It is well established that diseases in adult life such as hypertension and CVD, and the subsequent development of the metabolic syndrome emerge as a consequence of interplay between genetic and environmental factors. Recent research has concentrated on undernutrition in pregnancy and the role that it may play in the onset of disease(Reference Gluckman and Hanson1–Reference Eriksson, Forsen, Tuomilehto, Osmond and Barker3). It has been postulated that undernutrition programmes long-term changes in gene expression, which alter metabolism in the developing fetus and result in cardiovascular abnormalities in later life(Reference Barker, Gluckman, Godfrey, Harding, Owens and Robinson4). Whilst the origins of the metabolic syndrome are clearly multifactorial, nutrition in early life may have a profound influence upon risk and upon the responses of the individual to environmental and lifestyle-related risk factors in adulthood. The expression of genes that either predispose or protect against these conditions will be further modified by interactions between the genotype, early life nutrition and the postnatal environment(Reference Langley-Evans5).
Reports of epidemiological associations between birth weight and CVD(Reference Barker, Gluckman, Godfrey, Harding, Owens and Robinson4, Reference Barker, Eriksson, Forsen and Osmond6, Reference Langley-Evans, Gardner and Jackson7) are supported by animal experiments showing that both undernutrition and overnutrition in fetal life can programme adult blood pressure(Reference Bergel and Belizan8–Reference Langley and Jackson11). We have demonstrated that, in rats, the feeding of a maternal low-protein diet (MLP) during gestation programmes a lifelong elevation of blood pressure(Reference Langley-Evans, Gardner and Jackson7, Reference Langley and Jackson11–Reference Langley-Evans, Welham, Sherman and Jackson14). Evidence from both human and animal studies indicates that the kidney, and specifically nephron number, may play an important role in the programming of hypertension(Reference Mackenzie and Brenner15). In the MLP rat model of nutritional programming, fetal exposure to undernutrition reduces nephron number by as much as 30 %(Reference Langley-Evans, Welham and Jackson16).
A number of studies have demonstrated that prenatal undernutrition has the capacity to modulate the epigenetic regulation of gene expression(Reference Jaenisch and Bird17, Reference Burdge, Hanson, Slater-Jefferies and Lillycrop19). This raises the prospect that periods of undernutrition can establish heritable changes to the epigenome and, as such, the disease programming effects of undernutrition in the fetal period may not be limited to the first generation. Indeed, there is emerging evidence from studies of humans and animals to suggest that transgenerational effects may occur, whereby the consequences of deficits in maternal nutrition are subsequently passed onto the grandchildren(Reference Beach, Gershwin and Hurley20–Reference Pembrey22). The aim of the present study was to assess the potential for both a prenatal insult of maternal protein restriction and a postnatal high-fat (HF) challenge to impact across several generations.
Materials and methods
Animal protocols
The experiments described in the present report were performed under license from the Home Office in accordance with the 1986 Animals (Scientific Procedures) Act. The study used rats of the Wistar strain, and all animals were housed in plastic cages and subjected to a 12 h light–dark cycle at a temperature of 20–22°C. The animals had ad libitum access to food and water at all times.
Maternal procedure
Fig. 1 summarises the overall design of the present experiment. Twenty virgin female Wistar rats (Charles River UK Ltd, Margate, Kent, UK) were mated with a single stud male at between 180 and 220 g. Upon confirmation of mating by the presence of a semen plug on the cage floor, the rats were allocated to be fed either an 18 % (w/w) casein (control; CON) or a 9 % (w/w) casein (MLP) diet throughout gestation, as described previously(Reference Langley-Evans, Welham, Sherman and Jackson14). The diets were isoenergetic, the difference in protein-derived energy between the two diets being made up by the addition of carbohydrate (starch–sucrose, 2:1, w/w). During pregnancy the animals were weighed and food intake was recorded daily. At the time of birth (day 22) the animals were transferred to a standard laboratory chow diet (B&K Universal Ltd, Hull, UK).
Schematic showing study design. CON, control; MLP, maternal low-protein; experimental endpoints, determination of body composition.
