This chapter tries to summarise one of the main objectives of Section N1 from the 2017 CICM Primary Syllabus, which expects the trainee to "describe the storage, synthetic, metabolic, immunological and excretory functions of the liver." Just the synthetic functions alone were enough to merit their own chapter. There is no specific question addressing this in the CICM past papers, but SAQs asking about liver function more broadly are numerous:
Trying to separate the functions of the liver into "storage, synthetic, metabolic, immunological" is basically impossible without either duplicating a lot of content or missing large chunks of important material, as for the purposes of learning this would be an illogical classification of liver function. It would have been much better to separate the functions of the liver into physiological roles, eg. "metabolism of carbohydrates" or "production of blood proteins". For example, the liver metabolises amino acids, synthesises glucose out of them, biotransforms it into fat and glycogen for storage, and then releases it on demand (again as glucose)- and it is much more difficult to develop an appreciation of this metabolic pathway when one encounters its dismembered pieces strewn across several articles. Fortunately, a lot of the critical metabolic pathways involving the liver are sufficiently important to get their own page, and the following summary points to them as a directory of sorts.
Synthetic functions of the liver include:
- Synthesis of ~90% of blood proteins (everything other than immunoglobulins)
- Synthesis of regulatory molecules
- Synthesis of nutrients
- Synthesis of glucose (gluconeogenesis) - 120 mmol/hr
- Synthesis of ketones from free fatty acids - 250 mmol/hr
- Synthesis of non-essential amino acids (30-40g/day, i.e. a minor contribution in comparison to dietary intake)
- Synthesis of lipids (~50g VLDL/day)
- Synthesis of cholesterol (50-85% of total cholesterol synthesis)
- Synthesis of bile (600-800ml/day)
- Biosynthesis of bile acids from cholesterol, and their export as conjugated bile salts via the biliary canalicular membrane of hepatocytes
- Concentration and modification of bile by cholangiocytes
- Synthesis of bilirubin
- Bilirubin is generated from haem in the reticuloendothelial system, to which the liver is the major contributor (Kupffer cells are 80-90% of this system)
- Circulating bilirubin is actively and passively transported into hepatocytes for conjugation
- It is conjugated with glucouronide to make it water-soluble, and excreted via the bile
- Synthesis of urea (the urea cycle)
- Sources of ammonia:
- Organ and tissue production
- Gut microorganisms
- Deamination of amino acids (mainly glutamine)
- Ammonia is then incorporated into the urea cycle
- About 420 mmol/day (~25g) of urea is produced in this way
Unsurprisingly, with so much ground to cover, there was no single reliable reference for this topic. The best single resource was probably Kuntz & Kuntz (2008), except one could not possibly recommend it, as it is not available for free, not referenced particularly well, nor an easy read (at over forty pages). What follows was collected over some serial database searches, yielding occasional obscure and specialised articles.
The protein content of blood plasma is described in garish detail elsewhere. It will suffice to remind the reader that the plasma is a fluid full of dissolved and suspended proteins (70-90g/L on average), the vast majority of which originate in the liver (with the notable exception of immunoglobulins). The rest of the body tissues may occasionally export some protein or peptide to circulate through the bloodstream, but this would be a departure from the norm: when most normal cells make a protein, that protein can generally be expected to remain inside that cell. In contrast, virtually none of the hepatic plasma proteins are stored inside the hepatocyte, except in the course of their pre-exocytosis modification in the ER and Golgi apparatus.
Often textbooks make mention of "three hundred proteins" or "five hundred protein synthesis pathways" at around this point. These numbers come from nowhere and mean nothing. For the reader to be suitably impressed by the synthetic capacity of the liver, it would probably be unnecessary to count and itemise all the proteins produced by the liver, as nobody would be surprised that they are numerous. For example, here's a list of just the acute phase proteins from Gabay & Kushner (1999) via Fuhrman et al (2004). It is useful because it classifies the proteins functionally, by whether they increase or decrease in response to critical illness:
- Acute phase reactants:
- Complement system
- Factor B
- C1 inhibitor
- C4b-binding protein
- Mannose-binding lectin
- Coagulation and fibrinolytic system
- Tissue plasminogen activator
- Protein S
- Plasminogen-activator inhibitor
- α1-Protease inhibitor
- Pancreatic secretory trypsin inhibitor
- Inter-α-trypsin inhibitors
- Transport proteins
- Participants in inflammatory responses
- Secreted phospholipase A2
- Lipopolysaccharide-binding protein
- Interleukin-1-receptor antagonist
- Granulocyte colony-stimulating factor
- C-reactive protein
- Serum amyloid A
- α1-Acid glycoprotein
- Negative acute phase reactants:
- Transthyretin (prealbumin)
- α2-HS glycoprotein
- Thyroxin-binding globulin
- Insulin-like growth factor I
- Factor XII
In case you're wondering, the healthy liver can produce a huge amount of protein. Barle et al (1997), using radiolabeled substrate, determined that it can crank out about 100mg/kg/day of albumin alone (i.e. 7g/day for a normal-sized person). Obviously, the rate of this production is closely controlled; the liver does not just dumbly squirt a constant stream of protein into the bloodstream, squandering nutrients like an idiot. Careful consideration is given to every molecule, and complex regulatory systems interact to ensure the right proteins are prioritised. For example, in the setting of life-threatening sepsis, the synthesis of acute phase reactants is more important because these proteins (many of them opsonins or immunomodulatory actors) have some kind of essential antibacterial function. Even skeletal muscle is catabolised to make amino acids available for the production of acute phase reactants. At the same time, negative acute phase reactants are deprioritised, and the levels of these proteins decrease in critical illness. An example of hepatic protein synthesis regulation which might be instantly recognisable is the production of C-reactive protein in response to the release of interleukin-6, which is then suppressed when you give the patient a dose of tocilizumab.
