- Top of page
Cytokines are emerging as important in developmental processes. They may induce alterations in normal gene expression patterns, activate angiotensinogen transcription, or alter expression of the renin–angiotensin system (RAS). To determine whether prenatal exposure to interleukin-6 (IL-6) influences gene expression of the intrarenal RAS and contributes to renal dysfunction and hypertension in adulthood, we exposed female rats to IL-6 early (EIL-6 females) and late (LIL-6 females) in pregnancy and analysed blood pressure in the offspring at 5–20 weeks of age. Renal fluid and electrolyte excretion was assessed in clearance experiments, mRNA expression by real-time PCR, and protein levels by Western blot. Systolic pressure was increased at 5 weeks in IL-6 females and at 11 weeks in males. Circulatory RAS levels were increased in all IL-6 females, but angiotensin-1-converting enzyme (ACE) activity was increased only in LIL-6 females. LIL-6 males and IL-6 females showed decreased urinary flow rate and urinary sodium and potassium excretion. Dopamine excretion was decreased IL-6 females. In adult renal cortex, renin expression was increased in all IL-6 females, but angiotensinogen mRNA was increased only in LIL-6 females; AT1 receptor (AT1-R) mRNA and protein levels were increased in LIL-6 females, whereas AT2 receptor (AT2-R) levels were decreased in LIL-6 females and EIL-6 males. In adult renal medulla, AT1-R protein levels were increased in LIL-6 females, and AT2-R mRNA and protein levels were decreased in EIL-6 males and LIL-6 females. Prenatal IL-6 exposure may cause hypertension by altering the renal and circulatory RAS and renal fluid and electrolyte excretion, especially in females.
Alterations of the intrauterine environment can cause cardiovascular and metabolic disease in adult offspring (Barker, 1990). Hypertension, the most extensively studied disease that is affected by prenatal programming, has been linked to intrauterine growth retardation (IUGR) in humans (Law & Shiell, 1996) and rats (Vehaskari et al. 2001). Studies in animal, most commonly of maternal protein (Woods et al. 2001) and calorie (Woodall et al. 1996) restriction and maternal glucocorticoid (GC) exposure (Langley-Evans, 1997), have suggested several mechanisms of prenatal programming, including dysregulation of the hypothalamic–pituitary–adrenal (HPA) axis (O'Regan et al. 2001).
Intrauterine influences appear to cause adult hypertension by permanently altering renal gene expression and function (Dodic et al. 2002). The circulatory renin–angiotensin system (RAS) is important in regulating fluid and electrolyte balance, arterial pressure and cardiac and renal function, and tissue-specific RAS pathways have been described (Carey & Siragy, 2003). The RAS is important in renal development. All RAS components are detectable in kidney by embryonic day 12–17, and their levels are higher in fetal and newborn rats than in adults (Gomez & Norwood, 1995). Human infants with IUGR have elevated renin and angiotensin II (AngII) levels in cord blood (Kingdom et al. 1993) and evidence of increased renin gene expression in the kidney, suggesting elevation of both the intrarenal and peripheral RAS (Kingdom et al. 1999), which may regulate blood pressure (BP) independently (Price et al. 1999).
AngII regulates renal blood flow, glomerular filtration rate and sodium and potassium reabsorption (Dworkin et al. 1983). Rat kidney has three types of angiotensin receptor (AT-R), which have opposing effects on BP (Siragy, 2002). AT1A-R and AT1B-R increase BP by promoting sodium reabsorption and vasoconstriction (Allen et al. 2000), but the hypertensive effects of AngII appear to be mediated largely by AT1A-R (Ito et al. 1995; Chen et al. 1997). The second type of AT-R, AT2-R, decreases BP by mediating bradykinin, prostaglandin and nitric oxide release, causing vasodilatation and decreased peripheral resistance (Carey & Siragy, 2003). AT2-R expression is high in fetal mesenchymal tissue and rapidly diminishes postnatally but is detectable in adult kidney and blood vessels (Ozono et al. 1997). In rats born to protein-restricted mothers, up-regulation of renal RAS activity may increase BP by altering the balance of receptor expression (McMullen et al. 2004). These effects appear to be sex specific and permanently prevented by treatment with an AT1-R antagonist or angiotensin-1-converting enzyme (ACE) inhibitor during the first post-natal month (Sherman & Langley-Evans, 1998, 2000).
