Metabolic functions of the liver

This chapter is relevant to Section N1(i) from the 2017 CICM Primary Syllabus, which expects the trainee to "describe the storage, synthetic, metabolic, immunological and excretory functions of the liver."  The biggest challenge for the author has been to find the conceptual gap which separates the synthetic functions from the metabolic, which were clearly intended to be viewed as distinct chemical activities by the examiners. Before the angry reader snorts that this was not the product of thoughtful syllabus design but rather a lazy cut-and-paste from page 36 the 2004 ANZCA manual, they are reminded that the ANZCA and CICM examiners are for our intents and purposes invested with papal infallibility, and that moreover other reputable educational institutions reproduce exactly the same wording in their own syllabus documents. 

So: let's try to work with the syllabus wording on this one. Whereas synthesis might imply a creation of more complex wholes from elemental parts (for example, the creation of  rather complex proteins from amino acids), metabolism refers more broadly to the chemical processes which sustain life, and sometimes more specifically to the transformation of molecules (often into simpler or smaller ones, with some liberation of energy, though not necessarily always so).

Questions asking about the functions of the liver have never specifically asked to list and categorise the metabolic function per se, but - unsurprisingly - a lot of liver questions touched on functions that could be described as "metabolic" by a generous reader, and these can be used as an unofficial syllabus of sorts. For example:

• Question 4 from the second paper of 2020 asked about the metabolism of dietary energy sources in the liver
• Question 16 from the second paper of 2015 asked about ammonia metabolism (which is basically asking about the urea cycle)
• Question 12 from the second paper of 2011 asked about the metabolism of paracetamol, i.e. the biotransformation of xenobiotics
• Question 8(p.2) from the first paper of 2010 also asked about the urea cycle, but more directly ("Explain the role of urea in the body").

From these, we can at least identify the important metabolic functions which the college examiners wish to emphasise (because how better to draw the trainee's attention to a subject, than to include it in a past exam paper). In summary:

Fortunately, this topic is well supported by academic writers. From literally the first page of search results one can pick among Mitra & Metcalf (2012), Mitra & Metcalf (2009), Campbell (2006) and Alamri (2012), all of which basically cover the same concepts. Interestingly, all of these authors framed their articles as rapid revision summaries, basically keeping to a limit of 2-3 pages, which makes them all equally valid recommendations for the time-poor exam candidate. Unfortunately, being summaries of textbook chapters, these papers never reference anything. Fortunately specific metabolic areas have some excellent review articles with extensive bibliographies, and the reader would surely feel cheated if they were not referred to these peer-reviewed resources early in this chapter:

• For carbohydrate metabolism, Adeva-Andany et al (2016) 
• For lipids, Huang et al (2014)
• For protein,  Mortimore et al (1989)  - and - if you really want - you could delve into Mammalian Protein Metabolism, volumes I to IV, by Munro (1970)
• For ammonia, Adevo et al (2012)
• Specifically for the urea cycle, Watford (2002)
• For lactate, Rabinowitz & Enerbäck (2020)
• For biotransformation of drugs, McKindley et al (1998).

Carbohydrate metabolism in the liver

In summary:

• Circulating glucose and dietary glucose is delivered to the liver via the portal venous and hepatic arterial blood
• The hepatocytes absorb this glucose via the GLUT2 membrane transporter, which is not insulin-sensitive (i.e the rate of glucose entry into the hepatocyte is completely unregulated and is totally related to the concentration of glucose)
• In the hepatocyte, the glucose is rapidly phosphorylated into glucose-6-phosphate by hexokinase, which is insulin sensitive. This traps the glucose in the hepatocyte. 
• Of this trapped glucose,
  • Some will be metabolised to pyruvate, and then transformed into ATP for local use by the hepatocyte
  • Some will be stored as glycogen
  • Some (an excess) will be stored as fat

In short, the only unique thing the liver does with carbohydrates is store them. Apart from packaging glucose as glycogen and fat, and then accessing them later to release glucose, the hepatocytes don't do anything especially imaginative with their sugar. Sure, they can burn it as fuel, making CO₂ and deriving ATP from the process, but that is something all tissues can do, and in any case hepatocytes tend to prefer lipid metabolism for their energy, particularly in the fasted state where the liver is expected to be a net exporter of glucose.

