OUP user menu

The V2 vasopressin receptor mutations and fluid homeostasis

Mariel Birnbaumer
DOI: http://dx.doi.org/10.1016/S0008-6363(01)00337-6 409-415 First published online: 15 August 2001

Abstract

Although three different G-protein coupled receptors have been identified for arginine vasopressin, a significant physiological role has been recognized only for the V2 subtype that controls water homeostasis. Identification of the gene encoding the V2 vasopressin (or antidiuretic hormone) receptor enabled researchers to test the hypothesis that mutations of this gene were responsible for X-linked recessive nephrogenic diabetes insipidus. The affected patients are unable to concentrate their urine and as a consequence live in constant danger of dehydration that can cause death, particularly in infancy, or lead to severe hypernatremia that can impair their intellectual and physical development. The danger of severe dehydration diminishes in the adult patients, although they remain highly susceptible to this condition for the rest of their lives.

Keywords
  • Diabetes
  • Hormones
  • Intra/extracellular ions
  • Receptors
  • Vasoactive agents

Time for primary review 7 days.

1 Introduction

The role of the antidiuretic hormone (ADH or arginine vasopressin, AVP) in maintaining water homeostasis has been known for many years. Although the name of the peptide stems from the pharmacological effect of posterior pituitary extracts from which the peptide was later extracted and identified, the possible role of AVP on blood pressure control has never been characterized. The advances in molecular endocrinology of the last 10 years provided significant information about the receptors that mediate cellular responses to AVP. Three different G-protein coupled receptor subtypes have been identified for AVP receptor subtypes designated V1a, V1b and V2 [1–4].

The V1a and V1b receptors activate G proteins of the Gq/11 family; thus, in response to AVP, they increase the hydrolytic activity of phospholipase Cβ on phosphatidyl-inositolbisphosphate. The products of this enzymatic reaction are inositol trisphosphate and diacylglycerol that in turn promote release of Ca2+ from intracellular stores and stimulate the activity of protein kinase C, respectively [5]. The V1a receptor is ubiquitous and particularly abundant in liver and vascular smooth muscle cells. The cellular effects mediated by the V1aR are similar to those of the α adrenergic receptors present in liver cells and the ATI angiotensin receptors present in smooth muscle cells, among others. The presence of additional receptors with identical signaling properties in the same cells has hindered the ability of researchers to identify the physiological role of the V1a receptor, if any, in those tissues. It has been easier to assign a physiological role to the V1b receptor that is present in the corticotrophs of the anterior pituitary and can mediate the AVP-stimulated release of ACTH. In these pituitary cells, AVP enhances the release of ACTH promoted by the corticosterone releasing factor (CRF) through its receptor [6]. Though distribution of the V1b receptor is more restricted than that of the V1a, the presence of the CRF receptor in the corticotroph suggests that the V1bR is also physiologically dispensable. The third receptor is the V2 subtype that activates the heterotrimeric G protein Gs promoting a stimulation of adenylyl cyclase activity.

2 The V2 vasopressin receptor

The V2 receptor is expressed exclusively in the principal cells of the collecting duct, the last portion of the nephron. It is co-expressed with the cAMP regulated water channel aquaporin2 and plays a key role in the maintenance of water homeostasis [7–10]. In the presence of low levels of circulating AVP the luminal surface of the collecting duct has minimal permeability to water, a condition that changes drastically when the blood levels of AVP increase. Fig. 1 provides a conceptual summary of the cellular events that take place in the principal cell of the collecting duct. These cells have V2 receptors and the non-regulated water channels aquaporin3 and aquaporin4 on their basolateral surfaces. They also synthesize the AVP-regulated water channel aquaporin2 (AQP2). Most AQP2 molecules are synthesized and stored in small vesicles called ‘aggregophores’ that accumulate in the cytoplasm near the apical surface of the principal cells [9,11,12]. Upon vasopressin stimulated increases in cellular levels of cAMP, there is an increase in the catalytic activity of protein kinase A, and the phosphorylation cascade that follows promotes the movement of the aggregophores toward the apical surface followed by fusion with this membrane. These insertions cause a massive increase in the number of AQP2 molecules present in the luminal surface of the nephron and a fast and significant increase in the water permeability of the apical plasma membrane. The increase in permeability allows water reabsorption driven by the differences in salt and urea concentration between the hypotonic content of the collecting ducts and the iso-osmotic interstitium of the cortex or the hypertonic interstitium of the kidney medulla [13]. The blood levels of AVP are tightly regulated by osmolar sensors located in the third ventricle that inhibit hormone release once water reabsorption has restored blood osmolality to normal values [14]. AVP also plays a role in the establishment and maintenance of hyperosmolarity in the medullary interstitium. It enhances Na+ transport in the thick ascending limb of the loop of Henle, a major contributor to the establishment of the counter current multiplication system and increases the activity of the urea transporter in the inner medullary collecting duct, probably by interacting with a V2 type receptor [15]. Cranial trauma and surgery that alter the regulation of vasopressin release have profound consequences for fluid homeostasis, demonstrating that the most important physiological effects of AVP are those mediated by the V2 receptor.

