On the mechanism of thiazide-induced hypocalciuria

Back from the 2009 ASN Meeting in San Diego!  I'll admit to a few lapses in blogging, but the meeting kept me pretty busy.  Thanks to everybody who stopped by the Renal Fellow Network poster and their encouraging words.

The 2009 ASN Homer Smith Award went to Rene Bindels, a master physiologist from the Netherlands who has made great advances in our understanding of renal calcium and magnesium handling via the Trp channels.  Amongst the highlights of his Saturday morning talk was a good explanation for why hypocalciuria results from thiazide diuretic treatment.  

The hypocalciuric effect of thiazide diuretics is well-known:  we frequently put patients prone to develop calcium nephrolithiasis on thiazides with this effect in mind, and patients with Gitelman's Syndrome (caused by mutations in the thiazide-sensitive Na channel, NCC) also have significant hypocalciuria.  There were two general mechanisms postulated to account for this.  Either:

(1).  blockage of NCC results in increased active Ca uptake via TrpV5 channels in the distal tubule, or

(2). thiazide-induced hypovolemia results in enhanced passive Ca uptake in the proximal convoluted tubule.  

The generation of TrpV5 knockout mice provided an opportunity to see which of these possibilities was correct.  Interestingly, these mice still developed hypocalciuria when given thiazide diuretics, indicating that hypocalciuria does NOT require TrpV5 (that is, the 2nd explanation above is true while the 1st explanation is not).  Just another example of how the physiology in one segment of the nephron strongly influences the ion uptake in another segment.  

Renal Cilia Length Increases with ATN

Get your flu shots, everyone!  Renal Fellow Network recently had to take an H1N1-sponsored weekend off, and I strongly suggest getting the vaccine if available to you; this bug is no fun at all. 

I've written about the ciliary hypothesis of cystic kidney disease before, and it is one of my personal research interests:  defects in renal primary cilia are the common pathogenesis of renal cyst formation. Despite large advantages in understanding cystic kidney diseases, however, there remains one really large, looming question:  what exactly does the cilia do in normal physiology?  Some studies suggest that the cilium is a rigid antennae, functioning to sense urine flow, but it is still not clear how this connects with the pathogenesis of cystic kidney disease. 

One potential clue to cilia function has come from recent studies which demonstrate that the cilia grows longer in response to kidney injury.  Previously, these results have been limited to animal studies; a recent article in JASN by Verghese et al however illustrates that cilia lengthening occurs in human kidneys subjected to ATN!  Their study design was fairly clever:  they performed cilia staining on serial biopsies of donor kidneys at the time of harvest (prior to any ATN changes having taken place) and then at 6-days post-transplant (after which ATN-like changes were present).  They found a striking doubling in average cilia length following ATN (compare the image on the right to the one on the left; the cilia is the green thread-like structure sticking into the lumen of the tubule).  

Why is cilia length increasing?  Is it simply an abnormal reaction to tubular injury?  Or could increasing the length of the cilium somehow be an adaptive response which aids in the recovery of nephron function?  The answer is not clear from this study alone, but understanding this phenomenon could turn out to be an important step forward in figuring out the precise role of renal cilia.

Acid-Base Issues Related to Eating Disorders

There are some interesting acid-base perturbations that can take place in the setting of eating disorders and/or surreptitious use of substances used to intentionally lose weight. Due to the surreptitious nature of many of these disorders, some lab & physical exam sleuthwork may be necessary to make the diagnosis.

Self-induced vomiting is a common feature of the eating disorders anorexia nervosa and bulimia. Because these individuals are losing HCl from stomach secretions, it is not uncommon for them to develop a metabolic alkalosis. Sometimes physical exam clues can be important in making this diagnosis: vomiting induced by sticking a finger in the back of one's throat can result in scarring on the dorsum of the hand, the formation of oral ulcers, dental erosions due to chronic gastric acid exposure, and puffy cheeks as a result of salivary gland hypertrophy. Furthermore the urine chloride is often profoundly depleted; this is classically a "chloride-responsive metabolic alkalosis."