Export of glucose from the liver

Between meals the liver supplies metabolic fuel to the tissues by slowly releasing stored glucose. Glucose issued from the liver can have two possible origins:

• From the breakdown of glycogen (glycogenolysis)
• From de novo synthesis (gluconeogenesis)

Hepatic glycogenolysis is the main mechanism of maintaining blood glucose concentrations between meals. Other tissues also contain glycogen, and do use it as their own personal reserve of metabolic fuel, but they are unable to release it into the systemic circulation because they lack glucose-6-phosphatase, the key final step in the pathway. Glucose-6-phosphate is trapped in cells by its polarity and by a lack of suitable transport protein; glucose-6-phosphatase separates it into phosphate and glucose, which can then be exported from the hepatocytes by various transporters (probably mainly GLUT1). 

Lipid metabolism in the liver

The liver, though normally storing no more than about 600 kcal of fat within its own cells, contributes indirectly to the total lipid storage in the body by being involved in the traffic of fat to and from the adipose tissue. This function can be described in a stepwise progression of logical points, which unfortunately involves covering some of the same ground as would be expected from the metabolism and nutrition chapters. The reader who develops an avid interest in the odyssey of lipid through the circulatory system is directed to an incredibly lucid paper by Spector ("Plasma lipid transport", 1984). For the reader who has no such enthusiasm, the barebones diagram offered below will probably suffice.

In short, fat is an attractively dense energy storage medium. Unlike glycogen, which only contains about 4kcal per gram, and each gram of which ends up being stored in a hydrated form along with 2-3g of water, fat produces about 7kcal per gram and is stored in lipid droplets which are almost entirely water-free.  It would be convenient to store it in the liver, but unfortunately the liver cannot store too much fat, as it is an encapsulated organ of finite size, and besides becoming filled with fat vacuoles tends to make it function less efficiently (to say nothing of the cirrhosis and tumours). Ergo, fat needs to be stored at some distant site, preferably one with ample space. Adipose tissue is this convenient distant store of fat, and its capacity can be near-infinite. Thus, fat needs to be transported to the adipose tissue, and then retrieved from it as needed, so that it can act as a reservoir of metabolic fuel. This transport is in the form of free fatty acids and triglycerides.

Free fatty acids are an immediately useable form of metabolic fuel which are not especially water-soluble and which are mainly transported bound to albumin. They are unfortunately somewhat toxic, mainly because they are chemically active. You intend for them to just politely lay there and wait to be burned as metabolic fuel, but instead they worm their way into all sorts of intracellular machinery, produce oxidative stress, act as unexpected signalling molecules and secondary messengers, and rudely embed themselves as droplets in the endoplasmic reticulum, interfering with all sorts of biosynthesis pathways. This translates to cell dysfunction and even cell death: for example, there is evidence of high free fatty acids levels producing apoptosis of pancreatic β-cells and inhibition of fibroblast proliferation. For this reason it is impossible to transport all the lipid in the body as free fatty acids, and normal free fatty acid levels do not tend to exceed 0.5 mmol/L (though a range of 0.18-1.65 mmol/L is occasionally given by older authors). Their role is really just to supply fuel to cells, and this is reflected in their half-life, which is usually less than 2 minutes (i.e. as soon as they are released they end up sucked into the vortex of cellular metabolism). Obviously, if you want to send a lot of lipid from place to place, you're going to have to package it into a form that is less biologically active.

Triglycerides are this less toxic method of transporting large amounts of lipid around the body. As a form of lipid storage, these are pretty good - they have a similar energy density to free fatty acids, as each triglyceride molecule simply consists of three fatty acids of variable length, attached to a glycerol backbone. Unlike free fatty acids, these are rather benign from a chemical standpoint. They're certainly not going to interfere with cellular function if they end up loose in the cytosol. Like the free fatty acids, they are basically insoluble in water, and end up packaged in chylomicrons or VLDL in order to sail the bloodstream (chylomicrons originate in enterocytes and VLDL originates in the liver). These particles are massive (weighing up to 30 million Daltons) and are therefore a much more efficient method of transporting lipid in bulk. Their role also includes supplying metabolic lipid to cells, and for many their fate is to be eroded by endothelial lipase in the capillaries, releasing free fatty acids locally (where they are rapidly taken up by tissues). Where these fat oil tankers stop at the adipocytes, their cargo disembarks by means of lipoprotein lipase which liberates the fatty acids and triglycerides and allows them to become incorporated into the body fat store. 

Handling of circulating lipids by the liver

In summary, what does the liver do with the two major forms of circulating lipids can be broadly described as "biotransformation and repurposing". It can make ketones from them, which is a valuable service unique to the liver, and it can repackage them as VLDL (thereby returning them to the adipose tissue).