Fig. 1

Principal cell of the kidney collecting duct. The intracellular localization of the elements participating in urine concentration in the last segment of the nephron are shown. The V2 vasopressin receptor (V2R) located in the basolateral cell surface binds to AVP present in the interstitial fluid, and activates the heterotrimeric protein Gs. Activated Gs increases the activity of the membrane-bound enzyme adenylyl cyclase (AC) causing an augmentation of the intracellular levels of cAMP. Increased levels of cAMP stimulate the activity of the cAMP-dependent protein kinase A (PKA), triggering a yet undefined phosphorylation cascade that promotes the insertion of the ‘aggregophores’ containing aquaporin2 (AQP2) into the apical membrane of the cell. AVP-regulated AQP2 increases the water permeability of the apical membrane and allows the reabsorption of water from the hypotonic processed filtrate into the surrounding hypertonic interstitium. The water can exit the cell through the aquaporin3 and aquaporin4 water channels, constitutively present in the basolateral surface of the cells.

3 Nephrogenic diabetes insipidus

A dramatic alteration of body water homeostasis is observed in individuals suffering from diabetes insipidus, a disorder characterized by the inability of the kidney to concentrate urine. Diabetes insipidus may be caused by reduced production of AVP (neurogenic DI) [16,17], or by the inability of the kidney to respond to circulating AVP (nephrogenic DI) [18]. Adult onset nephrogenic DI can be triggered by toxic damage to the kidneys, whereas the neonatal onset is due to the lack of activity of a defective gene. In approximately 5% of the cases, the gene encoding the AVP regulated water channel (aquaporin 2) is affected and the presentation is autosomal recessive with some instances of autosomal dominant inheritance [19]. The majority of the nephrogenic DI patients present with the X-linked recessive form distinguished by the almost exclusive appearance of symptoms in males. Isolation of the cDNA encoding the V2 receptor provided a crucial step toward identifying the molecular basis of this disease.

