Cerebellar Morphology and Behavioral Profiles in Mice Lacking Heparan Sulfate Ndst Gene Function

Disruption of the Heparan sulfate (HS)-biosynthetic gene N-acetylglucosamine N-Deacetylase/N-sulfotransferase 1 (Ndst1) during nervous system development causes malformations that are composites of those caused by mutations of multiple HS binding growth factors and morphogens. However, the role of Ndst function in adult brain physiology is less explored. Therefore, we generated mice bearing a Purkinje-cell-specific deletion in Ndst1 gene function by using Cre/loxP technology under the control of the Purkinje cell protein 2 (Pcp2/L7) promotor, which results in HS undersulfation. We observed that mutant mice did not show overt changes in the density or organization of Purkinje cells in the adult cerebellum, and behavioral tests also demonstrated normal cerebellar function. This suggested that postnatal Purkinje cell development and homeostasis are independent of Ndst1 function, or that impaired HS sulfation upon deletion of Ndst1 function may be compensated for by other Purkinje cell-expressed Ndst isoforms. To test the latter possibility, we additionally deleted the second Purkinje-cell expressed Ndst family member, Ndst2. This selectively abolished reproductive capacity of compound mutant female, but not male, mice, suggesting that ovulation, gestation, or female reproductive behavior specifically depends on Ndst-dependent HS sulfation in cells types that express Cre under Pcp2/L7 promotor control.


Introduction
Heparan sulfate (HS) is produced by most mammalian cells as part of membrane and extracellular matrix proteoglycans [1]. The chain grows by exostosin (Ext) copolymerization of GlcAβ1,4 and GlcNAcα1,4 and is modified by one or more of the four Ndst isozymes; the N-deacetylase activity of Ndsts removes acetyl groups from GlcNAc residues, which are then converted to GlcNS through the N-sulfotransferase activity. Subsequent modifications of the HS chain by most O-sulfotransferases and a GlcA C5-epimerase depend on the presence of GlcNS residues, making the Ndsts responsible for the generation of sulfated ligand binding sites in HS [2]. Ndst1 and Ndst2 mRNA are expressed in all embryonic and adult tissues examined, whereas Ndst3 and Ndst4 transcripts are predominantly expressed during embryonic development [3]. Behavior: For behavioral tests, male and female wildtype (n = 12,9), heterozygous (n = 13, 11) and mutant mice (n = 9, 13) were compared. Mice were kept under a 12 h/12 h light dark cycle for 3 weeks before testing began. Food (10 H 10, Nohrlin GmbH, Bad Salzuflen, Germany) and water were available ad libitum. Mice were housed in standard (37 (l) × 21 (w) × 15 (h) cm) cages, with up to five sex-matched animals per cage. Ear punctures allowed the discrimination of individual mice but behavioral experiments were carried out with the experimenter being unaware of the genotypes of the subjects. The cages contained a thin layer of wood shavings and paper towels as nesting material. Mice were looked after daily, but handled only once a week while transferring them to clean cages. All procedures and protocols met the guidelines for animal care and experiments in accordance with national (Animal research permit number A32/2006) and European (86/609/EEC) legislation. General health and neurological status were assessed using a protocol including tests as described elsewhere [44]. Animals were inspected for physical appearance and underwent neurological testing including acoustic startle, visual placing, grip strength and reflex functions to ensure that behavioral findings were not the result of deteriorating physical conditions of the animals. The Rotarod test was employed to assess balance and motor function (TSE RotaRod, TSE Systems GmbH, Bad Homburg, Germany). Mice were placed on the rotating cylinder of 3 cm diameter, and latency to fall off the rotating cylinder was measured automatically. Initial speed set to 4 rpm was accelerated to 400 rpm Behavior: For behavioral tests, male and female wildtype (n = 12, 9), heterozygous (n = 13, 11) and mutant mice (n = 9, 13) were compared. Mice were kept under a 12 h/12 h light dark cycle for 3 weeks before testing began. Food (10 H 10, Nohrlin GmbH, Bad Salzuflen, Germany) and water were available ad libitum. Mice were housed in standard (37 (l) × 21 (w) × 15 (h) cm) cages, with up to five sex-matched animals per cage. Ear punctures allowed the discrimination of individual mice but behavioral experiments were carried out with the experimenter being unaware of the genotypes of the subjects. The cages contained a thin layer of wood shavings and paper towels as nesting material. Mice were looked after daily, but handled only once a week while transferring them to clean cages. All