Intranasal drug delivery has gained increasing interest as a route for drug administration, in particular for targeting of the central nervous system (CNS). However, the nasal cavity is not a uniform structure as it is covered by four different types of epithelia, the respiratory epithelium and the olfactory epithelium, but also squamous and transitional epithelium (Doty, 2003). The respiratory mucosa consists of respiratory epithelium and lamina propria with glands and immune cells, is well-perfused and covers about 90% of the lower nasal cavity in humans and about 50% of the rostral parts in rodents (Lochhead and Thorne, 2012). Consequently, the respiratory mucosa is an easily accessible region for nasal sprays. In fact, it has already been used in clinical practice for the targeting of drugs into the systemic circulation (Illum, 2012; Stützle et al., 2015; Gänger and Schindowski, 2018). In contrast, the olfactory mucosa that contains the olfactory epithelium is a rather well-hidden structure that is harder to access in both humans and rodents. In humans, it is located at the roof of the nasal cavity where it makes up less than 10% of the total mucosa surface (Carmichael et al., 1994). In rodents, however, it covers up to 40% of the caudal side of the rodent nasal cavity (Carmichael et al., 1994). The olfactory epithelium is composed of olfactory sensory neurons, sustentacular or supporting cells and basal cells. The axons of the olfactory sensory neurons form neuronal bundles that extend from the lamina propria to the olfactory bulb and provide a direct connection between the olfactory mucosa and the CNS (Gänger and Schindowski, 2018). It has previously been reported that small molecule drugs and biopharmaceuticals delivered to the olfactory mucosa can pass the so-called nose-brain barrier and reach the olfactory bulb, from where they can diffuse to other parts of the brain (Kumar et al., 2018; Maigler et al., 2021). Thus, it is believed that the olfactory mucosa mediates predominantly drug delivery to the CNS while the respiratory mucosa mediates drug absorption into the blood stream. Hence, nose-to-brain (N2B) delivery via the olfactory mucosa may be a suitable delivery route for bypassing the blood brain barrier (Illum, 2004; Lochhead and Thorne, 2014; Engelhardt et al., 2016).

In order to use the great potential of intranasal N2B drug delivery, many aspects are yet to be elucidated. For a successful clinical translation, quantitative studies on efficacy, safety, reproducibility, pharmacokinetics, and drug transport mechanisms are needed (Stützle et al., 2015). Although there has been a rise in publications on intranasal drug delivery experiments performed in rodents, only a minority of 3.1% delivers quantitative, robust data that are suitable for a clinical development of this drug delivery route (Kozlovskaya et al., 2014). Furthermore, even though these studies may provide information on a quantitative readout, the highly heterogenous and non-standardized intranasal application procedures that are used to administer the drugs limit the interpretation of the data. For instance, one common intranasal administration approach is performed by pouring a drop of drug solution onto, or into, the nostril of a mouse or rat (Figure 1A) (Thorne et al., 1995; Francis et al., 2008; Thorne et al., 2004). In this scenario, even the administration of a 20 µL volume (the approximate volume of a common drop) leads to a flooding of the murine nasal cavity, so that the total amount of deposited drug remains unknown. Additionally, the described administration approach is accompanied by safety issues for the laboratory animals such as the swallowing and the inhalation of the administered drug solution. Notably, rodents are obligate nose breathers, but there is no study data available on the percentage of animals that were asphyxiated by the intranasal delivery of overly high volumes (Harkema et al., 2006). Finally, with the pipette-based conventional method as demonstrated in Figure 1A it is impossible to discriminate among the regions of the nose, and hence to target a specific kind of nasal mucosa (Hanson et al., 2013). As described above, the nasal region and thereby the targeted mucosa is responsible for whether the drug is directed to the systemic circulation via uptake by blood vessels, or to the CNS along neuronal connections. Consequently, the lack of a precise and region-specific application technique for rodents is a major drawback for a meaningful clinical development of N2B delivery.

For the reasons discussed above, we have previously presented a refined intranasal administration technique that uses a catheter to specifically reach the olfactory or the respiratory region (Flamm et al., 2018). Region-specific in this context means that the liquid drug solution is applied exclusively to the olfactory mucosa without wetting of the respiratory mucosa. This approach was developed and validated with 3D-printed casts of a murine nasal cavity under 3R (replace, reduce, refine) criteria.

The aim of the present study was to transfer the region-specific technique from in vitro (3D casts (Flamm et al., 2018)) to mice and to determine the maximum application volume as well as the suitable insertion depth of the catheter. Both criteria were investigated in dependence of the bodyweight of the individual mice. In summary, we could successfully transfer and validate the region-specific catheter-based application technique from murine 3D casts to mice and were able to determine the average volume and insertion depth for olfactory targeting. While demonstrating here the establishment of this region-specific method in mice, we could recently provide first successful proof-of-concept studies using this technique (Maigler et al., 2021).