• Handling of free fatty acids by the liver:
  • The liver absorbs circulating free fatty acids
  • Some of these are used as the metabolic energy source for the hepatocytes
  • Under conditions of fasting, these are transformed into ketones and released into the circulation
  • In the postprandial state, these free fatty acids are re-esterified into triglycerides, packaged with lipoproteins, and released into the circulation as VLDL
  • The liver is usually not a source of free fatty acids
• Handling of triglycerides by the liver
  • The liver is a source of triglycerides, in the form of VLDL, which it makes from free fatty acids and from stores of triglycerides it creates from excess carbohydrate
  • VLDL undergoes the same fate as chylomicrons in the tissues: they are hydrolysed by lipoprotein lipase, releasing free fatty acids, and are incorporated into adipocytes
  • Lipoprotein fragments return to the liver, where they are absorbed into hepatocytes by endocytosis, and then either
    • hydrolysed into free fatty acids and released (unchanged or as ketones)
    • hydrolysed into free fatty acids and burned by the hepatocytes as a metabolic fuel source
    • incorporated into hepatic fat stores as triglycerides

Protein metabolism in the liver

The liver is the destination for the majority of circulating proteins and peptides (as it is also the origin for most of them). The roles of the liver in their processing can be summarised as:

• Catabolism of circulating proteins and peptides
• Interconversion of amino acids
• Deamination of amino acids (eg. for use in gluconeogenesis)
• Urea synthesis

Most of what follows is the upshot of the fact that, when one ingests protein, there is no option to store it, as there is with carbohydrate or fat. One cannot simply form a depot of pork proteins in one's liver after consuming a rack of ribs. There's not even a solid mechanism for storing amino acids. All ingested and internally degraded protein and peptide material must therefore either be immediately synthesised into some sort of human body protein, or be deaminated and converted into a metabolic fuel precursor. The latter step produces ammonia, the noxious product of deamination, which needs to be removed from the body in the form of urea. It appears that about half of the daily ingested protein load (about 20g/meal)  is incorporated into protein synthesis and the rest is converted into metabolic fuel.

Dietary protein handling is discussed elsewhere. In summary, whole protein disintegrates in the intestine, and then the portal venous blood brings this ruin to the liver in the form of individual amino acids and oligopeptide. This is one source of this substrate.

Protein catabolism is another source of raw amino acids. It is more strictly the job of the reticuloendothelial system, which the liver happens to contain the bulk of (in the form of Kupffer cells). Large circulating proteins and protein complexes undergo endocytosis by these macrophages and are degraded into amino acids and peptides within their lysosomes. Hepatocytes have receptors for specific proteins (eg. lipoproteins) and are therefore themselves able to extract proteins from the circulation by receptor-mediated endocytosis. 

Though officially there is no such thing as a protein store, it is pretty clear that the liver and skeletal muscle are the main sites from where protein is mobilised (for example in times of starvation or stress). Liver protein is used first: after 48 hours of fasting, most of the protein loss is from intracellular liver protein, and the skeletal muscle remains mostly intact, whereas after seven days the vast majority of the losses (66%) are from muscle and only 16% from the liver (Addis et al, 1936). Even during periods of normal nutrition there is still a constant hepatic protein turnover, where some sort of transient pool of peptides and protein synthesis intermediaries is broken and reformed cyclically over very short timeframes. To give the reader a sense of scale, Mortimore et al (1989) report that this pool accounts for only 0.6% of total liver protein, and that these short-lived fragments have a half-life of about 10 minutes.

Amino acids are the final common pathway: both internal protein catabolism and dietary protein ingestion ultimately just supply amino acids to the liver. There's also a small amount of circulating amino acids, which everybody always seems to forget about, as their concentration in the bloodstream is relatively small. These are also a substrate for hepatic amino acid metabolism, and are transported into hepatocytes by a variety of mainly active transport mechanisms (in case anybody needs to know this in detail, it is covered well by Hou et al, 2020).  Revelling in the obscurity of his reference, the author offers this table from Luc Cynober (2002), in case anyone out there ever needs to know how much citrulline they are supposed to have.

With the understandable reluctance of basically everybody to cannulate the healthy human portal vein, we have only a very limited idea of what the hepatic uptake of amino acids is like. It is somehow ethically more acceptable to cannulate the portal veins of non-consenting pigs, and studies from the Danish Pig Research Centre (eg. Kristensen & Wu, 2012) have indicated that amino acid uptake is unequal and related to demand (for example, half of all delivered alanine and asparagine is voraciously absorbed from the portal blood, whereas the porcine liver appears to have minimal interest in aspartate and glutamate, and is a net exporter of these molecules).

In short, the liver is presented with a bunch of amino acids, and it needs to do something with them, because it has nowhere to put them. Its options are:

• Incorporate them into protein 
• Transform them into another amino acid 
• Deaminate or oxidise them, transforming them into a molecule capable of acting into a metabolic fuel and an ammonia molecule.