procedures and protocols met the guidelines for animal care and experiments in accordance with national (Animal research permit number A32/2006) and European (86/609/EEC) legislation. General health and neurological status were assessed using a protocol including tests as described elsewhere [44]. Animals were inspected for physical appearance and underwent neurological testing including acoustic startle, visual placing, grip strength and reflex functions to ensure that behavioral findings were not the result of deteriorating physical conditions of the animals. The Rotarod test was employed to assess balance and motor function (TSE RotaRod, TSE Systems GmbH, Bad Homburg, Germany). Mice were placed on the rotating cylinder of 3 cm diameter, and latency to fall off the rotating cylinder was measured automatically. Initial speed set to 4 rpm was accelerated to 400 rpm within a maximum session time of 300 s. Mice conducted two trials per day on 5 consecutive days. In the open-field test mice had the opportunity to explore a square arena (80 by 80 cm, walls 40 cm high) for 5 min. Locomotor activity and the ratio between exploration and fear of open space, as measured by the time spent near the walls or in the center of the arena, were assessed automatically by a tracking system [Lewejohann, L. Digital Image Processing in Behavioral Sciences. http://www.phenotyping.com/digital-image-processing.html. Date of Access: 03/07/2020]. In the Dark-Light test, anxiety-related behavior, e.g., the tendency of mice to stay in dark sheltered areas, was measured. The test apparatus consisted of a standard cage with one-third of the area being painted in black and covered by an opaque plastic ceiling. The dark compartment was separated from the light compartment by an opaque wall with a guillotine door. Each mouse was placed into the dark compartment for 1 min before the door was opened. Latency to enter the light compartment as well as the time spent in the light compartment were measured for 5 min using a Palm handheld computer and proprietary software for behavioral recording. Anxiety-related behavior was further assessed using the Elevated plus-maze [45]. Here, mice had the choice of moving into opposing arms, which were either shielded or open. The maze was elevated 50 cm above the floor and arms were 30 cm long and 5 cm wide. At the beginning of a trial, mice were placed into the center of the maze, facing one of the shielded arms. Each entry into an open or shielded arm was counted and the time animals spent in either type of arm was measured automatically using a tracking system [Lewejohann, L. Digital Image Processing in Behavioral Sciences. http://www.phenotyping.com/digital-image-processing.html. Date of Access: 03/07/2020]. Data analysis was conducted using the statistical software "R" (http://www.r-project.org/). Pairwise comparison of datasets that showed a non-Gaussian distribution was done two-tailed using the two-sample Wilcoxon test. A repeated-measures ANOVA was calculated to analyze performance over time in the Rotarod. The Chi2 test was used to determine statistical significance between litter sizes of Ndst1;Ndst2 compound mutant mice. A p-value <0.05 was considered as statistically significant.
Histology and detection of mRNA expression. Mouse brains were fixed in 4% paraformaldehyde overnight, dehydrated, embedded in paraffin and sectioned at 8 µm. Sections were stained with hematoxylin and eosin for histological analysis. For in situ hybridization, 700 base probes against the most variable N-terminal regions of Ndst1 and Ndst4, or directed against the 3 UTR of Ndst2 and the 5 UTR of Ndst3 were employed (DIG RNA Labeling Kit, Roche, Mannheim, Germany).

Ndst1 Expression in the Mouse Cerebellum
Aberrant sulfation of HS upon Ndst1 deletion in mice affects several neurodevelopmental processes and is perinatal/postnatal lethal [9,10,12,13,21]. To study the roles of HS in adult brain homeostasis in vivo, mice harboring a conditional, loxP-flanked allele of Ndst1 (Ndst1 f/f ) were thus crossed with transgenic mice expressing the bacteriophage recombinase Cre under the control of the Purkinje cell protein 2 (Pcp-2 or L7) promotor, which is predominantly expressed in postnatal cerebellar Purkinje cells [43] ( Figure 1A). Postnatal (P30) Purkinje cells in the cerebellum strongly express Ndst1, and only limited amounts of Ndst2, as demonstrated by in situ hybridization ( Figure 1B,D). Ndst3 and Ndst4 expression was weak or absent ( Figure 1E,F). This identifies Purkinje neurons are the main HS-producing cell type in this tissue, making them ideal to study the role of neuronal HS in adult brain function. To this end, the well-established method of cell-type specific Cre-mediated Ndst1 deletion was employed [9,[46][47][48].

