Based upon previous expression studies demonstrating that Advillin is expressed in a somatosensory-specific manner [8, 9], we generated a BAC transgenic Cre-driver mouse line that expresses Cre-recombinase from the Advillin locus. We characterized this line by means of copy number analysis, X-gal staining and pain behavior. We found that Advillin is highly expressed in sensory ganglia such as TG and DRG, with sparser expression in vestibular ganglia, superior cervical ganglia, nodose ganglia, taste buds, secondary sensory nuclei in the brainstem, and the VTA in the midbrain. Finally, acute nociceptive responses to thermal and mechanical stimuli were detected to be normal.
Transgenic mice generated using BAC technology are widely used in biology to study gene function. BACs are excellent tools for maintaining large DNA fragments and are extensively used to generate transgenic Cre recombinase driver lines [26, 27]. BAC-based transgenes have the advantage of usually directing gene expression at physiological levels with the same developmental timing and expression patterns as endogenous genes . Generation thereof is less time consuming than embryonic stem (ES) cell mediated transgenesis by knock-in, with comparable end results . With this in mind we chose to use a BAC transgenic-based approach to generate our Advillin-Cre transgenic mouse line. Of note, a Cre driver line has been described that utilized a knock-in strategy into the Advillin locus . As yet, there is no reported characterization of the developmental or adult expression profile of Cre in these mice, nor any behavioral analysis to the best of our knowledge.
BAC transgenes usually contain sufficient regulatory information to recapitulate endogenous gene expression patterns , and the optimal transgene drives expression in a cell-specific, copy-dependent fashion [19, 20]. Usually one to five BAC transgene concatemers integrate into a single locus of the genome [20, 21, 32] and an increased BAC transgene copy number correlates with increased BAC transgene expression [19, 20]. However, rare cases have been reported, in which more than five copies were integrated [22, 23]. Our copy number estimates are based on a standard curve of known BAC copy number standards, which is a useful means to confirm transgene expression and stable integration. Quantification by qPCR revealed that our Advillin-Cre mice carry five copies of the transgenes in their genomes. Based on a hallmark publication, copy numbers usually remain fixed in subsequent generations and among littermates .
We assessed the expression of the Cre gene in the Advillin locus using X-gal staining after breeding with R26R mice and a CFP reporter line. Functional Cre could be found in more than 82% of all neurons of these ganglia indicating that it was expressed by both, mechanoreceptors and nociceptors. The remaining Advillin-negative expressing neurons were not further characterized. However, secondary neurons of dorsal root afferents in the dorsal horn of the spinal cord were Advillin-negative, indicating that this Cre-line targets only the peripheral part of the somatosensory pathway. We were not able to detect Advillin-Cre expression in DRG at embryonic stages, while X-gal staining was clearly detectable in neonatal mice at P1. In previous studies, Advillin expression has been reported starting at embryonic stage E12.5 with peak expression between E14.5 and E16.5 [9, 12]. Consistent with our negative spinal cord staining, expression of Advillin in the spinal cord has not been reported previously.
The trigeminal ganglion, which contains primary sensory neurons innervating the face, projects to trigeminal sensory nuclei (principal trigeminal nucleus and spinal trigeminal nucleus) in the brainstem (reviewed by ). We found X-gal staining also in these second-order neurons, however in adult mice only.
Although TG progenitors are generated at early developmental stages (E8.25 - E9) in the mouse, and brainstem nuclei progenitor cells are formed by stage E15.5 , we detected X-gal staining for functional Cre in trigeminal ganglia at embryonic stage E16.5 and E18.5 but not at E12.5. Advillin expression has been described at these earlier stages [9, 12] and Hasegawa and colleagues  found scattered Advillin expression in TG at stage E11.5. In line with our study, they also reported adult Advillin expression in higher-order trigeminal nuclei .
In agreement with previous reports, we observed Advillin mediated Cre expression in the vestibular ganglion of adult mice . In addition, we detected functional Cre expression in second-order brainstem neurons, and in the medial and spinal vestibular nuclei of adult mice. Moreover, during development, we observed Advillin expression in the vestibular ganglion at E18.5. Positive X-gal staining also occurred in taste buds of both the circumvallate and foliate papilla, which has been previously reported . However, taste buds in the fungiform papillae, which are not innervated by the glossopharyngeal nerve but by the greater superficial petrosal nerve , were Cre negative. As for the TG and vestibular nuclei, we detected staining in second-order neurons of the gustatory pathway, in the nucleus of the solitary tract of the brainstem . At embryonic stages, we did not detect any staining in these areas. The only other brain region where we found positive (albeit sparse) staining in adults was the ventral tegmental area (VTA) of the midbrain. This region contains mainly dopaminergic neurons and functions in the reward circuitry of the brain. From here, neurons project to numerous areas of the brain, such as the prefrontal cortex . Besides the above-mentioned regions, other tissues and brain areas did not express functional Cre in our Advillin-Cre animals. In summary, Advillin is expressed in several sensory ganglia where it is found in primary and some secondary neurons. Advillin-Cre mice may therefore be useful driver lines for gene targeting in other sensory areas in addition to the somatosensory system.
To rule out a putative affect on nociceptive behavior response caused by expression of Cre in primary sensory neuron populations, we obtained data from paradigms testing thermal and mechanical nociceptive thresholds (hot plate test, dynamic plantar test) [36, 37]. Responses to these stimuli were similar in Advillin-Cre mice and their wildtype littermates (Figure 5A). In addition, transgenic animals had normal overall gross motor and sensory function, assessed by the SHIRPA test . Thus our data indicate that expression of Cre from the Advillin locus does not alter baseline behavioral responses.
Taken together, this is the first report of a sensory neuron specific Cre driver line with expression covering both mechanoreceptive and nociceptive neurons. Furthermore the late developmental onset of Cre expression will be advantageous in avoiding developmental defects and lethality associated with some genes. Advillin-Cre mice have been deposited at the EMMA mouse repository.