Alzheimer's disease (AD) presents one of the greatest global healthcare challenges. Amyloid b (Ab), produced by cleavage via β-secretase of amyloid precursor protein (APP), is an established AD biomarker. Previous work has shown that low intracellular pH (pHi) correlates with increased β-secretase activity; low extracellular pH (pHo) correlates with increased Ab plaque load. Moreover, aging, an AD risk factor, is associated with metabolic acidosis (MAc: ¯HCO 3 − o/pHo). Thus, defending against pHi/pHo decreases could delay AD onset and slow progression. However, pHi regulation in the AD brain, particularly in the hippocampus (HC), is poorly understood. Here we test the hypothesis that pathological human APP mutations—Swedish (NL), Arctic (G), and Iberian (F)—knocked into mice, modify HC pHi homeostasis in culture long before overt AD onset in APPNL-G-F living mice (cortical plaque formation: 2 mo., memory loss: 6 mo.). We co-culture HC neurons and astrocytes from P0-P2 wild-type (WT) or APPNL-G-F mouse pups. The APPNL-G-F knock-in mice express normal WT levels of APP but eventually accumulate pathological levels of Ab, thereby recapitulating AD pathology without APP overexpression. At 14 to 28 days post-isolation, cultures are loaded with 2',7'-bis-(2-carboxyethyl)-5-(and-6)-carboxyfluorescein, acetoxymethyl ester (BCECF-AM). After hydrolysis releases the pH-sensitive BCECF, we perform digital fluorescence imaging to monitor pHi. With cells exposed to saline containing 5% CO 2 /22 mM HCO 3 − (pHo=7.4), baseline pHi is significantly more alkaline in APPNL-G-F neurons than WTs. When first challenged with 5% CO 2 /14 mM HCO 3 − (pHo=7.2; MAc1), all WT and APPNL-G-F neurons are MAc “sensitive” ((ΔpHi)/(ΔpHo) >40%). However, APPNL-G-F neurons defend pHi less effectively, exhibiting significantly larger ΔpHi. After 10-min recovery in 5% CO 2 /22 mM HCO 3 − and a second MAc pulse (MAc2), WT neurons demonstrate a “consistent behavior” ((ΔpHi)MAc2 within ±12.2% of (ΔpHi)MAc1), whereas APPNL-G-F neurons trend toward an “adaptation behavior” ((ΔpHi)MAc2 ~12.2% smaller than (ΔpHi)MAc1). Results with co-cultured astrocytes parallel the neuronal responses: both genotypes are MAc sensitive, but APPNL-G-F astrocytes exhibit larger (ΔpHi)MAc1 and trend toward adaptation, whereas WT astrocytes are “consistent.” These findings show that pathogenic APP mutations dysregulate hippocampal pHi homeostasis in culture before substantial Ab deposition would have occurred in vivo. Elevated baseline pHi and adaptive MAc1-MAc2 behavior—paradoxically coupled with impaired MAc1 defense—suggest altered acid-base transporter expression or regulation. The alkaline steady-state may reflect compensatory upregulation of acid-extrusion mechanisms (NBCe1, NBCn1, NHEs) or AE3 downregulation, yet this compensation fails during the first of two acute MAc challenges, indicating either insufficient reserve of acid-extruding transporters or altered sensitivity to extracellular CO 2 /HCO 3 − signals. The parallel astrocytic phenotype, despite minimal APP expression in astrocytes vs. neurons, raises the potential of neuron-to-astrocyte crosstalk. At 14–28 days in culture, when Ab burden is negligible in vivo, these abnormalities likely represent direct mutant-APP effects on pH-regulatory machinery rather than secondary plaque responses. Identifying transporters or acid-base sensors modified by mutant APP, and whether correcting early pHi abnormalities would mitigate Ab accumulation, represents a promising presymptomatic intervention target for familial AD. This abstract was presented at the American Physiology Summit 2026 and is only available in HTML format. There is no downloadable file or PDF version. The Physiology editorial board was not involved in the peer review process.
Pan et al. (2026) studied this question.