Which of these it chooses to do depends on a myriad of factors too vast to enumerate here, except to say that in times of starvation, deamination and incorporation into gluconeogenesis appears to be the standard response. Conversely, in times of nutritional plenty only about 5% of the total body urea production is contributed by the hepatic deamination of amino acids (again from Kristensen & Wu, 2012). Interconversion of one amino acid into another is also regulated by various homeostatic mechanisms which maintain the availability of free amino acids for the organs and tissues.

Deamination of amino acids as a source of metabolic fuel

Amino acids are not a suitable metabolic fuel, in their original state; i.e no organism is able to derive energy directly from amino acids. Amino acids need to first be transformed into a "carbon skeleton" of themselves, an oxoacid. This is achieved by two main mechanisms:

• Deamination is a direct pathway taken by some amino acids which have a specific metabolic enzyme dedicated to their degradation. For example:
  • Glutamate is oxidised to alpha-ketoglutarate
  • Glycine is oxidised to glyoxylate
  • Serine is deaminated to pyruvate
• Transamination is what happens to the amino acids which do not have a dedicated enzyme. This reaction transfers an amino group from an amino acid to a keto-acid, leaving behind an oxoacid carbon skeleton of the original amino acid and creating a new amino acid that is amenable to direct deamination (usually it is glutamate).

As you can see,  the end results of these reactions are molecules that can plug into the citric acid cycle without too much further manipulation (for example alpha-ketoglutarate and pyruvate). This is a great way to get rid of unwanted amino acids, but it certainly produces a whole pile of ammonia  (NH₃) and ammonium (NH₄⁺), which are hideously toxic and which need to be disposed of in some way. Which brings us to:

Ammonia metabolism and the urea cycle

Ammonia, the gas, and ammonium, the positively charged ion produced when this gas is dissolved in water, can be considered a single entity for the purposes of human biology, once you've ejected all the pedants from the lecture theatre. The two species are in equilibrium:

NH₃ (gas) + H⁺ ⇌  NH₄⁺

This equilibrium has a pKa of about 9.0-9.3, which means that at body fluid pH the reaction is pushed hard to the right, and 99% of the ammonia ends up in the form of NH₄⁺. Unlike NH₃, this charged molecule is not able to cross lipid bilayers, and is therefore dependent on various sorts of carrier-mediated transport.

The human body produces a remarkable amount of this stuff because there's plenty of amino acids being metabolised all over the place, with a total daily turnover of something like 300-400g of protein. That number is obviously a bullshit statistic used for various unscientific theatrics, but it probably represents some sort of ballpark estimate. If you dig deep enough you'll find it comes from various reasonable oversimplifications made in reputable peer-reviewed publications, such as Tomé & Bos (2000) or Waterlow (1981).  The bottom line is that ammonia production is a vast and necessary evil. 

The origins of total body ammonia, from Huizenga et al (1996), are:

Organ/tissue	Ammonia release in mmol/day
Intestine	71
Kidney	13
Pancreas	10
Exercising muscle *	130

*Resting muscles are a net absorber of ammonia, and only exercising muscles produce enough ammonia for it to matter on a whole-body scale

In case a grainy scan from an original paper is what you need to activate your limbic trust centres, here it is. In general all tissues produce and consume some trivial amounts of ammonia, but as you can see, the intestine is the most important source of ammonia in the body, and this is reflected in the high ammonia concentration of the portal venous blood (usually about 90 μmoI/L - three times greater than the concentration of normal arterial blood). The reason for this is the activity of various ammonia-producing microorganisms in the gut.

As the result, the intestine is the main source of hepatic ammonia supply. According to Huizenga et al, every day the liver will receive approximately 137 mmol of portal venous ammonia and 13 mmol of systemic arterial ammonia, for a total daily load of approximately 150 mmol. In addition to this extrahepatic ammonia, there is also a massive and hard-to-estimate flux of intrahepatic ammoniagenesis which results from the deamination of amino acids in the liver. Then, the hepatic venous blood leaving the liver has a relatively normal arterial-like concentration of ammonia (about 40 μmoI/L), which means the liver is able to dispose of the excess ammonia in some way. The main mechanisms for this are the production of glutamine and the synthesis of urea.

