺ ]rise during chloroplast illumination.

Three other enzymes essential to the Calvin cycle's operation are subject to another type of regulation by light. Ribulose-5-phosphate kinase, fructose-1,6-bisphosphatase, and sedoheptulose-1,7-bisphosphatase can exist in either of two forms, differing in the oxidation state of Cys residues essential to their catalytic activity. When these Cys residues are oxidized as disulfide bonds, the enzymes are inactive; this is the normal situation in the dark. With illumination, electrons flow from photosystem I to ferredoxin (see Fig. 18-44), which passes electrons to a small, soluble, disulfide-containing protein called thioredoxin (Fig. 19-33). Thioredoxin donates electrons for the reduction of the disulfide bridges of these light-activated enzymes and is then reactivated in a disulfide-exchange reaction catalyzed by thioredoxin reductase.

Gluconeogenesis and Glycolysis Are Reciprocally Regulated in Plants

The possibility of futile cycling by the simultaneous operation of glycolysis and gluconeogenesis exists in plants as in animals. Plant cells (the chloroplast stroma and cytosol) have all the enzymes of glycolysis, and during dark periods glycolytic breakdown of starch is a major source of energy. During periods of illumination, photosynthetic plant cells produce the triose phosphate intermediates common to glycolysis and gluconeogenesis, and convert these into hexoses, sucrose, and starch as we have just seen. Futile cycling through these two paths is prevented in plants by the regulation of key enzymes of each cytosolic pathway by fructose-2,6-bisphosphate, whose concentration reflects the level of photosynthetic activity.

The concentration of fructose-2,6-bisphosphate varies inversely with the rate of photosynthesis in higher plants (Fig. 19-34). The enzyme phosphofructokinase-2, responsible for fructose-2,6-bisphosphate synthesis, is inhibited by dihydroxyacetone phosphate or 3-phosphoglycerate and stimulated by P_i. During active photosynthesis, dihydroxyacetone phosphate is produced and P_i is consumed, resulting in inhibition of PFK-2 and lowered concentrations of fructose2,6-bisphosphate.

As in animals (Figs. 19-7, 19-8), fructose-2,6-bisphosphate in plants is an activator of glycolysis and an inhibitor of gluconeogenesis (Fig. 19-34). It slows gluconeogenesis by inhibiting the cytosolic fructose-1,6-bisphosphatase, which catalyzes a rate-limiting step in the synthesis of fructose-6-phosphate. It stimulates glycolysis by activating the PP; dependent form of phosphofructokinase-1 (p. 435); this enzyme, a critical control point in glycolysis, is virtually inactive in the absence of fructose-2,6-bisphosphate.
Thus, when photosynthesis ceases, the concentration of fructose2,6-bisphosphate rises, stimulating glycolysis and inhibiting gluconeogenesis; glycolysis (combined with mitochondrial oxidative phosphorylation) provides energy in the dark. In the light, the fructose-2,6bisphosphate concentration drops, glycolysis is inhibited, gluconeogenesis is turned on, and the energy and precursors produced by photosynthesis are used to make hexoses to be transported as sucrose or stored as starch.

Sucrose and Starch Synthesis Are Coordinately Regulated

In most plant cells, triose phosphates produced during CO₂ fixation are converted primarily to sucrose and starch. The balance between these processes is tightly regulated, and both processes must be coordinated with the rate of carbon fixation. Communication and coordination between sucrose synthesis in the cytosol and carbon fixation and starch synthesis in the chloroplast are mediated by the P_i-triose phosphate antiport system. If sucrose synthesis is too rapid, the transport of the excess P_i into the chloroplast will result in the removal of too much triose phosphate; this will have a deleterious effect on the rate of carbon fixation because five of every six triose phosphate molecules produced in the Calvin cycle are needed to regenerate ribulose-1,5bisphosphate and complete the cycle. If sucrose synthesis is too slow, insufficient P_i will be made available to the chloroplast for triose phosphate synthesis.

Sucrose synthesis is regulated primarily at three steps, those catalyzed by fructose-1,6-bisphosphatase, sucrose-6-phosphate synthase, and sucrose phosphate phosphatase. When light first strikes a leaf in the morning, triose phosphate levels in the cytosol (derived from carbon fixation in the chloroplast) rise. The triose phosphates inhibit the activity of phosphofructokinase-2, and the levels of fructose-2,6bisphosphate fall. This releases the inhibition of fructose-1,6bisphosphatase, permitting the synthesis of fructose-6-phosphate and thus other hexose phosphates, including glucose-6-phosphate. Glucose-6-phosphate is an allosteric activator of sucrose-6-phosphate synthase. Sucrose phosphate phosphatase catalyzes the final step in sucrose synthesis as its substrate becomes available. The key regulatory enzyme in starch synthesis is ADP-glucose pyrophosphorylase (Fig. 19-16), which is indirectly activated as cytosolic sucrose levels rise. Sucrose-6-phosphate synthase and sucrose phosphate phosphatase are both inhibited to some degree by sucrose. This inhibition becomes effective as sucrose levels increase, leading to an increase in the cytosolic pool of hexose phosphates that effectively sequesters much of the available phosphate in a form that cannot be transported back to the chloroplast. The levels of free Pi in the cytosol, and the resulting decrease in the activity of the Pi-triose phosphate antiporter, lead to a decrease in Pi and an increase in three-carbon compounds in the chloroplast. ADP-glucose pyrophosphorylase is inhibited by Pi and stimulated by 3-phosphoglycerate, so that one effect of the increase in cytosolic sucrose concentration is an increase in starch synthesis in the chloroplast. In the steady state, all of these activities are balanced so that sucrose and starch synthesis each consume about 50?0 of the triose phosphate produced during carbon fixation, and the rates of both processes are regulated so that precursors and intermediates required for carbon fixation are maintained at optimal levels.