The liver is an endocrine organ, the origin and the destination of numerous hormones and prohormones. To borrow from Rhyu & Yu (2021), these functions can be split into:
It would be impossible expand much more on this without digressing into the description of what those hormones do or how the liver manages them, which is perhaps excessive. Suffice to say you miss these functions when they are gone. Endocrine dysregulation in liver disease leads to gynaecomastia, hypogonadism, osteoporosis, glucose intolerance, and something that looks like sick euthyroid syndrome. Weirdly, angiotensinogen production is apparently preserved until the absolute final stages.
It's probably important to mention that the liver is responsible for the transformation of various random substrates into amino acids, which can then be made available to the rest of the organism (Hou et al, 2020). This can't really fit into the "nutrient" category because these amino acids are firstly not intended as a source of energy, and secondly because their rate of synthesis is not particularly great, accounting for only a small fraction of the total protein turnover in the body. According to the extremely poorly referenced textbook chapter by Kuntz & Kuntz (2008), 30-40g/day of non-essential amino acids are contributed by this mechanism to the total amino acids pool. By definition, these are non-essential amino acids (i.e. you are capable of making your own), whereas the essential ones need to be made available in the diet.
If one's organism was being constantly supplied with sources of biologically available metabolic fuel, one could argue that no complex liver would be required, and a direct pipeline could connect the organs of absorption to the tissues that consume the fuel. However, the human organism did not evolve under these conditions. Our ancestors routinely had to ingest weird and seemingly inedible things which were completely unprepared for our metabolic pathways, and moreover they would ingest them so infrequently that obviously some kind of storage and retrieval system was required, to smooth the peaks and trough of metabolic substrate availability. Because of this, we have a liver that is adapted to:
Apart from accessing the reserves of stored glycogen, the liver can create glucose de novo by gluconeogenesis, out of various precursors (fructose, alanine, glycerol, lactate, etc). Apart from the proximal tubule, hepatocytes are the only tissue type that can do this (surprisingly, tubule output can account for up to 40% of total systemic gluconeogenesis under conditions of fasting).
Gluconeogenesis is resorted to when glycogen reserves are depleted, and after two days of starvation "new" glucose accounts for over 90% of the total glucose released by the liver (Chandramouli et al, 1997). With enough metabolic substrate availability, the capacity for gluconeogenesis could be massive - Exton & Park (1967) reported each gram of rat liver producing glucose at a rate of 120 μmol/hr, so long as it was fed with a steady supply of fructose. Without encroaching on the type of content that might be normally expected from the nutrition and metabolism section, it will suffice to say that all of this is regulated by the balance of insulin and glucagon.
Alanine and lactate are mentioned specifically here because they are the key molecules involved in eponymous cycles, designed to shuttle carbohydrates back to the liver. Specifically:
A great temptation always exists, wherever these pathways are discussed, to bring out some sort of enzyme flowchart full of arrows. So as to avoid extending over this metabolic ledge, the author will instead defer to Felig (1973) and Carl Cori himself (1981). In case the reader is really interested, Current Topics in Cellular Regulation had an entire volume dedicated to biological cycles in 1981 (in honor of Hans Krebs' 80th birthday), and it is an absolute goldmine, in spite of its age. For the average critical care trainee, it will suffice to know that the liver is responsible for the majority of gluconeogenesis, and of the raw substrate, some originates from the intestine and some is recirculated from the rest of the body.
Ketones, the fallback carbohydrate substitutes made available during periods of starvation, are produced almost exclusively by the liver. Again, taking great effort not to degenerate into wild branching metabolic diagrams, the origin and role of ketones can be summarised as follows:
HMG-CoA synthase, the enzyme responsible for making acetoacetate out of fatty acids, is the rate-limiting step of ketogenesis, and is only available in sufficient quantities in the liver - whereas other parts of that pathway are also expressed in the heart, the kidney and the intestine. The capacity for production is vast: McGarry & Foster (1980), reporting their own unpublished data, claimed that each gram homogenised hepatocyte suspension can crank out 250 μmol of acetoacetate per hour (ie. twice the rate of gluconeogenesis). In other words, if your liver is even remotely as efficient as this experimental pâté, it could produce its own weight in ketones every 48 hours.