Surreptitious diuretic use is also surprisingly common, and metabolic alkalosis is felt to derive from multiple contributing mechanisms: secondary hyperaldosteronism often develops due to volume depletion, renal chloride loss, or a contraction alkalosis; chloride-unresponsiveness may also develop due to a profound K depletion which may result from either chronic thiazide or loop diuretic exposure. Tests which assess the concentration of common diuretics in the urine by chromatography are available and may be necessary to cement the diagnosis.

Surreptitious laxative abuse can result in either a non-anion gap metabolic acidosis (similar to patients with chronic diarrhea) OR a hypochloremic metabolic alkalosis which results from hypokalemia, increased renal bicarbonate reabsorption, and volume contraction due to profound loss of sodium and water in the stool. These patients will often present with a fictitious diarrhea, and is found with greatest frequency in females who are related to the health care field.

The Proteinuria Controversy

One of the largest controversies in the field of proteinuria/nephrotic syndrome research derives from a 2007 Kidney International paper published by Russo et al. Briefly stated, the authors suggest the paradigm-shifting idea that the glomerulus filters massive amounts of albumin, and that nephrotic syndrome is a defect in tubular reabsorption of albumin. This flies in the face of decades worth of research on proteinuria, which based on a combination of micropuncture and other physiologic experiments in mice and man has led to the conventional model in which albumin is prevented from entering Bowman's space by the charge-selectivity of the glomerular filtration barrier, and nephrotic syndrome results from a breakdown of this barrier. What are we to make of such a debate? Who is right and who is wrong? The following is a (hopefully unbiased) list of the pros and cons of each side of the argument.

Pro: Normal glomeruli filter nephrotic levels of albumin.

The major piece of data in support of this hypothesis comes from the Russo et al paper in which the authors use a relatively new imaging technique, intravital 2-photon microscopy, which enables in vivo imaging of the kidney using injected fluorescent compounds at a resolution previously unachievable. Essentially, the crucial experiment involved injection of a fluorescently-labeled albumin into the vasculature of rats; by quantifying the degree of fluorescence in the plasma compared to Bowman's space, the authors were able to calculate a "sieving coefficient" for albumin. The sieving coefficient they arrived at was about .02--which is orders of magnitude higher than the previous value obtained via the micropuncture method, about .0006. This implies that nephrotic levels of proteinuria are being filtered from normal glomeruli, and the authors postulate a proximal tubular-based mechanism of rapid albumin reclamation. The nature of such a mechanism is unclear, but the authors suggest that vesicles of intact albumin are transcytosed through proximal tubular cells, providing evidence of this by showing an electron micrograph of endogenous albumin within proximal tubular cells. Finally, in a 2008 JASN article by Dr. Wayne Comper, the author sites various methodologic problems with some of the initial experiments used to demonstrate the "charge selectivity" model of the glomerulus. A recent follow-up paper in JASN by Russo et al furthermore suggests that it is an impaired tubular uptake of filtered albumin which accounts for the changes seen in early diabetic nephropathy.

Con: The case against the "tubular proteinuria" model.

