Kidney 11β-HSD2 Is Inhibited by Glycyrrhetinic Acid-Like Factors in Human Urine
We have previously shown that human urine contains substances that, like glycyrrhetinic acid, inhibit 11β-HSD1. We have named these substances “glycyrrhetinic acid-like factors” or GALFs. We now have found that human urine contains measurable quantities of both 11β(HSD1)- and 11β(HSD2)-GALF inhibitory substances. Both are markedly elevated in pregnancy. Their chemical and high-performance liquid chromatography (HPLC) characteristics suggest that several of the GALFs are steroidal. Large quantities of neutral 11β(HSD1)- and 11β(HSD2)-GALFs can be extracted directly from urine into ethyl acetate, yielding fraction EA1. Hydrolysis of the GALFs remaining in the aqueous phase by β-glucuronidase markedly increases the total amounts of GALFs, with the majority now being ethyl acetate extractable (fraction EA2). These EA2 posthydrolysis GALFs can be separated by HPLC resulting in at least six components with inhibitory activity against each isoenzyme. Only two GALF peaks are active against both 11β-HSD1 and 11β-HSD2. The others are peaks with specific 11β(HSD 1)- and 11β(HSD2)-GALF inhibitory activity. The GALFs in the same posthydrolysis EA2 extract are also inhibitory toward the 11β-HSD1 that is present in vascular smooth muscle where they may play a role in the mechanisms controlling blood pressure. We have also found that 11β-HSD2 is selectively inhibited by 5α-(but not by 5β-) reduced steroids. GC-MS analysis of the 11β(HSD2)-GALFs in EA2 is now being performed to determine whether this group includes 3α,5α-ring A-tetrahydro-reduced derivatives of steroids.
- 11β-HSD = 11β-hydroxysteroid dehydrogenase
- 11-dehydrpo-B = 11-dehydrocorticosterone
- AME = apparent mineralocorticoid excess
- CBX = carbenoxolone sodium
- DOC = deoxycorticosterone
- EA = ethyl acetate
- GA = glycyrrhetinic acid
- GALF = glycyrrhetinic acid-like factor
- HPLC = high-performance liquid chromatography
- MC = mineralocorticoids
- MR = mineralocorticoid receptor
- VSM = vascular smooth muscle
We have previously suggested that endogenous substances might exist in humans and other mammalian species that, like the licorice derivative GA, inhibit the steroid inactivating enzymes 11β-HSD and 5β-reductase.1 This followed from the important work of Ulick, New, Monder, and coworkers2–4⇓⇓ who demonstrated that hypertensive children with the syndrome of AME lack both 11β-HSD and 5β-reductase enzyme activity. Diminished activity of these enzymatic steroid metabolic pathways results in changes in the peripheral metabolism of the glucocorticoid cortisol. It has been postulated that the resulting higher intrarenal concentrations of cortisol may then interact with mineralocorticoid receptors and promote Na+ reabsorption.3–5⇓⇓ Edwards et al6 and Funder et al7 proposed that in vivo renal mineralocorticoid receptors remain aldosterone-specific because the enzyme 11β-HSD metabolizes cortisol and corticosterone (B) to cortisone and 11-dehydro-B, respectively. These 11-dehydro products, which have low binding affinities for MR,7 do not elicit MC-like effects and are considered inactive.