Sodium balance and BP are also regulated by dopamine, which acts directly on renal and intestinal epithelial ion transport, interacts with other receptors, and modulates aldosterone, catecholamine and renin secretion (Aperia, 2000). Renal dopamine production or receptor function may be defective in arterial hypertension (Aperia, 2000). It is interesting that AngII-mediated increases in sodium reabsorption and BP counteract the effect of dopamine (Aperia, 2000).
GCs are important regulators of the RAS (Guo et al. 1995), and fetal overexposure to GCs may be central to BP programming, perhaps through interactions with renal AT-Rs (Moritz et al. 2002). Not all studies support this hypothesis and there may be two independent mechanisms that act in a sex-specific manner (McMullen & Langley-Evans, 2005).
Although traditionally perceived as immune system messengers, cytokines are emerging as important in organogenesis and developmental processes (Borish & Rosenwasser, 1996). In organogenesis, cytokines may alter normal gene expression patterns, modulate the extracellular matrix, influence the levels of other cytokines, and interact with transcription factors (Cale, 1999). It is interesting that macrophages are among the first haematopoietic cells found in the kidney during renal development (Morris et al. 1991). Their numbers and activation status are closely regulated by cytokines, and evidence is accumulating that macrophages can induce cell death during organogenesis and thereby directly influence tissue remodelling (Borish & Rosenwasser, 1996). Thus, cytokines expressed in the developing fetus at specific sites, such as in the metanephros, may be chemo-attractants for macrophages, and alterations of cytokine levels may have an effect on the inflammatory response (Cale, 1999). Stimulated macrophages also produce a wide range of cytokines that inhibit renal development and alter gene expression (Gordon, 1995). Both epidemiological and experimental studies have revealed a close association between cytokines, and in particular interleukin-6 (IL-6), and essential hypertension (Mahmud & Feely, 2005), and IL-6 seems to be an independent risk factor for high blood pressure in apparently healthy subjects (Bautista et al. 2005). Recent studies (Lee et al., 2006)with IL-6-knockout mice also show that Ang II-induced hypertension is IL-6-dependent. These data suggest that there is a powerful effect of IL-6 in mediating the rightward shift in the renal pressure–natriuresis relationship caused by Ang II and high-salt intake and provide evidence for an important long-term blood pressure effect of IL-6 (Lee et al. 2006). IL-6 has also been showed to be linked to hypertensive glomerular injury (Kusaka et al. 2002; Ruiz-Ortega et al. 2002).
Expression of the angiotensinogen gene and probably genes encoding other parts of the RAS appears to be under coordinate developmental, tissue-specific, hormonal and cytokine control (Brasier & Li, 1996). Inflammation activates angiotensinogen transcription as a result of the macrophage-derived cytokines interleukin-1 (IL-1) and tumour necrosis factor-α (TNF-α) (Brasier & Li, 1996). These pro-inflammatory cytokines are likely to be the key activators of hepatic angiotensinogen expression; however, IL-6 can activate angiotensinogen expression in vitro (Brasier & Li, 1996). In another study, TNF-α stimulated both angiotensinogen expression and AngII secretion in a concentration-dependent manner in human adipocytes (Harte et al. 2005).
In this study, we sought to determine whether prenatal IL-6 exposure alters gene expression of the intrarenal RAS during development and thereby affects renal fluid and electrolyte excretion, dopamine production and BP levels in adulthood. We also assessed the circulatory levels of renin, aldosterone, ACE and corticosterone. In order to evaluate the importance of critical windows in fetal development in relation to programming effects we used two different time points of IL-6 exposure, early and late pregnancy.
