Skip to main content
NCBI home page
Science Advances logoLink to Science Advances
. 2025 Jul 25;11(30):eadq7105. doi: 10.1126/sciadv.adq7105

Role of HAUS7 as a DOCK3 binding partner in facilitating axon regeneration

Naoki Kiyota 1,2,†,‡, Youichi Shinozaki 1,†, Xiaoli Guo 1, Atsuko Kimura 1, Kazuto Kawamura 1,§, Euido Nishijima 1, Sari Honda 1, Chikako Harada 1, Toru Nakazawa 2, Kazuhiko Namekata 1,*, Takayuki Harada 1,*
PMCID: PMC12292649  PMID: 40712007

Abstract

The molecular mechanisms involved in reconstructing the eye-to-brain connection and functional recovery following optic nerve damage remain unclear. This study revealed that HAUS augmin-like complex subunit 7 (HAUS7) is a molecule that binds to dedicator of cytokinesis 3 (DOCK3), a regulator of neurotrophic factor signaling and axon regeneration. We observed a distribution pattern of HAUS7 expression, suggesting that neuronal HAUS7 is transported from the cell body to the growth cone under the control of DOCK3. In addition, phosphorylation of DOCK3 at Y562 by tropomyosin receptor kinase B signaling leads to the dissociation of HAUS7, which is considered an important step for microtubule assembly. Deletion of Haus7 in mice significantly reduced microtubule formation and axon regeneration following optic nerve crush (ONC). Transcriptome analysis suggested that HAUS7 levels decrease in glaucoma and after the ONC, while retinal ganglion cells actively regenerating their axons express high levels of HAUS7. In summary, HAUS7 is a binding partner of DOCK3 necessary for axon elongation.

The molecular interaction between DOCK3 and HAUS7 is crucial for optic nerve axon regeneration and microtubule branching.

INTRODUCTION

Vision serves as the primary source of information among human senses (1). Photoreceptors receive visual information, which is then relayed by interneurons to retinal ganglion cells (RGCs), the output neurons of the eye (2). RGC axons, which constitute the optic nerve, sprout from RGCs and transmit visual information to the brain, making them crucial for maintaining sight. Unfortunately, the optic nerve is vulnerable and susceptible to damage from injuries and diseases, including glaucoma, the leading cause of irreversible blindness (3). In the mature central nervous system and the optic nerve, neuronal axons exhibit limited regenerative potential, which has prompted researchers to explore methods for promoting axon regrowth and reestablishing proper synaptic connections (4–6). Previous studies have indicated that both extrinsic and intrinsic factors play a dynamic role in influencing the regenerative capacity of RGC axons (2). For instance, deletion of regeneration-inhibitory factors such as phosphatase and tensin homolog (PTEN) and suppressor of cytokine signaling 3 (SOCS3) can induce robust regeneration of the optic nerve (7). The enhancement of regeneration-promoting signaling pathways, such as mammalian target of rapamycin (mTOR), as well as brain-derived neurotrophic factor (BDNF) and its receptor, tropomyosin receptor kinase B (TrkB), has also shown promise for successful axon regeneration (8, 9). These findings underscore the importance of precise regulation of certain intrinsic signaling within RGCs as a crucial therapeutic target for axon regeneration and the restoration of visual function. Recent advancements in single-cell genomics have revealed a high degree of heterogeneity among RGCs (10). These intrinsic characteristics are pivotal in determining the regenerative capacity of axons (11, 12). Consequently, gaining a deeper understanding of the mechanisms that promote regeneration within RGCs holds great potential for future strategies aimed at visual recovery.

