All experiments utilizing animals were approved by the Institutional Animal Care and Use Committees of Northwestern University and in accordance with the National Institutes of Health guidelines. Animals were housed in Northwestern’s Center for Comparative Medicine facilities.

The Camsap3^{tm 1a (EUCOMM) Wtsi} (referred to as Camsap3-KD) mouse model was obtained from the Wellcome Trust Sanger Institute. The original Camsap3-KD line on the C57B6N background was re-derived on FVB murine backgrounds to increase their viability (Robinson et al., 2020). As shown in Figure 1A, a Camsap3-KD mouse was created using the “knockout first, conditional-ready” strategy (Skarnes et al., 2011). A targeted trap allele was inserted into the intron between the exon 6 and 7 of the Camsap3 gene to knock out Camsap3 expression through RNA processing (Testa et al., 2004; White et al., 2013) without deleting any exons. To create a cKO mouse model, Camsap3-KD mice were first crossed with B6-Tg(CAG-FLPe)36 (Kanki et al., 2006) to delete the neo and LacZ cassettes flanked by FRT sites and create Camsap3^{tm 1c} (floxed allele), a strain of pseudo-wildtype (WT) mice. This WT Camsap3^{tm 1c} floxed line was further crossed with the TgPax2^Cre mouse line, Tg(Pax2-Cre)1Akg/Mmnc (MMRRC stock number 10569) (Ohyama and Groves, 2004) to create a conditional knockout mouse model: TgPax2^Cre/+; Camsap3^{tm 1c/tm1c} and their WT littermates: TgPax2^Cre+/+;Camsap3^{tm 1c/tm1c}. We call this mouse model Camsap3^{tm 1d/tm1d} (null allele) Camsap3-cKO for short. All animals were maintained by heterotypic breeding, and the genotypes were determined by detecting the inserted cassettes in mouse tail samples performed by Transnetyx (Cordova, TN, United States).

Since Cre expression was limited in a few tissues including the inner ear, olfactory bulb, and kidneys of TgPax2^Cre mice, the exon 7 of camsap3 was expected to be fully removed from these tissues of Camsap3-cKO mice. To verify the exon 7 exclusion, genomic DNA was purified from the inner ear of WT and Camsap3-cKO mice using the QIAamp DNA FFPE Tissue Kit (Qiagen). PCR was used to verify the exon 7 exclusion using the PCR Master Mix (Thermo Fisher Scientific, A44647100) with forward primer L1L2-BactP-MD-F (GCTGGCGCCG GAACC) and reverse primer C3Floxed5-R (TTGGCCTGGGGAACATGAC). Conditions for the thermal cycler were: step 1–95°C, 2 min; step 2–95°C, 30 s; step 3–55°C, 30 s, step 4–72°C, 90 s; repeat steps 2–4, 35 × ; step 5–72°C, 10 min. Expected sizes were 1,307 base pairs (bp) for WT, and 199 bp for Camsap3-cKO mice. Both male and female mice were used in all experiments.

