Impaired EDHF-mediated vasodilatation in adult offspring of rats exposed to a fat-rich diet in pregnancy

Abstract

We recently reported vascular dysfunction in adult offspring of rats fed a fat-rich (animal lard) diet in pregnancy. This study reports further characterization of constrictor and dilator function in mesenteric and caudal femoral arteries from 180-day-old offspring of dams fed the high fat diet (OHF). Endothelium-dependent relaxation in response to acetylcholine (10⁻⁹–10⁻⁵m) was impaired in mesenteric small arteries from male and female OHF compared with offspring of dams fed normal chow (males (maximum percentage relaxation): OHF 67.92 ± 2.89, n = 8 versus control 92.08 ± 2.19, n = 8, P < 0.01). Substantial relaxation in response to acetycholine in control mesenteric arteries remained after inhibition of nitric oxide synthase, soluble guanylate cyclase and cyclo-oxygenase but was blocked by 25 mm potassium. This component of relaxation, attributed to EDHF, was significantly reduced in OHF mesenteric arteries compared with controls. However, EDHF played a minor role in acetylcholine-induced relaxation in both control and OHF femoral caudal arteries (male and female). In these arteries, in contrast to mesenteric vessels, acetylcholine-induced relaxation was significantly enhanced in OHF but only in males (ACh (maximum percentage relaxation): OHF 58.40 ± 4.39, n = 8 versus male controls 32.18 ± 6.36, P < 0.05). This was attributable to enhanced nitric oxide-mediated relaxation. In conclusion, reduced endothelium-dependent relaxation in OHF mesenteric arteries is due to impaired EDHF-mediated relaxation. This defect was not apparent in femoral arteries in which EDHF has a less prominent role.

Increasing evidence supports the concept that windows of vulnerability to adverse environmental stimuli in early life may predispose to adulthood disease (Gluckman & Hanson, 2004). In this study we have carried out a detailed investigation of functional abnormalities in small arteries from adult animals, acquired as a result of developmental ‘programming’ induced by maternal dietary imbalance.

Recent epidemiological (Roseboom et al. 2001) and animal studies (Ozaki et al. 2001; Ozanne & Hales, 2002; Khan et al. 2003) have suggested that the cardiovascular risk factors which cluster in the metabolic syndrome, hitherto attributed to genetic and adult environmental influences, can be acquired in utero. These include insulin resistance, hypertension, glucose intolerance, central adiposity and abnormal serum lipid profiles. We have recently developed an animal model in which adult offspring of rats fed a diet rich in saturated fat during pregnancy and suckling develop certain features of the metabolic syndrome including hypertension, dyslipidaemia, adiposity and altered glucose homeostasis (Khan et al. 2003, 2004). In addition, small mesenteric arteries of the offspring demonstrated marked reduction in relaxation in response to the endothelium-dependent agonist acetylcholine. Endothelial dysfunction has been implicated in insulin resistance and atherogenesis and reduced endothelium-dependent relaxation is an independent risk factor for cardiovascular disease and the metabolic syndrome (Bonora et al. 2003). To date, we have established that the defect in relaxation in the mesenteric small arteries cannot be attributable to altered vascular smooth muscle sensitivity to nitric oxide since relaxation in response to exogenously applied NO was unaffected (Khan et al. 2003).

The principal purpose of this study was to determine which of the different component pathways of endothelium-dependent dilatation contributes to the failure of endothelial function in the small mesenteric arteries from adult offspring of the fat-fed dams. The relative roles of nitric oxide, prostacyclin and the postulated endothelium-derived hyperpolarizing factor(s) (EDHF) have been studied. Further investigations of constrictor function were also undertaken including responses to angiotensin II, since altered activity of the renin–angiotensin axis has been implicated in other models of developmental programming of hypertension (Langley-Evans et al. 1996; Sahajpal & Ashton, 2003). In addition, in order to determine whether endothelial dysfunction was common to different vascular beds, arteries from the femoral circulation were studied.