Cerebellar Architecture after L7 Mediated Ndst1 Deletion
Next, we stained cerebellar sections of Ndst1 f/f ;L7-Cre + (mutant) and Ndst1 f/f (producing wildtype HS) mice with anti-Cre antibodies to confirm specific expression of the recombinase in cerebellar Purkinje cells ( Figure 2A). L7/pcp expresses Cre-recombinase starting from postnatal day (P)6, with strongest expression from P14 to P21 [43]. Consistent with this, we detected Cre-expression in Ndst1 f/f ;L7-Cre + mice at P20 and throughout adulthood. Thus, Ndst1 function is deleted during the time when murine cerebellar anatomy, synaptic connectivity and neuronal morphology are established (up to the third to fourth postnatal week). Indeed, in situ hybridization employing a 500 bp probe directed against the floxed exon 2 confirmed Ndst1 deletion in Cre-expressing Purkinje cells ( Figure 2C). Loss of Ndst1 protein in Cre-expressing double transgenic Purkinje cells of Ndst1 f/f ;L7-Cre + mice, but not in Ndst1 f/f mice, was further confirmed immunohistochemically ( Figure 2E).
Therefore, we next employed immunohistochemistry using anti-calbindin antibodies to assess possible changes in the number and morphology of Purkinje cells. However, no differences in cerebellar folia development, laminar organization or cross-sectional areas of the molecular or granule cell layers were found between control and Ndst1 f/f ;L7-Cre + mice, and no significant changes in the number, density or organization of Purkinje cells was observed. This suggested that survival of these projecting neurons in vivo does not require Ndst1 function ( Figure 3A,B). Calbindin staining also showed that primary and secondary trees of Purkinje cells were present and similarly branched in Ndst1 f/f ;L7-Cre + conditional mutant mice and Ndst1 +/+ ;L7-Cre + controls ( Figure 3C,D). Lastly, the number of Purkinje cells was determined in 3 months, 6 months and 12 months old mice to assess possible neuronal cell death. Again, no significant differences were detected between Ndst1 f/f ;L7-Cre + and Ndst1 +/+ ;L7-Cre + control mice ( Figure 3E). of these projecting neurons in vivo does not require Ndst1 function ( Figure 3A,B). Calbindin staining also showed that primary and secondary trees of Purkinje cells were present and similarly branched in Ndst1 f/f ;L7-Cre + conditional mutant mice and Ndst1 +/+ ;L7-Cre + controls ( Figure 3C,D). Lastly, the number of Purkinje cells was determined in 3 months, 6 months and 12 months old mice to assess possible neuronal cell death. Again, no significant differences were detected between Ndst1 f/f ;L7-Cre + and Ndst1 +/+ ;L7-Cre + control mice ( Figure 3E).  Because no histological abnormalities were found upon Purkinje-cell specific Ndst1 deletion, we next assessed motor function of Ndst1 f/f ;L7-Cre + mutant mice and controls as a functional read-out. Mice underwent a general health check to rule out that differences observed were due to deteriorating health of the mice, and the latency to fall off a Rotarod was used as a standard measure of motor function and coordination. Again, we found that Rotarod performance did not differ significantly between conditional Ndst1 f/f ;L7-Cre + mutant and Ndst1 +/+ ;L7-Cre + control mice ( Figure 4A). Both male and female groups showed improved performance over 10 trials (F(9580) = 10.9; p < 0.001) ( Figure 4B). ANOVA further revealed a significant effect of sex (F(1580) = 29.8; p < 0.001) but genotype differences were not significant (F(2580) = 1.26; p > 0.28), indicating that no functional deficiencies resulted from Ndst1 deletion in Purkinje cells in the adult brain. function and coordination. Again, we found that Rotarod performance did not differ significantly between conditional Ndst1 f/f ;L7-Cre + mutant and Ndst1 +/+ ;L7-Cre + control mice ( Figure 4A). Both male and female groups showed improved performance over 10 trials (F(9580) = 10.9; p < 0.001) ( Figure  4B). ANOVA further revealed a significant effect of sex (F(1580) = 29.8; p < 0.001) but genotype differences were not significant (F(2580) = 1.26; p > 0.28), indicating that no functional deficiencies resulted from Ndst1 deletion in Purkinje cells in the adult brain.  with genotype and sex as independent measures and trials as repeated factors revealed significant sex differences (F(1580) = 29.8; p < 0.001) and an overall improvement over trials (F(9580) = 10.9; p < 0.001) but no effect of genotype. f Ndst1 +/f ;L7-Cre + n = 11, f Ndst1 f/f ;L7-Cre + n = 13, f Ndst1 +/+ ;L7-Cre + n = 9, m Ndst1 +/f ;L7-Cre + n = 13, m Ndst1 f/f ;L7-Cre + n = 9, m Ndst1 +/+ ;L7-Cre + n = 12.