Our initial experiments1 demonstrated that partially purified extracts of urine from men and nonpregnant women did indeed contain substances that exhibited GA-like activity, inhibiting both 11β-HSD and 5β-reductase. We termed these inhibitory substances “GA-like factors” (GALFs). The level of GALF inhibitory activity was elevated in urine of pregnant women and shown to increase several-fold during the second and third trimesters of pregnancy. GALF inhibitory activity is also markedly elevated in patients with congestive heart failure.8 These earlier experiments used microsomal preparations of rat liver or kidney that predominantly contain 11β-HSD isoform 1. This isoform, 11β-HSD1, has a high Km for B (≈2 μmol/L), utilizes NADP+, and is bi-directional.9 Immediately thereafter, 11β-HSD isoform 2 (11β-HSD2) was discovered in MR containing cortical-collecting duct cells of kidney.10,11⇓ This was followed by the finding that AME patients exhibit mutations of this unidirectional NAD+-dependent, low Km for B (≈10 to 20 nmol/L) 11β-HSD isoform 2.12,13⇓ These experiments focused attention on renal 11β-HSD2 as the “guardian” enzyme that most likely confers MC specificity on MR-mediated mechanisms on Na+. 11β-HSD2 has been shown to be the major isoform of this enzyme in sheep and human kidney and has now been cloned from those tissues.14,15⇓
We have continued our efforts to isolate, further purify, and characterize the GALF substances in pregnant human urine. We now report our experiments utilizing sheep kidney microsomal 11β-HSD2 in addition to rat liver microsomal 11β-HSD1 for the quantitation and chemical characterization of these 11β-HSD inhibitory substances present in pregnant human urine.
Preparation of Urine Extracts
Urine samples were collected from five pregnant (6 to 8 months) and five age-matched nonpregnant women and stored at −20°C until assayed. They were desalted and partially purified by the Sep-Pak C18 solid-phase extraction method, which we have previously described.1 Briefly, Sep-Pak C18 cartridges (Waters Chromatography Division, Millipore Co) were primed with successive washes of 5 mL methanol and 5 mL water. Aliquots of 5 to 20 mL of urine were then passed through the cartridges. Following elution of unbound solutes with 5 mL of water, the compounds of interest were eluted with 3 mL of 100% methanol. These methanolic eluates were dried under reduced pressure in a Savant Speed-Vac system (Savant Instruments Inc) and then redissolved in water, to one fifth of the original volume. These aqueous preparations were designated Uo and were stored at 4°C until used.
Fractionation of U0 Extracts
The Uo preparations were extracted three times with equal volumes of EA, and the EA extracts (EA1) were dried under N2 and reconstituted with water to the volume of Uo used for the extraction. The remaining aqueous-phase extracts (Aq1) were incubated overnight at 37°C with β-glucuronidase (2 U/mL) in 0.1 mol/L Na acetate buffer (pH 6.8) and were then passed through Sep-Paks as described above to remove the enzyme and buffer (see flowsheet in Fig 1). Small aliquots of these samples, of known volume, were removed and stored at 4°C for later assay of 11β(HSD1)- and 11β(HSD2)-GALF inhibitory activity. The remainder of the β-glucuronidase-treated aqueous extracts (Aq1G) were each extracted three times with equal volumes of EA to yield the second EA extracts (EA2) and the second aqueous extracts (Aq2). As before, aliquots of the extracts from each step were taken and stored at 4°C before measurement of their 11β(HSD1)- and 11β(HSD2)-GALF inhibitory activity.
The abilities of these extracts to inhibit 11β-HSD activities in microsomal preparations from rat liver (11β-HSD1) and sheep kidney (11β-HSD2) were measured in duplicate as previously described.16 Briefly, for the assay of 11β-HSD1 enzymatic activity, rat liver microsomes (50 to 100 μg protein) were incubated at 37°C for 10 minutes with 5 μmol/L corticosterone (B) and 3H-corticosterone (1 μCi) as tracer in 50 mmol/L Tris-HCL buffer (pH 8.5) containing 3.4 mmol/L NADP+ in a total volume of 0.25 mL. For the assay of 11β-HSD2 enzymatic activity, sheep kidney microsomes (4 to 30 μg protein) were incubated at 37°C for 10 minutes with 50 mmol/L. 3H-corticosterone (1 μCi) as tracer and substrate in 50 mmol/L Tris-HCl buffer (pH 8.5), containing 200 μmol/L NAD+ in a total volume of 0.25 mL. 11β-HSD1 from homogenates of rat aorta VSM cell preparations was measured as previously described17 and also used for studies of inhibition. For all assays, an aliquot of either water (controls), urine extracts, or aqueous solutions of known quantities of CBX were added. The CBX was used for the construction of standard curves of inhibitory activity against which the inhibitory activities of the urine extracts could be compared. The reactions were terminated by addition of 0.75 mL of 100% methanol.