Glutamate and glutamine involvement in the metabolism of ammonia

One convenient way of ridding oneself of troublesome ammonia is to make amino acids back out of it, in essence reversing deamination. This is what happens through the actions of glutamate dehydrogenase and glutamine synthase. To maximally simplify the actions of these enzymes, the following minimalist diagram is offered to the reader, based on more detailed work in Adeva et al (2012)

As you can hopefully tell from the arrows, glutamate dehydrogenase is an enzyme that can catalyse both the dehydrogenation of glutamate into α-ketoglutarate, and the reverse reaction (which happens to consume an ammonium ion). If the reaction runs in that direction, the next step can be the conversion of glutamate into glutamine by glutamine synthetase, which consumes another ammonium ion. Voila: you got rid of two ammonium ions, and now you have glutamine, the most abundant amino acid in the body. And you don't even need to rely on the liver for this: fully one third of all glutamine synthesis is extrahepatic, eg. even your astrocytes can do it. 

That might sound great, except 1) what are you planning to do with all that osmotically active glutamine piling up in your astrocytes, and 2) the answer better not be "metabolise it as fuel", because that would liberate ammonia and defeat the whole purpose of this process. Sure, the extrahepatic tissues can cheat (and do), exporting glutamine and glutamate into the bloodstream - but this is also not a solution, as this simply shifts the responsibility for the disposal of ammonia to being the liver's problem. Clearly this is not a complete process for the removal of excess ammonia from the body. In order to finish the job, ammonia needs to be exported from the body via the kidneys, as urea.

Urea synthesis in the liver

The very practical reason for packaging ammonia as urea, as well as the mechanisms of handling urea in the kidney, are discussed elsewhere. The interested reader is redirected to an excellent article by Malcolm Watford (2003).  It will suffice to summarise that for land animals there is no practical means of excreting ammonia directly into the environment, which means they need to eliminate it via the urine - but it would be impossible to do so without either concentrating urinary ammonia to the point of nightmarish urethra-incinerating toxicity, or producing comically large amounts of urine. Urea is therefore a convenient chemically benign vehicle which can carry ammonia out of the body. The conversion of ammonia into urea is mediated by the urea cycle enzymes, and is described in detail elsewhere because of the author's arbitrary decision to label this as a "synthetic" rather than "metabolic" step. 

Lactate metabolism and the Cori cycle

The metabolic origins and metabolic fate of lactate are discussed in considerable detail in several chapters of the acid-base section, where lactate and lactic acidosis are carefully and methodically dissected. Without duplicating any content, the reader is reminded of the following facts:

• Lactate is constantly produced by various tissues, especially those that either lack a normal means of making energy from oxidative phosphorylation (eg. red cells), or those that run out of substrate for it (exercising muscle). The production of lactate from glycolysis produces 2 ATP, which those tissues can use.
• The total rate of lactate production is vast, approaching 1.5 mol/day, or 0.8 mmol/kg/hr
• The liver extracts about 55% of this lactate from the blood that flows through it, with the result that the hepatic vein has the lowest venous lactate concentration in the body. These are extremely high-affinity enzymes, and the healthy liver has such a high capacity for the metabolism of lactate that the rate of lactate metabolism is almost entirely dependent on the rate of hepatic blood flow.
• In the liver, lactate dehydrogenase metabolises lactate into pyruvate, which can then be either burned locally in the service of supplying ATP to hepatocytes, or exported as glucose after being converted by gluconeogenesis enzymes, at a cost of 6 ATP per molecule.
• In this fashion, lactate can be used to shift the cost of tissue metabolism into the liver. 

Metabolism of drugs and toxins by the liver

For the vast majority of clinically relevant drugs, the liver is the major site of metabolism. It would surprise nobody to find that this, a topic with considerable exam relevance, has multiple pages to cover it all across Deranged Physiology. Some examples include:

• Hepatic clearance
• Biotransformation and metabolism of drugs
• Phase I and Phase II biotransformation reactions
• Importance of the liver for pharmacology and toxicology
• Paracetamol toxicity (which has attracted a lot of Part I and Part II SAQs)

Without reproducing an excessive amount of content, the liver's role in pharmacology can be summarised as follows:

• The role of the liver in pharmacokinetics is as an organ of clearance
• By convention, the metabolic functions of the liver are divided into Phase I and Phase II reactions
  • Examples of Phase I reactions:
    • Hydrolysis
    • Reduction
    • Oxidation.
  • Characteristics of Phase I reactions:
    • these reactions expose or introduce a functional group (–OH, –NH2, – SH or –COOH)
    • They usually result in a small increase in hydrophilicity.
  • Examples of Phase II reactions:
    • Glucouronidation
    • Sulfation
    • Acetylation
    • Methylation
    • Conjugation with glutathione
    • Conjugation with amino acids eg. taurine, glutamine, glycine
  • Characteristics of Phase II reactions:
    • The products are supposed to be significantly more hydrophilic than the original substrate