Not surprisingly, this newer model has been met with much resistance. In an article by Christensen et al forcefully entitled, "Controversies in nephrology: renal albumin handling, facts, and artifacts!", the authors describe their opposition to the idea that the glomerulus exhibits such large permeability to albumin. First, they point out several methodologic concerns with the technique of 2-photon microscopy, suggesting that the low fluorescent signal they observed is subject to misinterpretation; perhaps some of the "filtered albumin" seen in Bowman's space is fluorescent bleed-through from nearby blood vessels. The authors also point out a very logical question: if massive amounts of protein are retrieved by proximal tubular cells, why haven't we seen evidence of this throughout decades of research in this field? Many veterans of the field have simply not observed proximal tubular cells chock-full of vesicles containing endogenous albumin, and suggest that the published electron micrographs by Russo et al could be fixation artifact. Furthermore, the authors point out that mice lacking megalin function--thought to be a major player in protein uptake in proximal tubular cells--show only a mild degree of proteinuria, not nearly enough to be consistent with the massive amounts of albumin purported to be filtered by the glomerulus. Finally, virtually all of the mutations identified in patients with congenital nephrotic syndrome target genes known to be important in podocyte function: nephrin, podocin, alpha-actinin 4, etc, which would seemingly point to the podocyte, rather than the proximal tubular cell, as the primary player in regulating proteinuria.

So: Who's right and who's wrong? I don't think we know yet, but for now my money is on the conventional explanation--in general, it seems that the "tubular etiology of nephrotic syndrome" is supported predominantly by a new technique, the limitations of which are not yet fully known, and in order to lend it further credence, alternative techniques which support this theory would be necessary. If it does turn out to be true, however, this would really represent a major paradigm shift in nephrology.

Basic Review: The Renin-Angiotensin-Aldosterone Axis

One of the coolest aspects of the renin-angiotensin-aldosterone system (RAAS) is that it involves multiple organ systems: the liver, lung, adrenal gland, kidney, and vasculature are all prominently involved. It never hurts to review basic physiologic principles, right? Listed are the three main components of the RAAS and their main mechanisms of action.

1. Renin is a peptide hormone secreted from the juxtaglomerular cells of the afferent arteriole in response to 3 main stimuli: (a) renal hypoperfusion, (b) decreased distal chloride delivery to the macula densa, and (c) increased sympathetic activity. Renin antagonists such as aliskiren are presently being tested as antihypertensive agenst with thus far promising results.

2. Angiotensinogen--which is synthesized and secreted from the liver--is cleaved by renin in the systemic circulation to form angiotensin I.

Angiotensin I is cleaved to form angiotensin II by angiotensin converting enzyme (ACE), which is found predominantly within lung endothelium. ACE-inhibitors, as their name implies, targets the ACE enzyme and is one of the most potent anti-hypertensives (and GFR-preserving) therapies available.

Angiotensin II has the following physiologic effects, which it carries out via binding to AT1 and AT2 receptors. Drugs which block the ability of angiotensin II to bind to its receptors ("angiotensin receptors blockers", or ARBs) make up another highly successful and renoprotective antihypertensive therapy. Angiotensin II binding to its receptors have the following major effects:

a) angiotensin II acts as a systemic vasoconstrictor.
b) angiotensin II causes renal efferent arterial vasoconstriction. Acutely, efferent vasoconstriction should increase GFR; however, over time the increased glomerular pressure leads to glomerular damage and, ultimately, renal injury.
c) angiotensin II increases secretion of aldosterone from the zona glomerulosa of adrenal cortex.

3. Aldosterone: in cortical collecting duct cells, aldosterone diffuses into the cell and interacts with the mineralocorticoid receptor, which upon binding translocates to the nucleus and increases expression of ENac. The end result of aldosterone action is sodium reabsorption and potassium & hydrogen secretion. In addition to angiotensin II, hyperkalemia can also stimulate aldosterone secretion. The drug spironolactone interferes with aldosterone interacting with its receptors, and can be effective in the treatment of hypertension.

DDx for Elevated BUN or Cr without Low GFR

Most of the time when we are consulted for an inpatient with a rising creatinine, the assumption is that the kidney is to blame. However, there is an important list of non-renal conditions which can increase either BUN or creatinine independently of low GFR.

An elevated BUN without AKI can occur in the following situations:
1. G.I. bleeding (because digested blood is a source of urea nitrogen).
2. excessive protein intake or a catabolic state.
3. steroid use.
4. TPN.
5. multiple boluses of iv albumin given.