Methods

Animal husbandry and experimental diets

Female Sprague-Dawley (100–120 days) rats were fed ad libitum, for 10 days prior to mating and throughout pregnancy, either a control diet of a standard laboratory chow (5.3% fat (corn oil), 21% protein, 57.4% carbohydrate, 4.6% fibre, vitamins and minerals: Rat and Mouse Diet no. 3; Special Diet Services, Witham, Essex, UK) or an experimental diet consisting of the standard chow supplemented 20% (w/w) with animal lard with 20% additional vitamins and minerals, protein, inositol and choline to correct for the dilution (final composition 25.7% fat (palmitic acid 4.50%, stearic acid 1.99%, palmitoleic acid 0.12%, oleic acid 6.86%, linoleic acid 2.58%, α-linolenic acid 0.21%, arachidonic acid 0.19%), 19.5% protein, 41.3% carbohydrate, 3.5% fibre; Special Diet Services) (Taylor et al. 2003). The efficacy of supplementation was confirmed by independent analysis of the diets (Eclipse Scientific Group, Cambridge, UK). At birth all litters were reduced to eight pups (4 male, 4 female). All animals were fed a normal balanced diet from weaning and were maintained under controlled conditions of temparature and humidity on a 12 h light–dark cycle.

Mesenteric and femoral artery functional reactivity studies

Isolated mesenteric and femoral artery vascular function was assessed in 180-day-old male and female offspring of control (OC) and fat-fed rats (OHF). Rats were killed by cervical dislocation. One male and one female offspring from each litter were studied. Third order branches of the mesenteric arcade and sections of the caudal femoral artery were dissected and mounted on a small vessel myograph and bathed in (physiological salt solution) PSS as previously described (Khan et al. 2003). Concentration–response responses were carried out in both caudal femoral and mesenteric arteries, to depolarizing potassium solution in PSS (10–125 mm), noradrenaline (NA; 10⁻⁷–10⁻⁵m), phenylephrine (PE; 10⁻¹¹–10⁻⁵m), angiotensin II (10⁻¹¹–10⁻⁵m) and, in arteries submaximally constricted with NA (80% of maximal concentration), to acetylcholine (ACh; 10⁻⁹–10⁻⁵m), and nitric oxide (NO; 10⁻⁸–10⁻⁵m).

Determination of the components of endothelium-dependent relaxation

To evaluate the contribution of cyclo-oxygenase products to ACh-mediated EDR, arteries were pretreated with the cyclo-oxygenase (COX) inhibitor indomethacin (10 μm, 30 min) and an ACh concentration–response curve was plotted in arteries submaximally preconstricted with NA. To establish the role of nitric oxide in EDR, the nitric oxide synthase (NOS) inhibitor N^ω-nitro-l-arginine methyl ester (l-NAME 0.1 mm, 30 min) and the soluble guanylate cyclase (sGC) inhibitor (1H-[1,2,4]oxadiazolo[4,3-a]quinoxaloin-1-one, ODQ, 1 μm, 30 min) were added in the continued presence of indomethacin and relaxation responses to ACh again determined in NA preconstricted arteries. In order to block the residual EDHF-mediated response, the arteries were pre-constricted with 25 mm potassium in physiological salt solution (KPSS) and NA (to a concentration required to achieve similar tone to the two previous pre-constrictions) in the continued presence of the NOS and COX inhibitors, and a further ACh concentration–response was then carried out.

Statistical analysis

All values are given as mean ± s.e.m. Pharmacological concentration–response curves were defined by the log concentration that produced half the maximum effect (EC₅₀) and by the maximum response (V_max). Sigmoidal functions were modelled to individual dose–response curves (Prism, GraphPad Software, San Diego CA, USA) and statistical comparisons of EC₅₀ values and maximum responses were made by ANOVA with Bonferroni correction for multiple comparisons, and statistical significance was assumed if P < 0.05. The study was powered for differences in vascular function based on previous studies (Khan et al. 2003). One male and one female were studied from each litter, and n refers the number of litters studied.

Results

Maternal weight and food intake

Maternal body weights and food intake for the control dams and dams fed a fat-rich diet in pregnancy and suckling have been reported previously (Khan et al. 2004). Maternal weight was greater in fat-fed dams until day 16 of gestation and also from birth until day 8 postpartum (P < 0.05). Maternal dietary intake increased during the suckling period in both groups, although reduced dietary intake in the fat-fed dams during gestation, previously reported in this model (Taylor et al. 2003), did not attain significance.