HPLC Assay of Enzyme Activity
The conversion of B to 11-dehydrocorticosterone (A) was measured by separating the compounds by HPLC and detecting and quantifying them with an on-line system. Aliquots of the methanol extracts from the incubation media were diluted with water to 45% methanol and chromatographed on Du Pont Zorbax C8 reverse-phase columns at 44°C using isocratic 62% aqueous methanol. This system allowed the adequate separation of radioactive peaks (A at 6.8 minutes and B at 8.8 minutes). The quantities of A and B were measured from the radioactivity detected and integrated by the on-line β-detection system (Radiomatic model FLO-ONE\Beta, radiochromatography detector, Packard Instrument Co). The percentage conversion was calculated and used as a measure of enzyme activity. Inhibitory activity was expressed as the percentage reduction in enzyme activity resulting from the presence of either known amounts of CBX or known volumes of urine extract in the incubation mixture. The percent of inhibition produced by varing quantities of CBX was plotted against its concentrations as a reference curve, and the inhibitory activity of the urine samples was expressed as the nanograms or micrograms of CBX that produced the same amount of inhibition under the standardized conditions. Since the IC50 for CBX against rat-liver 11β-HSD1 is two orders of magnitude smaller than that against sheep-kidney 11β-HSD2, independent GALF CBX units for the two isoenzymes are required. These differ greatly. The range used for the CBX standard curve for rat-liver 11β-HSD1 is 2 ng/mL to 32 ng/mL, while that used for sheep-kidney 11β-HSD2 is 200 ng/mL to 3200 ng/mL. These separate GALF units represent only the relative inhibitory potencies of the GALFs in the samples compared with CBX. They are not an absolute measure of the quantities of the inhibitory substances.
It should be noted that similar patterns of inhibitory activity were found when 3H-cortisol was used as a substrate to determine 11β-HSD2 activity.
Generation of HPLC Inhibitor Graphs
In our initial experiments to separate the GALF substances, we found that there was more than one 11β-HSD inhibitor in each of the urines. Therefore, we have chromatographed the individual EA1 and EA2 extracts on a Du Pont Zorbax C8 reverse-phase column, eluting the individual fractions with a methanol gradient beginning with 52% aqueous methanol that increased concavely to 100% methanol by 35 minutes. Each milliliter of the eluant was collected into a borosilicate test tube using an automated fraction collector and was later tested for its effects on rat-liver (11β-HSD1) and sheep-kidney (11β-HSD2) enzyme activity. Radiolabeled cortisol (F), corticosterone (B), DOC, progesterone (P), and unlabelled 3α,5β-tetrahydro-progesterone (detected at 191 nm absorbance) were added to the chromatographic samples as internal standards and were used for comparisons of the retention times of the GALFs.
The results are plotted as a graph of inhibitory activity against fraction number. This is termed the “inhibitor graph” and shows the inhibitory profile for each EA1 and EA2 sample.
[1,2-3H] was obtained from Du Pont New England Nuclear. Methanol (HPLC-grade) was obtained from Fisher Scientific. NADPH, Tris-HCI buffer, NADP+, NAD+, B, and 11 dehydro-B were obtained from Sigma Chemical Co.
Measurement of 11β-HSD2 GALF in Uo Extracts from Pregnant and Nonpregnant (Control) Human Urine
When Uo extracts (Fig 1) from pregnant and control human urines were assayed for their inhibitory activities against sheep-kidney 11β-HSD2, the amounts of 11β(HSD2)-GALF inhibitory activity found in the pregnant urine extracts were approximately four times higher than those in control extracts (Table).