4 Cloning of the human V2R

When the cloning project was started, there was no information about the amino acid composition of the V2R, and as a consequence it was not possible to design olinucleotides for library screening. Only the cDNAs encoding rhodopsin and some of the neurotransmitter receptors were known, and the similarity between the transmembrane domains of these receptors and those of the V2R was insufficient to obtain partial cDNA segments by the polymerase chain reaction. Thus, the only method feasible was expression cloning. This cloning methodology relied on the abundance of the Alu I repetitive sequences in the human genome and it had been previously employed in the isolation of the human transferrin and nerve growth factor receptor cDNAs [20,21]. The methodology, summarized in Fig. 2, requires the use of a recipient cell line from a mammalian species containing a different set of repetitive sequences (i.e., no Alu I sequences), and the ability to identify the transfected cells that are expressing the gene of interest. The well established Ltk murine cell line was the appropriate recipient cell, bearing a different set of repetitive sequences. The identification of the cells expressing the receptor of interest had been previously accomplished with good quality antibodies against the human transferrin and nerve growth factor receptors, but this tool was not available to identify the V2R. Ltk cells were stably transfected with human genomic DNA along with a selectable marker plasmid containing the thymidine kinase gene, then exposed to growth medium containing hypoxhantine, aminopterin and thymidine (HAT) to select for successfully co-transfected cells. The resulting clones were analyzed for receptor expression by measuring in-situ adenylyl cyclase activity in the presence of AVP [22]. Once identified, the responsive cells were cloned and used to extract genomic DNA that was transfected into naive murine L cells. Each transfection step further diluted the human genomic DNA in the background of murine DNA [23]. After two additional consecutive transfections, it was possible to identify a ‘tertiary’ transformed cell line that expressed the V2R and contained a significantly reduced amount of human repetitive sequences. A lambda gt11 library was prepared with genomic DNA from these cells and screened by standard molecular biology techniques using radioactively-labeled human genomic DNA as the probe. The abundance and slight polymorphisms of Alu I sequences made human genomic DNA the probe of choice for the screening. A small number of positive phages were tested for the presence of the V2R gene by co-transfection with the thymidine kinase gene into the murine L cells. Even at this point, the identification of the receptor gene relied on the ability of the isolated segments to express the protein in transfected fibroblasts. Subsequent fragmentation identified a 2.2 kb fragment of human genomic DNA that encoded the receptor and this segment was used to identify a cDNA from a human kidney library. The cloning process was completed applying standard protocols [4], and the V2R gene was located to the q28–qter region of the human X-chromosome [24]. During the grueling struggle to bring the project to completion, the prospect of testing whether X-linked recessive nephrogenic diabetes insipidus was due to mutations in the V2R gene bolstered our determination.

Fig. 2

Expression cloning of the human V2R. A diagram of the steps involved in expression cloning of the human V2 receptor is shown. For further details consult references [4,21,22].

5 Identification of disease causing mutations

Once the composition of the V2 gene was available, the next step was to sequence the gene present in patients affected with nephrogenic diabetes insipidus. This phase of the research was expedited by the existence of a well studied group of patients affected with NDI. It was necessary to find a clinician interested in NDI that could provide blood samples of properly diagnosed patients. Daniel Bichet, a french nephrologist residing in Canada, had been intrigued by the possible correlation between X-linked NDI and the V2 receptor. He was not unique in that regard, but what was unique was the dedication with which he had collected blood samples from affected families and carefully documented their clinical parameters and pedigrees [25–27]. As a consequence, genomic DNA samples from well characterized families were waiting to be analyzed.

The analysis of the V2R gene was carried out on DNA segments amplified by the polymerase chain reaction using genomic DNA as template. The DNA was analyzed by sequencing, an approach facilitated by the small size of the coding region (2200 bp), and the first mutations were identified within a short time [28–30]. The decision to search for mutations by sequencing the gene instead of applying single-strand polymorphisms, a faster alternative method to identify mutations, was based on the higher reliability of the sequence analysis. Single strand polymorphism analyzes the migration of single-stranded DNA segments obtained by amplifying the gene under study from genomic DNA of healthy or affected individuals. The nucleotide sequence of the V2R gene is uncommonly rich in guanosine and cytosine nucleotides that cause the atypical migration of many DNA segments of the gene, a peculiarity that created significant difficulties during the elucidation of the composition by DNA sequencing. This experience guided our choice of complete nucleotide sequencing of the amplified fragments, and it is likely that similar difficulties may be encountered when analyzing other GC rich regions of the genome.

The majority of the mutations in the V2R gene are caused by base substitutions; both transitions (A↔T) and transversions (A or T↔G or C) have been identified. Base deletions or insertions have been found at lower frequency. Fig. 3 is a two-dimensional representation of the primary amino acid composition of the V2R, including the topological distribution of the different segments of the receptor. The open circles represent mutations identified in patients affected with the full NDI phenotype while the grey diamonds identify mutations that cause a partial NDI phenotype. The V2R gene resides in the human X chromosome near the genes that encode the rhodopsin pigments that mediate the perception of red and green color, two genes notorious for the great number of mutations they can suffer, as attested by the frequency of color blindness among males in the general population. Cumulative experience suggests that this region may be very susceptible to mutagenesis, opening the possibility that there may be other mutations of the V2R that remain unrecognized as long as they do not interfere with protein function and yield a phenotype.

Fig. 3

Location of V2R mutations causing NDI. A two dimensional representation of the V2R identifies the topology of the receptor across the membrane. The location of the missense mutations that result in a full NDI phenotype is shown by the ○; the location of the mutations that cause a partial phenotype is shown by the gray diamonds.

