AD is classically characterized by the presence of neurofibrillary tangles (NFTs), containing aggregates of the microtubule binding protein tau and plaque deposits composed of aggregates of the amyloid beta peptide (Aβ) [
1,
2]. The amyloid cascade hypothesis has evolved over the past 30 years as one of the leading theories of disease, which states that the secretion of Aβ leads to the formation of toxic oligomers that play a causal role in AD [
3]. However, the cellular and molecular mechanism by which Aβ aggregates induce these pathological damages and cause dementia remains unclear. Semagacestat, an inhibitor of γ-secretase which creates the carboxyl terminus of Aβ accelerated the decline in cognitive function in clinical trials [
4]. We have proposed an alternative formulation of the amyloid hypothesis, based on the observations that FAD PS mutations interfere with the carboxyl-terminal trimming of the initial “long” Aβ cleavage products leading to their aggregation and accumulation in neurons as intraneuronal amyloid [
5]. Several transgenic animal models of Aβ deposition that contain FAD PS mutations demonstrate robust intraneuronal amyloid aggregates that occur prior to plaque deposition and in human brain is associated with early stages of AD [
6,
7]. The intraneuronal amyloid is aggregated in an amyloid-like structure because it reacts with aggregation-specific antibodies and or thioflavin S [
5,
6,
8]. We hypothesized that the intraneuronal amyloid forms the core of the neuritic plaque after the death of the neuron [
5]. The evidence for this includes the observation that the core of neuritic plaques contains the same intraneuronal aggregated amyloid immunoreactivity and stains with dyes specific for DNA [
5]. Some of the neuritic plaques also contain NeuN immunoreactivity, providing further evidence of their neuronal origin [
5]. Since neuritic plaques are typically surrounded by a halo of microglia, we also have hypothesized that neuritic plaques and cored plaques are developmentally related with microglia removing neuronal debris to produce cored plaques [
5]. In order to test critical aspects of these hypotheses, we examined the effect of early ablation of microglia in the 5XFAD mouse model that exhibits widespread and robust intraneuronal amyloid immunoreactivity beginning at 1.5 months, prior to neuritic plaque deposition that begins at 2 months and is robust at 5 months [
6]. We pharmacologically ablated microglia with PLX3397, a previously described orally-administered selective CSF1R/c-kit inhibitor that is able to cross the blood-brain barrier [
9,
10]. Previous studies have determined the optimal dosing and the time-window for PLX3397-induced microglial ablation and it was reported that a 2 month treatment with PLX3397 was reversible in a short period of 3–14 days and had no effect on cognition and behavior of 5xFAD mice [
9]. Microglia repopulation was shown to depend on the internal pool of remaining microglia, which in turn is dependent on IL-1 receptor signaling [
11]. In contrast to previous studies starting treatment in ten-month-old 5XFAD mice with no effect on Aβ levels or plaque load [
10], we started PLX3397 treatment in mice with an age of 2 months. We found that this early treatment with PLX3397 profoundly inhibited not only neuritic plaque formation but also intraneuronal amyloid accumulation. This observation indicates that CSF1R signaling or the presence of microglia is necessary for neuritic plaque formation and that intraneuronal amyloid accumulation and neuritic plaque formation are developmentally linked in a causal series with intraneuronal amyloid upstream of neuritic plaques, consistent with the hypothesis that intraneuronal amyloid is the penultimate source for the amyloid in the center of neuritic plaques.