General anesthesia causes long-lasting disturbances in the ultrastructural properties of developing synapses in young rats

Abstract

Common general anesthetics administered to young rats at the peak of brain development cause widespread apoptotic neurodegeneration in their immature brain. Behavioral studies have shown that this leads to learning and memory deficiencies later in life. The subiculum, a part of the hippocampus proper and Papez's circuit, is involved in cognitive development and is vulnerable to anesthesia-induced developmental neurodegeneration. This degeneration is manifested by acute substantial neuroapoptotic damage and permanent neuronal loss in later stages of synaptogenesis. Since synapse formation is a critical component of brain development, we examined the effects of highly neurotoxic anesthesia combination (isoflurane, nitrous oxide, and midazolam) on ultrastructural development of synapses in the rat subiculum. We found that this anesthesia, when administered at the peak of synaptogenesis, causes long-lasting injury to the subicular neuropil. This is manifested as neuropil scarcity and disarray, morphological changes indicative of mitochondria degeneration, a decrease in the number of neuronal profiles with multiple synaptic boutons and significant decreases in synapse volumetric densities. We believe that observed morphological disturbances of developing synapses may, at least in part, contribute to the learning and memory deficits that occur later in life after exposure of the immature brain to general anesthesia.

Fig. 1

Anesthesia did not cause significant differences in weight gain between control and experimental rats during the first 21 days of postnatal life. Daily weight measurements before and after anesthesia exposure (marked with an arrow as Treatment) show no difference in weight gain between control and experimental rats up to 14 days after anesthesia (PND 21) (marked with an arrow as Sacrifice) (n = 36 control and 36 experimental animals total)

Fig. 2

Anesthesia caused long-lasting ultrastructural disarray in the subicular neuropil (A, upper panel) and mitochondria (B, lower panel) of 21-day-old rats. a The pyramidal layer of a subiculum from a control rat shows the abundance of tightly packed neuro-glial profiles with numerous synaptic contacts (arrows). b, c The pyramidal layers of subiculums from two different rats from different litters show a striking scarcity of synaptic contacts (two synaptic contacts are indicated by arrows in b and c) and a gross loss of neuro-glial profiles. d Mitochondria in a control subiculum appear healthy with a clear double membrane and tight, orderly cristae. e, f Experimental mitochondria exhibited various stages of degeneration. In the early stage, they appeared swollen, with balloon-like cristae, but had normal-looking inner and outer membranes (e). In the later stage they appeared dark, condensed, and shrunken, with no clear outline between the inner and the outer membranes (f) (magnification × 12,000)

Fig. 3

Anesthesia caused significant decrease in volumetric density in all three subicular layers of 21-day-old rats. When the overall number of synapses per μm³ was quantified in the pyramidal, polymorphic, and molecular layers, there was significantly less synapse density in each subicular layer of anesthesia-treated animals (on average 30–40%) as compared to controls (*pyramidal layer P <0.001; *polymorphic layer and molecular layers, P <0.01) (n = 4 control and 4 experimental pups from two litters)

Fig. 4

Anesthesia-treated animals had scarce multiple synaptic boutons (MSBs). A Shows the abundance of MSBs (arrows) on the neuronal profile of a control subiculum. In the experimental subiculum (B) there is, in contrast, a substantial lack of synaptic contacts (magnification × 12,000)

Fig. 5

Morphologically intact-looking synapses were found on degenerating neuronal profiles in anesthesia-treated 21-day-old rats. Multiple synapses (black arrows, left and right panels) with numerous docked vesicles were engaged on condensed neuronal profiles that were undergoing degeneration (magnification × 12,000)

Fig. 6

Anesthesia did not cause preferential loss of different types of synaptic contacts (upper panel) or preferential loss of excitatory versus inhibitory synapses (lower panel). There was no difference in the relative ratio of excitatory, inhibitory, and undetermined synapses in any subicular layer. The ratio among excitatory, inhibitory, and undetermined synapses spared by anesthesia remained the same as in controls; i.e., approximately 75, 5, and 20%, respectively. Similarly, there was no difference in the relative ratios among axo-spinous, axo-dendritic, and other types of synapses compared to the ratios in controls (on average 70–80%, 15–20%, and 5–10% of synapses were defined as axo-spinal, axo-dendritic, and others, respectively, in both controls and experimentals) (n = 4 controls and 4 experimental pups from two litters)