F1 offspring procedure
At birth, litters were sexed and weighed. All litters were culled to eight pups (four male and four female) to ensure a standard plane of nutrition. At approximately 3 weeks of age the offspring were micro-chipped using the Avid micro-chipping system and allocated into two sub-groups, breeders and nutritional challenge (NC) (two males and two females per group). Breeders and one male and one female from the NC group were allocated to the standard chow diet (B&K Universal Ltd). The remaining NC animals were allocated to be fed a HF diet. All animals were housed in single-sex groups and had ad libitum access to food and water at all times.
At approximately 10 weeks of age, two males and two females (breeders sub-group) from each F1 litter were utilised in a breeding programme for production of the F2 generation. The four animals from each litter allocated for breeding were crossed with animals from another litter in order to produce the four possible crosses from the F1 generation (CON male × CON female, MLP male × MLP female, CON male × MLP female and MLP male × CON female). Upon confirmation of mating by the presence of a semen plug on the cage floor the female rats were singly housed and fed the standard chow diet until delivery of the litter.
F2 offspring procedure
At birth, litters were sexed and weighed. All litters were culled to eight pups (four male and four female) to ensure a standard plane of nutrition. At approximately 3 weeks of age the offspring were micro-chipped and allocated into breeders and NC sub-groups as described above. Breeders and one male and one female from the NC group were allocated to the standard chow diet (B&K Universal Ltd). The remaining NC animals were allocated to be fed a HF diet. The animals were housed in single-sex groups and had ad libitum access to food and water at all times.
At approximately 10 weeks of age, one male and one female (breeders sub-group) from each litter were crossed with animals from another litter in order to produce the following crosses based upon the dietary exposures of their parents in the F1 generation: CON male × CON female and MLP male × MLP female. Upon confirmation of mating by the presence of a semen plug on the cage floor the female rats were singly housed and fed standard chow until delivery of the litter. Male rats were placed back into single-sexed housing with their littermates, until studied later in life.
F3 offspring procedure
At birth, litters were sexed and weighed. All litters were culled to eight pups (four male and four female) to ensure a standard plane of nutrition. At approximately 3 weeks of age the offspring were micro-chipped and allocated to the standard chow diet (B&K Universal Ltd). One male and one female animal were allocated to be fed a HF diet. The animals were housed in single-sex groups and had ad libitum access to food and water at all times.
High-fat feeding procedure
Rats (where possible one male and one female offspring from each NC sub-group per litter in each generation) were allocated to a HF feeding protocol to assess the impact of this NC upon subsequent weight gain and body composition, from approximately 4 weeks of age. The HF diet was identical to that used in our previous work(Reference Erhuma, Bellinger, Langley-Evans and Bennett23) and contained 29·5 % fat (w/w) in the form of lard, and comprised 20 % (w/w) protein (casein). The gross energy content of the diet was 25·12 MJ/kg. A further group (consisting of one male and one female offspring from each NC group per litter in each generation) were fed the standard laboratory chow diet as a control. This diet contained 19 % protein, 4·7 % fat and had a gross energy content of 16·39 MJ/kg. All animals from this group were housed in pairs and provided with ad libitum access to the HF diet or standard laboratory chow. At 5, 7 and 9 weeks of age the animals were singly housed and food intake and body weight was monitored for a period of 3 d. Food intake data are shown corrected for body weight as in our previous studies of appetite in this model(Reference Bellinger, Lilley and Langley-Evans24). This corrects for any influence of body weight upon food intake, and allows comparison between sexes.
Determination of blood pressure
Blood pressure was determined in all animals using an indirect tail-cuff method, as described previously(Reference Sherman and Langley-Evans25). Measurements were made using the IITC Life Sciences Model 229 Blood Pressure amplifier/pump (IITC Inc., Woodlands, CA, USA). This instrument allows measurements at a temperature of 27°C, thereby avoiding heat stress of the animals. All rats were housed at this temperature, in the room where the measurements were made for at least 2 h before testing. Measurements were made at the same time of day following standard procedures to minimise variation due to diurnal changes in blood pressure. Blood pressure was determined at 4, 6 and 8 weeks of age. Measurements were taken in triplicate and an average value was derived for each animal.