With the aid of a 3D cast of the murine nasal cavity we have recently determined the criteria for a refined region-specific intranasal delivery technique (Flamm et al., 2018). While with the conventional pipette-based method the whole nasal cavity is targeted, our refined technique is able to discriminate between the olfactory and respiratory region (Figure 1B).

A comprehensive refinement of the intranasal delivery in rodents has recently been demonstrated based on 3D anatomical cast models of the murine nose and nasal cavity (Flamm et al., 2018). The use of a catheter instead of a pipette tip for delivery enables region-specific targeting (Hanson et al., 2013; Gänger and Schindowski, 2018). Thereby, the exclusive targeting of the olfactory region for a standardized, reproducible N2B delivery becomes feasible. Also, the usage of the catheter in combination with a lipophilic ointment (petroleum jelly or a dexpanthenol-containing nasal balm) was shown to reduce the traction and to limit the spreading of the applied drug to the lower parts of the nasal cavity. Although there has been remarkable progress in the accuracy of 3D printing and prototyping processes in the last few years (Liska et al., 2007), the scanning and printing of small structures such as murine nasal turbinates and meatuses still proves to be a challenge. A further issue with considerable influence on the characteristics of a 3D printed prototype is the choice of material used in the printing process (Flamm et al., 2018). While a transparent and flexible silicone cast successfully mimics the cartilage-like texture of the murine skull, the hydrophobic surface characteristics of silicone are far from that of a mucosal surface. Nevertheless, the use of 3D printing is a promising approach to refine and advance drug delivery technologies under the 3R criteria (Engelhardt et al., 2016). Despite the above-mentioned limitations, the murine 3D casts have provided a good basis for the development and training of the catheter-based refined technique. In the present study, we have transferred this refined application method to mice and determined the relevant parameters for a standardized and reproducible methodology for future pharmacokinetic and bioactivity studies.

During the intranasal application, the nostril of the mouse poses the first important hurdle as it is the narrowest structure to be passed. Here, we suggest to use an angle of 20-25° relative to the noseline and to push the catheter forward by gentle rotations (Figures 2C–E). The mean penetration depth of the catheter was marginally longer in the application in the mouse than what was determined previously in the 3D printed casts. The differences in insertion depth may be due to the material properties of the 3D-printed casts. Probably, the delicate and fine structures of the ethmoid turbinates as they occur in mice could not be accurately resolved by the CT scans of the mice, or during the prototyping of the casts. Importantly, we suggest to use the insertion depth of the catheter respective to the mice’s weight as a means to control the application (Figures 1C,D and Figure 2A). While the penetration depth is limited when the catheter is inserted to the dorsal meatus (Figure 2F), an insertion to the middle meatus directs the catheter through the nasal passage further down to the lower airways or the esophagus, so that no apparent limitation in insertion depth occurs. By controlling the insertion depth, misapplications can be avoided.

As demonstrated by the sagittal sections prepared from the murine heads (Figures 3A–F), the refined technique resulted in a region-specific application to the ethmoid turbinate, which is the major part of the olfactory region (Doty, 2003). The rostral parts of the murine nasal cavity, which are predominantly covered with respiratory mucosa, did not display any fluorescence when the fluorescein solution was administered to the ethmoid turbinates (Figures 3A–F). The use of a lipophilic ointment achieved that not even traces of the fluorescein solution could appear at the respiratory regions, despite this area being part of the catheter’s path. Dexpanthenol-containing nasal balms are already approved for clinical use at the nasal mucosa and are hence as an ointment well suitable for the refined technique in rodents. The balm used in this study contains paraffin and petroleum jelly, which are both of hydrophobic character. Consequently, the ointment is able to hinder hydrophilic formulations and water-based solutions from spreading and prevents them from draining to the lower parts of the nasal cavity covered with respiratory mucosa (Peltier et al., 2020). Other formulations such as water-based hydrogels would reduce the surface tension of the water-based drops at the tip of the catheter and, hence, could possibly drain to the rostral parts covered with respiratory mucosa. In a recent study we demonstrated that a region-specific administration is highly important for CNS targeting (Maigler et al., 2021). Finally, the ointment also serves as lubricant to avoid irritations and lesions on the sensitive mucosa that may be caused by the catheter.