Condensation of O₂ with Ribulose-1,5-Bisphosphate Initiates Photorespiration

Rubisco is not absolutely specific for CO₂ as a substrate; O₂ competes with CO₂ at the active site, and rubisco catalyzes the condensation of O ₂ with ribulose-1,5-bisphosphate to form one molecule of 3-phosphoglycerate and one of phosphoglycolate (Fig. 19-35). This is the enzyme's oxygenase activity, evident in its name: RuBP carboxylase/oxygenase. The metabolic function of this reaction is not clear. It results in no fixation of carbon and appears to be a net liability to the cell in which it occurs; salvaging the carbons from phosphoglycolate, which is not a useful metabolite, uses cellular energy. The condensation of 02 1with ribulose-1,5-bisphosphate occurs concurrently with C02 fixation, with the latter predominating by a factor of about three.
The salvage pathway (Fig. 19-36) involves the conversion of two molecules of phosphoglycolate into a molecule of serine (which has three carbons) and a molecule of C02. The oxygenase activity of rubisco combined with the salvage pathway consumes 02 and produces CO₂, a process called photorespiration. Unlike mitochondrial respiration, however, this process does not conserve energy

Apparently the evolution of rubisco produced an active site not able to discriminate well between CO2 and O2, perhaps because much evolution occurred before O2 was an important component of the atmosphere. The Km for CO2 is about 20μM, and for O2 is about 200 μM. (A solution that is in equilibrium with air at room temperature contains about 10 μM CO2 and 250 μM O2.) The modern atmosphere contains about 20% O2 and only 0.04% CO2, proportions that allow O2 "fixation" by rubisco to constitute a significant waste of energy. During photosynthesis CO2 is consumed in the fixation reactions, altering the CO2 to O2 ratio in the air spaces around the leaf in favor of O2. In addition, the affinity of rubisco for CO2 decreases with increasing temperature, exacerbating the tendency of the enzyme to catalyze the wasteful oxygenase reaction. Photorespiration may inhibit net biomass formation as much as 50%. As we shall see, this has led to adaptations in the process by which carbon fixation takes place, particularly in plants that live in warm climates. Some Plants Have a Mechanism to Prevent Photorespiration Most plants in the tropics, as well as temperate-zone crop plants native to the tropics, such as corn, sugar cane, and sorghum, have evolved a mechanism to circumvent the problem of wasteful photorespiration. The ultimate step of CO2 fixation into a three-carbon product, 3-phosphoglycerate, is preceded by several steps, one of which is a preliminary fixation of CO2 into a compound with four carbon atoms. These plants are referred to as C4 plants. Plants in which the first step in carbon fixation is reaction of CO2 with ribulose-1,5-bisphosphate to form 3-phosphoglycerate-as we have described thus far-are called C3 plants.

C₄ plants, which typically grow in high light intensity and temperatures, have several important characteristics: high photosynthetic rates, high growth rates, low photorespiration rates, low rates of water loss, and an unusual leaf structure. Photosynthesis in the leaves of C₄ plants involves two cell types: mesophyll and bundle-sheath cells (Fig. 19-37). There are three known patterns of C₄ metabolism. The best-understood pathway, worked out in the 1960s by two plant biochemists, Marshall Hatch and Rodger Slack, is described here.
In plants of tropical origin the first intermediate in which 14CO₂ is fixed is not 3-phosphoglycerate but oxaloacetate, a four-carbon compound. This reaction, which occurs in leaf mesophyll cells (Fig. 19-37), is catalyzed by phosphoenolpyruvate carboxylase:

Phosphoenolpyruvate + HCO₃^-- ==== oxaloacetate + P_i

This enzyme does not occur in animal tissues and is not to be confused with phosphoenolpyruvate carboxykinase (p. 603), which catalyzes the gluconeogenic reaction

Oxaloacetate + GTP ===== phosphoenolpyruvate + CO₂ + GDP

The oxaloacetate formed in the mesophyll cells is reduced to malate at the expense of NADPH:

Oxaloacetate + NADPH + H⁺; L-malate +NADP⁺

Alternatively, oxaloacetate may be converted to aspartate by transamination: Oxaloacetate + α-amino acid ===== L-aspartate + α-keto acid The malate or aspartate formed in the mesophyll cells, which contains the fixed CO2, is transferred into the neighboring bundle-sheath cells via specialjunctions (plasmodesmata; p. 49) between the cells (Fig. 1937). In the bundle-sheath cells the malate is oxidized and decarboxylated to yield pyruvate and CO2 by the action of malic enzyme: L-Malate + NADP+ ===== pyruvate + CO2 + NADPH + H+