Culling of animals
At approximately 10 weeks of age the NC sub-group animals were killed to coincide with matings for the next generation.(using a rising concentration of CO2 and cervical dislocation). Liver, heart, kidneys, lungs, spleen, thymus, hippocampus, hypothalamus, gonadal fat and perirenal fat were removed, accurately weighed and then snap-frozen in liquid N2 and stored at −80°C until used for further analysis. The left kidney was fixed in formalin for later determination of nephron number.
Nephron number determination
Nephron number was determined using an adaptation of the acid maceration method of Welham et al. (Reference Welham, Wade and Woolf26). Although determination of nephron complement via stereology(Reference Bertram27) is the gold standard method of analysis, we used maceration in the present study due to the high throughput required. The present results and overall effect match those obtained via stereology in studies of the MLP diet in the F1 generation(Reference Zimanyi, Bertram and Black28). Formalin fixed kidneys were weighed, cut in half and one portion was then incubated in 1 m-hydrochloric acid at 37°C for 30 min. Acid was removed and replaced with 5 ml 50 mm-PBS (pH 7·4). The tissue was then homogenised using a bench-top homogeniser (Polytron) and a further 5 ml PBS were added to give a final volume of 10 ml. The sample was mixed thoroughly by inversion and a 20 μl sample was taken and placed on a microscope slide and overlain with a cover slip. A × 10 objective lens was used to count all glomeruli in the sample. This was carried out in triplicate for each tissue sample. The values obtained were averaged and used to calculate the total number of glomeruli per kidney using the following equation:
Determination of circulating metabolites
Cholesterol and TAG assays
Total plasma cholesterol and plasma TAG concentrations were determined using commercially available kits, following the manufacturer's instructions (Thermo Life Sciences, Basingstoke, Hants, UK).
Glucose assay
Plasma glucose concentrations were determined in samples from non-fasted animals, using an adaptation of the glucose oxidase method of Trinder(Reference Trinder29). A standard curve was produced by making serial dilutions of the glucose standard (0–2 μg glucose). Samples were diluted 1:5 with phosphate buffer and loaded with standards in duplicate onto a microtitre plate in 10 μl quantities; 200 μl of glucose reagent was then added to each well. The plate was then incubated at 37°C for 15 min and then read at an absorbance of 620 nm using Magellan, version 4.0 software and plate reader (SunriseTM; Tecan Group Ltd, Männedorf, Switzerland). The inter-assay CV was 2·99 %.
Statistical analysis
All data were analysed using the Statistical Package for Social Sciences (version 14; SPSS, Inc., Chicago, IL, USA). Differences between groups were assessed using a mixed model ANOVA (fixed factors, maternal diet, sex and age), unless otherwise indicated in the text. Values are expressed as mean values with their standard errors. P < 0·05 was considered as significant. As multiple pups from the same dam were used throughout the present study, litter of origin was included as a fixed nested factor in all analyses(Reference Festing30). Analyses were performed within generations, with no consideration of influences between generations, except for a comparison of blood pressure across the three generations. Within each generation the 18 % casein CON diet (F1) or the CON × CON breeding cross (F2 and F3) was used as the control group.
Results
Birth outcomes
The number of successful pregnancies, litter size, birth weight and the ratio of male and female offspring was unaffected by maternal diet or cross in all three generations when compared with CON animals (data not shown). Average litter size was fourteen, with a ratio of seven males to six female offspring.
Blood pressure
F1 generation
Systolic blood pressure at 8 weeks of age was significantly increased in males fed either the HF (Fig. 2 (b)) diet (19 mmHg higher) or the standard chow (Fig. 2 (a)) diet (9 mmHg higher) postnatally, in the offspring (F1) of animals that had been subjected to maternal protein restriction during pregnancy (P < 0·001), when compared with CON animals of the same age. This effect was also noted in the females (Figs. 2 (c) and (d)) with an increase of 20 mmHg in animals on the HF diet and an increase of 3 mmHg (P < 0·001) in animals maintained on the standard chow diet. This trend of increased blood pressure was also apparent at 6 weeks of age (P < 0·001) (data not shown). Blood pressure at 4 weeks of age (data not shown) in the F1 generation was significantly lower in animals exposed to maternal protein restriction, relative to CON animals (P < 0·001). This result occurred independent of sex, although a postnatal HF diet did substantially reduce systolic blood pressure at this age (P < 0·002).
