In an earlier study, male fetuses from pregnant ewes fed 50% of liveweight maintenance requirements from day 0 of gestation exhibited reduced fetal weight at day 110 of gestation  but no reduction in testicular size; the results of the current study confirm these previous observations. In contrast, day 110 female fetuses, from the earlier study, showed no treatment effect on either fetal weight or mean ovary weight . The impact of maternal undernutrition on male fetal weight probably reflects the inability of the ewe to respond to the increasing nutrient demand of the faster-growing fetus, at this stage of gestation. The mean loss of 0.5 condition score units in the dams of the current study subjected to restricted nutrition, from 0 to 110 days, is consistent with that known to occur in poorly nourished animals, in practice, and is indicative of the nutritional demands on the dams.
The observation that the proliferation marker (Ki67) is localised to Sertoli and germ cells in the day 110 fetal testis supports our previous published study using another marker of proliferation (proliferating cell nuclear antigen: PCNA) which localised to the same testicular cell types . Ki67 is a better marker of cell proliferation than PCNA since it is expressed at all stages of the cell cycle and is rapidly degraded when proliferation ceases, unlike PCNA which has a half-life of 20 hours and is also involved in DNA repair [24, 25].
The localisation to germ cells of both the pro-apoptotic factor, Bax, and its antagonist Mcl-1, indicates that these factors play a role in the control of germ cell apoptosis. Indeed, apoptosis is reported to play a crucial role in both the control of germ cell numbers and elimination of defective germ cells during development . Since the homodimerisation or heterodimerisation of Mcl-1 with Bax determines the overall effect on apoptosis, the ratio of these factors in germ cells is likely to determine germ cell survival. The c-kit/SCF system is also critical for germ cell survival; mouse knock outs for either of these factors result in testes devoid of germ cells [26, 27] and other mouse, in vitro studies have shown that the interaction between germ cell c-kit and SCF is required for proliferation and migration of germ cells (reviewed in , ). In addition, murine in vitro studies indicate that the effects of SCF-c-kit interaction on germ cell survival operate through the Bcl-2 family .
In the current study, although c-kit, Bax and Mcl-1 were detected in testicular germ cells, these genes were not affected by maternal undernutrition. In contrast to the fetal testis, fetal ovarian development was shown to be perturbed in previous studies using the same experimental sheep model (50% maintenance feeding regime). The change was in the form of a delay in ovarian folliculogenesis at day 110 and this was associated with reduced germ cell proliferation at day 65 and an increased expression of Bax in granulosa cells at day 110 [4, 23]. Since there was no effect of maternal undernutrition on these genes in the fetal testis, it is concluded that they are nutritionally-sensitive only in the developing fetal ovary.
We have reported, previously, that the testes of fetuses from underfed ewes exhibit a transient increase in the expression of the cholesterol transporter, STAR, at day 50 and that maternal undernutrition reduces GnRH sensitivity at day 115 [9, 29]. This suggests that the timing of the nutritional insult may be crucial and/or that other components of the hypothalamic pituitary gonadal axis may be affected. It is possible, also, that other testicular genes not included in this study may be perturbed. However, given that the same nutritional treatment perturbed the expression of Ki67, Bax and Mcl-1 in the day 110 fetal ovary  and that the markers are related to proliferation, apoptosis and development, we suggest that there are few other likely candidate genes at this stage of gestation. However, this does not rule out the possibility of effects on these genes, or alternative genes, in the developing testis later in gestation or during the post-natal period, or the expression of a transient effect such as that seen at an earlier stage of gestation .
In keeping with the lack of an effect on developmental genes, the current study indicates that maternal undernutrition for the first 110 days of gestation has no effect on Sertoli cell number. The 0 to 110 day window of exposure to maternal undernutrition encompasses sexual differentiation (day 30), the onset of pituitary function and gonadotrophin secretion (day 80) and a substantial part of the period of Sertoli cell proliferation which occurs from 70 days through to parturition [10, 11]. Our data indicate, for the first time, that developmental stages falling in this gestational period remained largely unaffected in terms of fetal testis Sertoli cell numbers and testicular developmental gene expression. This is consistent with our previous observations of the effects of maternal undernutrition, from mating to day 95 of gestation, on testis size in male offspring at 6 weeks or 10 months of age . In contrast, others have reported that underfeeding of the ewe from 70 days to parturition reduced Sertoli cell numbers and the absolute volume of cords in the newborn lamb  while maternal feed restriction from 31 to 100 days of gestation reduced Sertoli cell numbers and seminiferous cord size in 10 month old lambs . These apparently conflicting results may reflect the developmental stage investigated. In the current study, in which the effects of maternal undernutrition were examined only on the pre-natal day 110 testis, the possibility remains of post-natal expression of pre-natal undernutrition effects on testicular structure or function.
Support for the concepts of action at the hypothalamic level and expression of effects at later developmental stages is provided by findings based on the adolescent sheep model (2x maintenance diet throughout gestation; placental growth restriction and reduced lamb birth weight) . As in the underfeeding model, testes from day 103 fetuses showed no changes in Sertoli cell numbers or the number of seminiferous cords  whereas pubertal lambs of 28 to 35 weeks of age from the same experimental model, had reduced testosterone concentrations and testicular volume and a delayed seasonal increase in testosterone . Similarly, 20 month old male offspring from ewes fed a 50% maintenance diet from mating to day 95 of gestation exhibited increased FSH levels  and, in a separate study, maternal undernutrition from 31 to 100 days increased the FSH response to a GnRH challenge in 10 month old lambs . Collectively, these findings suggest that fetal undernutrition may impact on the expression of genes which regulate the onset of postnatal hypothalamic-pituitary activity at puberty. The finding that suppression of the ovine pituitary testis axis during fetal life with a GnRH agonist reduces plasma testosterone concentrations in 28 week old lambs  supports this suggestion.