Aq1 and EA1 extracts (prepared as described above) from both pregnant and control urines were also assayed for inhibitory activity. The amount of 11β(HSD2)-GALF inhibitory activity was again approximately four times higher in both fractions (Aq1 and EA1) from pregnant urine extracts (Table). However, the majority (>70%) of the 11β(HSD2)-GALF inhibitory activity was recovered in the EA1 extracts of both groups.
Hydrolysis by β-Glucuronidase
There was a considerable increase in inhibitory activity when the Aq1 extracts from both pregnant and control urines were hydrolyzed with β-glucuronidase.
Compared with Aq1, inhibitory activity of the posthydrolysis Aq1G was increased 2 to 3 times in pregnant and 4 to 5 times in control urines. Following EA extraction, the majority (>75%) of the increased inhibitory activity in Aq1G was recovered in the EA2 fractions in both pregnant and control urines and is comparable in amount to that in the EA1 extracts. Therefore, the majority of the originally water-soluble polar GALF substances in the Aq1 fractions must be glucuronides that, after β-glucuronidase hydrolysis became aglycones, which partitioned into the EA phase (EA2) (Table). The 11β(HSD2)-GALF inhibitory activity present in EA2 fractions of pregnant urine was approximately 2.5 to 3 times higher than that in control EA2 fractions.
Measurement of 11β(HSD1)-GALF in Pregnant and Nonpregnant (Control) Human Urine
When Uo extracts (Fig 1) from pregnant and control urines were tested against rat-liver 11β-HSD1, the 11β(HSD1)-GALF inhibitory activity was approximately 8 times higher in the pregnant urine extracts than in the controls (Table).
Aq1 and EA1 extracts were also tested for 11β(HSD1)-GALF inhibitory activity. Both of the extracts from the pregnant urine contained approximately 6 to 7 times more inhibitory activity than that in the corresponding extracts from controls (Table). As with 11β(HSD2)-GALF, the bulk of the 11β(HSD1)-GALF inhibitory activity in both pregnant and control urine extracts was found in the EA1 fraction.
Hydrolysis by β-Glucuronidase
After the Aq1 fractions were hydrolyzed by β-glucuronidase, there was an increase of approximately 6 times in the 11β(HSD1)-GALF inhibitory activity in the pregnant urine Aq1G extracts and approximately 10 times in the equivalent control extracts (Table). As in the case of 11β(HSD2)-GALF substances, the majority (>75%) of the originally polar 11β(HSD1)-GALF substances then partitioned into the EA phase (EA2) after β-glucuronidase hydrolysis. 11β(HSD1)-GALF inhibitory activity became significantly higher in the EA2 extracts when compared with the EA1 extracts in both pregnant and control groups. The final 11β(HSD1)-GALF inhibitory activity in pregnant urine EA2 fractions was approximately 3 times larger than those of the EA2 extracts from the control urines.
Comparison of 11β(HSD2)-GALF Inhibitor Graphs From Pregnant and Control Urine
The 11β(HSD2)-GALF substances in EA1 and EA2 fractions were further separated into a series of “polar” and “less polar” inhibitory fractions by HPLC reversephase chromatography as described in “Methods.” This solvent system reproducibly separates the steroid standards: cortisol, corticosterone, DOC, progesterone, and 3α5β-tetrahydroprogesterone with retention times of 11.8, 17.4, 24.5, 31.3, and 34.2 minutes, respectively. EA1 and EA2 extracts, obtained from comparable volumes of Uo from pregnant and control urines, were chromatographed and each 1 mL fraction assayed for GALF inhibitory activity as previously described. The results are plotted as the inhibitor graphs in Figs 2 through 5⇓⇓⇓.