In plants that employ aspartate as the carrier of CO₂, aspartate is first transaminated to form oxaloacetate then reduced to malate in bundlesheath cells before the release of CO₂₂ by malic enzyme. The free CO₂ formed in the bundle-sheath cells is the same CO₂ molecule that was originally fixed into oxaloacetate in the mesophyll cells.
In the bundle-sheath cells the CO₂arising from the decarboxylation of malate is fixed again (Fig. 19-37), this time by rubisco, in exactly the same reaction that occurs in C₃ plants, leading to incorporation of CO₂ into the C-1 of 3-phosphoglycerate. The pyruvate formed by the decarboxylation of malate in the bundle-sheath cells is transferred back to the mesophyll cells, where it is converted into PEP by an unusual enzymatic reaction, catalyzed by the enzyme pyruvate phosphate dikinase:

Pyruvate + ATP + P_i======= phosphoenolpyruvate + AMP + PP_i

This enzyme is called a dikinase because two different molecules are simultaneously phosphorylated by one molecule of ATP: pyruvate is phosphorylated to PEP, and phosphate is phosphorylated to pyrophosphate. The pyrophosphate is subsequently hydrolyzed to phosphate, so two high-energy phosphate groups of ATP are used in regenerating PEP. The PEP is now ready to fix another molecule of CO₂ in the mesophyll cell.
The PEP carboxylase of mesophyll cells has a high affinity for HCO₃^-- and can fix CO₂ more efficiently than can rubisco. Unlike rubisco, PEP carboxylase does not use O₂ as an alternative substrate, so there is no competition between CO₂ and O₂ with this enzyme. This reaction serves to fix and concentrate CO₂ in the form of malate. Release of CO₂ from malate in the bundle-sheath cells yields a sufficiently high local concentration of CO₂ for rubisco to function near its maximal rate, with relatively little of the competitive side reaction with O₂.
Once CO₂ is fixed into 3-phosphoglycerate in the bundle-sheath cells, all the other reactions of the C₃ or Calvin cycle take place exactly as described earlier (Figs. 19-20, 19-22, 19-23). Thus in C₄ plants the mesophyll cells carry out CO₂ fixation by the C₄ pathway, but starch and sucrose biosynthesis occur by the C₃ pathway in the bundle-sheath cells.
The pathway of CO₂ fixation in C₄ plants has a greater energy cost than in C₃ plants. For each molecule of CO₂ fixed in the C4 pathway, a molecule of PEP must be regenerated at the expense of two highenergy phosphate groups of ATP. Thus C₄ plants need a total of five ATPs to fix one molecule of CO₂, whereas C₃ plants need only three. As the temperature increases (and the affinity of rubisco for CO₂ decreases, as noted above), a point is reached at about 28° to 30°? where the gain in efficiency from the elimination of photorespiration in C₄ plants more than compensates for this energetic cost. C₄ plants (crabgrass, for example) outgrow most C₃ plants during the summer, as any experienced gardener can attest!

Each Triose Phosphate Synthesized from CO2 Costs Six NADPH and Nine ATP

The net result of the Calvin cycle is the conversion of three molecules of CO₂ and one molecule of phosphate into a molecule of triose phosphate. The stoichiometry of the overall path from CO₂ to triose phosphate, with the regeneration of ribulose-1,5-bisphosphate, is shown in Figure 19-27. Three molecules of ribulose-1,5-bisphosphate (a total of 15 carbons) condense with three CO₂ (three carbons) to form six molecules of 3-phosphoglycerate (18 carbons). These six molecules of 3-phosphoglycerate are reduced to six molecules of glyceraldehyde-3phosphate, with the expenditure of six ATP (in the synthesis of 1,3-bisphosphoglycerate) and six NADPH (in the reduction of 1,3bisphosphoglycerate to glyceraldehyde-3-phosphate). One of these molecules of glyceraldehyde-3-phosphate is the net product of the process. The other five glyceraldehyde-3-phosphate molecules (15 carbons) are rearranged in steps l to 8 of Figure 19-23 to form three molecules of ribulose-1,5-bisphosphate (15 carbons). The last step in this conversion requires one ATP per ribulose-1,5-bisphosphate, or a total of three ATP. Thus, for every molecule of triose phosphate produced by photosynthetic CO₂ fixation, six NADPH and nine ATP are required.
The source of ATP and NADPH for these reactions is the lightdriven reactions of photophosphorylation (Chapter 18). Of the nine ATP molecules converted to ADP and phosphate in the generation of a molecule of triose phosphate, eight of the phosphates are released as P; and combined with eight ADP to regenerate ATP. The ninth phosphate is incorporated into the triose phosphate itself. To convert the ninth ADP to ATP, a molecule of P; must be imported from the cytosol, as we will see later.
In the dark, the production of ATP and NADPH by photophosphorylation ceases, and the incorporation of CO₂ into triose phosphate (by the so-called "dark reactions") also stops. The "dark reactions" of photosynthesis are so named to distinguish them from the primary lightdriven reactions of electron transfer to NADP+ and synthesis of ATP, described in Chapter 18. They do not, in fact, occur at significant rates in the dark in photosynthetic organisms (although they do in chemotrophic organisms). Therefore, these reactions are more appropriately called the carbon fixation reactions of photosynthesis.
Fixed carbon generated in the chloroplast is also stored there in significant amounts. Within the chloroplast stroma are all the enzymes necessary to convert the triose phosphates produced by CO₂ fixation (glyceraldehyde-3-phosphate and dihydroxyacetone phosphate) into starch, which is stored in the chloroplast as insoluble granules. Aldolase condenses the trioses to fructose-1,6-bisphosphate, fructose-1,6-bisphosphatase produces fructose-6-phosphate, phosphohexose isomerase yields glucose-6-phosphate, and phosphoglucomutase produces glucose-1-phosphate, the starting material for starch synthesis.
All the reactions of the Calvin cycle except the first, catalyzed by rubisco, and the last, catalyzed by ribulose-5-phosphate kinase, also take place in animal tissues. For lack of these two enzymes animals cannot carry out net conversion of CO₂ into glucose.

