The proximal tubule reabsorbs about 70 percent of the fluid filtered from the blood by the glomerular capillaries. As there is no such thing as active transport for water (no water pump), it accomplishes this by reabsorbing electrolytes by active transport and thus dragging water across from the urine and into the blood by osmosis. This is essentially the same mechanism used by water absorbing (gut) and water secreting (salivary glands) epithelia throughout the body. The proximal tubule even gets a helping hand in this process because the blood that circulates in the capillaries near the proximal tubule is the very same blood that was filtered in the glomerular capillaries and is thus slightly hyperosmotic to begin with. Anyway, there is a limitation to this mechanism for water reabsorption. Not volume, it is possible to reabsorb vast quantities of fluid using this sort of mechanism, but rather concentration. The problem with this sort of mechanism for fluid absorption is that it works on a very small osmotic gradient and consequently, water absorption is practically isotonic. This means that the urine will also be isotonic. It won't contain the same substances as plasma (e.g. no glucose or bicarbonate etc. and more PAH or DDT etc) but the overall osmolarity will be the same. Suppose the body has taken on an excess sodium chloride load and is desperate to get rid of it because it is disturbing the osmolarity of the body. If the urine is isotonic, the salt can't be got rid of. Think about it. If the urine is always isotonic with the plasma you can adjust the amount of salt and water in the body but not the concentration. To get rid of excess salt, you would have to drink sufficient water to restore normal osmolarity and then excrete the excess salt with the excess water.
To dispose of an excess salt load, you must be able to generate a concentrated urine. Enter the renal medulla and the loop of Henle. The medulla is very poorly drained (more on that later) and so it is possible to set up a large osmotic gradient in the medulla and keep it there (anywhere else and the blood would wash the gradient away). The other part of the problem is how to create a large osmotic gradient. Active transport is the key, but even the mighty sodium pump can't shift enough sodium to create a worthwhile gradient by itself. The other part of the solution, and the reason that the loop of Henle is in fact a loop lies in the principle of countercurrent multiplication. In a countercurrent multiplier, the combined action of active pumping and circulation and re-circulation of solutes around the loop of Henle create an osmotic gradient with the following properties.
Ok. Now we have a huge concentration gradient extending into the medulla, The osmolarity of the medulla may be as high as 1400mOsm/l, compared to the normal plasma osmolarity of 300 mOsm/l, but it isn't doing much good in concentrating the urine. True the urine gets very concentrated as it descends the loop of Henle into the medulla but it gets progressively weaker again as it ascends back out into the cortex. The latter part of the loop of Henle and the distal tubule are sometimes known as the 'diluting segment' of the nephron because the urine can in fact be more dilute as it enters the distal tubule than it was leaving the proximal tubule. This used to drive renal physiologists nuts. They knew that the loop of Henle was involved in concentrating the urine somehow but so far as they could tell, all it did was dilute it. The answer is in the collecting ducts. After going through the distal tubule, the urine passes into the collecting ducts. The collecting ducts travel through the medulla on their way out of the kidney. Right past the high osmolarity part of the medulla in fact. If the collecting ducts are water permeable then the huge concentration gradient between the medulla and the collecting ducts will drag water out of the urine into the medulla. Finally we have a concentrated urine. Next stop after the collecting ducts is the bladder. The extent to which the urine is concentrated depends mainly on how water permeable the collecting ducts are. The water permeability of the collecting ducts is controlled by anti-diuretic hormone (ADH). In the absence of ADH, the collecting ducts are water impermeable, no water is reabsorbed and a large volume of dilute urine is produced. In the presence of a high concentration of ADH the collecting ducts are highly water permeable, a lot of water is reabsorbed and a small volume of very concentrated urine is produced. ADH, working via cAMP as a second messenger stimulates insertion of water channels (aquaporins) into the plasma membrane. Way back in the first plenary, I mentioned that aquaporins are not gated, i.e. they are always open. The way to get variable water permeability is therefore to change the number of aquaporins in the membrane. When the water permeability must be low, collecting duct cells move the aquaporins from the membrane by exocytosis and keep them in little membrane vesicles inside the cell; when water permeability must be high, they reverse the process. Activation of V2 receptors by ADH also stimulates synthesis of additional receptors. The V2 receptors for ADH is yet another example of a 7-membrane-spanning-domain-G-protein-coupled receptor, but more about that later. All intermediate stages of water permeability are possible so that the volume and the concentration of urine can respond to the changing needs of the body in regulating fluid and electrolyte homeostasis.
The biophysics of how countercurrent multiplication creates these monster gradients is probably the most difficult physiological concept to grasp. I don't intend to try to explain it here, if for no other reason that you don't need to know. What is much more important is that you understand the consequences of the huge gradients in the medulla and how these gradients allow formation of a concentrated urine. Re-visit the PowerPoint animation... see if that helps.
Before I forget. One of the requirements to build a countercurrent multiplier is to keep it away from the blood supply, or else the blood flow will wash the gradients away before they get started. The capillaries that accompany the long loops of Henle on their trip to the medulla and back are called the vasa recta. These capillaries also form long loops and run in parallel to the loop of Henle. Most importantly, the arterial and venous ends of the capillaries run in parallel with each other (in other words a hairpin loop). This allows a process called countercurrent exchange to occur. In short, countercurrent exchange allows exchange of solute between the ascending and descending limbs of the capillaries and prevents them from dissipating the medullary concentration gradient. Among the solutes exchanged in this process are oxygen and carbon dioxide (e.g. oxygen moves from the descending limb of the vasa recta to the ascending limb without travelling around the loop, carbon dioxide moves in the opposite direction). There is probably a lot of anarobic metabolism going on in the renal medulla.