- Top of page
In this study, prenatal IL-6 exposure caused hypertension in male and female rats in adulthood. Urinary flow rate and electrolyte and dopamine secretion were reduced in IL-6 females, which also had increased plasma renin and aldosterone levels at 20 weeks of age. LIL-6 males had reductions in urinary flow and electrolyte and dopamine secretion at 20 weeks. In LIL-6 females, renin and angiotensinogen mRNA levels were increased in renal cortex. AT1-R mRNA and protein levels were also increased in medulla, whereas AT2-R levels were decreased in cortex and medulla. In IL-6 males, however, these mRNA and protein levels were unchanged, except in the EIL-6 group, which had decreased AT2-R mRNA and protein levels in cortex and medulla. Thus, IL-6 exposure changes the renal and circulatory RAS and alters electrolyte and dopamine secretion, especially in LIL-6 females.
Programming, hypertension and the RAS
Fetal and perinatal programming increases the risk for hypertension or other cardiovascular disease in adulthood (Wintour et al. 2003). In most studies, low birth weight increases the risk of hypertension in adulthood (Law & Shiell, 1996). However, adult-onset hypertension can be programmed by prenatal treatments (low protein diet and exposure to GCs) that do not decrease birth weight, suggesting that IUGR is an indirect marker of BP programming. In our rat models, endotoxin and pro-inflammatory cytokines had strong reprogramming effects on insulin sensitivity, BP, the HPA axis and stress sensitivity without reducing birth weight (Dahlgren et al. 2001; Samuelsson et al. 2004). Although their mechanisms of action in the developmental processes have not yet been fully delineated, cytokines appear to be important in organogenesis, regulation of transcription factors, and gene expression patterns (Cale, 1999).
In mice and rats, nephrogenesis begins around day 10–12 after conception (Jensen, 2004). However, maternal low-protein diets can reduce cell numbers in the inner cell mass of pre-implantation embryos and can programme hypertension in male offspring (Kwong et al. 2000); there is also some evidence that this treatment affects the rate of apoptosis in the mesenteric mesenchyme. These studies suggest that programming stimuli may alter cell turnover and gene expression before nephrons and glomeruli have begun to form (Welham et al. 2002, 2005). These observations are in line with the hypothesis that epigenetic alterations – stable alterations of gene expression through covalent modifications of DNA and core histones – in early embryos may be carried forward to subsequent developmental stages. Thus, times of exposure that do not strictly coincide with organogenesis might also be of interest (Waterland & Jirtle, 2004).
IL-6 can activate angiotensinogen expression (Brasier & Li, 1996). Disruption of carefully regulated processes in kidney development, where cytokines such as IL-6 might have an important role, may alter the intrarenal or circulatory RAS, leading to hypertension or renal dysfunction.
Circulatory and renal RAS and hypertension
Plasma aldosterone levels were elevated in adult IL-6 females, indicating increased circulatory RAS activity, but ACE levels were increased only in LIL-6 females. Plasma renin was elevated at 20 weeks in IL-6 females and reduced at 5 weeks in EIL-6 males. The intrarenal RAS was most altered in LIL-6 females, which had increased mRNA levels of renin and angiotensinogen in renal cortex and increased AT1-R and decreased AT2-R mRNA and protein levels in cortex and medulla; EIL-6 females only had increased renin mRNA expression in the cortex. However, all IL-6 females had elevated BP at 20 weeks, with no difference between the early and late exposure groups. Male EIL-6 rats had decreased AT2-R mRNA and protein in both cortex and medulla, and all IL-6 males had significantly higher BP than controls.