Offspring body weight and food intake

Offspring body weights and food intake for the control and fat fed dams have been reported previously (Khan et al. 2004). There were no differences in body weight or food intake between offspring of controls and offspring of fat-fed dams up to 180 days of age. In male and female offspring of fat-fed dams central adiposity was increased compared with controls, as assessed by the combined weight of abdominal and visceral fat lobes (fat mass as percentage body weight; males: OHF 4.97 ± 0.38, n = 10, versus OC, 3.60 ± 0.32, n = 10, P < 0.05; females: OHF, 4.44 ± 0.41 versus OC, 2.16 ± 0.33, n = 10, P < 0.001).

Vascular endothelium-dependent relaxation (EDR)

Acetylcholine-induced relaxation was significantly reduced in the mesenteric small arteries in male and female OHF when compared with controls (Table 1, Figs 1 and 2). In the femoral arteries, acetylcholine-induced relaxation was significantly enhanced in male OHF but unaltered in female OHF when compared with controls (Table 1, Figs 3 and 4).

Table 1.

Vascular function in isolated mesenteric and femoral small arteries from adult offspring of control (OC) and high-fat-fed (OHF) dams at 180 days of age

	OC male	OHF male	OC female	OHF female
Vascular parameters
Mesenteric	(n = 8)	(n = 8)	(n = 8)	(n = 8)
Lumen diameter (μm)	295.93 ± 15.23	277.13 ± 21.30	244.60 ± 20.18	262.71 ± 13.27
Maximal contraction
Noradrenaline (mN mm−1)	4.05 ± 0.23	4.06 ± 0.28	4.00 ± 0.66	3.90 ± 0.34
Phenylephrine (mN mm−1)	4.63 ± 0.41	4.06 ± 0.61	3.85 ± 0.41	3.47 ± 0.25
AII (mN mm−1)	1.35 ± 0.57	2.40 ± 0.47	0.96 ± 0.41	1.34 ± 0.59
K+ (mN mm−1)	5.47 ± 0.50	4.42 ± 0.68	4.60 ± 0.53	4.08 ± 0.35
Maximal relaxation
ACh (% relaxation)	92.08 ± 2.19	67.92 ± 2.89 B	93.74 ± 2.08	75.24 ± 7.56a
NO (% relaxation)	95.24 ± 1.97	95.20 ± 2.34	98.36 ± 0.82	96.22 ± 2.81
EC50
Phenlyephrine	− 5.70 ± 0.04	− 5.67 ± 0.06	− 5.74 ± 0.04	− 5.66 ± 0.04
AII	− 6.76 ± 0.10	− 6.68 ± 0.10	− 6.61 ± 0.18	− 6.38 ± 0.11
ACh	− 7.31 ± 0.04	− 7.44 ± 0.06	− 7.08 ± 0.07	− 7.19 ± 0.08
NO	− 6.73 ± 0.08	− 6.62 ± 0.05	− 7.08 ± 0.08	− 6.83 ± 0.06
Femoral	(n = 8)	(n = 8)	(n = 10)	(n = 12)
Lumen diameter (μm)	331.19 ± 11.99	307.93 ± 20.52	311.04 ± 15.92	279.14 ± 13.22
Maximal contraction
Noradrenaline (mN mm−1)
Phenylephrine (mN mm−1)	5.31 ± 0.33	4.75 ± 0.88	4.96 ± 0.53	5.17 ± 0.57
All (mN mm−1)	3.01 ± 0.42	1.56 ± 0.52 a	3.60 ± 0.40	2.53 ± 0.27 a
K+ (mN mm−1)	6.43 ± 0.46	6.02 ± 1.07	5.60 ± 0.55	5.83 ± 0.69
Maximal relaxation
ACh (% relaxation)	32.18 ± 6.36	58.40 ± 4.39 a	31.42 ± 4.08	25.45 ± 6.10
NO (% relaxation)	87.80 ± 5.00	92.64 ± 3.92	85.75 ± 4.31	87.71 ± 4.24
EC50
Phenylephrine	− 5.53 ± 0.06	− 5.57 ± 0.05	− 5.68 ± 0.07	− 5.72 ± 0.08
AII	− 5.96 ± 0.23	− 6.30 ± 0.28	− 6.62 ± 0.22	− 6.69 ± 0.18
ACh	− 7.32 ± 0.12	− 7.40 ± 0.10	− 7.10 ± 0.16	− 6.94 ± 0.14
NO	− 6.53 ± 0.07	− 6.56 ± 0.07	− 6.93 ± 0.27	− 6.52 ± 0.08

Values expressed as mean ± s.e.m.