A Transport System Exports Triose Phosphates and Imports Phosphate

The inner chloroplast membrane is impermeable to most phosphorylated compounds, including fructose-6-phosphate, glucose-6-phosphate, and fructose-1,6-bisphosphate. There is, however, a specific transporter (antiporter) that catalyzes the one-for-one exchange of P_i; with triose phosphates, either dihydroxyacetone phosphate or 3-phosphoglycerate (Fig. 19-28; see also Fig. 19-22). This antiporter simultaneously moves triose phosphate out of the chloroplast to the cytosol and P_i into the chloroplast, where it is used in photophosphorylation.
Without this antiport system, CO₂ fixation in the chloroplast would quickly come to a halt. The net transport of triose phosphates out of the chloroplast serves the important function of removing the triose phosphate products of carbon fixation. In the cytosol, the triose phosphates are converted to sucrose by the pathways illustrated in Figures 19-22 and 19-17. Sucrose synthesis in the cytosol and starch synthesis in the chloroplast are the major pathways by which the excess triose phosphates are "harvested." The last step of sucrose synthesis yields one molecule of free P_i (Fig. 19-17). This is transported back into the chloroplast and used in the synthesis of ATP, effectively replacing the molecule of P_i that is used to generate triose phosphate, as described above. For every molecule of triose phosphate removed from the chloroplast, one P_i is transported into the chloroplast. If this exchange were blocked, triose phosphate synthesis would quickly deplete the available P_i in the chloroplast and prevent further CO₂ fixation.
This Pi-triose phosphate antiport system serves one additional function. ATP and reducing power are needed in the cytosol for a variety of synthetic and energy-requiring reactions. These requirements are met to an as yet undetermined degree by the mitochondria. A second potential source of energy is the ATP and NADPH generated in the stroma during the light reactions of photosynthesis; however, ATP and NADPH do not cross the chloroplast membrane. The antiport system has the indirect effect of moving ATP and reducing equivalents across the chloroplast membrane (Fig. 19-29). Dihydroxyacetone phosphate formed in the stroma by C0₂ fixation is transported to the cytosol, where it is converted by glycolytic enzymes to 3-phosphoglycerate, generating ATP and NADH. 3-Phosphoglycerate reenters the chloroplast, completing the cycle. The net effect is transport of NADPH/NADH and ATP from the chloroplast to the cytosol.

Photosynthetic Carbohydrate Synthesis

The synthesis of carbohydrates in animal cells always employs precursors having at least three carbons, all at an oxidation state lower than that of carbon in CO₂. Photosynthetic organisms, by contrast, can make carbohydrates from CO₂ and water. They synthesize glucose, sucrose, and other carbohydrates by reducing CO₂ at the expense of energy furnished by the ATP and NADPH generated in photosynthetic electron transfer. This process represents a fundamental difference between autotrophic (phototrophic or chemotrophic) and heterotrophic organisms. Autotrophs can use CO₂ as the sole source of all the carbon atoms required for biosynthesis not only of cellulose and starch, but also of the lipids and proteins and all of the many other organic components of plant cells. By contrast, heterotrophic organisms in general are unable to bring about the net reduction of CO₂ to form "new" glucose in any significant amounts. Carbon dioxide can be taken up by animal tissues, as in the pyruvate carboxylase reaction during gluconeogenesis, but the CO₂ molecule incorporated into oxaloacetate is lost in a subsequent reaction step (Fig. 19-3). Similarly, the CO₂ taken up by acetyl-CoA carboxylase during fatty acid synthesis in animal tissues (Chapter 20) or by carbamoyl phosphate synthetase I during urea formation (Chapter 17) is lost in later steps.
Green plants contain in their chloroplasts unique enzymatic machinery to catalyze the conversion of CO₂ into simple (reduced) organic compounds, a process called CO₂ fixation, or carbon fixation. Plants convert these simple products of photosynthesis into more complex biomolecules, including sugars, polysaccharides, and metabolites derived from them, using metabolic pathways similar to those in animals. The reactions that result in CO₂ fixation make up a cyclic pathway in which key intermediates are constantly regenerated. The pathway was elucidated in the early 1950s by Melvin Calvin and coworkers, and is often called the Calvin cycle.