In recent studies in which a maternal low-protein diet during rat pregnancy was found to programme BP, renal AT2-R expression was reduced, but AT1-R expression was unaltered (McMullen et al. 2004; McMullen & Langley-Evans, 2005). In another study in which protein restriction induced hypertension, only AT1-R expression was increased (Vehaskari et al. 2004). Cross-talk between the Ang II receptor subtypes has been demonstrated by AT1-R blockade, which significantly increased AT2-R expression (Cosentino et al. 2005). In EIL-6 male and LIL-6 female rats, an unbalanced expression of renal angiotensin receptor subtypes might have an important role in the development of hypertension.
However, extrarenal AT1-Rs also contribute to BP homeostasis (Crowley et al. 2005), with vascular, CNS and adrenal gland AT-Rs probably making the major kidney-independent contributions (Ito et al. 1995; Davisson et al. 2000). In adrenal cortex, the AT1-R stimulates aldosterone release (Aguilera, 1992), promoting sodium reabsorption in the mineralocorticoid-responsive segments of the distal nephron (Aguilera, 1992). Thus, the increased aldosterone levels in IL-6 females might affect BP levels and exert an additive effect through the extrarenal AT1-R-dependent action of mineralocorticoids in the adrenal gland.
Renal fluid and electrolyte excretion and hypertension
IL-6 females and LIL-6 males showed decreases in urinary flow rate and excretion of dopamine (females only), sodium and potassium. Primary renal dysfunction probably causes hypertension by increasing sodium retention, leading to expanded extracellular fluid volume. Among its renal actions, AngII stimulates vasoconstriction and tubular sodium reabsorption, effects attributed to AT1-Rs (Dworkin et al. 1983) and opposed by AT2-Rs, which therefore may be protective (Rasch et al. 2004). In stroke-prone spontaneously hypertensive rats, the circulatory and intrarenal RAS may be inappropriately activated, resulting in decreased AT2-R and increased AT1-R expression perinatally and in adulthood (Goto et al. 2002). This would lead to increased sodium reabsorption and increased blood volume, which could increase cardiac output, stroke volume and BP (Dodic et al. 2002).
Nevertheless, our data on renal Na+ and K+ handling cannot be explained solely by regulation mediated by the RAS and aldosterone. When activated, this system causes Na+ reabsorption by stimulating the apical membrane epithelial sodium channel and the sodium chloride cotransporter (NCC) as well as basolateral Na+–K+-ATPase in the distal nephron segment. Also, aldosterone activates apical plasma membrane inwardly rectifying potassium channels (ROMK), the rate-limiting step for transepithelial potassium transfer, which together with Na+–K+-ATPase activation and depolarization, causes K+ secretion (Lang et al. 2005). A number of factors regulate these transporters more or less independently of aldosterone (for review see Meneton et al. 2004).
We speculate that alterations in some of these pathways in response to prenatal IL-6 exposure may explain the effects on renal Na+ and K+ handling in the offspring. These additional regulatory pathways include hormones (e.g. natriuretic peptides), locally produced factors (e.g. prostaglandin E22) and regulatory proteins (e.g. serum- and glucocorticoid-inducible kinase, protein tyrosine kinases and the WNK (with-no-lysine-kinase) serine–threonine kinase). For example, alterations in renal WNK kinase expression/activity could potentially explain the altered renal Na+/K+ excretion. Mutations in the WNK 1 and WNK 4 genes cause Na+ retention, hypertension and hyperkalaemia (and thus decreased urinary potassium secretion) without increased RAS activity (Wilson et al. 2001), probably through increased NCC-mediated Na+ reabsorption and decreased K+ secretion mediated by inhibition of ROMK activity.
Therefore, decreased WNK expression/activity in the distal nephron in response to prenatal IL-6 exposure may explain the hypertension and the retention of Na+ and K+ (and thus their decreased urinary excretion) in both males and females. Furthermore, in LIL-6 males, this change was not accompanied by marked activation of the RAS, explaining hypertension and Na+ retention despite a largely normal RAS. In females, in addition to the suggested down-regulation of renal WNK, the RAS was clearly activated; however, the increase in RAS does not normalize K+ secretion, which might be explained by a WNK defect (not measured in the present study) in the distal nephron. This postulated ‘double defect’ in the female might explain the earlier onset of hypertension in females than in males.