^aP < 0.05 OC versus OHF;

^bP < 0.01 OC versus OHF. All comparisons within same sex.

Figure 1. Endothelium-dependent relaxation to acetylcholine in third order mesenteric arteries of male and female offspring of dams fed a control diet during pregnancy and lactation.

In PSS (○, n = 8), after incubation and in continued presence of indomethacin (▵, n = 5), after incubation and in continued presence of indomethacin, l-NAME and ODQ (▿, n = 5) and after incubation with indomethacin, l-NAME and ODQ and 25 mm K⁺ (□, n = 5). Data given as mean ± s.e.m.

Figure 2. Endothelium-dependent relaxation to acetylcholine in third order mesenteric arteries of male and female offspring of dams fed the fat-rich diet throughout pregnancy and lactation.

In PSS (•, n = 8), after incubation and in continued presence of indomethacin (▴, n = 5), after incubation and in continued presence of indomethacin, l-NAME and ODQ (▾, n = 5) and after incubation with indomethacin, l-NAME, ODQ and 25 mm K⁺ (▪, n = 5). Data given as mean ± s.e.m.

Figure 3. Endothelium-dependent relaxation to acetylcholine in third order femoral arteries of male and female offspring of dams fed a control diet during pregnancy and lactation.

In PSS (OC,○, n = 8), after incubation and in continued presence of indomethacin (▵, n = 5), after incubation and in continued presence of indomethacin, l-NAME and ODQ (▿, n = 5) and after incubation with indomethacin, l-NAME, ODQ and 25 mm K⁺ (□, n = 5). Data given as mean ± s.e.m.

Figure 4. Endothelium-dependent relaxation to acetylcholine in third order femoral arteries of male and female offspring of dams fed the fat-rich diet throughout pregnancy and lactation.

In PSS (•, n = 8), after incubation and in continued presence of indomethacin (▴, n = 5), after incubation and in continued presence of indomethacin, l-NAME and ODQ (▾, n = 5) and, after incubation with indomethacin, l-NAME, ODQ and 25 mm K⁺ (▪, n = 5). Data given as mean ± s.e.m.

Investigation of components of endothelium-dependent relaxation

Endothelium-dependent relaxation in the mesenteric vessels of control males and females was largely attributable to EDHF(s) as incubation with indomethacin, l-NAME and ODQ resulted in only a 25% reduction in maximal relaxation in males and a 37% reduction in females (Table 2, Fig. 1). In contrast, the endothelium-dependent relaxation in OHF mesenteric arteries which was blunted compared with controls, was largely mediated by COX and NO pathways. The reduced relaxation in response to ACh occurred as a result of a diminished EDHF component which was considerably reduced in OHF compared with controls, accounting for only 10% of maximal relaxation in both males and females (Table 2, Fig. 2).

Table 2.

Maximal relaxation to ACh in the presence and absence of pharmacological inhibitors in third order male and female mesenteric arteries in offspring of control (OC) and fat fed (OHF) groups at 180 days of age

Maximal relaxation (%)	OC male	OHF male	OC female	OHF female
ACh (n = 8)	92.08 ± 2.19	77.92 ± 2.87 b	93.74 ± 2.08	75.23 ± 7.56 a
ACh + indomethacin (n = 5)	91.07 ± 3.08	45.87 ± 5.29 cd	68.92 ± 3.01 d	38.91 ± 9.21 bd
ACh + indomethacin +l-NAME + ODQ (n = 5)	75.64 ± 1.64 e	11.91 ± 3.09 ce	63.89 ± 5.22 e	7.94 ± 4.35 ce
ACh + indomethacin, l-NAME + ODQ +25 mm K+(n = 5)	− 2.5 ± 2.5 fgh	− 2.67 ± 2.68 fgh	− 0.03 ± 1.2 fgh	− 2.00 ± 2.00 fgh

Values given as mean ± s.e.m.

