Distal tubule and collecting duct

This chapter seems relevant to the aims of Section H1(ii) from the 2017 CICM Primary Syllabus, which expects the exam candidate to "describe glomerular filtration and tubular function". Regionally specific questions about the nephron (eg. "what happens in this specific tubule") are virtually unknown in the CICM First Part exam, which means the trainees can safely ignore this entire chapter. It is made available here mainly for completeness.

In summary:

The best article for this, and - for that matter - for all of the structure and function of the kidney - is "Structural organisation of the mammalian kidney" by Kriz & Kaissling. For something that deals specifically with the DCT, one cannot go past  Subramanya & Ellison (2014), and for a clear explanation of what intercalated cells do, one can read Roy et al (2015). It s of course pointless to do any of that, because CICM have never asked for this much detail. 

Macula densa and the juxtaglomerular apparatus

At the distal end of the loop of Henle, as it reenters the cortical labyrinth after touring the pickled hell of the inner medulla, a mural "plaque" of cells is usually seen, which thickens the already thick ascending limb where it comes close to its glomerulus of origin. The most famous image of this area comes from a 1992 paper by Kriz & Kaissling; it is a particularly lucky section through the vascular root of a glomerulus, a version of which has been vandalised below:

These are basically sensors that detect increased salt delivery to the distal tubule and use that signal to coordinate a series of regional and systemic responses which aim to decrease the aforementioned salt delivery. The sensor-effector mechanism itself is hellishly complex, and those who are interested are directed to a detailed overview in Bell et al (2003). To cut a long story short:

• Sodium and chloride are delivered to the thick ascending limb (and the concentration for either of them should not be especially great, as the osmolality at this part of the tubule should be something like 100 mOsm/kg, i.e you'd be lucky to have 40-50  mmol/L of sodium here.
• Sodium potassium and chloride then enter the macula densa cells via a Na/K/2Cl cotransporter
• Chloride leaves the cell via a basolateral channel
• The net gain in positive charge depolarises the macula densa cell 
• This, by a calcium-mediated mechanism, results in the release of ATP into the perivascular spaces in the vicinity of the afferent and efferent tubules
• The ATP then either activates specific purine receptors on the afferent arteriole, or is converted to adenosine (which then acts on A₁-adenosine receptors).
• The net effect is that increased salt delivery to the nephron results in decreased glomerular blood flow, which decreases salt delivery (i.e. this is a negative feedback mechanism

That is the basis of tubuloglomerular feedback, which is described better in an excellent paper by Volker Vallon (2003) and in the chapter on renal blood flow. The macula densa is also implicated in the humoral control of blood pressure, by mediating the release of renin through some apparently unknown mechanisms, where decreased salt intake produces the release of renin, and vice versa. For this structure-and-function-oriented chapter, it will suffice to say that this functionally important structure caps the topmost end of the thick ascending limb of the loop of Henle, just before it abruptly turns into the distal convoluted tubule. 

Distal convoluted tubule

This thing, though convoluted, is quite short. Even unfurled, Crayen & Thoenes (1978) got it to stretch about 1mm. The walls are even thicker than the thick ascending limb, made up of tall cuboidal cells full of mitochondria. Like the thick ascending limb and the proximal tubule, the basolateral part of these cells is topographically very complex, with multiple interdigitations. There's probably at least four different types of cells and they appear variably scattered in this segment, making it quite difficult to describe its boundaries histologically. Most early researchers described the extent of the distal convoluted tubule in terms of what it is not; i.e. it is whatever tubule is seen between the macula densa and whatever the next tubule is. 

The main characteristic features of how this segment manages the tubule contents are explained systematically in an excellent review by Subramanya & Ellison (2014)

• Water: the epithelium of the distal convoluted tubule does not reabsorb water. Towards the most distal end of the distal tubule, the bit which is sometimes referred to as the "connecting tubule" can also express aquaporins, similar to the collecting duct, but the proximal DCT is supposed to be completely water-impermeable. This means the dilute product of the thick ascending limb should remain dilute for most of the length of the distal convoluted tubule. This appears to be supported by experimental data: Clapp & Robinson (1966) micropunctured dog tubules and determined that the osmolality of DCT fluid remained hypotonic in relation to plasma for its entire length, somewhere around 60-120 mOsm/kg.
• Sodium and chloride: the main salt transporter is different; this tubule reabsorbs 5-10% of the total filtered salt load, but the NKCC2 (the target of loop diuretics) is replaced by the NCC,  the thiazide-sensitive sodium chloride cotransporter. In fact, in the modern definition of the distal convoluted tubule, the beginning and end of NCC expression define the boundaries of this tubule segment. Additionally, the distal part of the distal convoluted tubule can express the ENaC channels, like the collecting duct.
• Potassium is almost completely reabsorbed by the time it gets here (the proximal tubule and the thick ascending limb of the loop of Henle have taken care of that), so only about 10% of the filtered potassium load ends up at the distal convoluted tubule. Then the concentration of potassium in the tubular lumen increases over its transit through the DCT because it leaks out via the ROMK (Renal Outer Medulla Potassium) channels, driven probably by a voltage-dependent mechanism. The more sodium is reabsorbed, the more potassium needs to leak out, to maintain electroneutrality.
• Essential to magnesium transport: this is where active transcellular mechanism reabsorb the magnesium, and it is generally thought that this part of the nephron is the most important in determining the final magnesium concentration in the urine
• Essential to calcium transport: together with the calcium channels in the collecting duct, the distal tubule contributes to the total renal calcium handling.

This area has some notable drug targets, which brings a bit of ICU-related clinical relevance to an otherwise forgettable chapter.

