Hypertension genetics is of interest to different health care professionals: The clinician is often embarrassed by patient questioning on the origins of the blood pressure (BP) elevation in the absence of risk factors and in the clinic, signs indicating the presence of a rare monogenic hypertensive syndrome are important to be recognized. The clinical-trialist can find proof for causality between BP and for example, target organ damage in Mendelian randomization studies. BP is of interest to the scientist in genetic or genomic medicine as it is a classic quantitative trait in the population and monogenic disease in rare families.
Hypertension (HTN) or BP genetics has been proceeding at two separate paces for primary hypertension and the rare familial forms of monogenic hypertension. The former requires genotyping of hundreds of thousands of variants that only became practical with microarrays and the implementation of genome-wide association studies (GWAS). Genes underlying monogenic family traits can be identified with a few hundred genetic markers, and the identification of causal genes was therefore feasible much earlier. Both types of experiments have largely contributed to our understanding of the architecture of BP genetics.
The Contribution of Genetics to the Blood Pressure Distribution
The contribution of genetics to the BP distribution is of two types: Rare mutations segregating in families drive up BP substantially in many cases and make affected individuals outliers in the BP distribution. This is secondary hypertension caused by single genes and is discussed in more detail in the first part of this chapter. The first such defect was described in 1991 and the latest was described in 2015, such monogenic hypertension is a typical example of classic medical genetics.
On the other hand, the distribution of systolic BP (SBP) and diastolic BP (DBP) in the general population has a skewed, but otherwise close to normal distribution, and is a classic quantitative trait. BP in the general population has surprising high heritability at 30% to 50%, opening an opportunity to an improved understanding of the interindividual differences in BP levels by understanding the origins of the heritability observed. The nature of the genetic architecture of primary HTN has been the subject of the combative controversy between Robert Platt and George Pickering around 1950, where Dr. Platt advocated a monogenic dominant disease and Dr. Pickering multigenic inheritance and a continuous trait. Today Dr. Pickering’s model of primary hypertension is clearly documented by a large body of data. Because HTN is defined as an arbitrary threshold of BP, causes that explain the interindividual variability of BP values also explain HTN (or primary hypertension when other specific causes of HTN are excluded). BP (continuous phenotype) is preferred over HTN (dichotomous phenotype) in many genetic experiments because the use of a continuous phenotype has greater precision and therefore greater statistical power. The second part of this chapter will describe in more detail the advances made over the last decade to better describe the genetic architecture of primary hypertension.
Monogenic (Secondary) Hypertension
Monogenic hypertension should be considered secondary hypertension because an underlying genetic defect is clearly identifiable. The genetic defects that are necessary and sufficient for monogenic hypertension have distinctive characteristics that make them different from genetic variants underlying primary hypertension ( Table 6.1 ). Eight different monogenic hypertensive syndromes (MHS) have been described and are summarized in Table 6.2 . Three MHS have typically elevated aldosterone levels and are listed above the two MHS with typically low aldosterone. Three additional MHS have special features (occurring in pregnancy, brachydactily, or virilization features). Among the three groups there is considerable overlap.
| Characteristic | Monogenic Hypertension | Primary Hypertension |
|---|---|---|
| Allele frequency in the population | rare (<1/1000) | ∼30% |
| Effect size per genetic variant | Large (likely average ∼20 mm Hg) | Small (average ∼0.5-1 mm Hg so far) |
| Total number of known genes (loci) involved | 13 | ∼90 |
| Estimated number of all genes (loci) involved | Likely ∼15-20 | >500 |
| Short Disease Name | Complete Disease Name | Omim Number | Genes | Renin Blood Level | Aldosterone Blood Level | Inheritance |
|---|---|---|---|---|---|---|
| Elevated Aldosterone | ||||||
| GRA | glucocorticoid remediable aldosteronism = familial hyperaldosteronism type I = glucocorticoid suppressible hyperaldosteronism | #103900 | CYP11B2 | ↓ | ↑ | AD |
| Gordon syndrome | = pseudohypoaldosteronism type II (PHA2) = Gordon hyperkalemia-hypertension syndrome = familial hyperkalemic hypertension (FHHt) | %145260 | WNK1, WNK4 KLHL3 CUL3 | ↓ | ↑ | AR and AD |
| FH III | Familial hyperaldosteronism type III | #613677 | KCNJ5 | ↓ | ↑ | AD |
| Low Aldosterone | ||||||
| Liddle syndrome | = pseudoaldosteronism | #177200 | SCNN1B , SCNN1G | ↓ | ↓ | AD |
| AME | cortisol 11-beta-ketoreductase deficiency = syndrome of apparent mineralocorticoid excess | #218030 | HSD11B2 | ↓ | ↓ | AR |
| Low Aldosterone and Associated Features | ||||||
| HTNB | hypertension and brachydactyly syndrome = Bilginturan syndrome | #112410 | PDE3A | ↓ | ↓ | AD |
| Autosomal dominant hypertension with exacerbation in pregnancy | hypertension, early-onset, autosomal dominant, with exacerbation in pregnancy | #605115 | NR3C2 | ↓ | ↓ | AD |
| CAH | CAH type IV (congenital adrenal hyperplasia, because of 11-beta-hydroxylase deficiency) and CAH type V (congenital adrenal hyperplasia, because of 17-alpha-hydroxylase deficiency) | #202010 #202110 | CYP11B1 CYP17A1 | ↓ | ↓ | AR |
Even collectively, monogenic familial hypertension is thought to be rare with an incidence of likely below 1/5000 in the general population. But these estimations have been challenged and pathologic mutations might occur more frequently than previously thought, definite proof of significance of these genes for the general population is outstanding. Even though likely rare, the genetic variants underlying MHS are important in two respects:
- 1.