The inner ears, olfactory bulbs, and olfactory mucosa in nasal cavities were collected from WT mice and their Camsap3-cKO littermates for immunofluorescence as previously described (Sekerka et al., 2011; Robinson et al., 2020). Briefly, samples were immersion fixed in 4% formaldehyde in phosphate buffered solution (PBS). Some samples underwent whole mount preparation, in which the utricles and cristae were further dissected out and used for immunostaining. Other samples were cryosectioned before staining. These samples were first decalcified in 0.125M EDTA for 2 days. Decalcified samples were then placed in a series of sucrose solutions in 1X PBS (10–30%) then two changes of Tissue-Tek Optimal Cutting Temperature (OCT) Embedding Medium (Sakura Finetek, 4583), and embedded in fresh OCT. The inner ear samples were then sectioned into 12 micrometer thick slices, and post-fixed in 2% formaldehyde for 10 min and blocked at room temperature for 1 h in blocking solution: 5% goat serum, 2% Triton X-100 in Tris-buffered saline (TBS). The following primary antibodies were used for immunostaining: anti-Camsap3 (Robinson et al., 2020) at 1:2500, anti-acetylated tubulin at 1:500 (Thermo Fisher Scientific 32-2700, AB_2533073). Samples were incubated with primary antibodies at 4°C overnight. The next day samples were then washed in PBS and incubated with appropriate fluorophore-conjugated secondary antibodies, including goat anti-rabbit Alexa 488 at 1:500 (Thermo Fisher Scientific 32-2700, AB_143165), goat anti-mouse Alexa Fluor 488 (RRID: AB_2556548), goat anti-mouse IgG2b Alexa 647 at 1:500 (Thermo Fisher Scientific AB_143165), Hoechst at 1:1000 (Thermo Fisher Scientific, H3570), and Phalloidin-Alexa 568 at 1:2000 (Thermo Fisher Scientific 21838, AB_2532159), or Phalloidin-Alexa 647 at 1:400 (Thermo Fisher Scientific A22287) for 2.5 h at room temperature. The stained samples were subsequently mounted in Fluoromount Aqueous Mounting Medium (Sigma, F4680). When mounting the whole mount utricle samples, a 0.12 mm deep Secure-Seal spacer (Thermo Fisher Scientific, S24735) was placed between the slide and the cover glass in order to avoid destruction of the vestibular kinocilia on hair cells. Immunostained samples were imaged using a Nikon C2 or A1R+ confocal microscope using objects of a plan Apo VC 20×/0.75 DIC N2 (Nikon) and a plan Apo 60×/1.4 oil (Nikon). All images were taken as 12-bit, which allows for 0–4,095 range. For vestibular kinocilia length measurement, Z-stacks images were captured from the apical surface of utricle hair cells to the tip of vestibular kinocilia (from the position “A” extended to “C-D” shown in Figure 3B) using the optical section (0.5 μm). These Z-stack images were then reconstructed into 3D images, which were analyzed using Imaris 8 (Bitplane) and Nikon NIS Element software. Mean gray value, the sum of the gray values of all the pixels in the selection divided by the number of pixels, was measured using FIJI.

Utricles from WT mice and their Camsap3-cKO littermates were dissected out and immersion fixed in 0.1 M sodium cacodylate buffer pH 7.3, containing 2% paraformaldehyde and 2.5% glutaraldehyde for at least 24 h and post-fixed with 2% osmium tetroxide followed by 3% uranyl acetate. Tissues were then dehydrated in ascending grades of ethanol, transitioned with propylene oxide and embedded in resin mixture from the Embed 812 kit, cured in a 60°C oven. Samples were sectioned on a UCT ultramicrotome (Leica Microsystems). To identify regions of interest 1 μm thick sections were collected, stained with Toluidine Blue O and examined by light microscopy. 70 nm tissue sections were collected on 200 mesh copper grids and stained with uranyl acetate and Reynolds lead citrate, and examined using a FEI Tecnai G2 Spirit Transmission Electron Microscope. The number of MTs in kinocilium was counted and sorted by structure: “9+2,” “9+0,” “8+2,” or “8+0.” The first number indicates the numbers of MT doublets outlying, and the second number implies the amount of MTs in the center.

The Buried Food Test was used to evaluate olfaction in mice (Yang and Crawley, 2009; Robinson et al., 2020). WT mice and their Camsap3-cKO littermates were food restricted but provided water ad libitum for 24 h prior to testing. Mice were given a piece of a Nutter Butter cookie during this day to familiarize the mice with its scent. Testing was performed during daylight, starting at 9 AM. The base of a Sterilite translucent storage box (cage) [47 cm (L) × 37.8 cm (W) × 28.3 cm (H)] was filled with bedding to a depth of 3 cm. Each mouse was placed in an acclimation bin which was identical to the testing bin in every way, with the exception of no food being present, and were given 5 min to adjust to the enclosure. Mice were then transferred individually to the testing cage, which had ∼ 0.5 g of Nutter Butter cookie hidden in the center of the cage under the bedding. As soon as the mouse entered the testing bin a stopwatch was started and the latency to find (LTF) the cookie was recorded. Following the experiment, mice were returned to their home cages and given their full allotment of food.

