Clinically relevant hormone treatments fail to induce spinogenesis in prefrontal cortex of aged female rhesus monkeys

Abstract

Preclinical animal models have provided strong evidence that estrogen (E) therapy (ET) enhances cognition and induces spinogenesis in neuronal circuits. However, clinical studies have been inconsistent, with some studies revealing adverse effects of ET, including an increased risk of dementia. In an effort to bridge this disconnect between the preclinical and clinical data, we have developed a nonhuman primate (NHP) model of ET combined with high-resolution dendritic spine analysis of dorsolateral prefrontal cortical (dlPFC) neurons. Previously, we reported cyclic ET in aged, ovariectomized NHPs increased spine density on dlPFC neurons. Here, we report that monkeys treated with cyclic E treatment paired with cyclic progesterone (P), continuous E combined with P (either cyclic or continuous), or unopposed continuous E failed to increase spines on dlPFC neurons. Given that the most prevalent form of ET prescribed to women is a combined and continuous E and P, these data bring into convergence the human neuropsychological findings and preclinical neurobiological evidence that standard hormone therapy in women is unlikely to yield the synaptic benefit presumed to underlie the cognitive enhancement reported in animal models.

Figure 1.

Experimental design of HT. Groups included: vehicle control, continuous E, continuous E plus continuous P, cyclic E every 28 d (black-to-gray gradient bars) plus cyclic P for 10 d beginning 10 d after E injection (red bars), and continuous E plus cyclic P 10 d of every 28. See Results for treatment efficacy.

Figure 2.

dlPFC neuronal reconstruction methodology. A, Confocal image of a layer III monkey dlPFC neuron used in the present study. B, Dendritic reconstruction of the same neuron. C, Deconvolved confocal images of apical dendritic segments used in the present study. Scale bars: A, 50 μm; C, 5 μm.

Figure 3.

HT and dendritic spine density. A, Overall dendritic spine density on aged monkey dlPFC neurons. ANOVA found no significant differences in overall spine density between groups. B–D, Spine densities divided by spine types: thin (B), mushroom (C), and stubby (D). ANOVA found no main effect on condition for any spine type density. See Results for details.

Figure 4.

HT and dendritic spine morphology. A, Cumulative frequency plots of individual spine head diameters. Kolmogorov–Smirnov tests found no differences between groups (p > 0.7 for all pairwise comparisons). Similarly, mean spine head diameter was also unaffected by treatment (data not shown). B, Cumulative frequency plots of individual spine head volume. Kolmogorov–Smirnov tests found no differences between groups (p > 0.8 for all pairwise comparisons). Similarly, mean spine volume did not differ between groups (data not shown). See Results for details.