The BDNF/TrkB signaling pathway stands out as one of the most promising candidates for promoting axon regeneration. In a recent study, it was shown that constitutively active TrkB signaling induces robust axon regeneration following optic nerve crush (ONC) (9). DOCK3, a member of the Dock family of guanine nucleotide exchange factors, serves as an essential regulator and a key downstream target of TrkB-mediated axon regeneration and represents a key downstream target (13). A recent study has shown that a DOCK3 activator successfully enhanced axon regeneration following ONC (14). Members of the family share two evolutionarily conserved domains: Dock homology region 1 (DHR1) and Dock homology region 2 (DHR2) (15). Notably, the DHR1 domain of DOCK3 (DOCK3-DHR1) interacts with lipids such as phosphatidylinositol (3,4,5)-triphosphate, resulting in the recruitment of DOCK to the cell membrane (15, 16), where various protein-protein interactions are actively regulated in response to extracellular stimuli such as BDNF. Previous study has shown that DOCK3-DHR1 binds to WAVE1, a protein that promotes actin polymerization and enhances axon regeneration (13). Therefore, DHR1 is likely to be the key to the identification of additional DOCK3-interacting partners.

Because actin polymerization and microtubule elongation at the growth cone are essential for driving the extension of regenerating axons, it was postulated that there might be other factors that are involved in DOCK3-mediated axon regeneration. The HAUS members play a critical role in regulating tubulin assembly and microtubule elongation (17), particularly in the formation of the γ-tubulin ring complex, which acts as a template for initiating microtubule polymerization and caps the minus end of microtubules (18). The HAUS-mediated microtubule assembly operates independently of the centrosome, the classical microtubule organizing center (19). Neuronal axon extension, reliant on microtubule assembly, also proceeds independently of centrosomal microtubule nucleation (20). HAUS governs noncentrosomal microtubule nucleation, playing a crucial role in polarized axon elongation (21). Despite these previous findings highlighting the significance of HAUS in microtubule assembly and axon elongation, the upstream regulatory system and its involvement in pathological conditions have remained unclear. This study provides a comprehensive understanding of the regulatory mechanisms of HAUS augmin-like complex subunit 7 (HAUS7) in combination with DOCK3 and its role in axon regeneration following ONC.

RESULTS

HAUS7 is a binding partner of DOCK3

To uncover binding partners of DOCK3, we performed a yeast two-hybrid assay as part of our screening process (fig. S1). Given that the DOCK3-DHR1 domain has the potential to interact with multiple molecules, including WAVE1 (13) and phosphatidylinositol (3,4,5)-triphosphate (15), we used the DHR1 domain as bait to identify previously unknown binding partners of DOCK3. We could detect two positive colonies (fig. S1, A to C), and these obtained clones contained the DNA fragment of Haus7. To confirm the interaction between HAUS7 and DOCK3, we conducted pull-down assays. Cos-7 cells were transfected with myc-tagged full-length DOCK3 (myc-DOCK3) and glutathione S-transferase–tagged HAUS7 (GST-HAUS7) plasmids. Subsequently, we isolated the protein complex using glutathione sepharose beads [Fig. 1A (a)]. Immunoblotting confirmed the presence of both myc-DOCK3 and GST-HAUS7 in the isolated complex, indicating their interaction. A similar pull-down assay involving Cos-7 cell lysates transfected with myc-HAUS7 and GST-DOCK3 also demonstrated the binding between HAUS7 and DOCK3 [Fig. 1A(b)]. As DOCK3 is predominantly expressed in the nervous system (22), we further explored the spatial expression patterns of DOCK3 and HAUS7 in the retina by immunohistochemistry. Both DOCK3 and HAUS7 were expressed in several retinal layers, including the nerve fiber layer (NFL), ganglion cell layer (GCL), inner and outer plexiform layers (IPL and OPL, respectively), and inner nuclear layer (INL) (Fig. 1B). Previous study has shown that DOCK3 is expressed in RGCs (13), prompting us to investigate the expression of HAUS7 in RGCs. The HAUS7 expression level was relatively high in the inner retinal layers (NFL to INL). Punctate HAUS7 signals were observed in the GCL and INL, suggesting the expression of HAUS7 in the neuronal and/or glial cell bodies as previously reported by HAUS2 immunostaining (23). The punctate HAUS7 signals at the GCL colocalized with Rbpms, a marker for RGCs. In addition, the HAUS7 expression pattern varied among RGCs; some RGCs showed high expression levels (arrowheads in Fig. 1B), while others displayed relatively low levels.