Inhibitor Graphs From EA1 Extracts
The 11β(HSD2)-GALF inhibitor graphs from EA1 extracts of both pregnant and control urines (Fig 2) have three common peaks of inhibitory activity in the more polar region (fraction numbers 11, 13, and 18) and three common peaks in the less polar region (fraction numbers 32, 35, and 39). An additional highly polar peak of 11β(HSD2)-GALF inhibitory activity (fraction number 5) was present in the control urine but not in the pregnant urine EA1 extracts. The areas under the peaks of fraction numbers 11, 13, and 18 were higher in pregnant urine extracts, whereas those of fraction numbers 32, 35, and 39 were reduced compared with controls.
Inhibitor Graphs From EA2 Extracts
The profile of 11β(HSD2)-GALF inhibitory activity in the EA2 extracts from pregnant urine (Fig 3) showed the presence of 6 principal peaks together with at least 6 secondary peaks of inhibitory activity. By contrast, HPLC of control urine EA2 extracts showed the presence of two principal peaks together with 6 secondary peaks of 11β(HSD2)-GALF inhibitory activity. Although the major peaks eluted in the same region of the solvent gradient as cortisol, corticosterone, and DOC, they did not exhibit the UV characteristics of these steroid standards. Large differences were observed in the areas under these peaks when the inhibitor graph of the pregnant urine EA2 extracts was compared with those of the controls. Since the total quantity of inhibitory activity is less in the EA2 sample from the control urine than in the EA2 extracts from the pregnant urines, all of the control peaks were smaller than those from the pregnant urine. However, the peaks eluting at 16, 18, 25, and 28 minutes were disproportionately markedly reduced (Fig 3). Fraction numbers 23 and 28 in EA2, which contain both 11β(HSD1)- and 11β(HSD2)-GALF inhibitory activities (Fig 4), are not present in EA1 extracts of either control or pregnant urines.
Comparison of 11β(HSD1)- and 11β(HSD2)-GALF Inhibitor Graphs of EA2 Extracts From Pregnant Urine
Fractions of the EA2 extracts from pregnant urines separated by HPLC were also tested for both 11β(HSD1)-GALF and 11β(HSD2)-GALF inhibitory activity. Considerable differences in the resulting inhibitor graphs were observed and are shown in Fig 4. With the exception of the two common peaks of inhibitory material in fraction numbers 23 and 27/28, the major peaks of 11β(HSD2)-GALF inhibitory activity appeared in the polar region (left of the steroid standard DOC), whereas those of 11β(HSD1)-GALF appeared in the nonpolar region (right of DOC) of the chromatograph. However, both sets of GALFs elute in the same region of the HPLC solvent gradient that encompasses the steroid standards. As indicated by the differences in the inhibitory activity of the HPLC separated peaks of GALF substances against 11β-HSD1 and 11β-HSD2, two peaks inhibit both 11β-HSD1 and 11β-HSD2 activity, while the others differentially inhibit either 11β-HSD1 or 11β-HSD2.
We also used rat VSM cell preparations as a second source of 11β-HSD1. Other experiments in our laboratories17 have shown that rat aorta preparations and VSM cells cultured from rat aorta contain a bi-directional NADP+-dependent 11β-HSD1, which has a smaller Km for B (100 nmol/L) than that for rat-liver microsomal preparations. When Uo extracts (Fig 1) from pregnant and control urines were assayed against homogenates of VSM cells,17 both types of urine showed measurable quantities of 11β(HSD1)VSM-GALF inhibitory activity.
HPLC fractions of the EA2 extracts from pregnant urines were also tested for 11β(HSD1)VSM-GALF inhibitory activity for comparison with their inhibitory activity against 11β-HSD2. The profile of the inhibitor graph is similar to that obtained by when rat liver microsomal preparations of 11β-HSD1 were used (Fig 5).