^aP < 0.05 OC versus OHF;

^bP < 0.01 OC versus OHF;

^cP < 0.001 OC versus OHF;

^dP < 0.01 for response to ACh versus response to ACh in the presence of indomethacin within the same experimental group;

^eP < 0.01 for response to ACh versus response to ACh in the presence of indomethacin, l-NAME and ODQ within the same experimental group;

^fP < 0.001 for response to ACh versus response to ACh in the presence of indomethacin, l-NAME and ODQ and 25 mm K⁺ within the same experimental group;

^gP < 0.001 for response to ACh in the presence of indomethcin versus response to ACh in the presence of indomethacin, l-NAME and ODQ and 25 mm K⁺ within the same experimental groups;

^hP < 0.05 for response to ACh in the presence of indomethcin l-NAME and ODQ versus response to ACh in the presence of indomethacin, l-NAME and ODQ and 25 mm K⁺ within the same experimental groups, by ANOVA with Bonferroni post hoc test for multiple comparisons.

The relative contribution of each of the components of endothelium-dependent relaxation in the control femoral arteries were different from the mesenteric vessels. The response to ACh in the control femoral arteries from both male and female animals was biphasic, with relaxation at lower concentrations and contraction at higher concentrations. The addition of indomethacin led to the inhibition of the constrictor component of the response, indicating ACh-induced stimulation of constrictor prostanoid synthesis. The majority of the relaxation response to ACh was evoked by NO since the addition of l-NAME and ODQ almost entirely inhibited relaxation and the EDHF component, the residual relaxation inhibitable by K⁺, was subsequently very small (Table 3, Fig. 3).

Table 3.

Maximal relaxation to ACh in the presence and absence of pharmacological inhibitors in third order male and female femoral arteries in offspring of control (OC) and fat-fed (OHF) groups at 180 days of age

Maximal relaxation (%)	OC male	OHF male	OC female	OHF female
ACh (n = 8)	32.18 ± 6.36	58.40 ± 4.38 b	31.42 ± 4.08	25.45 ± 6.10
ACh + Indomethacin (n = 5)	43.43 ± 3.57	57.23 ± 5.20	45.61 ± 4.71	18.84 ± 4.40 a
ACh + Indomethacin, l-NAME + ODQ (n = 5)	9.25 ± 3.15 d	10.59 ± 4.15d	10.91 ± 1.81d	11.99 ± 7.25
ACh + Indomethacin, l-NAME + ODQ +25 mm K+(n = 5)	2.25 ± 2.59eg	− 3.29 ± 4.20egf	− 0.18 ± 1.41 egf	1.64 ± 0.64eg

Values given as mean ± s.e.m.

^aP < 0.05 OC versus OHF;

^bP < 0.01 OC versus OHF;

^cP < 0.01 for response to ACh versus response to ACh in the presence of indomethacin within the same experimental group;

^dP < 0.01 for response to ACh versus response to ACh in the presence of indomethacin, l-NAME and ODQ within the same experimental group;

^eP < 0.01 for response to ACh versus response to ACh in the presence of indomethacin, l-NAME and ODQ and 25 mm K⁺ of same group;

^fP < 0.05 for response to ACh in the presence of indomethacin versus response to ACh in the presence of indomethacin, l-NAME and ODQ and 25 mm K⁺ within the same experimental groups;

^gP < 0.05 for response to ACh in the presence of indomethcin, l-NAME and ODQ versus response to ACh in the presence of indomethacin, l-NAME and ODQ and 25 mm K⁺ within the same experimental group, by ANOVA with Bonferroni post test for multiple comparisons.

Femoral arteries from male OHF, which showed an increase in ACh-mediated relaxation compared with controls, also demonstrated a biphasic constrictor response to higher concentrations of ACh, which was similarly blocked by indomethacin. This biphasic response was absent in female OHF, and the addition of indomethacin consequently reduced the relaxation in response to ACh. In both male and female femoral arteries, addition of l-NAME and ODQ almost entirely inhibited relaxation. The greater relaxation observed in male OHF compared to controls was the result of an enhanced NO-mediated component of relaxation (Table 3, Fig. 4).

Endothelium-independent relaxation

Relaxation in response to aqueous nitric oxide was similar in control and OHF groups in both mesenteric arteries and caudal femoral arteries and between males and females (Table 1).