Analysis of Ndst1 Expression and Function in the Adult Brain
L7-Cre is also expressed in various brain areas outside of the cerebellum, such as the olfactory bulb, cerebral cortex, caudate putamen, hippocampus, thalamus, midbrain and medulla oblongata [49]. Here ( Figure 5A,C), as well as in Purkinje neurons ( Figure 5E,F), robust Ndst1 and Ndst2 expression were detected, indicating that tissues other than the cerebellum may be functionally affected in Ndst1 f/f ;L7-Cre + mice. Because Ndst3-deficient mice have been reported to show subtle changes in anxiety-like behavior [21], and because Ndst3 mutations have been linked to shizophrenia and bipolar disorder in humans [22] that both include anxiety as a symptom, we next assessed anxiety, anxiety-related behavior and exploratory behavior in Ndst1 f/f ;L7-Cre + mice and Ndst1 +/+ ;L7-Cre + control mice. expression were detected, indicating that tissues other than the cerebellum may be functionally affected in Ndst1 f/f ;L7-Cre + mice. Because Ndst3-deficient mice have been reported to show subtle changes in anxiety-like behavior [21], and because Ndst3 mutations have been linked to shizophrenia and bipolar disorder in humans [22] that both include anxiety as a symptom, we next assessed anxiety, anxiety-related behavior and exploratory behavior in Ndst1 f/f ;L7-Cre + mice and Ndst1 +/+ ;L7-Cre + control mice. We first conducted the dark-light test as an unforced measure of anxiety ( Figure 6). Mice were placed into a dark compartment with the opportunity to enter and explore an illuminated compartment. The latency to enter the light compartment and the percentage of entries into the light compartment were measured. As shown in Figure 6A, Ndst1 f/f ;L7-Cre + mutant males showed enhanced latency to enter the light compartment if compared to Ndst1 f/f ;L7-Cre + mutant females (Exact Wilcoxon rank sum test, W = 25, p = 0.02338 (*)). We also found a significant difference between Ndst1 +/+ ;L7-Cre + females and males, showing that females entered the light compartment significantly earlier than the male mice (Exact Wilcoxon rank sum test, W = 26.5, p = 0.04883 (*)). However, no genotype differences were determined, indicating normal anxiety in this experimental setting. Likewise, as shown in Figure 6B, Ndst1 +/+ ;L7-Cre + females and males differed significantly in time spent in the light compartment. Here, females spent significantly more time in the light compartment than male mice (Exact Wilcoxon rank sum test, W = 90.5, p = 0.007223 (**)). However, genotype differences were not detected. We first conducted the dark-light test as an unforced measure of anxiety ( Figure 6). Mice were placed into a dark compartment with the opportunity to enter and explore an illuminated compartment. The latency to enter the light compartment and the percentage of entries into the light compartment were measured. As shown in Figure 6A, Ndst1 f/f ;L7-Cre + mutant males showed enhanced latency to enter the light compartment if compared to Ndst1 f/f ;L7-Cre + mutant females (Exact Wilcoxon rank sum test, W = 25, p = 0.02338 (*)). We also found a significant difference between Ndst1 +/+ ;L7-Cre + females and males, showing that females entered the light compartment significantly earlier than the male mice (Exact Wilcoxon rank sum test, W = 26.5, p = 0.04883 (*)). However, no genotype differences were determined, indicating normal anxiety in this experimental setting. Likewise, as shown in Figure 6B, Ndst1 +/+ ;L7-Cre + females and males differed significantly in time spent in the light compartment. Here, females spent significantly more time in the light compartment than male mice (Exact Wilcoxon rank sum test, W = 90.5, p = 0.007223 (**)). However, genotype differences were not detected.  