For the occasional patient with hypertension who carries a pathogenic monogenic hypertension variant, the recognition of the syndrome is important because in some cases, specific treatment approaches exist that can have spectacular treatment effects and because the recognition of the familiarity makes cascade screening possible. In MHS, untreated hypertension is often very elevated and can be severe with target organ damage occurring early in life, precocious death by stroke is observed in some cases.
- 2.
It is without question that the pathways and mechanisms illuminated by the defects induced by monogenic hypertension have permitted great advances in the understanding of general BP pathways. All but one monogenic hypertension gene act either in the kidney or in the steroid metabolism or at the mineralocorticoid receptor ( Fig. 6.1 ). The one exception is the latest identified member of the monogenic hypertension genes, PDE3A, a phosphodiesterase that likely mediates the hypertensive effect in the vasculature. Many of the 13 genes in which mutations can cause monogenic hypertension have been described by the group of Dr. Richard Lifton and consequently the genes are also referred to as “Lifton genes.” Gene mutations found in families leading to low blood pressure have also been described and these are not discussed in more detail here. Note that although classically the renin levels are always low and aldosterone levels high for some entities and low for others, levels are often borderline or normal. Features that should prompt the clinician to suspect a monogenic form of hypertension are summarized in Table 6.3 and the family history is of particular importance. Once a monogenic hypertension syndrome is identified, there are special treatment approaches available for some forms that permit, in general, to obtain large treatment effects. The entities in which specific treatment is possible are summarized in Table 6.4 .
FIG. 6.1
Tissue and pathway localization of monogenic hypertension genes. Blue stars indicate the location of monogenic hypertension genes.
(From Ehret GB and Caulfield, MJ. Genes for blood pressure: an opportunity to understand hypertension. Eur Heart J. 2013;34:951-61.)
TABLE 6.3
Clinical Recognition of Monogenic Hypertension
Characteristic
Typically Encountered in Monogenic Hypertension
Renin level
Always low
Family history
Usually positive for early-onset hypertension
Patient age
Usually young
Blood pressure elevation
Often important
TABLE 6.4
Monogenic Hypertension Responsive to Special Treatment
Monogenic Hypertensive Disease
Treatment with Usually Large Effect
GRA
Glucocorticoid at physiologic doses or mineralocorticoid receptor antagonist
Gordon syndrome
Low-salt diet or thiazide
Liddle syndrome
Amiloride or triamterene
AME
High doses of mineralocorticoid antagonists, glucocorticoids (long term treatment with important side effects)
Glucocorticoid-Remediable Aldosteronism
Through unequal crossover, a chimeric gene is formed between portions of the 11-beta-hydroxylase gene and the aldosterone synthase gene in such a unique way that adrenocorticotropic hormone (ACTH) stimulates aldosterone synthesis. Similar to other monogenic hypertensive disease, the pattern of inheritance is autosomal dominant (see Table 6.2 ) and therefore the disease is usually readily apparent in families. Hypertension is often observed at a young age, in one study all affected members of a large pedigree were diagnosed with hypertension before the age of 21 and hypokalemia is not usually present. The diagnosis can be made by demonstrating the overproduction of the cortisol C-18 oxidation products in the urine. When defining the disease by criteria based on steroids, it is rare with about 100 cases described worldwide, but affected individuals might have mild hypertension and normal electrolyte levels, making the entity difficult to distinguish from primary hypertension, potentially leading to underdiagnosis. The therapeutic approach is a physiologic dose of an intermediary-acting glucocorticoid (e.g., prednisone) administered at bedtime to suppress the early morning surge of ACTH. An alternative approach is treatment with mineralocorticoid receptor antagonists that may be just as effective and avoids the potential disruption of the hypothalamic-pituitary-adrenal axis and risk of iatrogenic side effects.