The discovery that ingestion of licorice by humans mimics the clinical features and changes in cortisol metabolism observed in patients with AME18 led to our hypothesis that GA (in licorice) may mimic the physiological activity of endogenous factor(s), GALFs. It was hypothesized that these GALF substances would serve to cause glucocorticoids, and possibly other steroids, to elicit Na+ retention by MR-mediated mechanisms and so augment, either naturally or in disease states, the Na+ retaining actions of aldosterone.
We had earlier shown in vivo that the succinate derivative of GA CBX conferred renal MR-mediated Na+ retention on cortisol and corticosterone (B) in adrenalectomized rats.19,20⇓ We have shown that endogenous substances such as the bile acid, chenodeoxycholic acid, and several progesterone derivatives likewise conferred Na+ retention on B.21,22⇓ However, the original 11β-HSD “guardian” hypothesis first proposed by Edwards et al,6 Funder et al,7 and others2 was later refined by Rusvai and Naray-FejesToth,11 who proposed that the 11β-HSD isoform 2 (11β-HSD2) in the distal portions of the kidney acts as the “protective” mechanism that normally prevents glucocorticoids from accessing renal MR. The more recent finding that AME patients exhibit mutations of the NAD+-dependent 11β-HSD2+-dependent 11β-HSD212,13⇓ has offered considerable evidence supporting the “guardian” role of 11β-HSD2. The role of 11β-HSD1, which is present predominantly in liver but also in several other tissues such as VSM,17 is less clear at this time.
In this report, we demonstrate that human urine contains, in addition to 11β(HSD1)-GALF, measurable quantities of 11β(HSD2)-GALF inhibitory substances and that both are markedly elevated in pregnancy. Using HPLC, we have now separated several peaks of GALF inhibitory substances, some of which possess specific 11β(HSD2)-GALF, while others possess only 11β(HSD1)-GALF inhibitory activity. Two HPLC peaks, which elute in fraction numbers 23 and 28 (Fig 4), possess both 11β(HSD1)-GALF and 11β(HSD2)-GALF inhibitory activity. Each isolated GALF-containing peak from urine of pregnant humans was markedly elevated when compared with those isolated from nonpregnant females of similar age.
Significant quantities of both 11β(HSD1)- and 11β(HSD2)-GALF inhibitory substances were extractable from the partially purified urine extracts with EA (EA1, see flow sheet, Fig 1). Interestingly, when the GALF inhibitory substances then remaining in the aqueous phase were subjected to hydrolysis by a β-glucuronidase preparation (devoid of any sulfatase activity), not only did the measurable quantities of 11β(HSD1)- and 11β(HSD2)-GALF inhibitory substances markedly increase by at least 2- to 10-fold, but the majority then became EA (EA2) extractable. The HPLC inhibitor graphs indicated that the GALFs released after hydrolysis (and EA2 soluble) did not possess the same elution times as those isolated and separated by HPLC of the EA1 fraction. Thus, significant quantities of neutral (EA soluble) GALF substances (aglycones) are released, which are more active as 11β(HSD1)- and 11β(HSD2)-GALFs than their glucuronide conjugate forms. Monder and White9 had previously shown reduced potency of glycyrrhizin (the glycoside) compared with the 11β-HSD inhibitory activity of GA itself. The GALF inhibitory substances remaining in the aqueous phase (Aq2G) are now being studied to determine the proporation and chemical identities of those that may be present as sulfates and be susceptible to hydrolysis by pure sulfatase enzyme preparations.
Several of the 11β(HSD1)- and 11β(HSD2)-GALF inhibitory substances in the EA1 extractable fraction, and in the EA2 extractable fraction (following β-glucuronidase hydrolysis), yielded separated peaks of activity after HPLC, with retention times in the vicinity of known steroid hormone standards (eg. cortisol. corticosterone. DOC. and progesterone). However, they did not exhibit the UV characteristics of these steroid standards.
