Enhancing cardiovascular artificial intelligence (AI) research in the Netherlands: CVON-AI consortium

Background

Machine learning (ML) allows the exploration and progressive improvement of very complex high-dimensional data patterns that can be utilised to optimise specific classification and prediction tasks, outperforming traditional statistical approaches. An enormous acceleration of ready-to-use tools and artificial intelligence (AI) applications, shaped by the emergence, refinement, and application of powerful ML algorithms in several areas of knowledge, is ongoing. Although such progress has begun to permeate the medical sciences and clinical medicine, implementation in cardiovascular medicine and research is still in its infancy.

Objectives

To lay out the theoretical framework, purpose, and structure of a novel AI consortium.

Methods

We have established a new Dutch research consortium, the CVON-AI, supported by the Netherlands Heart Foundation, to catalyse and facilitate the development and utilisation of AI solutions for existing and emerging cardiovascular research initiatives and to raise AI awareness in the cardiovascular research community. CVON-AI will connect to previously established CVON consortia and apply a cloud-based AI platform to supplement their planned traditional data-analysis approach.

Results

A pilot experiment on the CVON-AI cloud was conducted using cardiac magnetic resonance data. It demonstrated the feasibility of the platform and documented excellent correlation between AI-generated ventricular function estimates as compared to expert manual annotations. The resulting AI solution was then integrated in a web application.

Conclusion

CVON-AI is a new consortium meant to facilitate the implementation and raise awareness of AI in cardiovascular research in the Netherlands. CVON-AI will create an accessible cloud-based platform for cardiovascular researchers, demonstrate the clinical applicability of AI, optimise the analytical methodology of other ongoing CVON consortia, and promote AI awareness through education and training.

Lectin-induced inhibition of desensitization of the kainate receptor GluR6 depends on the activation state and can be mediated by a single native or ectopic N-linked carbohydrate side chain

The ionotropic glutamate receptor GluR6 exhibits strongly and rapidly desensitizing current responses. Treatment of heterologically expressed GluR6 with the lectin concanavalin A (ConA) in Xenopus oocytes as well as in human embryonic kidney-293 cells results in a considerable increase of the steady-state current, presumably by inhibiting receptor desensitization. In the present study, we investigated the molecular basis of this effect using a systematic mutagenesis approach. We found that although N-glycosylation is an absolute prerequisite for the lectin-mediated inhibition of desensitization, no single one of the nine extracellular consensus sites for N-glycosylation of GluR6 is required. Rather, each of the nine N-linked carbohydrate side chains is independently capable of modulatory interaction with the lectin. Moreover, even artificially introduced N-glycosylation sites can substitute for native sites. Thus, the specific site of the lectin binding does not appear to be important for its desensitization-inhibiting action. Furthermore, we show that the extent of the receptor's ConA sensitivity depends on its state of activation, because the desensitized GluR6 exhibits significantly lower lectin sensitivity than the nondesensitized receptor. We conclude that binding of ConA "locks" the receptor in the activatable state, thereby inhibiting conformational changes required to shift the receptor to the desensitized state.

Pharmacological differentiation between neuronal and recombinant glutamate receptor channels expressed in Xenopus oocytes

To determine the molecular components of neuronal glutamate receptors, it is important to identify pharmacological tools that allow differentiation between different glutamate receptor types. Here, we utilized the naphthalene derivative Evans Blue (EB) and a collection of other subtype-specific compounds (polyamine toxins, concanavalin A, cyclothiazide) to compare the pharmacological profile of neuronal and recombinant glutamate receptors GluR1-GluR6 expressed in Xenopus oocytes. Submicromolar concentrations of EB selectively reduced the activity of homomeric glutamate receptors GluR1, GluR2(Q) and GluR4. Applied at concentrations above 100 microM, EB potentiated kainate responses of receptors GluR1, GluR3 and GluR4, while receptors GluR2(Q) and GluR6(Q) were completely blocked. Similar experiments were performed on identified neurones in brain slices and after injection of rat brain RNA in Xenopus oocytes. Neuronal kainate responses were (i) potentiated by 100 microM cyclothiazide, (ii) slightly blocked after preincubation in 10 microM concanavalin A, and (iii) not significantly affected by either low (< 1 microM) or high (> 100 microM) concentrations of EB. Their pharmacological properties were markedly different from those of recombinant glutamate receptor channels GluR1-GluR6 investigated in heterologous expression systems.

N-Glycosylation is not a prerequisite for glutamate receptor function but Is essential for lectin modulation

All ionotropic glutamate receptor (iGluR) subunits analyzed so far are heavily N-glycosylated at multiple sites on their amino-terminal extracellular domains. Although the exact functional significance of this glycosylation remains to be determined, it has been suggested that N-glycosylation may be a precondition for the formation of functional ion channels. In particular, it has been argued that N-glycosylation is required for the formation of functional ligand binding sites. We analyzed heterologously expressed recombinant glutamate receptors (GluRs) of all three pharmacological subclasses of glutamate receptors, N-methyl-D-aspartate (NMDA), alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid, and kainate receptors. By expressing the GluR subunits in tunicamycin-treated, nonglycosylating Xenopus laevis oocytes, we determined that in neither case is N-glycosylation required for ion channel function, although for NMDA receptors, functional expression in the absence of N-glycosylation is very low. Furthermore, we analyzed and compared the interaction of the desensitization-inhibiting lectin concanavalin A (ConA) with all functional GluR subunits. We show that although ConA has its most pronounced effects on kainate receptors, it potentiates currents at most other receptor subtypes as well, including certain NMDA receptor subunits, although to a much lesser extent. One notable exception is the alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptor GluR2, which is not affected by ConA. Furthermore, we show that ConA acts directly via binding to the carbohydrate side chains of the receptor protein.