As previously reported, the global KD mice, Camsap3-KD, were born with smaller bodies along with sub-infertility, hydrocephalus, anosmia or hyposmia, sinusitis, respiratory distress, otitis media, and hearing loss (Perez-Garcia et al., 2018; Ingham et al., 2019; Robinson et al., 2020). In contrast to Camsap3-KD, Camsap3-cKO had normal fertility and a normal appearance comparable to their WT littermates regardless of their sex prior to 4 months of age, The normal appearance of Camsap3-cKO was not surprising as Camsap3 was expected to be eliminated only in the inner ear, midbrain, cerebellum, olfactory bulb, and the kidneys (Ohyama and Groves, 2004). To test whether the exon 7 of Camsap3 was indeed removed from the genome of affected cells, we performed PCR genotyping using genomic DNA isolated from the inner ear, where Cre was expressed to remove the floxed exon 7 of Camsap3. As shown in Figure 1B, in all Camsap3-cKO mice (Pax2-Cre positive), the expected short fragments (199 bps) were detected, but not in mice without the Pax2-Cre allele (Pax2-Cre negative) or original homozygous Camsap3^{tm 1c} mice. We noted that the long PCR band, 1307 bp, was also detected in Camsap3-cKO mice with low intensity. As we did not perform cardiac perfusion before extracting genomic DNA from the inner ear, we suspected that cells from the circular bloodstream in the inner ear could contribute to the long PCR bands as the genomic DNA from these cells were not affected by Pax2-Cre expression. It is also possible that the exon 7 of Camsap3 was not completely removed from all cells in the inner ear. To verify whether Camsap3 was expressed in the vestibular hair cells, we performed immunofluorescence using vestibular organs collected from Camsap3-cKO and their WT littermates, with the ages of postnatally day (P) 23 to 260. In WT, CAMSAP3 expression was found in the vestibular organ including the hair cells in cristae (Figure 1C), utricle (Figure 1F), and saccule (data not shown). CAMSAP3 expression in Camsap3-cKO was absent or reduced in vestibular cells from Camsap3-cKO mice as shown in Figures 1D,G. CAMSAP3 staining was expressed in both hair cells and their surrounding supporting cells in vestibular organs from WT samples. CAMSAP3 signals were more concentrated in the apical cortical area of the WT vestibular organ (Figure 1C). Such distribution patterns are unsurprising as CAMSAP3 is needed at apical cortical area of epithelial cells to establish the intercellular connection and orient the apical-to-basal polarity of microtubule arrays (Meng et al., 2008; Toya et al., 2016). We further quantified CAMSAP3 staining intensities in the apical cortical area that included both vestibular hair cells and supporting cells. The mean gray value (intensity) of CAMSAP3 staining in these areas were measured using FIJI software. As shown in Figure 1E, CAMSAP3 expression was significantly decreased in the hair cell area and supporting cells from Camsap3-cKO mice, with CAMSAP3 expression in Camsap3-cKO mice equaling 48.74% of expression in WT mice at P75 (p = 0.031, unpaired t-test, n = 4). In sum, we created a cKO model that significantly reduced CAMSAP3 proteins in the vestibular system.

To investigate CAMSAP3’s role in establishing kinocilia on the vestibular hair cells, utricles collected from Camsap3-cKO mice (3 male) and their WT littermates (3 male) around 1-year old (P296-P384) were examined using immunofluorescent staining of anti-acetylated tubulin and phalloidin, which labeled vestibular kinocilia (green) and stereocilia (red, dash lines), respectively (Figure 2A). We measured the length of kinocilia on vestibular hair cells of utricles from Camsap3-cKO mice and their WT littermates at the most anterior portion for consistency (Figure 2B). From each mouse sample, the length of kinocilium was measured from base to tip in micrometers. Great care was taken to exclude vestibular kinocilia whose full length was not captured in the confines of the reconstructed 3D image. The average length of the vestibular kinocilia, in mean ± SD, were 20.39 μm ± 3.21 for WT and 17.16 μm ± 3.03 for Camsap3-cKO (Figure 2C). An unpaired t-test revealed a significant reduction in vestibular kinocilia length in Camsap3-cKO mice when compared with their WT littermates (p < 0.0001).