Investigation of vascular contractile function

Sensitivity (EC₅₀) and maximal responses to NA and PE were not different from OC in male or female OHF in mesenteric or femoral arteries, when expressed either as absolute tension or as a percentage of maximal constriction to 5 μm NA and 125 mm potassium substituted PSS. Responses to AII were significantly reduced in the femoral arteries of male and female OHF, whereas no significant differences were observed in the mesenteric vessels (Fig. 5). Responses to increasing potassium concentrations suggested a reduced contractility in the male and female OHF mesentery, but this did not reach significance (Table 1).

Figure 5. Responses to All in mesenteric (A) and femoral (B) resistance arteries of male and female offspring.

Control dams (OC,○, n = 8) or offspring of fat-fed dams (OHF, •, n = 8). Data are expressed as mean (mN mm⁻¹) ± s.e.m. *P < 0.05 offspring of control dams versus offspring of fat-fed dams for maximal response.

Discussion

The principal finding of the current study is that the defect in endothelium-dependent relaxation reported previously in mesenteric small arteries from adult offspring of fat-fed dams (Khan et al. 2003) is most likely attributable to reduced EDHF synthesis or responsiveness. This study has also shown that the defect in relaxation in response to acetylcholine was not apparent in femoral arteries, and that this is likely to reflect the relative absence of an EDHF-mediated component of relaxation in the femoral vessels studied.

Acetylcholine evokes relaxation through release of NO (Moncada & Higgs, 1991), prostacyclin (PGI₂) (Vane & Corin, 2003) and a putative EDHF (Garland & McPherson, 1992). In this study, the defect in maximal relaxation in response to acetylcholine in the mesenteric circulation of offspring of fat-fed dams was observed in the presence of cyclo-oxygenase blockade and also in combination with nitric oxide synthase and soluble guanylyl cyclase inhibition. Residual relaxation in the presence of cyclo-oxygenase (COX) and NOS blockade was totally inhibited in both experimental and control groups by depolarization to potassium, thus demonstrating that the component of relaxation dependent upon smooth muscle hyperpolarization (presumed to be EDHF) was lower in the offspring of the fat-fed dams. The contribution of EDHF to endothelium-dependent relaxation generally increases with decreasing artery size (Hill et al. 2001) such that in small arteries, which play a major role in peripheral resistance and blood pressure EDHF is generally considered to be the principal endothelium-derived dilator (Busse et al. 2002).

A generalized defect of EDHF-mediated relaxation in the peripheral vasculature, as observed in the mesenteric small arteries, may have ‘global’ functional consequences for cardiovascular and metabolic homeostasis. Reduced EDHF-mediated vasodilatation has been associated with experimental animal models of hypertension (Bussemaker et al. 2003) and could contribute to the hypertension observed in our model. However, this would not explain why only the female offspring in this model are hypertensive (Khan et al. 2003, 2004), although the increased endothelium-dependent relaxation in the femoral arteries of male offspring of fat-fed dams may reduce resistance in the hindlimb circulation and afford some protection against a rise in blood pressure. A defect in EDHF may arise from the abnormal glucose homeostasis we have observed in these offspring (Khan et al. 2003, 2004), since EDHF-mediated relaxation is impaired in insulin-resistant rats (Katakam et al. 1999) and STZ diabetic rats (Wigg et al. 2001; Matsumoto et al. 2003).

In the mesenteric small arteries of male and female controls maximal relaxation in response to acetylcholine was only modestly affected by COX, NOS and sGC inhibitors whereas relaxation in offspring of fat-fed dams was almost totally absent after these additions. This may suggest compensatory up-regulation of PGI₂ and NO pathways in the face of reduced EDHF-mediated relaxation. We have reported a similar interaction between the NO–PGI₂ pathways and EDHF in pregnant animals in which acute inhibition of NOS was associated with an increase in EDHF-mediated relaxation (Gerber et al. 1998). Others have demonstrated that, under physiological conditions, the production of EDHF can be reduced by over production of NO (Bauersachs et al. 1996; McCulloch et al. 1997). It may be relevant that up-regulation of nitric oxide synthesis occurs with the inflammatory response (Holm et al. 2002), a feature of the metabolic syndrome (Bonora et al. 2003). The observation of an increased NO-mediated component of relaxation in the femoral arteries of the male offspring of fat-ded dams would lend some support to this suggestion.