Systolic blood pressure at 8 weeks of age in (a) male offspring fed a standard chow diet, (b) male offspring fed a high-fat diet, (c) female offspring fed the standard chow diet and (d) female offspring fed the high-fat diet. Data are means (n 7 to 35 observations per group), with standard errors represented by vertical bars. CON, maternal control diet; MLP, maternal low-protein diet. ANOVA indicated significant effects of maternal diet (F1) (P < 0·001) and cross (F2) (P < 0·001).
F2 generation
The systolic blood pressures of the F2 generation were significantly influenced by the original F0 dietary intervention with prenatal MLP exposure of either male or female F1 parents increasing blood pressure (Figs. 2 (a), (b), (c) and (d)). Systolic blood pressure at 8 weeks of age was significantly increased (P < 0·001) in the offspring from all of the breeding crosses where the parents were originally subjected to protein restriction during pregnancy, when compared with CON animals. The increase in blood pressure was also apparent at 6 weeks of age (P < 0·005; data not shown). There was no effect on systolic blood pressure observed at 4 weeks for this generation. There was no influence of sex or any interaction of postnatal diet upon systolic blood pressure at 8 weeks of age.
F3 generation
The systolic blood pressures of the F3 generation at 8 weeks of age were unaffected by the protein restriction of the original F0 dams. It was noticeable that blood pressures of F3 animals derived from the F0 dams fed the CON diet in pregnancy were substantially higher than observed in their F1 equivalent group (Figs. 2 (a), (b), (c) and (d)).
Nephron number
In animals of the F1 generation, total nephron number (Fig. 3) was significantly reduced by 33 % in males and 35 % in females among the offspring of rats that were subjected to maternal protein restriction during pregnancy (P < 0·001). The F2 generation were similarly affected, exhibiting a reduction in nephron number of between 37 and 47 % in males and between 31and 39 % in females (P < 0·05) derived from the breeding crosses where the parents were originally subjected to protein restriction during fetal life. No reduction in nephron number was observed in the F3 generation.
Nephron number at 10 weeks of age in male (□) and female (■) offspring. Data are means (n 8 to 23 observations per group), with standard errors represented by vertical bars. CON, maternal control diet; MLP, maternal low-protein diet. ANOVA indicated significant effects of maternal diet (F1, P < 0·001) and cross (F2, P < 0·05).
Body composition
F1 generation
Among males of the F1 generation, body weight (Table 1) at cull was similar in CON and MLP-exposed rats that were fed the chow diet in both male and female animals. The HF diet did not increase body weight in the CON or MLP group in females; however, MLP-exposed male animals exhibited a significantly decreased body weight. All organs (data not shown) and fat pads (Table 1) were of similar size relative to body weight in the chow-fed male and female animals, irrespective of maternal dietary exposure. HF feeding in the F1 generation increased the size of the thymus in males from both the CON and MLP groups, and in the MLP group produced an enlarged spleen. Heart size relative to body weight was significantly reduced by HF feeding (data not shown). HF exposure postnatally in females substantially increased liver (P < 0·05), lung (P < 0·05), spleen (P < 0·05), gonadal fat pad (P < 0·05) and perirenal fat pad (P < 0·05) size relative to body weight in both CON and MLP groups (data not shown).
(Mean values with their standard errors)
CON, control; HF, high-fat; MLP, maternal low-protein.
* For F2 and F3 crosses the dietary origin of the male parent is shown before the female parent; for example, CON × MLP indicates a cross between a male exposed to the control diet and a female exposed to the MLP diet in utero.
F2 generation
F2 generation body composition did not differ greatly from the F1 generation. Body weight at cull was similar among all groups of male animals on standard chow and similar to the F1 generation. HF feeding did not increase body weight in the CON group; however, there was a trend towards a decreased body weight in the MLP-exposed animals although this was not statistically significant. F2 female body weight did not differ significantly with regards to parental cross in animals exposed to the postnatal chow diet; however, HF feeding postnatally generally decreased body weight (Table 1) in all crosses with the exception of the CON × MLP cross. All organs (data not shown) and fat pads were of similar size relative to body weight in the chow-fed male animals. However, HF feeding (Table 1) in both the males and females significantly decreased perirenal and gonadal fat deposit size relative to body weight in the animals of all cross groups involving a MLP-exposed parent (P < 0·05).