Carbon Dioxide Fixation Occurs in Three Stages

The first stage in the fixation of CO₂ into organic linkage (Fig. 19-19) is its condensation with a five-carbon acceptor, ribulose-1,5-bisphosphate, to form two molecules of 3-phosphoglycerate. (Note that Figure 19-19 shows the number of molecules reacting to give net formation of one molecule of triose-this takes three molecules of CO₂.) In the second stage the 3-phosphoglycerate is reduced to glyceraldehyde-3-phosphate: three molecules of CO₂ are fixed to three molecules of ribulose-1,5-bisphosphate to form six molecules of glyceraldehyde-3phosphate (18 carbons). One molecule of this triose phosphate (three carbons) can either be used for energy production via glycolysis and the citric acid cycle, or condensed to hexose phosphates to be used in the synthesis of starch or sucrose. In the third stage, five of the six molecules of glyceraldehyde-3-phosphate (15 carbons) are used to regenerate three molecules of ribulose-1,5-bisphosphate, the starting material
Thus the overall process is cyclical and allows the continuous conversion of CO₂ into triose and hexose phosphates. Fructose-6-phosphate is a key intermediate in stage 3. The pathway from hexose phosphate to pentose bisphosphate involves the same reactions used in animal cells for the conversion of pentose phosphates to hexose phosphates during the operation of the pentose phosphate pathway, an alternative route for glucose oxidation (Fig. 14-22). In photosynthetic fixation of CO₂, this pathway operates in the opposite direction, converting hexose phosphates into pentose phosphates.
Stage 1: Fixation of CO₂ into 3-Phosphoglycerate An important clue to the nature of the CO₂ fixation mechanisms in photosynthetic organisms first came in the late 1940s when Calvin and his associates illuminated a suspension of green algae in the presence of radioactive carbon dioxide (¹⁴CO₂) for only a few seconds. They quickly killed the cells, extracted their contents, and with the help of chromatographic methods searched for the metabolites in which the labeled carbon first appeared. The first compound that became labeled was 3-phosphoglycerate, with the ¹⁴C predominantly located in the carboxyl carbon atom. This atom does not become labeled rapidly in animal tissues in the presence of radioactive CO₂. These experiments strongly suggested that 3-phosphoglycerate is an early intermediate in photosynthesis. The enzyme in green leaf extracts that catalyzes incorporation of CO₂ into organic form is ribulose-1,5-bisphosphate carboxylase or RuBP carboxylase/oxygenase (often called rubisco for short). (The enzyme's oxygenase activity is discussed later in this chapter.) As a carboxylase, rubisco catalyzes the covalent attachment of CO₂ to the five-carbon sugar ribulose-1,5-bisphosphate and the cleavage of the unstable six-carbon intermediate to form two molecules of 3-phosphoglycerate, one of which bears the new carbon introduced as CO₂ in its carboxyl group (Fig. 19-20). The rubisco of plants (Mr 550,000) has a complex structure (Fig. 19-21a, b). There are eight large subunits (each of Mr 56,000), each containing an active site, and eight small subunits (each of Mr 14,000), whose function is not well understood. The subunit structure of rubisco of photosynthetic bacteria is quite different, with two subunits that resemble the large subunits of the plant enzyme in many respects (Fig. 19-21c). The plant enzyme is located in the chloroplast stroma, where it makes up about 50% of the total chloroplast protein. Rubisco, which does not occur in animals, is the most abundant enzyme in the biosphere; it is the key enzyme in the production of biomass from CO₂.
Rubisco is subject to regulation by the covalent addition of a molecule of CO₂ to the e-amino group of a certain Lys residue to form a carbamate (Fig. 19-21). This carbamate binds to a Mg²⁺ ion; the resulting complex is essential for activity, as we shall see later (see Fig. 19-30) and may function directly in catalysis. Note that this is not the same molecule of CO₂ that is attached to ribulose-1,5-bisphosphate in the reaction catalyzed by this enzyme.
Stage 2: Conuersion o f 3-Phosphoglycerate to Glyceraldehyde-3-Phosphate 3-Phosphoglycerate is converted into glyceraldehyde-3-phosphate in two steps that are essentially the reversal of the corresponding steps in glycolysis, with one exception: the nucleotide cofactor for the reduction of 1,3-bisphosphoglycerate is NADPH, not NADH (Fig. 19-22, p. 622). The chloroplast stroma contains the full complement ofglycolytic enzymes. These stromal enzymes are isozymes (see Box 14-3) of those found in the cytosol; both sets of enzymes catalyze the same reactions, but they are products of different genes.
In the first step of the sequence, 3-phosphoglycerate kinase in the stroma catalyzes the transfer of phosphate from ATP to 3-phosphoglycerate, yielding 1,3-bisphosphoglycerate (Fig. 19-22). Then NADPH donates electrons in a reduction catalyzed by glyceraldehyde-3phosphate dehydrogenase, producing glyceraldehyde-3-phosphate. In addition to its role as an intermediate in CO₂ fixation, glyceraldehyde-3-phosphate has several possible fates in the plant cell. It may be oxidized via glycolysis for energy production or used for hexose synthesis
Stage 3: Regeneration of Ribulose-1,5-Bisphosphate from Triose Phosphates As we have seen, the first reaction in the fixation of CO₂ into triose phosphates consumes ribulose- 1,5- bisphosphate. For continuous flow of CO₂ into carbohydrate, ribulose-1,5 - bisphosphate must be constantly regenerated. Plant cells solve this problem with a series of reactions that, together with stages 1 and 2 discussed above, form a cyclic pathway (Fig. 19-23). By this pathway, the product of the first reaction (3-phosphoglycerate) passes through a series of transformations that eventually lead to the regeneration of the starting material, ribulose-1,5-bisphosphate.
The regeneration of ribulose-1,5-bisphosphate involves rearrangements of the carbon skeletons of glyceraldehyde-3-phosphate and dihydroxyacetone phosphate produced in the first two stages of carbon fixation. The intermediates in the pathway include three-, four-, five-, six-, and seven-carbon sugars. In the following discussion, all step numbers refer to Figure 19-23.
Step l is catalyzed by the enzyme transketolase, which contains thiamine pyrophosphate (TPP) as its prosthetic group (see Fig. 14-9a) and requires Mg²⁺. Transketolase catalyzes the reversible transfer of a ketol (CH₂OH-CO-) group from a ketose phosphate donor, fructose6-phosphate, to an aldose phosphate acceptor, glyceraldehyde-3-phosphate (Fig. 19-24b). The products are xylulose-5-phosphate (a pentose) and the four-carbon sugar erythrose-4-phosphate.
Step 2 is promoted by an aldolase similar to that which acts in glycolysis; it catalyzes the reversible condensation of an aldehyde, erythrose-4-phosphate, with dihydroxyacetone phosphate, yielding the seven-carbon sedoheptulose-1,7-bisphosphate. After removal of the C-1 phosphate group by sedoheptulose-1,7-bisphosphatase (step 3), the product sedoheptulose-7-phosphate is split by transketolase (step 4 (Fig. 19-24c) into a pentose phosphate (ribose-5-phosphate) and a two-carbon fragment, carried on TPP (Fig. 19-25). This two-carbon fragment is condensed with the three carbons of glyceraldehyde-3-phosphate (step 5 ) to form another molecule of xylulose-5-phosphate in a reaction also catalyzed by transketolase.
The pentose phosphates-ribose-5-phosphate and xylulose-5phosphate-are converted into ribulose-5-phosphate (steps 6 and 7 ), which in the final step of the cycle (step 8 is phosphorylated to ribulose-1,5-bisphosphate by ribulose-5-phosphate kinase (Fig. 19-26).
This pathway is essentially the reversal of the oxidative pentose phosphate pathway described in Chapter 14, and employs the same enzymes to interconvert hexose and pentose phosphates. To highlight the similarity between the two pathways, the pathway of Figure 19-23 is sometimes called the reductive pentose phosphate pathway, even though no chemical reduction steps occur among these reactions. In the oxidative sequence, hexose phosphates are regenerated from pentose phosphates; here, hexose phosphates are converted into pentose phosphates.