Gordon Syndrome
Clinical hallmarks of this entity are hypertension, hyperkalemia, and metabolic acidosis. Because of the hyperkalemia, aldosterone levels are classically elevated despite the volume overload. Around 100 individuals with Gordon syndrome have been reported worldwide, the precise prevalence is unknown. In one large French pedigree all affected adults were hypertensive whereas all affected children had normal blood pressure. The mean age of hypertensives with Gordon syndrome was 27 years in another report. The causal mutations for Gordon syndrome have been in part only recently identified: Mutations of the genes encoding the WNK kinases 1 and 4 or the KLHL3 and CUL3 genes result in increased chloride and sodium reabsorption in the kidney with consequent volume expansion. The increased chloride reabsorption leads to potassium retention and hyperkalemia through a reduction in luminal electronegativity. Blood pressure can usually be rapidly corrected by thiazide diuretics or, more slowly, by a low-salt diet.
Familial Hyperaldosteronism Type III
This entity is very rare and is because of loss of function mutations in the potassium channel KCNJ5 (inwardly-rectifying channel, subfamily J, member 5). Pathogenic mutations result in membrane depolarization of the zona glomerulosa in the adrenal cortex, opening of voltage-activated calcium channels triggering inappropriate aldosterone biosynthesis. The pattern of inheritance is dominant. Typically, there is severe hypokalemia with hypertension in childhood and elevated aldosterone blood levels. Enlarged adrenal glands can be observed, in part with massive enlargement. Treatment is either medical, identical to other cases of primary hyperaldosteronism, or surgical.
Liddle Syndrome
This is also a rare entity with close to 100 cases reported in the literature worldwide. The causal defect is a gain-of-function mutation in one of two subunits (SCNN1B, SCNN1G) of the ENaC sodium channel. Pathogenic mutations lead to a large increase in sodium transport and loss of inhibition of channel activity by elevated levels of intracellular sodium. Hypertension is usually severe with an onset in young adulthood, but some cases are only diagnosed later. Usually there is prominent hypokalemia, metabolic alkalosis, and associated low plasma renin and aldosterone levels, but some patients do not have all those characteristics and can have near to normal potassium levels. Children usually have normal blood pressure. The diagnosis is suspected based on the typical clinical, blood, and urinary features and a positive family history (dominant inheritance, see Table 6.2 ), although sporadic cases have been described. Diagnostic confirmation is by genetic testing. Targeted treatment options are amiloride or triamterene that inhibit ENaC activity. Other antihypertensive therapies are largely inefficient. The patients should also follow a low-salt diet.
Syndrome of Apparent Mineralocorticoid Excess
This is one of the rare recessive forms of monogenic hypertension, with less than 100 cases reported worldwide. The causal defects are loss-of-function mutations or deletions of the HSD11B2 gene, encoding for the kidney isoform of the 11-β-hydroxysteroid dehydrogenase usually responsible for the conversion of cortisol to cortisone, permitting cortisol to activate the mineralocorticoid receptor. Cortisol has an affinity for the mineralocorticoid receptor similar to aldosterone and is therefore converted to cortisone at aldosterone sensitive sites such as the kidney. Clinically the disease is of usually very early onset with severe hypertension, low renin and aldosterone levels, hypokalemia, alkalosis, and often nephrocalcinosis. Strokes in children with apparent mineralocorticoid excess have been observed. But much milder forms also exist that manifest themselves later in life. As treatment approaches spironolactone or other mineralocorticoid antagonists are usually used at high doses, combined with thiazides to prevent nephrocalcinosis. The mineralocorticoid blockers other than spironolactone might be preferable because less elevated doses can be used with fewer side-effects. Exogenous glucocorticoids can be used to decrease the endogenous secretion of cortisol, but a long term treatment has important side effects. Often additional nonspecific antihypertensive medications are required. The syndrome of apparent mineralocorticoid excess can be mimicked by the chronic ingestion of high doses of licorice: glycyrrhetinic acid, contained in licorice, inhibits the 11-β-hydroxysteroid dehydrogenase.