Fig. 1. HAUS7 binds to DOCK3-DHR1 independently of WAVE1.

Fig. 1.

Open in a new tab

(A) Interaction of HAUS7 with full-length DOCK3. (a) Cos-7 cells were transfected with myc-DOCK3 and GST-HAUS7. Cell lysates were precipitated with glutathione sepharose and immunoblotted with anti-myc or anti-GST antibodies. (b) Transfection of myc-HAUS7 and GST-DOCK3 into Cos-7 cells, followed by precipitation of lysates with an anti-GST antibody. Both myc and GST were detected by immunoblotting (IB). MW, molecular weight. (B) Immunohistochemical analysis of DOCK3 and HAUS7 in the retina of 8-week-old male WT mice. ONL, outer nuclear layer; IS/OS, inner and outer segments of photoreceptors. Scale bars, 10 μm. DAPI, 4′,6-diamidino-2-phenylindole. (C) Interaction between HAUS7 and DHR1 of DOCK3. (a) Cos-7 lysates transfected with myc-DHR1 and HAUS7-His were precipitated, followed by immunoblotting with an anti-myc antibody. (b) Cell lysates transfected with HA-HAUS7 and DHR1-His were precipitated and immunoblotted with an anti-HA antibody. (D) Independence of WAVE1 in the DHR1-HAUS7 interaction. Cos-7 lysates transfected with myc-DHR1 and HAUS7-His were precipitated with an anti-His antibody and immunoblotted with an anti-myc antibody. (E) No direct interaction between HAUS7 and WAVE1. Transfection of Cos-7 cells with WAVE1 and either HA-DHR1-His or HAUS7-His showed that the His-precipitated sample in the latter case did not contain WAVE1. (F) Trimolecular complex formation of WAVE1, DOCK3, and HAUS7. Cos-7 lysates transfected with FLAG-WAVE1, myc-DHR1, and HA-GST-HAUS7 were precipitated with GST, followed by immunoblotting with anti-FLAG, anti-myc, and anti-HA antibodies. (G) Schematic of truncated DOCK3-DHR1 constructs. (H) HAUS7 binding to the C terminus of DHR1. Cos-7 cell lysates transfected with HA-HAUS7 or FLAG-WAVE1 were incubated with truncated forms of His-tagged DHR1 proteins. The pull-down assay using an anti-His antibody showed the interaction between (a) DHR1-C and HAUS7 and (b) DHC1-N and WAVE1. (I) Schematic diagram illustrating the formation of the DHR1-WAVE1-HAUS7 complex.

Given the importance of the DHR1 domain in molecular interactions of DOCK3 (13, 24), we investigated whether the binding between HAUS7 and DOCK3 was mediated by DOCK3-DHR1. Pull-down assays using an anti-histidine (His) antibody with Cos-7 cell lysates transfected with myc-DHR1 and HAUS7-His confirmed that both were present in the complex [Fig. 1C(a)]. A similar result was obtained when hemagglutinin-tagged HAUS7 (HA-HAUS7) and DHR1-His were used [Fig. 1C(b)]. These findings provide clear evidence that DOCK3 binds to HAUS7 through the DHR1 domain. To determine whether HAUS7 interacts with other members of DOCK family, we conducted a pull-down assay using GST-HAUS7 and the DHR1 domain of various DOCK family members (i.e., DOCK1 to DOCK11) transfected into Cos-7 cells. Our results revealed that HAUS7 exhibited weak binding to DOCK4 and DOCK7, but a strong interaction was observed with DOCK3 (fig. S2).