In an earlier study from our laboratory we reported reduced acetylcholine-induced relaxation in the femoral arteries of the female offspring in 160-day-old animals (Ghosh et al. 2001), whilst no abnormality was found in the present study. Smaller litter sizes in the earlier study may in part explain this difference. Litters were reduced to five or six pups compared with eight in the present study, a subtle difference in protocol which may influence plasma lipid profiles (Hahn, 1984) and thereby impair endothelium-dependent relaxation. Indeed, plasma triglyceride concentrations were raised in the experimental group by 160 days in the earlier study (Ghosh et al. 2001) but are not significantly elevated until 1 year of age with the current protocol (Khan et al. 2003).

Given the specific EDHF defect observed in the mesenteric arteries, the lack of overt endothelial dysfunction in the femoral arteries was not surprising since EDHF makes a minor contribution to acetylcholine-mediated relaxation in these vessels. The small femoral arteries may not, therefore, be a faithful model for the peripheral vasculature where EDHF is proposed to be the predominant endothelium-derived vasodilator. Absence of EDHF-mediated relaxation in femoral arteries has also been reported by several other groups (Zygmunt et al. 1995; Wigg et al. 2001; Sandow et al. 2002), although a study in different branches of femoral arteries (profunda femoris) from Wistar rats, but of a similar size to those used in this study has reported a substantial EDHF component of relaxation (Savage et al. 2003).

Whilst the identity of EDHF is still unclear, it is considered to be one of three substances either an epoxide derivative of cytochrome P450 pathway (e.g. EET or HETE), the cation K⁺ (Edwards et al. 1998) or anandamide (Randall et al. 1996). More recently, it has been suggested that EDHF may be preformed stores of NO resistant to nitric oxide synthase inhibition (Chauhan et al. 2003). Therefore, in our model there could both be a defect in the EDHF component of relaxation and/or a reduced capacity for NO storage and release. The observation that the sGC inhibitor ODQ failed to block residual relaxation in response to acetylcholine could argue against a role for preformed NO stores. However, NO may also relax smooth muscle via sGC-independent pathways through direct activation of K⁺ channels (Bolotina et al. 1994; Cohen et al. 1997).

We have previously demonstrated a reduction in the aorta arachidonic acid content of female adult offspring of fat-fed dams (Ghosh et al. 2001). Should EDHF in the rat mesenteric circulation be a cytochrome P450 metabolite of arachidonic acid, a reduction of vascular arachidonic acid could reduce EDHF synthesis. It must be also be considered that reduced smooth muscle sensitivity to EDHF as well as reduced synthesis may contribute to the blunted EDHF component of relxation observed, as has been observed in ageing spontaneously hypertensive rats (SHR) (Bussemaker et al. 2003).

Reduced AII responses were observed in the femoral circulation in both male and female offspring of fat-fed dams compared with controls. Modification of the renin–angiotensin system (RAS) has been demonstrated in other models of fetal programming. However, in contrast to the reduced constrictor response we observed in the femoral circulation, offspring of low-protein-fed dams demonstrate enhanced blood pressure responses (McMullen et al. 2004) and a reduction in glomerular filtration rate when infused with AII (Sahajpal & Ashton, 2003). This may reflect differences in tissue AII receptor distribution or, more simply, difference in the offspring phenotypes of the two dietary intervention models.

In conclusion, this study has demonstrated that the defect in endothelium-mediated relaxation in the mesenteric small arteries previously reported results from reduced EDHF synthesis or sensitivity. The branches of the femoral artery studied did not show the same defect since EDHF has a smaller contribution to relaxation in these vessels.

Supplementary material

The online version of this paper can be accessed at:

10.1113/jphysiol.2004.018879

http://jp.physoc.org/cgi/content/full/jphysiol.2002.018879/DC1 and contains supplementary material entitled:

Predictive adaptive responses to maternal high fat diet prevent endothelial dysfunction but not hypertension in adult rat offspring.

This material can also be found at:

http://www.blackwellpublishing.com/products/journals/suppmat/tjp/tjp378/tjp378sm.htm