F3 generation
In the F3 generation, all organs (data not shown) and fat pads were of similar size relative to body weight in the chow-fed male and female animals. HF feeding decreased body weight, and increased left kidney and heart size relative to body weight relative to their CON counterparts.
Circulating metabolites
Plasma glucose (Table 2) concentrations were unaffected by the original manipulation of the maternal diet, by sex or by HF feeding in any of the three generations. Plasma cholesterol concentrations were also unaffected by maternal diet or cross in all three generations; however, HF feeding substantially increased plasma cholesterol concentrations in both the F1 and F3 CON and MLP male groups (P < 0·05). Plasma TAG concentrations in the F1 generation were increased with HF feeding in the MLP-exposed male group (0·93 mmol/l) compared with CON animals (0·57 mmol/l) (P < 0·05). Plasma TAG concentrations were higher in F2 groups derived from MLP-fed F0 dams when they were maintained on the standard chow diet, compared with the CON × CON group (P < 0·05). TAG concentrations were similar in the F3 generation irrespective of cross.
(Mean values with their standard errors)
CON, control; HF, high-fat; MLP, maternal low-protein.
* For F2 and F3 crosses the dietary origin of the male parent is shown before the female parent; for example, CON × MLP indicates a cross between a male exposed to the control diet and a female exposed to the MLP diet in utero.
Food intake
Food intake (Table 3) at 5 and 7 weeks of age (data not shown) was unaffected by maternal diet or cross in the F1 and F3 generations although there was an effect of postnatal diet (P < 0·001) in the F1 generation with rats consuming roughly double the weight of food per d per kg body weight when maintained on the standard chow diet. In the F2 generation (P < 0·001) it was apparent that offspring of the MLP × CON cross had an increased food intake when on a HF diet (P < 0·001), relative to all other groups at 5 weeks of age. By 9 weeks of age there was an effect of postnatal diet (P < 0·002) in all three generations whereby animals on standard chow consumed on average more g/d per kg body weight than those fed the HF diet.
(Mean values with their standard errors)
CON, control; HF, high-fat; MLP, maternal low-protein.
* For F2 and F3 crosses the dietary origin of the male parent is shown before the female parent; for example, CON × MLP indicates a cross between a male exposed to the control diet and a female exposed to the MLP diet in utero.
Discussion
In the present paper we have considered the potential for intra-uterine protein restriction to impact upon blood pressure, renal development and body composition across several generations. We have previously characterised the programming effects of feeding a MLP diet in rat pregnancy upon blood pressure(Reference Langley and Jackson11, Reference Langley-Evans, Phillips and Jackson12, Reference Langley-Evans, Welham, Sherman and Jackson14), renal function and development(Reference Langley-Evans, Welham and Jackson16, Reference Nwagwu, Cook and Langley-Evans31), body composition(Reference Bellinger, Sculley and Langley-Evans32), feeding behaviour(Reference Bellinger, Lilley and Langley-Evans24), lipid metabolism(Reference Erhuma, Bellinger, Langley-Evans and Bennett23, Reference Erhuma, Salter, Sculley, Langley-Evans and Bennett33) and body fatness(Reference Bellinger, Sculley and Langley-Evans32) in the first generation (F1). In the present study we have for the first time shown that the effects of a MLP diet during rat gestation may not be limited to the first generation. This unique study has combined measurements of blood pressure with nephron number determination to show that a relatively mild protein restriction has far-reaching consequences for further generations.