Biosynthesis of Glycogen, Starch, and Sucrose

In a wide range of organisms, excess glucose is converted into polymeric forms for storage and transport. The principal storage forms of glucose are glycogen in vertebrates and many microorganisms, and starch in plants. In vertebrates, glucose itself is generally transported in the blood, but the transport form in plants is sucrose, or its galactosylated derivatives.
Although the intermediates in glycolysis and gluconeogenesis are sugar phosphates, many of the reactions in which hexoses are transformed or polymerized involve a different type of activating group, a nucleotide bound to the anomeric hydroxyl of the sugar through a phosphate ester linkage. Sugar nucleotides are the substrates for polymerization into disaccharides, glycogen, starch, cellulose, and more complex extracellular polysaccharides. They are also key intermediates in the production of aminohexoses and deoxyhexoses found in some of these polysaccharides. The role of sugar nucleotides (specifically UDP-glucose) in the biosynthesis of glycogen and many other carbohydrate derivatives was discovered by Luis Leloir.

The suitability of sugar nucleotides for biosynthetic reactions stems from several properties: 1. Their formation by the condensation of a nucleoside triphosphate with a hexose phosphate splits one high-energy bond and releases PPi, which is further hydrolyzed by inorganic pyrophosphatase; there is a net cleavage of two high-energy bonds (Fig. 19-11). The resulting large, negative, free-energy change drives the synthetic reaction and reflects a strategy common to many biological polymerization reactions

2. Although the chemical transformations of sugar nucleotides do not involve the atoms of the nucleotide itself, the sugar nucleotide molecule offers many potential groups for noncovalent interactions with enzymes; the free energy of binding contributes significantly to the catalytic activity of the enzyme (Chapter 8; see also p. 353). 3. Like phosphate, the nucleotidyl group is an excellent leaving group, activating the sugar carbon to which it is attached so as to facilitate nucleophilic attack.
4. By "tagging" some hexoses with nucleotidyl groups, cells may set them aside for one purpose (glycogen synthesis, for example) in a pool separate from hexose phosphates earmarked for another purpose (such as glycolysis).