The majority of steroid hormones are excreted by humans as glucuronide and sulfate conjugates of their ring A-3α,5αTH- and 3α.5βTH-reduced derivatives. We have recently shown16 that sheep-kidney microsomal preparations of 11β-HSD2 are selectively inhibited by these 3α,5α-tetrahydro derivatives of steroids but not by their 3α,5β-tetrahydro derivatives. Several of the posthydrolysis 11β(HSD2)-GALF inhibitory substances in the EA2 fraction, chromatograph where known standards of 3α,5α-ring A reduced steroids, such as 3α,5α-tetrahydroxy-B and 3α,5α-tetrahydro-11-dehydro-B and 11β-hydroxyand 11-keto-3α,5α-tetrahydro derivatives of progesterone elute. Thus, if the 11β(HSD2)-GALFs present in EA2 extracts are of steroidal origin, they would most likely be 3α,5α-TH steroid derivatives.
We have recently demonstrated 5α-derivatives of the glucocorticoid B (eg, 5α-DHB and 3α,5α-THB) and 11β-OH-progesterone23 as well as its 3α,5α-tetrahydro derivative, which all potently inhibit 11β-HSD2, can in vivo confer Na+-retaining activity on B in adrenalectomized rats and can also cause increases in blood pressure when infused into usually normotensive Sprague-Dawley rats (private communication, G.W. Souness and D.J. Morris, 1996). We plan in future experiments to test whether these GALF substances in the EA2 extracts have similar biological activities.
Each of the HPLC-separated peaks of 11β(HSD1)- and 11β(HSD2)-GALF substances is currently being analyzed using GC-MS analysis. These studies will also determine whether the 11β(HSD2)-GALFs in EA2 include 3α,5αring A-tetrahydro-reduced derivatives of steroids. It is important to point out that the chemical structure and identity of the individual GALF substances have not been ascertained at this time. Several GALF substances possess characteristics similar to those of neutral steroids and steroid glucoronides, and several of the EA extractable substances chromatographed with retention times close to (but not identical with) those of known steroid standards. Nevertheless, the complexity and the number of substances excreted in urine of human pregnant individuals will make the identification of the individual 11β(HSD1)- and 11β(HSD2)-GALF substances difficult.
Considerable additional experimentation is now necessary to conclusively provide evidence as to the origin and physiological role(s) of the GALF substances. The present experiments clearly provide evidence for the existence of both endogenous 11β(HSD1)- and 11β(HSD2)-GALF substances and that one class is most likely steroidal in origin. 11β-HSD1 has been clearly shown to be present in VSM24,25⇓ and to play a functional role17,26⇓ in the regulation of blood pressure. The finding that the isolated GALF material(s) also inhibits rat VSM preparations of 11β-HSD1 possibly indicates an additional role for GALFs.
The recent experiments of Takeda et al27 showing elevated levels of 11β(HSD2)-GALF inhibitory substances in the urine of patients with low renin essential hypertension, and that they can be regulated by dietary Na+, are indeed extremely interesting. Similarly, the recent findings of Soro et al28 clearly demonstrate diminished 11β-HSD and 5β-reductase activity in patients with essential hypertension. Their findings confirm the earlier work of Kornel et al29,30⇓ that the 5α-ring A-reduced pathway is favored in these patients, with increases in the excretion of both 5αdihydro- and 3α,5α (allo)-tetrahydro derivatives of several steroid hormones. These are very interesting findings, even though initial studies using measurements of 11β(HSD1)-GALF have not yet found consistently elevated levels of total GALF in patients with essential hypertension.8,31⇓
Further studies are now most important to determine not only the chemical identities of the various families of substances that can function as GALFs but to reveal the source and regulation of their synthesis and their physiological role(s).
This work was supported by National Institutes of Health National Heart, Lung, and Blood Institute grant HL-52972 and by the Miriam Hospital Research Foundation. We thank Elizabeth Gifford for excellent secretarial assistance.