Our previous work demonstrated that CAMSAP3 was needed for de novo formation of the central MT pair in a “9+2” configuration observed in the axons of motile cilia (Robinson et al., 2020). To test whether CAMSAP3 was also needed for formation of the central MT pair in the kinocilia on vestibular hair cells, we first examined the connection between CAMSAP3 and basal bodies by immunofluorescence. The basal bodies that extend to 9 peripheral MT doublets were labeled by anti-γ-tubulin, and phalloidin was used to label the actin-based stereocilia that were adjacent to the kinocilium. As shown in Figure 3A, a group of Z-stack images taken from the apical surface of utricle hair cells (the position “A” shown in Figure 3B) showed two vestibular kinocilia labeled by γ-tubulin (arrows). A stereocilia bundle was next to one of kinocilia, indicating these cells were hair cells from an utricle. Similar to other epithelial cells, CAMSAP3 punctuates (green) were found on the apical surface of hair cells. Two CAMSAP3-stained green dots (arrows) were colocalized with two basal bodies (red dots). As the focus pointing moved step by step (0.5 μm/per section) from the apical surface toward the nucleus, green CAMSAP3 dots (arrows in Figure 3A) in both vestibular kinocilia disappeared before the red basal bodies punctuates, suggesting that CAMSAP3 was at the base of vestibular kinocilia just above the basal bodes, as illustrated in Figure 3B. Such a close connection between CAMSAP3 and basal bodies was of particular interest because CAMSAP3 is not a centrosome- or basal bodies-bound protein. In fact, CAMSAP3 was not observed in the basal bodies of primary cilia with “9+0” MT configuration (Robinson et al., 2020). More importantly, the close association between γ-tubulin and CAMSAP3 was comparable to that observed in motile cilia with “9+2” MT configuration (Robinson et al., 2020; Saito et al., 2021), suggesting that CAMSAP3 was also involved in de novo formation of the central MT pair in the kinocilia on the vestibular hair cells.

To further verify the role of CAMSAP3 in formation of the central MT pair in vestibular kinocilia, the MT structure differences between WT and Camsap3-cKO vestibular kinocilia were then evaluated using TEM (Figures 3C–E). We examined 42 axoneme structures (the position “C-D” shown in Figure 3B) from 8-month-old WT mice and 33 axonemes from their Camsap3-cKO littermates. The “9+2” configuration here includes 9 MT doublets peripherally and two MT singlets or amorphous material with higher electron density (darker) at the center of axonemes. This is demonstrated in a transverse section of WT vestibular kinocilia (Figure 3C, indicated by an arrow). The “9+0” configuration lacks the central MTs structure and electron-dense materials as seen in the Camsap3-cKO image in Figure 3D. More than 71% of vestibular kinocilia from WT mice have “9+2” configuration, while only 24% vestibular kinocilia from Camsap3-cKO mice showed similar structure. The majority of Camsap3-cKO vestibular kinocilia (55%) have the “9+0” configuration. In addition, irregular configurations, “8+2” and “8+0,” were found in both WT (12%) and Camsap3-cKO (21%). Distribution of the MT configurations between WT and KO was statistically significant different as analyzed using a Kolmogorov-Smirnov distribution comparison test (p = 0.0003). Our data suggests that vestibular kinocilia of Camsap3-cKO were more likely missing the central MT pair than WT vestibular kinocilia.

Kinocilia are known to be essential for establishing the orientation of stereocilia bundles on hair cells. Data collected from P23 to 13-month showed that vestibular kinocilia were present on the vestibular hair cells of Camsap3-cKO mice just as their WT littermates. In addition, the shape and orientation of stereocilia on vestibular hair cells from Camsap3-cKO were quantitatively similar to those on WT mice despite the vestibular kinocilia from Camsap3-cKO mice were shorter and more likely to miss the central MT pair (Figures 2, 3). We also investigated the impact of CAMSAP3 on the orientation of stereocilia bundles by comparing auditory hair cells from WT and Camsap3-cKO mice with matched ages and sexes. As shown in Supplementary Figure 3, the orientation of stereocilia bundles on auditory hair cells from Camsap3-cKO mice were comparable to those on WT auditory hair cells. This data is not surprising because most of kinocilia on auditory hair cells are with “9+0” configuration (Sobkowicz et al., 1995), and CAMSAP3 signals do not co-localize with the basal bodies where “9+0” axonemes are originated (Robinson et al., 2020). In summary, CAMSAP3 is unlikely to have a significant impact on cochlear kinocilia and the orientation of stereocilia.