An elevated creatinine (structure pictured above) without AKI can also occur, though less commonly than an elevated BUN without AKI. Recall that a percentage of creatinine clearance occurs via secretion; therefore, drugs which inhibit creatinine secretion such as Bactrim, probenecid, or Pepcid can increase serum creatinine without actually affecting the GFR. Furthermore, intense exercise can transiently increase the creatinine. Finally, some chemicals (e.g., nitromethane, used as a component of types of fuels and solvents) can interfere with the assay by which creatinine is detected in the laboratory.

Stranded on a Lifeboat

We all know that if you are stranded on a lifeboat in the middle of the ocean, you are not supposed to drink the salt water.  Why is this the case--shouldn't the kidney be smart enough to retain the water and excrete the salt?  

It turns out that the high osmolarity of seawater (usually >1000mosm/L) either approaches or exceeds the concentrating capacity of the kidney--and therefore you can't expect to retain any free water.  In addition, seawater contains high concentration of magnesium and sulfate-containing minerals which can result in an osmotic diarrhea when large quantities are ingested; this can exacerbate free water loss which may likely already be high due to high insensible losses from wind and sun.  

Some animals have impressive adaptive mechanisms to maintain homeostasis of osmolarity in a high salt environment;for instance,  the albatross' nasal gland excretes a highly-concentrated salt solution, and the shark rectal gland has the same ability as detailed in another post.  

 

Plus one fun (well, not so fun for the patient involved) link for the hell of it.  The headline says its all:  Tainted Castor Oil Used for Cosmetic Augmentation Causes Kidney Failure, but you can click the link for more juicy details.

alpha & beta intercalated cells of the collecting duct

Heard a great talk today from the 2007 recipient of the Homer Smith Award & editor of Kidney International: Qais Al-Awqati of Columbia. The full details of much of the talk can be found in this JASN article describing his acceptance talk at the ASN, but I will attempt to paraphrase.

The collecting duct has two types of cells: principal cells (the aldosterone-responsive, ENac-expressing cells which also mediate water reabsorption via ADH) and the intercalated cells. The intercalated cells come in two different varieties: alpha-intercalated cells (which secrete acid) and beta-intercalated cells (which secrete base). The alpha-intercalated cells (on the left in the figure) are tall, columnar epithelial cells which contain apical H+ ATPase, explaining its ability to secrete protons. The beta-intercalated cells (on the right) in contrast are shorter, flatter cells which contain an apical chloride-bicarbonate exchanger (called pendrin) which enables it to secrete base.

The interesting thing is that inducing metabolic acidosis results in the conversion of beta-intercalated cells to alpha-intercalated cells--giving the kidney a greater ability to secrete protons and return pH to the normal range. This conversion event is regulated by a secreted extracellular matrix molecule called hensin. Since these cells evidently retain the ability to switch between markedly different cellular phenotypes, it has become of interest to the ever-evolving field of stem cell biology.

Limitations of FENa in CKD

There is an insightful article in this month's C-JASN entitled "Misapplications of Commonly-Used Kidney Equations: Renal Physiology in Practice" by Nguyen et al which is kind of fun to read.

One of the clinical vignettes they present involves the use of the FENa calculation by a medical student to erroneously conclude that a CKD patient had ATN. I see the FENa quite commonly misused even by housestaff, so the discussion is pertinent.

To summarize the case: a woman with CKD and a baseline creatinine of 1.7mg/dL presents to the emergency room with a three day history of diarrhea. She had acute-on-chronic renal failure with a Cr of now 3.4 mg/dL, and despite the fact that her clinical exam was consistent with hypovolemia (orthostatic vital signs, skin tenting), she was erroneously diagnosed with ATN because her FENa was calculated to be 1.45%--above the magical cutoff of 1% we all learn as young internists.