Autosomal Dominant Hypertension With Brachydactyly
Initially described in 1976, the underlying gene for this syndrome has been identified very recently. Activating mutations in the cGMP-inhibited phosphodiesterase 3A (PDE3A) gene appear to lead to inhibition of phosphorylation-dependent vasodilation. The typical clinical presentation of severe, age-dependent hypertension is associated with brachydactyly type E (short fingers, predominantly because of malformation of the metacarpal bones). Affected family members are reported to die frequently before the age of 50 of stroke. Neurovascular compression as cause for the hypertension had been postulated, but remains unproven. Currently there is no specific treatment known for this syndrome.
Early-Onset Autosomal Dominant Hypertension With Exacerbation in Pregnancy
This condition is extremely rare with one pedigree identified so far. The defect has been mapped to the mineralocorticoid receptor gene (NR3C2) and activating mutations induce spontaneous activity and nonspecific activation of the receptor. Affected individuals present with early-onset hypertension, before 21 years in the initial description, and hypertension is largely exacerbated during pregnancy in women. Although delivery improves hypertension related to the condition in pregnancy, the BP elevation is largely resistant to standard antihypertensive therapy and no clear treatment algorithm could be established so far.
Congenital Adrenal Hyperplasia
There are two forms of the congenital adrenal hyperplasias that lead to hypertension, both because of reduced cortisol production that leads to an ACTH-mediated stimulation of the adrenal gland. The consequent increase of steroid precursors with a mineralocorticoid effect leads to hypertension and hypokalemia while the aldosterone levels are low. One type is attributed to 11-beta-hydroxylase deficiency, the other is because of 17-alpha-hydroxylase deficiency. In the former mutations in the CYP11B1 gene also induce variable degrees of virilization and often occur during the first years of life, but can also be observed later. In the forms attributed to CYP17A1 mutations, hypertension frequently presents together with hypogonadism. These distinctive associated features are associated with steroid metabolism imbalances that can be found in the urine permit usually a diagnosis.
Genomics of Primary Hypertension
The rarity of the monogenic hypertension syndromes implies that they can at most explain very little of primary hypertension with a population prevalence around 30%, quantitatively the most important cardiovascular risk factor of our times. As BP in the population is moderately heritable as outlined above, there is a great interest to also understand the genetic basis of primary hypertension.
Linkage and candidate-gene studies performed in large numbers over the past 30 years have only yielded few reproducible genetic results. Building on modern microarray platforms, millions of genetic variants can be genotyped permitting to interrogate close to the entire genome for association with a trait such as BP. The most frequent type of variant is the single nucleotide polymorphism (SNP), but other types exist such as copy number polymorphisms and structural variants, methylation marks, and additional variability. As SNPs constitute, by far, the most common type of variant, it was therefore likely that they influence traits such as BP most significantly.
Key Challenges of Blood Pressure Genome-Wide Association Studies
When association statistics are calculated between many thousands of SNPs and BP traits in GWAS, low p -values are produced by multiple testing and consequently the p -values require to be adjusted for the number of tests. It is generally assumed for common variants that the effective number of tests is 1 × 10 6 , therefore the p -value significance threshold is 5 × 10 −8 when applying a multiple-test correction by Bonferroni.
When assessing the frequency of SNPs throughout the genome in GWAS, it becomes rapidly clear that there are many rarer SNPs than frequent SNPs. On the other hand it is also clear from many genetic studies, including the BP variants identified so far, that the effect size of a variant is in general inversely proportional to the frequency of the variant. For blood pressure genes this is exemplified by the rare monogenic familial gene-variants having large effect sizes (beyond 20 mm Hg in many cases), whereas the frequent BP-GWAS variants have low effect sizes (around 1 mm Hg), too little to be of significance individually ( Fig. 6.2 ). This has profound consequences on the design of genetic studies of blood pressure because statistical power depends on both, the effect size and the frequency of the variant.
Genome-wide association studies using hundreds of thousands of SNPs have changed the understanding of blood pressure genomics of the general population and demonstrated the presence of clearly reproducible BP loci although they have, by far, not yet explained the majority of blood pressure. The advantage of the method is the unbiased approach that is hypothesis generating. The disadvantage is overall low statistical power because of the multiple testing burden.
Current Findings From Blood Pressure Genome-Wide Association Studies Efforts
A number of BP GWAS studies have been published, starting in 2008, that identify loci and specific variants consistently associated with BP. All currently published BP loci with their sentinel SNPs are listed in Table 6.5 . From the GWAS studies on BP so far, the following overall conclusions on the origin of primary hypertension and the genomic architecture of BP in the general population can be drawn.