Much of the work previously undertaken on transgenerational programming has focused purely on a mechanistic viewpoint, without extensive effort to demonstrate that there is intergenerational transmission of any phenotype(Reference Anway, Rekow and Skinner34). The present paper is the first to examine the interplay between blood pressure and renal structure across several generations, and has demonstrated that the effects of a mild-moderate protein restriction and its resultant phenotype is transgenerationally passed from the F1 generation to the F2 generations via both the maternal and paternal lines. It is clear, from our findings on blood pressure and nephron number, that the F2 offspring of either male or female rats exposed to MLP diets in utero develop broadly the same traits as their parents. A major strength of the research design for the present study was the use of males from the F1 and F2 generations in the mating crosses to produce the successive generations. This allows us to assess the potential contribution of paternal as well as maternal factors in transgenerational programming. Other studies(Reference Zambrano, Martinez-Samayoa, Bautista, Deas, Guillen, Rodriguez-Gonzalez, Guzman, Larrea and Nathanielsz35, Reference Waterland, Travisano and Tahiliani36) that have found evidence of transgenerational programming have considered only maternal transmission of phenotypes, as they have utilised stud males unexposed to dietary or hormonal challenges in utero to breed from programmed females. One of the main problems with such studies is that they cannot conclusively demonstrate that maternal traits are transmitted to the next generation via programmed influences on the ovum as changes to the maternal environment and the composition of proteins in the developing embryo may play a key role. Metabolic traits such as glucose intolerance or cardiovascular traits such as raised blood pressure could be programmed as a response to the prevailing maternal environment during an F1 pregnancy. For example, it has been previously shown that MLP diets in rat pregnancy and lactation lead to glucose intolerance. F2 offspring derived from glucose-intolerant females also show this trait, acquired through glucose spill-over across the placenta(Reference Reusens and Remacle37). In the present study we can at least conclude that male animals must transmit a programming signal to their offspring via alterations to the sperm genome.
The feeding of a MLP diet during rat gestation has been consistently shown to produce a systolic blood pressure rise of between 7 and 30 mmHg(Reference Langley-Evans, Phillips and Jackson12) by 4 weeks of age. This increase in blood pressure appears permanent, remaining elevated well into adult life(Reference Langley-Evans and Jackson38). This phenotype was exhibited by the F1 generation within the present study regardless of their postnatal diet, and was subsequently passed to a second generation (F2) via both the maternal and paternal lines (as all F2 crosses involving low protein result in high blood pressure). The decreased blood pressure observed in the F1 MLP-exposed animals at 4 weeks of age was atypical of previous research using the MLP diet. However, similar outcomes have been observed following maternal restriction of Fe(Reference Crowe, Dandekar, Fox, Dhingra, Bennet and Hanson9, Reference Gambling, Dunford, Wallace, Zurr, Solanky, Srai and McArdle10), where high blood pressure follows a period of lower blood pressure around the time of weaning.
Current thinking is that the transgenerational passage of phenotypic information is likely to involve epigenetic gene regulation. Although the exact mechanism which underpins this phenomenon is poorly understood, DNA methylation(Reference Burdge, Hanson, Slater-Jefferies and Lillycrop19, Reference Waterland, Travisano and Tahiliani36, Reference Bird39, Reference Van den Veyver40) and histone modification(Reference Jaenisch and Bird17, Reference Waterland and Jirtle18) remain key potential candidates. Currently there is a large and varied volume of work within the programming field that is focused on these epigenetic mechanisms(Reference Burdge, Hanson, Slater-Jefferies and Lillycrop41). Lillycrop et al. (Reference Lillycrop, Slater-Jefferies, Hanson, Godfrey, Jackson and Burdge42) have, for example, shown that the MLP diet leads to hypomethylation of specific gene loci and hence leads to their overexpression. There is evidence that these methylation changes also appear in F2 offspring(Reference Burdge, Hanson, Slater-Jefferies and Lillycrop41). Although there is a great deal of controversy surrounding epigenetics it is important to emphasise that the results within the present study show that it is the most likely mechanism through which programming has occurred in the male line. Epidemiological evidence from Swedish cohort studies supports this hypothesis(Reference Kaati, Bygren and Edvinsson43). Arguments suggesting that maternal blood pressure tracking (familial aggregation) may produce these effects fall short of the mark, as high blood pressure was passed down both the maternal and paternal lines. This argument also cannot explain the decrease in nephron number found in both the F1 and F2 populations. It has been hypothesised that there may be the potential for some form of selection, whereby only fetuses with elevated blood pressures survive the pregnancy insult. This would potentially select for a hypertensive genotype. This ‘survivor’ effect is not plausible within the present study, as pregnancy outcomes were similar in rats fed the CON and MLP diets.
