Reply to Tremolizzo et al., 'Glutamate may be the soluble cerebrospinal fluid factor that induces calcium dysregulation in cultured astrocytes in HIV dementia'

We appreciate the comments of Tremolizzo and coworkers on our paper published in AIDS last year [¹]. We reported that in a functional assay evaluating intracellular calcium regulation in cultured astrocytes we were able to identify cerebrospinal fluid (CSF) samples from HIV-1-seropositive patients suffering from HIV-1-associated dementia complex (HADC). HADC-CSF samples contain factors that reduce the glutamate-evoked calcium increase [¹]. In the paper by Ferrarese and coworkers that appeared in Neurology [²] while our manuscript was in press, the authors showed increased glutamate levels in HADC-CSF samples. So their analyses identified a molecule in HADC-CSF samples that may have interfered with the electrophysiological properties of astrocytes, thereby confirming our functional study. In their present letter they suggest that analysing CSF glutamate levels may be sufficient to predict HADC.

We previously investigated the potential mechanisms resulting in reduced glutamate responses in astrocytes [³], and agree that besides other factors the short-term exposure to low amounts of glutamate decreases calcium responses to a further glutamate application. However, we do not think that N-methyl-d-aspartate (NMDA) receptor-mediated calcium overload, which plays a crucial role in cytotoxicity in neurons, contributes to astroglial calcium dysregulation. NMDA receptors are not expressed on astrocytes in culture [⁴]. Moreover, astrocytes do not respond to NMDA, and in addition, glutamate responses are not reduced by NMDA receptor antagonists such as dl-2-amino-5-phosphonovaleric acid. Therefore, we think that desensitization of the AMPA/kainate type of glutamate receptor is the most likely underlying mechanism for reduced glutamate responses in glutamate pre-exposed astrocytes. Other mechanisms for reduced calcium responses upon glutamate application are alterations of the electrical, e.g. by cell depolarization, or chemical, e.g. increase of intracellular calcium concentrations, transmembrane gradients. Possible mechanisms for the reduced glutamate response, experimental evidence and examples of factors with relevance in the pathogenesis of HADC are summarized in Table 1.

Table 1

The most important advantage of a functional assay as described in our paper is that it may reveal synergistic effects of several factors that may not be accessible in a mere analytical test. Nath and coworkers [⁵] reported the synergistic neurotoxicity of HIV-1 proteins Tat and gp120. Very recently, we showed that HIV-1 Tat and the cytokine TNF-α synergistically induced astroglial calcium dysregulation [⁶]. Presuming that a variety of factors, including virus proteins, endogeneously generated cytokines, chemokines, neurotoxins and potentially glutamate contribute to the pathogenesis of HIV-1-associated encephalopathy [⁷], we consider a functional assay that independently confirms and may even extend analytical tests as a valuable diagnostic tool in such a complex disease as HADC.

Hubertus Köllera

Hans-Jürgen von Giesena

Heiner Schaalb

Gabriele Arendta