In our original measurements of young mice, ranging from 1 to 4 months old, there was no weight difference between Camsap3-cKO mice and their WT littermates (Figure 4A). This phenotype differs from KD mice, Camsap3-KD, which were born smaller than their WT littermates (Robinson et al., 2020). Since we did not notice obvious body weight differences like those observed in the global Camsap3-KD mice, we assumed that their weights remained comparable. Though there were no obvious weight differences between WT and Camsap3-cKO mice when mice were less than 4-month old, weight differences emerged in 8-month old mice (Figure 4A). For mice 8 months and older (P237–P388), the average male WT mouse weighed (mean ± SD) 39.61 g ± 8.703 (n = 21), and the average male Camsap3-cKO mouse weighed 26.49 g ± 4.904 (n = 13). For female mice older than 8 months, the average WT mouse weighed 30.56 g ± 4.687 (n = 8), and the average female Camsap3-cKO mouse weighed 22.38 g ± 1.727 (n = 7). WT mice older than 8-month were significantly larger than the Camsap3-cKO mice regardless their sex (Figure 4).

Since CAMSAP3 was essential to build or maintain normal structure of vestibular kinocilia (Figures 2, 3), we further investigated whether shorter kinocilia on vestibular hair cells caused vestibular dysfunction. Vestibular dysfunction can manifest as issues with balance or gait abnormalities. In their home cages, Camsap3-cKO mice did not show obvious head tilt, walked normally, and were indistinguishable from their WT littermates. To assess the vestibular function of Camsap3-cKO mice, we performed multiple methods aiming to detect more subtle abnormalities. The simplest methods of testing vestibular functions include a swim test and a right-reflex test (Hardisty-Hughes et al., 2010). The swim test aims to determine if the mice could properly swim in a straight line and keep themselves upright. After assessing 26 WT (14 male and 12 female), and 27 Camsap3-cKO (14 male and 13 female) mice, ages ranging from P22 to P388, we found that there were no notable differences between the Camsap3-cKO and WT mice regarding their ability to swim. The right-reflex test assesses whether the mice can re-right themselves when flipped upside down. The same group of 26 WT and 27 Camsap3-cKO mice completed the right-reflex test. We found no difference in performance between the WT and Camsap3-cKO mice as each of the mice was able to flip themselves back to upright immediately (within 1 s). Subsequently, we analyzed the balance ability of Camsap3-cKO mice by rotarod apparatus. On the Rotarod, we assessed 12 WT (6 male and 6 female) and 8 Camsap3-cKO (5 male and 3 female) mice, ages P37–P153, and found that there was no significant difference in TTF between the two groups (data not shown).

The vestibular function of Camsap3-cKO mice was also examined using the DigiGait Imaging System (Berryman et al., 2009). Since mice older than 8-month Camsap3-cKO mice were lighter than their WT littermates (Figure 4A), mice at 4 months old were selected for gait measurement. At this age, gait was fully developed (Akula et al., 2020; Rahn et al., 2021), and Camsap3-cKO mice were not significantly lighter than their WT littermates (Figure 4A). The assessed litter consisted of 6 male mice: 3 WT and 3 Camsap3-cKO. We collected data from these mice using the DigiGait™ Imaging System. Of all the aspects of gait analyzed by the DigiGait analysis software, most showed no significant difference between the WT and Camsap3-cKO mice. Some of the results were shown in the Supplementary Figures 1, 2). Two aspects of gait with significantly differences were paw area at peak stance and maximum dA/dt (Figure 5). Paw area (cm²) is the maximal paw area captured by the camera and corresponds with the time of “peak stance” (Supplementary Figure 1). When compared to their WT littermates, Camsap3-cKO mice had smaller areas of their paws in contact with the treadmill for both fore and hind limbs regardless of running speeds (Figures 5A,B). For example, at a speed of 17 cm/s, the paw area that Camsap3-cKO mice used to touch the treadmill was 62.1% (fore limbs) and 58.9% (hind limbs) of the paw area WT mice used. Maximum dA/dt (cm²/s) is the maximal rate of change of paw area in contact with the treadmill belt during the braking phase. At a low treadmill speed (10 cm/s), there was no substantial difference between the WT and Camsap3-cKO mice in their dA/dt for both fore and hind lambs. However, as the treadmill speeds increased, the difference between the WT and Camsap3-cKO mice became more apparent. At a high treadmill speed (24 cm/s), the maximum dA/dt decreased in Camsap3-cKO mice’s hind limbs (−10.5%), while WT mice had increased maximum dA/dt (+10.0%). Similar to paw area, maximum dA/dt in Camsap3-cKO mice were always slower than their WT littermates independent of walking speed. These differences exist despite there being no difference in the body length, width (Figures 5E,F), and weight (Figure 4A) between the WT and Camsap3-cKO mice.

