Characterization of a c-Fos reporter system for in vivo imaging in the mouse auditory cortex

Abstract

Introduction and objective: c-Fos is an immediate early gene: first genes expressed upon depolarisation. It is coding for a transcription factor controlling the expression of late genes, involved in many cellular processes like apoptosis, proliferation and differentiation. Since the eighties, it is a very popular tool to tag and manipulate active neurons in the brains. It has recently been linked to memory, as cells expressing the gene during fear conditioning, reactivated a few days later can elicit freezing in mice. However, the gene is unspecific to neurons and neuronal activity. Moreover, the causal link between this IEG and neuronal activity or memory remains unclear. Therefore, we wanted to investigate how tight is the correlation of c-Fos levels and neuronal activity. Method: To study the correlation between c-Fos expression and neuronal activity, we created a mouse model expressing a c-Fos reporter (tet-off system derived from TetTag mice: c-Fos-tTA + TRE-BFP) and the widely used calcium indicator GCaMP6m. To track the same cells over weeks, we used a nuclear marker, the fusion protein H2B-mCherry. DNA coding for the three markers were packaged into adeno-associated viral particles (AAV2/8) and injected in the auditory cortex of adult black six mice (C57BL/6J). GCaMP6m and H2B-mCherry were expressed specifically in neurons, thanks to the human synapsin promoter (hSyn). We chronically observed the same cells expressing the three markers through a cranial window, under a two-photon microscope, for up to three weeks, under basal conditioning, during sound presentation and during fear conditioning. Results: The model is not capturing the subtle dynamics of activity-induced c-Fos, because of unspecific triggering of gene expression (stress, inflammation and exposure to novel environment). In addition, the slow synthesis time due to two layers of TetTag system and slow decaying time of BFP result in a poor temporal resolution of c-Fos reporter. However, when classifying cells as c-Fos positive and negative, we could observe a systematic difference in neuronal activity. c-Fos positive cells are responsive to a broader range of sounds and present a higher rate of replay events in their spontaneous activity, once more suggesting a role of the IEG in memory. Conclusion: Rather than reflecting purely neuronal activity, c-Fos in sensory cortices may play a role in population coding to generalise stimuli and to create memory. Hence, the IEG is probably not suited to tag active neurons, but seems adequate to label engrams.