IL-6 females also had decreased dopamine excretion. In isolated tubular preparations, dopamine inhibits Na+–H+-exchanger and Na+–K+-ATPase activities, reducing tubular sodium reabsorption and increasing sodium excretion (Aperia, 2000). Deficiency in renal dopamine synthesis or secretion occurs in hypertensive humans (Aperia, 2000) and in spontaneously hypertensive and Dahl salt-sensitive rats, which have poor natriuretic and diuretic responses to sodium loading and volume expansion (Aperia, 2000). Defective D1-like dopamine receptor function has also been reported (Jose et al. 1992; Aperia, 2000). In young normotensive subjects with a strong family history of hypertension, dopamine activity was suppressed before increased BP was evident (Iimura & Shimamoto, 1990). Renal dopamine deficiency might therefore be a primary factor in primary hypertension.
Injection of IL-6 causes a small but significant increase in corticosterone levels in the maternal circulation at 4 h (Samuelsson et al. 2004). However, the fetus is probably protected by 11βHSD-2 (11 beta hydroxysteroid dehydrogenase type-1), as there is no evidence that its activity is reduced, the pups show no signs of IUGR, and birth weight is increased in male offspring. In addition, we recently showed that radiolabelled IL-6 passes the placental barrier, allowing direct exposure of the fetus (Dahlgren et al. 2006).
It is important to remember that the kidney is regulated by neural and hormonal inputs from the brain. Programmed alterations in the brain RAS have been reported, particularly in the hypothalamus (leading to higher water and salt intake) and medulla oblongata (affecting cardiovascular function chiefly through control of peripheral sympathetic drive) (Dodic et al. 2002; Wintour et al. 2003). These findings merit further investigation.
Sex differences and IL-6 programming
Changes in circulatory and intrarenal RAS and renal functions were most prominent in IL-6 females, which also became hypertensive sooner than males (Table 7). The more pronounced effects on renal Na+ and K+ handling in females might reflect alterations in the expression or activity of renal tubular transporters (e.g. WNK kinase). IL-6 females also had decreased dopamine excretion, suggesting that renal dopamine deficiency contributed to their earlier development of hypertension.
It has been reported that gender has a role in programming and that female sex might also impart a protective effect, especially in rat models of perinatal protein restriction (Woods et al. 2001, 2005). In that study, programming of renal and BP abnormalities appeared to require a longer or stronger stimulus in female offspring than in their male littermates. Thus, in comparing different studies, the species, timing and duration of programming stimuli, age of offspring, and phenotyping technique are important (McMullen & Langley-Evans, 2005). A recent study showed that maternal low-protein diet programmed hypertension through different sex-specific mechanisms. Males had GC-dependent hypertension without modulation of renal RAS, whereas females had GC-independent hypertension and reduced expression of renal AT2-R mRNA (McMullen & Langley-Evans, 2005). In male and female littermates with similarly low birth weights after prenatal dexamethasone exposure, only females developed hypertension with increases in circulatory RAS activity and hepatic angiotensinogen expression. Males had a decreased HPA axis feedback sensitivity with increased basal plasma corticosterone levels (O'Regan et al. 2004). These findings, together with our data, might indicate increased sensitivity to the reprogramming effects of alterations in the set-point and activity of the RAS in female rats.
Time of exposure
The prenatal IL-6 exposure model used in this study had two different time points of exposure: early and late pregnancy. The purpose of this was to evaluate the importance of critical windows in fetal development in relation to programming effects. It seems, however, that none of the programming effects are solely dependent on the time of exposure but instead are dependent on gender sensitivity for time effects (Table 7) as discussed above.
Prenatal exposure to IL-6 may contribute to hypertension in adult rats by altering the RAS set-point and reducing fluid and electrolyte secretion. These alterations were more pronounced in females, implicating sex-dependent sensitivity of this early programming effect.