Figure 7B) were determined. We found that Ndst1 f/f ;L7-Cre + females entered the open arms significantly more often than Ndst1 +/f ;L7-Cre + heterozygous females (Exact Wilcoxon rank sum test, W = 38, p = 0.0442 (*)); however, no differences were found among male mice or female Ndst1 f/f ;L7-Cre + and Ndst1 +/+ ;L7-Cre + mice. We also did not observe any significant differences in the time spent in the open arms between genotypes or between genders.  Figure 7B) were determined. We found that Ndst1 f/f ;L7-Cre + females entered the open arms significantly more often than Ndst1 +/f ;L7-Cre + heterozygous females (Exact Wilcoxon rank sum test, W = 38, p = 0.0442 (*)); however, no differences were found among male mice or female Ndst1 f/f ;L7-Cre + and Ndst1 +/+ ;L7-Cre + mice. We also did not observe any significant differences in the time spent in the open arms between genotypes or between genders. The open field test represents another type of anxiety test. Here, mice are forced into a potentially threatening environment without a possibility to hide. Path length (in meters) ( Figure 8A) and the percentage of time spent in center field (with 20 cm distance from the walls, Figure 8B) were measured. The time spent in the center field was then divided by the total time spent in the open field. This test detected a highly significant difference between heterozygous males and females, because females covered a significantly longer distance than male subjects of the same genotype (Exact Wilcoxon rank sum test, W = 118, p = 0.005943 (**)). A trend between heterozygous and Ndst1 f/f ;L7-Cre + mutant females was found, because heterozygous females covered a longer distance in the open field than Ndst1 f/f ;L7-Cre + females (Exact Wilcoxon rank sum test, W = 100.5, p = 0.096 (T)). However, again, no significant genotype differences could be detected. The open field test represents another type of anxiety test. Here, mice are forced into a potentially threatening environment without a possibility to hide. Path length (in meters) ( Figure 8A) and the percentage of time spent in center field (with 20 cm distance from the walls, Figure 8B) were measured. The time spent in the center field was then divided by the total time spent in the open field. This test detected a highly significant difference between heterozygous males and females, because females covered a significantly longer distance than male subjects of the same genotype (Exact Wilcoxon rank sum test, W = 118, p = 0.005943 (**)). A trend between heterozygous and Ndst1 f/f ;L7-Cre + mutant females was found, because heterozygous females covered a longer distance in the open field than Ndst1 f/f ;L7-Cre + females (Exact Wilcoxon rank sum test, W = 100.5, p = 0.096 (T)). However, again, no significant genotype differences could be detected.

Discussion
HS regulates embryonic development by simultaneously binding numerous growth factors and morphogens, including but not restricted to Fgfs and Shh [9,13,48,51]. Shh as well as Shh receptors Patched and Smoothened are expressed in cerebellar Purkinje cells [26,27], and Shh is known to play crucial roles in prenatal [52] and postnatal [53][54][55][56] development of the cerebellum. Therefore, and because of the vast number of HS-binding, soluble factors expressed in the adult brain, HS is likely important for unimpaired physiology of differentiated brain tissues as well. Indeed, in the past years, the expression of HS-proteoglycans was shown to control behavior and memory in mice [14][15][16]18,57]. Moreover, altered Ndst3 expression has been linked to shizophrenia and bipolar disorder [22], which may be linked to altered HS-dependent neuregulin-1 distribution and signaling [58]. Together, these established HS functions in the adult nervous system prompted us to investigate the relevance of Ndst1, a key HS-modifier essential for embryonic development [9], in the adult murine brain. To this end, we deleted Ndst1 function by L7/pcp-Cre expression in Purkinje neurons, and expected morphological and functional consequences based on the known essential role of these cells in motor coordination.