Here's why it's incorrect in this instance. Recall that FENa is simply the ratio of the Na excreted divided by the Na filtered. For the same amount of daily dietary Na intake, A CKD patient will have a much higher FENa than a patient with normal kidneys, as in order to achieve a steady-state they will have to excrete the same amount of Na but using a much lower GFR. The authors calculate that this patient's resting FENa on the typical American diet would be about 2.4%. Thus, the measured FENa of 1.45% probably represents an appropriate Na retention response to volume depletion.

The FENa is most useful in the setting of oliguric acute kidney injury, and in the absence of metabolic alkalosis, significant CKD, or diuretic use. A value <1%>1% suggests ATN.

Page Kidney

In 1939, Irwin Page performed an experiment in a dog model in which one kidney was wrapped in cellophane. The resultant external compression on the kidney resulted in elevated renin-angiotensin-aldosterone levels and severe hypertension. Over time, a loss of renal function occurs due to some degree of renal ischemia and constrictive perinephritis. Because there are high levels of circulating aldosterone, renal K-wasting and hypokalemia can also be present.

Now, "Page Kidney" refers to any condition in which external compression of the kidney leads to activation of the renin-angiotensin-aldosterone system resulting in hypertension. The most common cause of a Page kidney is a subcapsular hematoma, which can result from blunt trauma or iatrogenesis, such as a renal biopsy or ESWL for kidney stones. Rarely, large cysts or tumors can also cause Page kidney.

Renal Adaptations to Hyperkalemia

Are ESRD patients "protected" from the arrhythmogenic effects of hyperkalemia if they are chronically exposed to elevated potassium levels? I'm not sure if there is a clear answer to this question, but it is certainly a relevant question. What do you do with the ESRD patient who comes into the Emergency Room at 7pm after skipping dialysis who has a K of 6.3? On the one hand, dialysis is a sure-fire way to rapidly lower the K and prevent any cardiac complication, which have the potential to be lethal. On the other hand, calling in the dialysis nurse and fellow uses valuable resources, and there is a belief by some that individuals who "live at a potassium in the 5's & 6's" are protected from the negative effects of hyperkalemia. What does the data show?

There certainly appears to be some degree of renal adaptation to hyperkalemia in chronic kidney disease. Early animal studies in which rats were exposed to varying degrees of infused potassium concentrations demonstrate that hyperkalemia dramatically regulates the degree of potassium secretion, primarily via aldosterone-mediated mechanism in the collecting duct. In addition, this 2002 Kidney International review argues that the hyperkalemia seen in CKD is an adaptive, physiologic response, reflecting a new steady state in which extracellular K rises to the level needed to stimulate K excretion so that it matches intake. There is less data as to what adaptive mechanisms are active in the complete absence of renal function, and furthermore, it's hard to know whether or not such adaptations to hyperkalemia would be operative in many ESRD patients where the K fluctuates dramatically with regards to pre- and post-K levels. One might expect different ion channels in the myocardium to be up/downregulated in order to guard against the cardiac effects of hyperkalemia, or perhaps key K transporters in skeletal muscle are upregulated to maintain the plasma K concentration within the appropriate range.

If anybody knows of other relevant studies looking at adaptations to hyperkalemia, let me know.

Basic Renal Anatomy

Corin and the Natriuretic Peptide System

Natriuresis is the process by which sodium is excreted in the urine; this process is regulated (in part at least) by the natriuretic peptides ANP and BNP (there is also a CNP, which appears to have direct effects on smooth muscle cells. Secreted by many organs (e.g., brain, kidney, and peripheral blood vessels) in response to high blood pressure or hypervolemic states, ANP and BNP peptides bind to a family of receptors (there are three: NPR-A, NPR-B, and NPR-C) where they generally work to decrease lower blood volume and blood pressure.

