The current study shows that: a) elevated levels of PRL induced by 2 pituitary homografts prevent the cognitive deficit induced by long-term OVX; b) these behavioral actions were clearly correlated to an increase of the mushroom spines percentage in CA1 pyramidal neurons, an effect blocked by BRC and c), the removal of ovaries induces changes in both dendrite complexity and dendritic spine morphology of hippocampal neurons.
The pituitary homografts surgery introduced by Adler et al. [11] generates approximately a tenfold increase in the circulating levels of PRL (an observation confirmed by the current study) and can prevent the cognition deficit observed at long-term in rats without ovaries [28]. This finding suggest that the removal of ovaries represents a useful approximation for the study of extra-ovarian hormones like PRL on cognition. Regarding this point, some authors have reported that high doses of PRL in female mice for three days improve the learning of a spatial memory task [8], while spatial and recognition memory impairment caused by 1,2-Diacetylbenzene or kainic acid, can be reversed by PRL treatment [29, 30]. In line with this, the low levels of this hormone found in PRL-null mice have been associated with a spatial learning deficit; an effect reverted by the chronic intrahippocampal infusion of PRL [3]. On the contrary, some authors have reported negative actions of PRL on cognition. For instance, induction of HPRL by subcutaneous osmotic pumps can impair memory in tests such as the Barnes maze [31] and, as above mentioned, HPRL male rats show a poor performance in the NORT but not in the Morris water maze, i.e., elevated concentrations of PRL impair recognition but not spatial learning [9].
As evident, the role of PRL on cognition is still controversial, however, data are showing that lactation is directly associated with neuroprotection and cognitive improvement in rats, effects that even appear to be long-lasting [32]. In this sense, it has been reported that nulliparous females, tested in the radial-arm maze show a poor performance in comparison to age-matched multiparous rats [33, 34]. Accordingly, our group reported that pregnancy improves the performance in the Morris water maze [35] and increases the number of dendritic spines in the CA1 hippocampal subfield [15]. A possible explanation regarding these discrepancies could be the different plasmatic concentrations of PRL present during memory testing. In support of this idea is the finding that patients with elevated PRL levels show verbal and working memory deficits, which can be alleviated by administering dopaminergic agonists [36, 37]. Therefore, it would be important to know the PRL levels achieved during the behavioral tests to clarify the true role of PRL on cognition.
Several mechanisms have been proposed to explain the PRL actions on learning. For instance, the addition of exogenous PRL to primary adult hippocampal cells produces an approximate 50% increase in the number of neurospheres [3], which is in line with the PRL-induced hippocampal neurogenesis previously reported [38, 39] in adult rodents. Thus, the pro-cognitive actions of PRL reported here could be partially explained by its capacity to promote neurogenesis.
On the other hand, it has been described that PRL decreases the expression of both the parvalbumin-positive cells and the β2/3 subunit of the GABAA receptor in the hippocampal CA1 region, which is in line with the hypothesis that the attenuation of GABAergic activity at this level could be responsible for the improvement of cognitive performance [31].
The neuroprotective effects of PRL have already been studied in silico and associated with an increased expression of various transcription factors and genes involved in learning and memory [29, 40]. Besides this later finding, other studies trying to explain the role of PRL on cognition include the PRL-releasing peptide, which seems to reduce the expression of inflammatory markers in the microglia by decreasing the activity of the NADPH oxidase-regulated NLRP3 inflammasome in rats [41], the reduction of Ca2 + input under conditions of Glu-induced excitotoxicity, the overexpression of anti-apoptotic proteins, etc. [42].
Taken together, all this evidence suggests that PRL possesses beneficial effects on learning and cognition, but its underlaying mechanisms remain to be elucidated. The cellular mechanisms involved in the pro-cognitive effects of PRL and its association with changes in the structural plasticity of regions related to learning such as the prefrontal cortex and hippocampus are still poorly understood. Nevertheless, some evidence has emerged suggesting a close relationship between these variables. For instance, Leuner and Gould [43] reported in mother rats an increase of dendritic spines density in pyramidal cells of the medial prefrontal cortex and hippocampal neurons from CA1 and linked this finding to an improvement of cognitive flexibility. More recently, Wang et al. [44] studying voles during the parental care stage found similar morphological changes in the same brain regions, which was close related to the PRL levels reached during this parental care stage in these animals. Such an increase in dendritic spines had already been observed during the lactation period, in tuberoinfundibular dopaminergic cells; a phenomenon associated with an intensification in excitatory inputs into this neuronal group [45].
In the piriform cortex and posterolateral cortical amygdala, an increased density of mushroom and stubby dendritic spines has also been found during early and late gestation [46]. At the same stage, PRL has been reported to promote increased progenitor cell expression in the subventricular region in mice [47]. The findings of the present research and data reported in the literature suggest that PRL not only promotes the appearance of plastic changes in different neuronal groups but also facilitates the process of neurogenesis.
It seems that the molecular events associated with changes in neuronal morphology after PRL exposure are related to its ability to induce, in addition to cell differentiation, neuronal and dendritic axonal growth through the expression of proteins associated with cytoskeleton modeling such as nestin and MAP2 at CA1 hippocampal level [48, 49]. Thus, it is likely that the high arborization together with the high density of dendritic spines, especially of the mushroom type, found in the dorsal hippocampal CA1 of HPRL rats may be due to PRL’s ability to induce the expression of cytoskeletal proteins together with neurotrophic factors such as GDNF and BDNF, as demonstrated by Arnold et al. [50], who used the same method of the current study (implanting two pituitary homografts), to induce HPRL rodents. In general, these morphological changes were blocked by BRC, which implies the important role of PRL in promoting plastic changes at the hippocampus and their close relationship with the improvement of both STM and LTM.
Our finding that the procognitive actions of PRL were not completely blocked by BRC may be due to the fact that the dose of BRC tested was not sufficient to fully impair the production of the hormone or only did it for a short period. Given that the dose used here had no impact on the rat´s motor skills (see Fig. 2), further experiments increasing the dose of BRC and blocking other neurotransmitter systems traditionally involved in the regulation of recognition memory are being carried out to clarify this point.