The kidney plays a major role in homeostatic regulation of blood pressure and, in rats, the MLP diet has been shown to reduce nephron complement by up to 30 %(Reference Langley-Evans, Welham and Jackson16, Reference Welham, Wade and Woolf26). The present study shows that this reduction in nephron number is passed to a second generation via both parental lines. It is hypothesised that to compensate for nephron deficits at birth and to maintain renal haemodynamic functions, blood pressure within the nephrons is increased to maintain glomerular perfusion. Thus a cycle of progressive nephron loss begins whereby increasing blood pressures result in the loss of more nephrons eventually resulting in hypertension(Reference Mackenzie and Brenner15). There is, however, emerging evidence that nephron number and the nutritional programming of blood pressure are independent processes acting in a sex-specific manner(Reference McMullen and Langley-Evans44). Although there is some debate over the relationship between nephron number and subsequent blood pressure, it is clear that the restriction of protein during gestation has the potential to generate a phenotype that features both reduced nephron complement and hypertension and that this phenotype is subsequently passed on to a second generation via both parental lines. Torrens et al. (Reference Torrens, Poston and Hanson45) have also recently shown, with a design powered only to assess maternal transmission, that high blood pressure can be transmitted to the F2 generation following maternal protein restriction. Within the study of Torrens et al. (Reference Torrens, Poston and Hanson45) there was evidence of an impaired vasodilatory response to acetylcholine, suggesting that in addition to renal programming, endothelial dysfunction may contribute to transgenerational phenomena. Both mechanisms are likely to operate in parallel. Further studies are required to identify the primary mechanisms that drive these physiological processes.
An interesting and unexpected result to arise from this investigation was that the blood pressures of CON animals in F3 were elevated, and nephron counts decreased, compared with the F1 equivalents. The reason for this is unknown. We would, however, suggest that these intergenerational differences may effectively be due to in-breeding of a small colony in the F2 and F3 generations. Furthermore, the fact that the present study shows that the experience of the mother and grandmother of each animal determines blood pressure and nephron complement may be of relevance. The grandmaternal influence on the F1 generation will have been different to the grandmaternal influence on the F3 generation, and it is this difference which could explain the changes.
The feeding of a HF diet within the present study was utilised to assess the impact of a ‘Western diet’ against a background of prenatal undernutrition. Typically a ‘Western diet’ has a high saturated fat content and is linked with obesity and the later development of adult diseases such as CVD(Reference Kris-Etherton, Eckel, Howard, St Jeor and Bazzarre46, Reference Adlercreutz47). The HF diet mirrored this composition as it had an overall fat content of 29·5 %. Interestingly, rather than becoming obese, the MLP animals within the present study exhibited a lean phenotype consisting of slightly reduced body weight and fat mass compared with chow-fed animals at 10 weeks of age. This result was not entirely unexpected, as our previous work(Reference Erhuma, Salter, Sculley, Langley-Evans and Bennett33) has demonstrated that up to 9 months of age MLP-exposed animals are resistant to obesity and exhibit similar plasma TAG, cholesterol, glucose and insulin concentrations to those of CON animals. By 18 months of age there is an abrupt change in metabolic profile with MLP-programmed hypertriacylglycerolaemia and insulin resistance. This indicates that prenatal protein restriction programmes the development of a metabolic syndrome-like phenotype resistant to obesity that develops with senescence(Reference Erhuma, Salter, Sculley, Langley-Evans and Bennett33). The present study, due to its complexity, was unable to examine the development of this ageing-related phenotype.
The work within the present study demonstrates that the feeding of a MLP diet throughout gestation results in a phenotype that is transgenerationally passed to a second generation via both the maternal and paternal lines. The impact of the initial period of maternal undernutrition was not observed in the F3 generation. Further work will take a mechanistic focus and attempt to identify possible gene targets involved in endothelial cell biology. Possible epigenetic modes of inheritance will be analysed in order to understand the molecular machinery behind the phenotype.
Acknowledgements
The present study was supported by the British Heart Foundation (studentship for M. H.). M. H. was responsible for the acquisition of data, statistical analysis, interpretation of data and is an author of the communication. S. L.-E. was responsible for study design, statistical analysis, discussion and interpretation of data and is an author of the communication.
The authors gratefully acknowledge the expert technical support of Mr R. Plant, Mrs C. Armett and Ms S. Kirkland. There are no conflicts of interest to declare.