Like many endocrine peptide hormones, ANP and BNP are synthesized as pre-propeptides; after the signal peptide is removed, an additional proteolytic cleavage event is necessary to activate the hormone. It has recently been determined that this cleavage event is mediated by corin, type of serine protease. Two key pieces of information suggest that corin is highly relevant to the field of nephrology and hypertension:

1. Mice lacking corin develop hypertension as a result of decreased conversion of pro-ANP to active ANP.

2. Human polymorphisms in the corin gene appear to be associated with salt-sensitive hypertension.

Perhaps these insights will lead to new drug therapies for the future...

Review: The Juxtaglomerular Apparatus

The juxtaglomerular apparatus (JGA) is located between the afferent arteriole and the returning distal convoluted tubule of the same nephron. It is responsible for regulating both intrarenal (tubuloglomerular feedback) and extrarenal (renin-angiotensin-aldosterone) mechanisms necessary to maintain both renal and entire body volume status.

The three components of the JGA are the following:

(1) the juxtaglomerular cells of the afferent arteriole, synthesize and store renin, which is secreted in response to specific stimuli (e.g., low blood flow, decreased NaCl delivery). The juxtaglomerular cells could be considered the "effector arm" of the renin-angiotensin-aldosterone axis.

(2) the macula densa, a region of the distal convoluted tubule characterized by tubular epithelial cells which are more densely-packed than in other regions of the nephron (and thereby leading to its characteristic appearance on light microscopy). The macula densa can be considered the "sensory arm" of the renin-angiotensin-aldosterone axis in that these are the cells which sense decreased Na Cl delivery which determines downstream function. They are also involved in the mechanism of tubuloglomerular feedback.

(3) mesangial cells, which form connections via actin and microtubules which allow for selective vasoconstriction/vasodilation of the renal afferent and efferent arterioles with mesangial cell contraction.

The Brenner Hypothesis

The Brenner Hypothesis--developed by Barry Brenner of the Brigham and Women's Hospital, author of the well-known Brenner & Rector "The Kidney" textbook--states that individuals with a congenital reduction in nephron number have a much greater likelihood of developing adult hypertension and subsequent renal failure. This hypothesis is supported by the observation that there is a strong epidemiologic relationship between intrauterine growth retardation (IUGR)/low birth weight and adult hypertension. The mechanistic explanation for this phenomenon is that compensatory hyperfiltration by the remaining nephrons--similar to what happens in diabetic nephropathy--results in accelerated renal decline. While this theory is appealing at several levels, it does not explain everything: for instance, why most kidney transplant donors, who instantaneously lose 50% of their nephron mass, for the most part do well post-transplant without developing hypertension or renal disease.

The Brenner Hypothesis is part of a larger body of work which posits that many common adult diseases (including diabetes and hypertension) are caused by factors which are established during early growth and development, the "fetals origins of adult disease" hypothesis.

How To Measure Renal Plasma Flow (RPF)

Question: How is renal plasma flow (RPF) measured?

Answer: By measuring para-aminohippurate (PAH) clearance.

How does it work? Ideally, one would simply choose a substance which is neither synthesized nor metabolized by the kidney; the amount of such a substance entering the kidney per unit time via the renal artery should equal the amount leaving via the renal vein. However, it is not trivial to obtain blood samples from the renal vein in humans and therefore another approach is needed.

PAH works because although it is not filtered at the glomerulus, it undergoes tubular secretion in an extremely efficient manner: about 90% of all PAH is removed from the plasma via secretion while 10% remains in the renal vein. This 10% is ignored, and it is then assumed that the amount entering the kidneys equals the amount showing up in the urine. Mathematically, this can be expressed as the formula:

RPF (in cc/min) x [PAH] in plasma = [PAH] in urine x urine flow rate V (in cc/min).

Rearranging, RPF = [PAH] in urine x urine flow rate V (in cc/min)/[PAH] in plasma.

Although this test is rarely used in clinical practice, it is essential for research studies in which the hemodynamic effects of various drugs can be measured. For example, ACE-inhibitors are associated with a measurable increase in RPF.