6 Molecular basis of nephrogenic diabetes insipidus

6.1 Reduced receptor expression

As represented in Fig. 3, the disease-causing mutations of the V2R seem clustered in the regions of the protein predicted to traverse the lipid bilayer. This apparent preferential distribution reflects the importance of preserving the amino acid composition of the protein to minimize interference with protein folding and transmembrane movement. Acquisition of the native conformation is required to generate the hormone-binding domain and to define the topology of the intracellular surfaces that interact with the heterotrimeric Gs protein facilitating the release of GDP and the subsequent entry of GTP into the nucleotide binding site. Appropriate folding of the receptor protein also plays a role in the threading of the molecule through the bilayer, with changes in the identity of the side chain hindering the successful transport of the protein to the surface of the cell. When expressed in transfected cells in the laboratory, most V2 mutant receptors fail to reach the cell surface, or do so in greatly reduced numbers [31]. Two independent lines of evidence have been used to examine this issue. One of them examined biochemically the state of glycosylation of the receptor, the other followed by confocal microscopy the progress of the V2R from the endoplasmic reticulum where it is synthesized to the plasma membrane of the cell [31].

The V2R acquires a core mannose complex on asparagine 22 (N-linked sugars), as a co-translational modification in the endoplasmic reticulum and serine and threonine linked (O-linked) sugars while traversing the last portions of the Golgi network [32,33]. The core mannose is first trimmed by enzymes present in the endoplasmic reticulum and then modified in the Golgi compartment where modified sugars are added to the mannose backbone, increasing its complexity. Before the addition of sialic acid in the Golgi compartment, the sugar complex can be cleaved from the protein by endoglycosidase H, whereas after the addition of sialic acid, it becomes resistant to this enzyme. Thus, a simple biochemical analysis testing for endo-H sensitivity can distinguish between proteins that can be successfully transported from the endoplasmic reticulum to the Golgi, and then to the cell surface and those that remain trapped in the ER or early Golgi. In addition to the process of the asparagine linked sugar, the Golgi compartment adds O-linked sugars to the amino terminus of the V2R dramatically altering the migration characteristics of the protein and facilitating the identification of the protein molecules that traverse the Golgi compartment and those that do not. Application of this analysis revealed that most mis-sense mutations reduce dramatically the number of receptors present on the cell surface, decreasing the sensitivity of the transfected cells to vasopressin by more than one hundred fold [34–36]. In addition, it is likely that the accumulation of improperly processed receptor in the principal cells could impair the appropriate synthesis of proteins required for their well-being.

6.2 Reduced receptor activity

The three diamonds shown in Fig. 3 identify three mutations that cause only a partial NDI phenotype characterized by a reduced response of the kidney to circulating AVP. The response approaches normality under mild dehydration when the blood levels of AVP are high. Patients affected by these phenotypes are typically diagnosed in late infancy or adolescence, and the predominant presenting symptom is failure to thrive. They do not suffer extreme polyuria, and often the first laboratory indication of their problem is the persistent lower value of urine osmolality. The D85N mutation was the first one identified with these characteristics. The only difference between wild type and mutant receptor is the presence of an amide group instead of a free carboxyl group of aspartic acid. This conserved amino acid change does not interfere with protein folding as assessed by ligand binding assays; cells expressing the wild type or the D85N mutant V2R are able to deliver the same number of receptors to the cell surface. Scatchard analysis of saturation binding assays performed with tritiated AVP detected a mild reduction in the affinity for AVP, whereas examination of the coupling efficiency of the mutant receptor to Gs as assessed by the method of Barber et al. [37], revealed that the mutant receptor was significantly impaired in its ability to activate the G-protein. It was determined in the same experiments that the G201D mutation reduced the number of receptor molecules that could be transported to the cell surface to 20% of the values detected for the wild type and D85N mutant V2Rs.

