Full-length and C-terminal neurogranin in Alzheimer's disease cerebrospinal fluid analyzed by novel ultrasensitive immunoassays

Neurogranin (Ng) is a neuron-specific and postsynaptic protein that is abundantly expressed in the brain, particularly in the dendritic spine of the hippocampus and cerebral cortex. The enzymatic cleavage of Ng produces fragments that are released into cerebrospinal (CSF), which have been shown to be elevated in Alzheimer's disease (AD) patients and predict cognitive decline. Thus, quantification of distinctive cleavage products of Ng could elucidate different features of the disease.

Assessments of the FL-Ng and CT-Ng levels in CSF with the novel sensitive immunoassays provide a high separation of AD from controls, even in early phase of the disease. The novel Ng assays are robust and highly sensitive and may be valuable tools to study synaptic alteration in AD, as well as to monitor the effect on synaptic integrity of novel drug candidates in clinical trials.

In this study, we developed novel ultrasensitive single molecule array (Simoa) assays for measurement of full-length neurogranin (FL-Ng) and C-terminal neurogranin (CT-Ng) fragments in CSF. The Ng Simoa assays were evaluated in CSF samples from AD patients (N = 23), mild cognitive impairment due to AD (MCI-AD) (N = 18), and from neurological controls (N = 26).

The intra-assay repeatability and inter-assay precision of the novel methods had coefficients of variation below 7% and 14%, respectively. CSF FL-Ng and CSF CT-Ng median concentrations were increased in AD patients (6.02 ng/L, P < 0.00001 and 452 ng/L, P = 0.00001, respectively) and in patients with MCI-AD (5.69 ng/L, P < 0.00001 and 566 ng/L, P < 0.00001) compared to neurological controls (0.644 ng/L and 145 ng/L). The median CSF ratio of CT-Ng/FL-Ng were decreased in AD patients (ratio = 101, P = 0.008) and in patients with MCI-AD (ratio = 115, P = 0.016) compared to neurological controls (ratio = 180). CSF of FL-Ng, CT-Ng, and ratio of CT-Ng/FL-Ng could each significantly differentiate AD patients from controls (FL-Ng, AUC = 0.907; CT-Ng, AUC = 0.913; CT-Ng/FL-Ng, AUC = 0.775) and patients with MCI-AD from controls (FL-Ng, AUC = 0.937; CT-Ng, AUC = 0.963; CT-Ng/FL-Ng, AUC = 0.785). Conclusions: Assessments of the FL-Ng and CT-Ng levels in CSF with the novel sensitive immunoassays provide a high separation of AD from controls, even in early phase of the disease. The novel Ng assays are robust and highly sensitive and may be valuable tools to study synaptic alteration in AD, as well as to monitor the effect on synaptic integrity of novel drug candidates in clinical trials.