• Thiazide diuretic drug target: the water and sodium-dumping effects of thiazides are mediated by the blockade of the NCC channel. The loss of sodium reabsorption here gradually produces hyponatremia. 
• Hypokalemic side-effects of frusemide are due to the way potassium and sodium are handled in the distal convoluted tubule. As sodium delivery is increased to this tubule, the NCC channel works harder, and more sodium ends up being reabsorbed. The presence of ENaC channels in the distal portion of the tubule also increases the absorption of sodium. This increases the net positive charge inside the cell, which pushes more potassium out via the ROMK channel. Ergo, anything that increases the delivery of sodium to the distal convoluted tubule will increase the voltage-mediated secretion of potassium. There are several possible reasons for an increase in the delivery of sodium to this part of the tubule, but the most important for exam purposes is the use of loop diuretics. When the NKCC2 cotransporter is blocked by frusemide, it cannot absorb sodium, and so the distal tubule ends up receiving a vast unexpected amount of it, which leads to a vigorous potassium leak and hypokalemia.  At the same time, sodium reabsorption is maintained. This is one of the reasons why frusemide-treated patients end up hypernatremic and hypokalemic.

Connecting tubule 

This forgotten and misunderstood segment of the nephron is a place where the last of the calcium absorption occurs.  In this day and age, its extent is defined by an overlap of apical proteins. There is the presence of the TPRV5 calcium channel and the aldosterone-regulated epithelial channel ENaC, which are already present in the distal part of the distal convoluted tubule, but there's now also the presence of vasopressin-mediated aquaporins which are present mainly in the cortical collecting duct. 

Calcium reabsorption happens here by means of the TPRV5 channel, the expression of which is regulated by parathyroid hormone and 1,25-dihydroxyvitamin D3. This does not happen in the collecting duct, which is the only thing that makes the connecting tubule somewhat unique.

The sodium potassium and water handling by this part of the tubule is so similar to what happens in the collecting duct that it would be better to discuss them all together as "aldosterone and  vasopressin-regulated distal nephron".

Collecting duct

The disappearance of calcium transport machinery (mainly the apical TPRV5 channel and the basolateral NCX exchange transporter) signals the transition of the connecting tubule into the collecting duct. This part of the tubule has couple of subdivisions, which have both geographical and functional differences:

• Cortical collecting duct is in the cortex (obviously), and is also quite urea-impermeable
• Outer and inner medullary collecting duct, which is highly urea-permeable and where the reabsorption of urea contributes to the mechanism of intrarenal urea recycling that maintains a high inner medullary urea concentration.

This part of the tubule is functionally characterised by its sensitivity to aldosterone and vasopressin, which regulate the sodium, water and urea reabsorption here:

Sodium reabsorption here mops up whatever is left behind after the proximal tubule, the thick ascending limb and DCT are done with the tubular fluid. This final stage is mediated by the ENaC channel, expression of which is controlled by aldosterone. This is the drug target of spironolactone (which blocks aforementioned aldosterone mediated expression) and amiloride (which directly blocks this channel). 

Potassium secretion in the collecting duct is also mediated by the ROMK channel, and like everywhere else in the distal nephron the potassium here is being pushed around by voltage gradients. If there is a significant aldosterone-mediated sodium reabsorption, potassium excretion is increased to maintain electroneutrality, and as the consequence potassium ends up being dumped along with the urine. The upshot of this is the association between hypernatremia and hypokalemia seen in hyperaldosteronism, and the association between hyponatremia and hyperkalemia seen in spironolactone use.

Water and urea reabsorption  in the collecting ducts is regulated by vasopressin, which activates the insertion of aquaporin and UT1-3 channels into the apical membrane. The former are prevcalent in the cortical duct, which allows the urea to be massively concentrated, and the latter are abundant in the innermost medullary sections of the duct, allowing urea to be reabsorbed into the inner medulla along the concentration gradient. This mechanism of urea recycling is essential for the maintenance of the medullary osmotic concentration gradient which is used to reclaim water from the tubule. More on this can be found in the chapters on renal handling of water,  renal handling of urea, and the countercurrent exchange/multiplier mechanisms.

Intercalated cells

The term "intercalated", meaning "shoved conspicuously in the middle of something else", describes these cells very well. They appear amid more normal-looking epithelia in the so-called "aldosterone sensitive tubule" segments (the distal D2 part of the convoluted tubule, the connecting tubule and the collecting duct).  They are clearly different morphologically to all the other cell types, and perform roles mainly related to the handling of ammonia, potassium and acid-base balance. For more detail, one could always read the review by Roy et al (2015), but a less interested person would probably be satisfied with the following stylised pictograph:

Or, in the form of words, the intercalated cells:

• Secrete ammonium (which traps hydrogen ions in the tubule). The ionized NH₄⁺, the combination of excreted NH₃ and H⁺, remains in the lumen of the tubule (where it is trapped by its charge).
• Secrete chloride.  This is a passive mechanism which is mainly driven by voltage; the positive charge of secreted ammonium creates a gradient for the removal of chloride. It can be reclaimed again by the activity of a chloride-bicarbonate exchanger, but this protein is only active if the intracellular pH is suitably high. 

This acid-base function has importance in the pathogenesis of Type 1 (distal) tubular acidosis, where there is a problem with ATP-powered H+ secretion, normally an acidity-regulated process. As pH drops, so the activity of this protein should increase, thus increasing the capacity for tubular ammonium trapping and chloride excretion. In distal RTA (particularly the recessive variant) the activity of this protein can remain sluggish even at a low systemic pH.