PET imaging for attention deficit preclinical drug testing in neurofibromatosis-1 mice

Abstract

Attention system abnormalities represent a significant barrier to scholastic achievement in children with neurofibromatosis-1 (NF1). Using a novel mouse model of NF1-associated attention deficit (ADD), we demonstrate a presynaptic defect in striatal dopaminergic homeostasis and leverage this finding to apply [(11)C]-raclopride positron-emission tomography (PET) in the intact animal. While methylphenidate and l-Deprenyl correct both striatal dopamine levels on PET imaging and defective attention system function in Nf1 mutant mice, pharmacologic agents that target de-regulated cyclic AMP and RAS signaling in these mice do not. These studies establish a robust preclinical model to evaluate promising agents for NF1-associated ADD.

Fig. 1. Nf1 CKO mice with attention system defects demonstrate a presynaptic DA defect, which can be visualized by PET imaging

(A) IHC reveals decreased DARPP32 phosphorylation (p-DARPP32) in the striatum of both male and female mice relative to WT littermates (p=.01; N=8). (B) Western blot shows a 5.8-fold decrease in p-DARPP32 (following normalization to total DARPP32 levels) in CKO compared to WT mice. In vitro quantitative receptor autoradiography demonstrates no change in postsynaptic D1, D2 and D3 DA receptor expression in CKO mice relative to control WT littermates (C), whereas presynaptic VMAT2 and DAT expression is reduced (D; ~10%; p=.03, VMAT2, p=.0004, DAT; N=6). (E) Representative Logan plots for WT and CKO mice (using the cerebellum as the reference region) are shown along with (F) representative [¹¹C]-raclopride transverse micro-PET images (summed across 5–60 minutes). The colorscale bar indicates the normalized peak uptake (percent injected dose per cubic centimeter tissue; %ID/cc). (G) In a cohort of WT and CKO mice, [¹¹C]-raclopride binding was increased in the striatum (Str) of CKO mice compared to control WT littermates on PET imaging (p=.03; N=4 per genotype). C_t = tissue radioactivity at time t; T = time point of each frame of PET scanning course.

Fig. 2. MPH and L-Deprenyl treatments restore [¹¹C]-raclopride binding and improve exploratory and attention behaviors

(A) Representative Logan plots and (B) [¹¹C]-raclopride transverse PET images (summed across 5–60 minutes) of vehicle-treated CKO mice and CKO mice following MPH and L-Deprenyl administration are shown. The colorscale bar indicates the normalized peak uptake (percent injected dose per cubic centimeter tissue; %ID/cc). (C) In these experiments, both MPH and Deprenyl reduced [¹¹C]-raclopride binding in the striatum to WT levels (p=.02, p=.005; N=4). C_t = tissue radioactivity at time t; T = time point of each frame of PET scanning course. During a 1h exploration of a novel environment, total ambulations (D; p=.02; N=8) and total rearings (a measure of non-selective attention) were increased in CKO mice (E; p=.0001; N=8) beyond Bonferonni correction (p=.05/4=.0125) following L-Deprenyl (CKO+D; 10mg/kg) treatment. All mice used for the PET imaging experiments also underwent behavioral testing. Additional independently-generated WT and CKO mice, which did not undergo PET imaging, were included in the behavioral experiments.

Fig. 3. Biologically-based neurofibromin therapies did not correct the DA defect or attentional deficit in CKO mice

(A) Rolipram (CKO+Rol; 5mg/kg/day × 2 weeks) treatment restored cAMP levels in the striatum of 3-month-old mice. (B) Lovastatin (Lov; 10mg/kg i.p.) reduced MAPK activation (p-MAPK) in the cortex (CTX) and hippocampus (Hip) of CKO mice following normalization to total MAPK expression (p=.001; N=3). All fold changes (relative pMAPK/MAPK levels) are relative to saline-treated (vehicle; V) hippocampal levels. Inset shows a representative Western blot for p-MAPK and MAPK in the hippocampus (Hip) following saline and Lovastatin administration. (C) Representative [¹¹C]-raclopride transverse micro-PET images (summed across 5–60 minutes) of CKO mice at baseline (vehicle-treated; V) and following Rolipram (Rol) and Lovastatin (Lov) treatment. The colorscale bar indicates the normalized peak uptake (percent injected dose per cubic centimeter tissue; %ID/cc). (D) Neither Rolipram nor Lovastatin reduced striatal [¹¹C]-raclopride binding in CKO mice (N=8, N=4). During a 1h exploration of a novel environment, total ambulations (E) and total rearings (F) show no improvement in CKO mice following Rolipram or Lovastatin treatment. All mice used for the PET imaging experiments also underwent behavioral testing. Additional independently-generated WT and CKO mice, which did not undergo PET imaging, were included in the behavioral experiments.