Synaptic distributions of pS214-tau in rhesus monkey prefrontal cortex are associated with spine density, but not with cognitive decline

Abstract

Female rhesus monkeys and women are subject to age- and menopause-related deficits in working memory, an executive function mediated by the dorsolateral prefrontal cortex (dlPFC). Long-term cyclic administration of 17β-estradiol improves working memory, and restores highly plastic axospinous synapses within layer III dlPFC of aged ovariectomized monkeys. In this study, we tested the hypothesis that synaptic distributions of tau protein phosphorylated at serine 214 (pS214-tau) are altered with age or estradiol treatment, and couple to working memory performance. First, ovariectormized young and aged monkeys received vehicle or estradiol treatment, and were tested on the delayed response (DR) test of working memory. Serial section electron microscopic immunocytochemistry was then performed to quantitatively assess the subcellular synaptic distributions of pS214-tau. Overall, the majority of synapses contained pS214-tau immunogold particles, which were predominantly localized to the cytoplasm of axon terminals. pS214-tau was also abundant within synaptic and cytoplasmic domains of dendritic spines. The density of pS214-tau immunogold within the active zone, cytoplasmic, and plasmalemmal domains of axon terminals, and subjacent to the postsynaptic density within the subsynaptic domains of dendritic spines, were each reduced with age. None of the variables examined were directly linked to cognitive status, but a high density of pS214-tau immunogold particles within presynaptic cytoplasmic and plasmalemmal domains, and within postsynaptic subsynaptic and plasmalemmal domains, accompanied high synapse density. Together, these data support a possible physiological, rather than pathological, role for pS214-tau in the modulation of synaptic morphology in monkey dlPFC.

Keywords: RRID:AB_2716719; RRID:AB_2750560; RRID:SCR_002865; RRID:SCR_014199; aging; area 46; delayed response; estradiol; menopause; pS214.

Figure 1.

Synapse morphology. Bar graphs of the average postsynaptic density length (left), and the average axon terminal diameter (middle) across age and treatment groups; and, the positive correlation between postsynaptic density length and axon terminal diameter for each monkey (right). V, vehicle; E, estradiol. Individual data points represent the average value for each monkey. Group results (bars) are expressed as the mean ± SEM. Young-V, n = 5; young-E, n = 3; aged-V, n = 4; aged-E, n = 4 monkeys.

Figure 2.

Subcellular synaptic representations of pS214-tau immunogold particles. (a) Schematic diagram illustrating the eight domains used to categorize the location of each pS214-tau immunogold particle. (b) Representative electron micrographs of five serial sections through a pS214-tau-containing perforated synapse spine (top row), and nonperforated synapse spine (bottom row). The postsynaptic densities are well defined, and easily distinguishable (black arrowheads). Black arrows point to the discontinuity in the postsynaptic density of the perforated synapse spine. pS214-tau immunogold particles (arrowheads) are shown localized to the presynaptic active zone (orange), mitochondrial (purple), and cytoplasmic (blue) domains; and, postsynaptic synaptic (magenta), subsynaptic (cyan), cytoplasmic (yellow), and plasmalemmal (green) domains. For each series, the third section (outlined with a black box) was used as a reference section, and all synapses containing a dendritic spine that possessed a readily apparent postsynaptic density in this section were marked, and followed throughout the series for subsequent morphological and immunolabeling analyses. Immunogold particles, specifically localized to the synaptic complex, are prominent in expanded versions of reference sections (insets). Scale bars, 250 nm. (c) Bar graph of the percentage of pS214-tau immunogold particles localized to each of the 8 synaptic domains averaged across all monkeys. V, vehicle; E, estradiol. At, axon terminal; sp, dendritic spine. Individual data points represent the average value for each monkey. Group results (bars) are expressed as the mean ± SEM. Young-V, n = 5; young-E, n = 3; aged-V, n = 4; aged-E, n = 4 monkeys.

Figure 3.

Density of pS214-tau immunogold particles. (a) Bar graphs of the average density of pS214-tau immunogold particles in the active zone (left), cytoplasmic (middle), and plasmalemmal (right) domains of axon terminals. (b) Bar graph of the average density of pS214-tau immunogold particles in the subsynaptic domain of dendritic spines. V, vehicle; E, estradiol. Individual data points represent the average value for each monkey. Group results (bars) are expressed as the mean ± SEM. Young-V, n = 5; young-E, n = 3; aged-V, n = 4; aged-E, n = 4 monkeys.

Figure 4.

Relationship of the density of pS214-tau immunogold particles to synapse density. (a) Positive correlation between the average density of pS214-tau immunogold particles in axon terminal cytoplasmic (top), and plasmalemmal (bottom) domains to synapse density for each monkey. (b) Positive correlation between the average density of pS214-tau immunogold particles in dendritic spine subsynaptic (top), and plasmalemmal (bottom) domains to synapse density for each monkey. V, vehicle; E, estradiol.

Figure 5.

Schematic diagram illustrating proposed relationships of synaptic distributions of pS214-tau to aging, estradiol, and spine density. (1) Small synapse spines are lost with age, and are partially restored to levels of young monkeys with estrogen treatment (Crimins et al., 2017; Dumitriu et al., 2010; Hao et al., 2007). (2) In axon terminals, pS214-tau is lost from the active zone, cytoplasmic, and plasmalemmal domains with age. (3) In dendritic spines, pS214-tau is lost from the subsynaptic domain. Such changes appear to target small and large spine types, and thus are depicted in both synaptic subclasses. Asterisks corresponding to the significance level for each Pearson’s correlation indicate domains where lower pS214-tau density strongly couples to synapse loss. At, axon terminal; az, active zone; cyt, cytoplasmic domain; pl, plasmalemmal domain; sp, spine; sub, subsynaptic domain. *p < 0.05; **p < 0.01.