Morris DJ, Semafuko WEB, Latif SA, Grimes CA, Sheff MF. Detection of glycyrrhetinic acid-like factors (GALFs) in human urine. Hypertension. 1992; 20 : 356 –360.
New MI, Oberfeld SE, Carey R, Grieg F, Ulick S, Levine LS. A genetic defect in cortisol metabolism as the basis for the syndrome of apparent mineralocorticoid excess syndrome. In: Mantero F, Biglieri C, Edwards CWR, eds. Endocrinology of Hypertension. New York, NY: Academic Press; 1982: 85 –101.
Monder C, Shackleton CHL, Bradlow HL, New MI, Stoner E, Iohan F, Lakshmi V. The syndrome of apparent mineralocorticoid excess: its association with 11β-dehydrogenase and 5β-reductase deficiency and some consequences for corticosteroid metabolism. J Clin Endocrinol Metab. 1986; 62 : 550 –557.
Funder JW, Pearce PT, Smith R, Smith AI. Mineralocorticoid action: target tissue is enzyme not receptor mediated. Science. 1988; 242 : 583 –585.
Semafuko WEB, Sheff MF, Grimes C, Latif SA, Sadaniantz A, Levinson P, Morris DJ. Inhibitors of 11β-hydroxysteroid dehydrogenase and 5β-steroid reductase (GALFs) in urine from patients with congestive heart failure. Ann Clin Lab Sci. 1993; 23 : 456 –461.
Rusvai E, Naray-Fejes-Toth A. A new isoform of 11β-hydroxysteroid dehydrogenase in aldosterone target cells. J Biol Chem. 1993; 268 : 1 –4.
Agarwal AK, Mune T, Monder C, White PC. NAD+-dependent isoform of 11β-hydroxysteroid dehydrogenase: cloning and characterization of cDNA from sheep kidney. J Biol Chem. 1994; 269 : 25959 –25962.
Latif SA, Sheff MF, Ribeiro CE, Morris DJ. 11β-Hydroxysteroid dehydrogenase (11β-HSD) isoform 2 from sheep kidney is selectively inhibited by 5α-metabolites (but not 5β-metabolites of adrenal cortex). Steroids. In press.
Brem AS, Bina B, King T, Morris DJ. Bidirectional activity of 11β-hydroxysteroid dehydrogenase in vascular smooth muscle cells. Steroids. 1995; 20 : 406 –410.
Souness GW, Morris DJ. 11α- and 11β-Hydroxyprogesterone, potent inhibitors of 11β-hydroxysteroid dehydrogenase, possess hypertensinogenic activity in the rat. Hypertension. 1996; 27 : 421 –425.
Ullian ME, Walsh LG. Corticosterone metabolism and effects on angiotensin II receptors in vascular smooth muscle. Circ Res. 1995; 77 : 702 –709.
Takeda Y, Miyamori I, Iki K, Inaba S, Furukawa K, Hatakeyama H, Yoneda T, Takeda R. Endogenous renal 11β-hydroxysteroid dehydrogenase inhibitory factors in patients with low-renin essential hypertension. Hypertension. 1996; 27 : 197 –201.
Soro A, Ingram MC, Tonolo G, Glorioso N, Fraser R. Evidence of coexisting changes in 11β-hydroxysteroid dehydrogenase and 5β-reductase activity in subjects with untreated essential hypertension. Hypertension. 1995; 25 : 67 –70.
Kornel L, Margolis E. Evidence for overall deficiency of cortisol 11β-hydroxy dehydrogenase in patients with essential hypertension. In: Proceedings of the 9th International Congress of Endocrinology. September 1992; Nice, France. Abstract P-04.31.123;142.
Walker BR, Williamson PM, Brown MA, Honour JW, Edwards CRW, Whitworth JA. 11β-Hydroxysteroid dehydrogenase and its inhibitors in hypertensive pregnancy. Hypertension. 1994; 25 : 626 –630.