The liver can manage lipids in two main ways:
In either case, it can do two main things with these free fatty acids:
The liver creates VLDL, LDL and HDL as a mechanism of exporting triglycerides and cholesterol. The liver does not seem to release lipid as free fatty acids, i.e. all hepatic lipid export is in this packaged lipoprotein form (Mashek, 2013). A typical daily rate of VLDL synthesis is apparently about 50g.
About 50% of the total body cholesterol is produced in the liver, as theoretically all cells in the body should be capable of producing it. Luo & Song (2020) put the figure at 50%, and other authors put it at 85%, suggesting that both values are probably wrong (but at least Luo & Song give a reference to support their assertion). It is a carefully regulated and very energy-expensive process, usually referred to as the mevalonate pathway, rate-limited by the step where hydroxyl-methyl glutaryl-coenzyme A (HMG-CoA) is converted into mevalonate by HMG-CoA reductase (the drug target of statins). The cholesterol which is created in this fashion is initially stored in cytosolic lipid droplets and eventually exported as LDL and VLDL. Cholesterol returning from the tissues to the liver is in the form of HDL; this cholesterol is reclaimed and either repurposed or excreted in the bile.
Speaking of bile:
It would be better not to get carried away with the molecular mechanisms here, as this is supposed to be a short pre-exam revision skim. Thus, in point form,
Hundt et al (2017) cover this quickly over two pages, and for the vast majority of people that will be more than enough. The intensivist rarely has the ability (or, for that matter, the need) to directly modify bile secretion, and our interest in its physiology is often limited to the role it plays in the elimination of fat-soluble wastes.
Bilirubin, the end product of haem metabolism, is produced at a rate of about 4mg/kg/day, or about 470 μmol/day for a normal-sized person, according to Kalakonda et al (2017). Like urea, bilirubin is a necessary step in the detoxification of haem, which is a highly reactive cytotoxin. The sites of its production are actually any organs that could be described as "reticuloendothelial", i.e. any tissue macrophages anywhere that feel like snacking on senescent red cells. Functionally, this means that the liver is the main site of its synthesis, because hepatic Kupffer cells account for something like 80-90% of the total tissue macrophage population (spleen and bone marrow being the other sites). To summarise, for rapid revision:
Handling of urea by the kidney, and the necessity of this molecule as an instrument of ammonia excretion, is explored in detail elsewhere. To simplify revision the reader will be reminded that there is an imperative to expel ammonia from the body, as it is an inevitable byproduct of amino acid deamination, and that the only convenient way of doing so is to transform it into something like urea (because the alternative is to either excrete an upsettingly large amount of urine or a corrosively high concentration of ammonia). The liver's job is to package ammonia as urea; the kidney's job is to eliminate the urea without losing too much water.
The urea cycle is annoying for the classification-obsessed, as it is not clear whether it fits into the "synthetic", "excretory" or "metabolic" part of the CICM syllabus. It's been left here because the biotransformation of ammonia is a synthesis step unique to the liver, and the excretion happens elsewhere. In short the "cycle" part of it is the recycling of ornithine, which is required by arginase to create urea (so in actual fact this thing should really be called the ornithine cycle, as the urea itself does no actual cycling. The ornithine liberated by the effects of arginase is returned to the hepatic mitochondria and is endlessly re-converted back into citrulline and argino-succinate, maintaining a constant supply of reagent to bind carbamoyl phosphate.
Ammonia metabolism is covered in some detail elsewhere. Its originates from several sources:
The glutamine to feed this cycle comes from the liver itself and from muscle, and was included in the simplified diagram above mainly because it is a major precursor. The other amino acids are obviously also fodder for the urea cycle, as they can be deaminated and therefore act as sources of ammonia; but glutamine is important because many tissues can create glutamine as temporary storage for their ammonia, exporting it into the bloodstream and therefore transferring the responsibility for urea metabolism to the liver. Because of this, glutamine is the most abundant amino acid in the bloodstream (and probably the body in general), with daily production often quoted as 40-80g/day, or around 50mmol/hr (Cruzat et al, 2018). Its bloodstream concentration (0.5-0.8mmol/L) accounts for about 20% of the total amino acid content of the blood.
In short, the total daily urea production is governed by the combination of ingested protein catabolism and systemic protein catabolism (all of which bring ammonia and amino acids to the liver), and under normal circumstances about 420 mmol/day of urea (25.5g) ends up produced and eliminated (Rudman et al, 1973).