UDP-Glucose Is the Substrate for Glycogen Synthesis

In animals and some microorganisms, excess glucose available from carbohydrates in the diet or from gluconeogenesis is stored as glycogen. Glycogen synthesis occurs in virtually all animal tissues but is especially prominent in the liver and skeletal muscles. In the liver, glycogen serves as a reservoir of glucose, readily converted into blood glucose for distribution to other tissues, whereas in muscle, glycogen is broken down via glycolysis to provide ATP energy for muscle contraction. The starting point for synthesis of glycogen is glucose-6-phosphate. This can be derived from free glucose by the hexokinase (in liver; glucokinase in muscle) reaction:

D-Glucose + ATP ----------------> D-glucose-6-phosphate + ADP

However, much of the glucose ingested during a meal takes a more roundabout path to glycogen. It is first taken up by erythrocytes in the bloodstream and converted to lactate glycolytically; the lactate is then taken up by the liver and converted to glucose-6-phosphate by gluconeogenesis.
To initiate glycogen synthesis, the glucose-6-phosphate is reversibly converted into glucose-1-phosphate by phosphoglucomutase:

Glucose-6-phosphate ---------------------> glucose-1-phosphate

The formation of UDP-glucose by the action of UDP-glucose pyrophosphorylase is a key reaction in glycogen biosynthesis:

Glucose-1-phosphate + UTP --------------------> UDP-glucose + PPi

This reaction proceeds in the direction of UDP-glucose formation because pyrophosphate is rapidly hydrolyzed to orthophosphate by inorganic pyrophosphatase (ΔG°' = -25 kJ/mol) (Fig. 19-11).
UDP-glucose, as we have already noted (see Fig. 14-13), is an intermediate in the conversion of galactose into glucose. UDP-glucose is the immediate donor of glucose residues in the enzymatic formation of glycogen by the action of glycogen synthase, which promotes the transfer of the glucosyl residue from UDP-glucose to a nonreducing end of the branched glycogen molecule (Fig. 19-12). The overall equilibrium of the path from glucose-6-phosphate to lengthened glycogen greatly favors synthesis of glycogen. Glycogen synthase requires as a primer an (α_1→4) polyglucose chain or branch having at least four glucose residues.
Glycogen synthase cannot make the (α_1→6) bonds found at the branch points of glycogen (see Fig. 11-15); instead, these are formed by a glycogen-branching enzyme, amylo (1→4) to (1→6) transglycosylase or glycosyl-(4→6)-transferase. Glycosyl-(4→6)-transferase catalyzes transfer of a terminal fragment of six or seven glucosyl residues from the nonreducing end of a glycogen branch having at least eleven residues to the C-6 hydroxyl group of a glucose residue of the same or another glycogen chain at a more interior point, thus creating a new branch (Fig. 19-13). Further glucosyl residues may be added to the new branch by glycogen synthase. The biological effect of branching is to make the glycogen molecule more soluble and to increase the number of nonreducing ends, thus making the glycogen more reactive to both glycogen phosphorylase and glycogen synthase.
requires as a primer an (α_1→4) polyglucose chain or branch having at least four glucose residues.
Glycogen synthase cannot make the (α_1→6) bonds found at the branch points of glycogen (see Fig. 11-15); instead, these are formed by a glycogen-branching enzyme, amylo (1→4) to (1→6) transglycosylase or glycosyl-(4→6)-transferase. Glycosyl-(4→6)-transferase catalyzes transfer of a terminal fragment of six or seven glucosyl residues from the nonreducing end of a glycogen branch having at least eleven residues to the C-6 hydroxyl group of a glucose residue of the same or another glycogen chain at a more interior point, thus creating a new branch (Fig. 19-13). Further glucosyl residues may be added to the new branch by glycogen synthase. The biological effect of branching is to make the glycogen molecule more soluble and to increase the number of nonreducing ends, thus making the glycogen more reactive to both glycogen phosphorylase and glycogen synthase.
If glycogen synthase requires a primer, how is a new glycogen molecule initiated? The answer lies in an intriguing protein called glycogenin (Mr 37,284), which itself acts as the primer to which the first glucose residue is attached and also as the catalyst for synthesis of a nascent glycogen molecule with up to eight glucose residues (Fig. 19-14). The first step is the covalent attachment of a glucose residue to Tyr194 of glycogenin, catalyzed by the protein's glucosyltransferase activity (step l ). Glycogenin then forms a tight l:l complex with glycogen synthase (step 2 ), within which the next few steps occur. The glucan chain is extended by the sequential addition of up to seven more glucose residues (step 3). Each new residue is derived from UDPglucose, and the reactions are autocatalytic (mediated by the glucosyltransferase of glycogenin). At this point, glycogen synthase takes over, extending the glycogen chain and dissociating from glycogenin (step 4) The combined action of glycogen synthase and the branching enzyme (step 5 ) completes the glycogen particle. Glycogenin remains buried within the particle, covalently attached to one end.

Glycogen Synthase and Glycogen Phosphorylase Are Reciprocally Regulated

Earlier we saw that the breakdown of glycogen is regulated by both covalent and allosteric modulation of glycogen phosphorylase (see Fig. 14-17). Phosphorylase a, the active form, which contains essential phosphorylated Ser residues, is dephosphorylated by phosphorylase a phosphatase to yield phosphorylase b, the relatively inactive form, which can be stimulated by AMP, its allosteric modulator. Phosphorylase b kinase can convert phosphorylase b back into the active phosphorylase a by phosphorylating the essential Ser residues.
Glycogen synthase also occurs in phosphorylated and dephosphorylated forms, but it is regulated in a reciprocal manner, opposite to that of glycogen phosphorylase (Fig. 19-15). Its active form, glycogen synthase a, is the dephosphorylated form. When it is phosphorylated at two Ser hydroxyl groups by a protein kinase, glycogen synthase a is converted into its less active form glycogen synthase b. The conversion of the less active glycogen synthase b back into the active form is promoted by phosphoprotein phosphatase, which removes the phosphate groups from the Ser residues. Glycogen phosphorylase and glycogen synthase are therefore reciprocally regulated by this phosphorylation-dephosphorylation cycle; when one is stimulated, the other is inhibited (Fig. 19-15). It appears that these two enzymes are never fully active simultaneously.