Contrary to our expectations, however, we found that cerebellar morphology and function in adult Ndst1 f/f ;L7-Cre + mice were unchanged, indicating a non-essential role for cell-autonomous Ndst1 expression in Purkinje cells. This may be explained by compensation of Ndst1 activity by other expressed Ndst family members, as has been observed for embryonic stem cells deficient in Ndst1 and Ndst2 [59]. One model of HS biosynthesis suggests that several of the biosynthetic enzymes are present in a multienzyme complex termed the GAGosome [60], and the GAGosome could vary in composition dependent on the HS-expressing cell type or tissue. Lack of Ndst1 expression may thus result in the altered composition of the GAGosome and functional compensation by another Ndst isoenzyme. Indeed, this possibility is supported by the observation that female mice heterozygous or homozygous for floxed Ndst1 on a systemic Ndst2-deficient background showed strongly reduced reproductive behavior, whereas Ndst1 f/f ;Ndst2 +/+ ;L7-Cre + (only Ndst1-deficient) and Ndst1 +/+ ;Ndst2 −/− ;L7-Cre + (only Ndst2-deficient) females were fertile. Remarkably, the observed insufficiency of one floxed Ndst1 allele in the absence of Ndst2 gene function (in Ndst +/f ;Ndst2 −/− ;L7-Cre + mice) seems at odds with the previous observation that mice carrying systemic deletions of Ndst1 and Ndst2 (Ndst1 +/− ;Ndst2 −/− ) are viable and fertile [50]. Although we do not know the reason for this discrepancy, we note that in the conditional Ndst1 mutant only the exon 2 (the first coding exon) is targeted, but the promotor, the transcription start site and much of the 5 untranslated region remained intact [9]. In contrast, the systemic Ndst1 mutant replaced a 6.8 kb genomic fragment containing exon 2 and much of the upstream sequence with a neomycin cassette [61]. We thus suggest that the presence of two genomic sites for transcription factors and other regulators in heterozygous floxed mice, only one of which would yield functional protein after Cre expression, may functionally differ from the presence of only one promotor and 5 untranslated region in the systemic mutant. Along the same line, the observed fertility variation between Ndst1 +/f ;Ndst2 −/− ;L7-Cre + female mice and Ndst1 f/f ;Ndst2 +/− ;L7-Cre + female mice may be due to different Ndst gene dosage (insufficient Ndst1 expression from the one functional allele if compared to Ndst2). Alternatively, these differences can be explained by the inability of Ndst1 to fully compensate for the loss of Ndst2 in some L7-Cre-expressing cell types, consistent with their specific HS biosynthetic profiles [3]. Regardless of the relative Ndst activities in the process, our finding that impaired HS biosynthesis as a consequence of L7-Cre mediated Ndst1 deletion in an Ndst2-deficient background in Purkinje neurons or other cell types that express Cre from the L7 promotor (the olfactory bulb, cerebral cortex, caudate putamen, hippocampus, thalamus, midbrain and medulla oblongata) affects breeding is consistent with impaired reproduction described in the hyperspiny mouse mutant (carrying a mutation that affects Purkinje cell morphology and function) [62] and in female HS 3O-sulfotransferase 1 mutant mice [63]. Therefore, together with these reports, our findings indicate the possibility that fully sulfated HS in cell types that express Cre under L7 promotor control plays a previously unsuspected role in the regulation of female reproductive behavior. Further support for essential HS functions in normal brain physiology comes from the observations that Ndst1 is a candidate gene for autosomal recessive intellectual disability in four unrelated families [64,65], that the Drosophila Ndst orthologue Sulfateless plays a role in normal social interactions and repetitive behavior in the fly [66], that Glypican 4 associates with hyperactivity and social interaction deficits in mice [67], and that a HS 6-O sulfotransferase gene variant associates with intellectual disability in humans [68].
Notably, we failed to observe any additional obvious problems in all seven obtained Ndst1 f/f ;Ndst2 −/− ;L7-Cre + compound mutant mice, such as in motor coordination and cerebellar architecture. In contrast, mice and tissues defective in both Ndst1 and Ndst2 function are usually strongly affected [50,59,69]. In addition to residual compensatory Ndst3 and Ndst4 activities that may explain this finding, we suggest that the rescue of impaired Purkinje cell HS biosynthesis and function may have occurred in trans by HS produced from other cell types, as has recently been described for VEGFR signaling [70]. Here, chimeric cultures of embryonic stem cells defective in either HS production or VEGFR-2 synthesis supported VEGF signaling in trans. Because cerebellar Purkinje cells constitute a single cell layer with elaborate, branched dendrites and long axons that make extensive contact with other cell types and the extracellular matrix, HS produced by other cell types in the cerebellum may have compensated for the loss of Purkinje cell HS in trans, at least to some part.
Taken together, our results show that, although Ndst1 plays crucial roles during development of the embryonic central nervous system, its cell-autonomous expression is not essential for L7-Cre expressing neurons in the adult mouse brain. Reduced reproductive capacity observed in Ndst1/Ndst2 compound mutant mice, however, confirms the essential yet specific roles of fully sulfated HS in adult physiology.