The hydrophobicity of the 3D casts’ surfaces was the reason for why it has not been feasible to determine the influence of breathing on the distribution of the intranasally applied solutions with the 3D printed prototypes. Preliminary unpublished data showed that the surface tensions of the administered drops rendered the spreading and wetting of the applied drug solution on the silicone surface as almost impossible. Therefore, any airstream applied with the aim of simulating the breathing pattern of a mouse in the 3D casts resulted in either a “sneezing”- or an “inhaling”-like behavior of the administered drops. Hence, the effect of breathing activity on intranasal drug distribution could not adequately be investigated in the 3D cast. Instead, the influence of breathing was now simulated in prepared mouse heads (Figure 3G) to determine the maximum delivery volume at the olfactory region and to avoid overapplication. Previous computer-based airflow simulations studies suggested that airflow in the murine olfactory region is mostly unidirectional (Coppola et al., 2017; Coppola et al., 2019). During inspiration, the air flows into the olfactory region via the dorsal meatus and then exits ventrally at the location of the nasopharynx. During expiration, the flow in the olfactory region is mostly stagnant, indicating that the olfactory region is continuously purged in a unidirectional manner. Although, in the present study the simulated aspiration of the fluoresceine solution occurred mainly during the first three initial simulated breathing cycles, we cannot exclude that the here used breathing simulation did not perfectly mimic the purged volume under in vivo conditions, which depends on the inspiratory airflow rate and the number of respiratory cycles. Probably, the remaining fluoresceine solution was trapped in the olfactory region due to high surface tension forces. Last but not least, in a recent in vivo study in mice we did not observe any significant levels of ingested fluoresceine with the here described technique while ingested fluoresceine was found with the pipette-based method (Maigler et al., 2021).

As expected, the maximum applicable volume exceeds the volume that remains in the olfactory region during breathing. The experimentally derived recommended volumes for intranasal application with the refined technique are displayed in relation to the mouse’s weight in Figure 3H and in Table 1. It should be noted that a suitable syringe is needed to apply such small volumes accurately. The parameters used in the present study correspond to the physiological conditions that occur in mice with an average weight of 31 g (Moshkin et al., 2014). It should be noted that the breathing frequency and intensity may change with stress or anesthesia during animal experimentation. Moshkin and co-workers determined the intranasal volume from one hemisphere of a mouse nose using magnetic resonance imaging and obtained a volume of 17 µL. The higher volume can be accounted to the fact that the anterior nasal cavity (covered with respiratory mucosa) was not included in our experiments. However, it can be assumed that volumes of at least 20 µL are delivered by the pipette-based method (Figure 1A), which will undoubtedly target the respiratory as well as the olfactory mucosa and will be swallowed, inhaled or exhaled, which altogether makes it impossible to generate reliable quantitative data (Coppola et al., 2014; Kozlovskaya et al., 2014). The here presented region-specific intranasal technique allows the administration of a small volume to the ethmoid turbinates, thanks to which it prevents the discussed issues of the conventional method and enables the performance of standardized and reproducible intranasal N2B studies. Finally, the here described technique has proven itself as successful in intranasal in vivo studies (Maigler et al., 2021).

As recently demonstrated, N2B transport via the olfactory region is an application route that allows hydrophilic and large molecules to enter the CNS by bypassing the blood-brain barrier (Maigler et al., 2021). In contrast, small lipophilic molecules (<500 Da) often cross the blood-brain barrier by diffusion. Therefore, our refined intranasal delivery technique via the olfactory region is highly suitable for larger hydrophilic drugs such as proteins. However, when administering lipophilic drugs to the olfactory mucosa, the drug formulation should be adapted to the pharmacokinetic properties of the drug. An oil-in-water (o/w), a water-in-oil (w/o) emulsion or a flowable cream with these properties could be suitable for lipophilic drugs. Presumably, it would be useful to substitute the here described hydrophobic dexpanthenol-containing ointment with a hydrophilic liquid or semi-solid formulation such as a Carbomer-based hydrogel. In this case, drug absorption via the respiratory region would also be reduced. However, the mucus layer covering the respiratory mucosa already has a mucin-based hydrogel so the effect should be less pronounced.

Last but not least, the here described delivery method also enables the region-specific delivery of viscous formulations, such as hydrogels. While the pipette-based method can only apply formulations with a low viscosity to the olfactory region by filling up the entire nasal cavity, our novel technique makes it possible to exclusively reach the ethmoid turbinates with gel formulations. Nevertheless, the application of gels must be performed carefully, slowly and steadily. This allows the gel emerging from the application device to gradually distribute in the nasal cavity, which ensures that little pressure is exerted on the nasal septum. Finally, when applying gels, it should be considered that mice are obligate nasal breathers (Harkema et al., 2006). In this context, care must be taken that the applied gel does not occlude the nasal airways. Therefore, it seems necessary to limit the application of intranasal gel formulations to one nasal hemisphere.