The G201D mutation slightly reduced the AVP binding affinity of the receptor but did not affect coupling efficiency to Gs. The retention of stimulatory activity in the presence of reduced levels of receptor expression explained the mild manifestation of NDI in these patients [38]. Although less well characterized until now, the P322S mutation that is associated with the mild NDI phenotype exemplifies the importance of maintaining proper receptor structure. Detailed analysis of the biochemical defect reducing the activity of this mutant receptor has not been reported; thus, the parameter that reduces vasopressin response has not yet been identified. The substitution of proline 322 by histidine was identified in patients from a different family bearing an inactive receptor that results in a full phenotype [39]. These findings confirmed the importance of the side chain of the amino acid replacing proline 322 for the preservation of structure in the V2R, and reinforced the notion that the impact of a missense mutation differs depending on the identity of the substituting amino acid.

7 Management of NDI patients

Diet modification and medications can be employed to reduce the volume of urine excreted by NDI patients. This strategy diminishes the symptoms and the danger of dehydration, and helps to control the excessive thirst. The recommended diets contain very low quantities of sodium and reduced quantities of protein and fat. Children's diets emphasize the reduction of salt intake and maintaining protein consumption at levels compatible with normal growth. Daily treatment with indomethacin and thiazides contribute to diminish urinary output by mechanisms that are not understood. These drugs were first shown to be effective in reducing the urinary volume of the Brattleboro rat which cannot release AVP from the posterior pituitary. Human trials demonstrated that a similar effect is achieved in humans. Recently, researchers have considered the possibility of using drugs that help the transit of mis-folded proteins to the cell surface to ameliorate the symptoms of diseases characterized by a reduction in the cell surface abundance of membrane proteins. Limited success has been reported in experiments analyzing the expression of disease associated mutant forms of the cystic fibrosis transmembrane conductance regulator (CFTR) in cultured cells. Similar compounds to those used for the CFTR experiments, as well as non-peptidic AVP analogues, are currently under investigation [40]. These experimental strategies seed hope for the future, although the translation of treatments designed for cells in culture to humans is bound to be a long process. The other approach has been to apply molecular biology techniques to attempt to rescue the proteins that are trapped in the intracellular compartments. These experiments have achieved some success when applied to cells in culture [41,42], but the treatment of NDI patients with similar methods faces the same daunting difficulties that affect all gene therapy protocols.

8 Water homeostasis and cardiovascular disease

The optimal physiological balance between cardiac and kidney function is seriously disrupted when cardiac output is reduced and the normal response of the nephrons to a reduction in sodium flow through the distal tubule increases renin release. The subsequent increase in circulating angiotensin II followed by augmented levels of aldosterone are aimed at restoring plasma volume but have the undesirable effect of increasing fluid retention, thus overloading the already impaired cardiovascular system [43]. In many cases the progression of this imbalance is worsened by the inappropriate secretion of AVP from the pituitary that promotes excessive water retention and hyponatremia. Currently furosemide- or amiloride-type diuretics are the agents of choice to diminish the volume of extracellular fluid in patients suffering congestive heart failure. The availability of an effective V2 receptor antagonist would expand the available choice of treatment. These compounds, able to cause selective loss of water, would mimic in the patient the symptoms of a mild NDI phenotype and allow the reduction of extracellular volume and management of hyponatremia by promoting loss of water in the last portions of the collecting tubules and medullary collecting ducts.

Previous efforts to develop these compounds used rats and other experimental animals to test lead compounds. Unfortunately, the differences in composition between the V2 receptor from different species caused significant pharmacological differences that destined these projects to failure. Cloning the cDNA of the human V2R and the development of kidney-derived cell lines expressing the receptor introduced the badly needed reagent to identify lead compounds and guide the efforts of the pharmacologists towards the development of an efficient antagonist of the human receptor.

The first non-peptidic compounds were identified by a Japanese company, and since then a variety of V2 antagonists have become available in Europe and the United States [44–47]. Most AVP antagonists are still at the developmental stage, but hopefully their incorporation into the arsenal of drugs used to manage fluid homeostasis in congestive heart failure should become a reality in the near future.

References

  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
  32. [32]
  33. [33]
  34. [34]
  35. [35]
  36. [36]
  37. [37]
  38. [38]
  39. [39]
  40. [40]
  41. [41]
  42. [42]
  43. [43]
  44. [44]
  45. [45]
  46. [46]
  47. [47]
View Abstract