In TgPax2^Cre mouse line, Cre was detected in the embryonic kidney since the 12-somite stage to remove the floxed allele (Ohyama and Groves, 2004). Since CAMSAP3 is essential to orienting the apical-to-basal polarity of MT arrays in epithelial cells (Toya et al., 2016), we suspected that kidney function might be affected in Camsap3-cKO mice. Therefore, both kidneys were prepared out of euthanized WT and cKO mice with ages of P22–P339. Since the kidney-to-body weight ratio has been commonly used to predict kidney function (Hughes et al., 1999; Mitsuhata et al., 2021), we weighed 19 WT mice and 7 Camsap3-cKO mice and their kidneys. Kidney weight as a proportion of body weight was recorded. Female mice that had previously given birth were excluded. Interestingly, we found that Camsap3-cKO mice, especially the males, had statistically significant heavier kidneys (as a function of body weight) than their WT littermates (Figure 6). In male WT mice, their kidneys made up (mean ± SD) 1.47 ± 0.13% of their body weight, and in male Camsap3-cKO mice, their kidneys equaled 3.90 ± 1.5% of their body weight (Figure 6A). In female WT mice, their kidneys made up 1.51 ± 0.34% of their body weight, while in female Camsap3-cKO mice, their kidneys equaled 2.23 ± 0.30% of their body weight (Figure 6B). Kidneys from the Camsap3-cKO mice were hypertrophic and discolored (Figure 6C). Such a phenotype was similar to those observed in CAMSAP3-mutant knocking in mice model Camsap3^dc/dc, in which exons 14–17 of Camsap3 were removed from the genome of Camsap3^dc/dc mice (Toya et al., 2016). Exons 14–17 encode the key MT binding domain of CAMSAP3, CKK domain. Because the mutant CAMSAP3 without the CKK domain is incapable of binding to MTs to form proper MT network in the epithelial cells in kidney, Camsap3^dc/dc mice developed malfunction of kidneys with cyst at the proximal convoluted tubules starting at E17.5 even though kidney had normal appearance at P21 (Mitsuhata et al., 2021). Since renal dysfunction often caused body weight loss (Hickman and Swan, 2010), the deterioration of kidneys observed in Camsap3-cKO may contribute to the small body size of Camsap3-cKO mice older than 8-month (Figure 4). In consequence, health issues arose, and these mice became thin, hunched, and moved slowly. These health issues prevented further investigation of CAMSAP3’s impact on vestibular function in Camsap3-cKO mice that were older than 13 months of age.

Since Cre was expressed in the olfactory bulb in TgPax2^Cre mouse line to remove the floxed Camsap3 exon 7, we investigated whether Camsap3-cKO had normal olfaction. Olfaction evaluation was performed for WT (11 male and 10 female) and Camsap3-cKO mice (12 male and 12 female) from 8 litters of mice, ages P55–P313. Mice were placed in a large container filled with bedding, and their latency to find (LTF) a buried Nutter Butter cookie was recorded. As shown in Figure 7A, WT and textitCamsap3-cKO mice took equal amounts of time to find the buried cookie. There was no significant difference between WT and Camsap3-cKO mice regardless their sex. We then examined CAMSAP3 expression in the olfactory bulb. Olfactory bulbs from P180 WT and Camsap3-cKO littermates were collected and stained with anti-Camsap3. As shown in Figures 7B,C, there was little CAMSAP3 expression in cells located in olfactory bulbs from both the WT and Camsap3-cKO mice, suggesting that CAMSAP3 has a minimal role in olfactory bulbs. We also compared CAMSAP3 expression in olfactory sensory neurons, where CAMSAP3 was abundantly expressed (Robinson et al., 2020). As shown in Figures 7D–G, CAMSAP3 was abundantly expressed in olfactory sensory neurons located in the nasal olfactory epithelia in both the WT and Camsap3-cKO mice. CAMSAP3 expression in olfactory sensory neurons is not affected in Camsap3-cKO mice and remained concentrated at dendritic knobs under the olfactory cilia (Figures 7F,G). In other words, CAMSAP3 expression in cells involved in olfaction were not changed in Camsap3-cKO mice compared to their WT littermates. Given the equivalent CAMSAP3 expression in cells involved in olfaction (Figures 7B–G), it was unsurprising that the Camsap3-cKO mice performed equally as well as WT mice on the open-field olfaction assessment.