The balance between glycogen synthesis and breakdown in liver is controlled by the hormones glucagon and insulin (Table 19-4). These hormones, by regulating the level of cAMP in their target tissues, determine the ratio of active to less active forms of glycogen phosphorylase and glycogen synthase. The hormones also regulate the concentration of fructose-2,6-bisphosphate and thereby the balance between gluconeogenesis and glycolysis. Epinephrine has effects similar to those of glucagon, but its target is primarily muscle whereas glucagon's primary action is on liver. The regulatory role of these hormones and the details of their action on target tissue will be considered further

ADP-Glucose Is the Substrate for Starch Synthesis in Plants

Starch, like glycogen, is a high molecular weight polymer of D-glucose in (α_1→4) linkage (see Fig. 11-15). Starch synthesis occurs in chloroplasts. The mechanism of hexose polymerization in starch synthesis is essentially similar to that in glycogen synthesis. An activated nucleotide sugar (ADP-glucose in this case) is formed by condensation of glucose-1-phosphate with ATP. Starch synthase then transfers glucose residues from ADP-glucose to the nonreducing end of preexisting starch molecules that act as primers (Fig. 19-16). The reaction involves displacement of the ADP of ADP-glucose by the attacking 4'hydroxyl of the primer, forming the characteristic (α_1→4) linkages of starch. The starch fraction called amylose is unbranched, but amylopectin has numerous (α_1→6)-linked branches like those of glycogen, from which it differs in molecular weight and extent of branching (Chapter 11). Chloroplasts contain a branching enzyme similar to that involved in glycogen synthesis (Fig. 19-13); it introduces the (α_1→6) branches of amylopectin.
With the hydrolysis by inorganic pyrophosphatase of PPi produced during ADP-glucose synthesis, the overall reaction for starch formation from glucose-1-phosphate is

Starch_n + glucose-1-phosphate + ATP -------------------> starch_n+1 + ADP + 2Pi

ΔG°' = -50 kJ/mol

Starch synthesis is regulated at the level of ADP-glucose formation (Fig. 19-16), as discussed later

UDP-Glucose Is the Substrate for Sucrose Synthesis in Plants

Most of the triose phosphates generated by CO2 fixation in plants are converted into sucrose (Fig. 19-17) or starch. Sucrose may have been selected during evolution as the transport form of carbon because its unusual linkage, which joins the anomeric C-1 of glucose and the anomeric C-2 of fructose, is not hydrolyzed by amylases or other common carbohydrate-cleaving enzymes. Sucrose is synthesized in the cytosol, beginning with dihydroxyacetone phosphate and glyceraldehyde-3-phosphate exported from the chloroplast. After condensation to fructose-1,6-bisphosphate (by aldolase), hydrolysis by fructose-1,6-bisphosphatase yields fructose-6-phosphate. Sucrose-6-phosphate synthase catalyzes the reaction of fructose-6-phosphate with UDPglucose to form sucrose-6-phosphate. Finally, sucrose-6-phosphate phosphatase removes the phosphate group, making sucrose available for export from the cell to other tissues of the plant. Sucrose synthesis is regulated and closely coordinated with starch synthesis, as we shall see.

Lactose Synthesis Is Regulated in a Unique Way

The synthesis of lactose in lactating mammary gland occurs by a mechanism similar to that for glycogen synthesis. However, the regulation of lactose synthesis is unusual. Most vertebrate tissues contain the enzyme galactosyl transferase (Fig. 19-18a), which promotes the transfer of an activated galactose residue in UDP-galactose to the monosaccharide N-acetylglucosamine:

UDP-D-galactose + N-acetyl-D-glucosamine ------------> D-galactosyl-N-acetyl-D-glucosamine + UDP

This reaction has no role in lactose synthesis; it is a step in the biosynthesis of the carbohydrate portion of galactose-containing glycoproteins in animal tissues.
In the lactating mammary gland, however, this same enzyme participates in lactose synthesis (Fig. 19-18b). Galactosyl transferase, which is very active with N-acetylglucosamine but only feebly active with glucose as galactosyl acceptor, is present in most tissues, including the mammary gland, as noted above. Immediately after a female gives birth, the specificity of galactosyl transferase in the mammary gland changes: it now transfers the galactosyl group of UDP-galactose to glucose at a very high rate, thus making lactose:

UDP-D-galactose + D-glucose --------------> D-lactose + UDP

This "new" enzyme is called lactose synthase (Fig. 19-18b).
The change in speciiicity of galactosyl transferase is caused by the synthesis of α-lactalbumin (Mr 13,500), a milk protein, whose function was long unknown. α-Lactalbumin has been found to be a specificity-modifying subunit; its synthesis in the mammary gland, which is regulated by the hormones promoting lactation, leads to the formation of an α-lactalbumin-galactosyl transferase complex, that is, lactose synthase