HAUS7 binds to DOCK3-DHR1 independently of WAVE1

Given that WAVE1 is known to bind DOCK3-DHR1 (13), we investigated whether WAVE1 competes with HAUS7 for this interaction. A pull-down assay demonstrated that the binding between myc-DHR1 and HAUS7-His was unaffected by the presence of FLAG-WAVE1 (Fig. 1D). We then explored whether HAUS7 and WAVE1 directly interact. In a pull-down assay with FLAG-WAVE1 and HA-DHR1-His, their interaction was confirmed, but no interaction was observed between FLAG-WAVE1 and HAUS7-His (Fig. 1E). To further determine whether DOCK3, WAVE1, and HAUS7 could form a trimolecular complex, we transfected Cos-7 cells with myc-DHR1, FLAG-WAVE1, and HA-GST-HAUS7. Cell lysates were precipitated and analyzed by immunoblotting with anti-myc, anti-FLAG, and anti-HA antibodies. The detection of all three proteins in the precipitated samples confirmed the formation of a trimeric complex (Fig. 1F). These findings strongly suggest that HAUS7 does not directly bind with WAVE1, but they interact with each other in the presence of DOCK3-DHR1.

Our findings indicate that both HAUS7 and WAVE1 bind to DOCK3-DHR1 without interfering with each other, suggesting that they attach to different regions of DHR1. To pinpoint the exact binding sites of WAVE1 and HAUS7 within DOCK3-DHR1, we generated two plasmids, one expressing the first half (1 to 119 amino acids) of DOCK3-DHR1 (DHR1-N) and the other expressing the second half (120 to 238 amino acids) of DOCK3-DHR1 (DHR1-C) (Fig. 1G). In the pull-down assay, HA-tagged HAUS7 coprecipitated with DOCK3-DHR1-C but not with DOCK3-DHR1-N [Fig. 1H(a)]. Conversely, FLAG-tagged WAVE1 coprecipitated with DOCK3-DHR1-N but not with DOCK3-DHR1-C [Fig. 1H(b)]. These results suggest that HAUS7 binds to the C-terminal side of DHR1, whereas WAVE1 binds to the N-terminal side of DHR1 (Fig. 1I).

HAUS7 is transported from the cell body to the growth cone by DOCK3

To elucidate the significance of the interaction between HAUS7 and DOCK3, we examined the subcellular localization of HAUS7 in primary cultured cortical neurons. Given that our previous data indicated that DOCK3 is transported to the growth cone, where it dynamically regulates axon regeneration (13, 24), we sought to determine the subcellular distribution of HAUS7. We generated a plasmid encoding myc-tagged HAUS7 fused with a DsRed-monomer fluorescent protein. Pull-down assays confirmed that DsRed-labeled HAUS7 effectively binds to DOCK3-DHR1 (Fig. 2A). To assess the impact of endogenous DOCK3 on the intracellular distribution of HAUS7, we conducted a knockdown of DOCK3 using the shDock3 plasmid in primary cultured neurons. Electroporation of the shDock3 plasmid markedly reduced endogenous DOCK3 expression, while HAUS7 levels remained unchanged (Fig. 2B). When neurons were transfected with Haus7-DsRed along with control short hairpin RNA (shRNA; shGfp-ZsGreen), red fluorescence signals were detected in the axon (Fig. 2C, upper left, arrowheads) and cell body (Fig. 2C, upper left, asterisk). In contrast, cotransfection with shDock3-ZsGreen diminished DsRed signals within the axon (Fig. 2C, upper right), although the fluorescence intensity in the cell body remained unaffected (Fig. 2D). By tracing DsRed fluorescence intensity along the axon (expressed as a percentage of the intensity at the cell body), we observed a proportional decrease in intensity with increasing distance from the cell body (Fig. 2E). Quantitative analysis of axons treated with shGfp or shDock3, which had comparable lengths (132.6 ± 8.8 μm versus 141.4 ± 9.2 μm), revealed a reduction in DsRed signal intensities at distances of less than 100 μm, between 100 and 200 μm, and between 200 and 300 μm from the cell body under the shDock3 condition (Fig. 2F). The rate of signal reduction was significantly greater in shDock3-treated neurons, highlighting the role of DOCK3 in transporting HAUS7 along the axon. Beyond the axon, DOCK3 knockdown also significantly decreased the presence of both DOCK3 and HAUS7 at the growth cone (Fig. 2G). Quantitative analysis exhibited that the DOCK3 signal intensity at the growth cone dropped by nearly half with shDock3 compared with shGfp-treated neurons [arbitrary units (a.u.), 350.4 ± 34.7 versus 206.4 ± 24.9 for shGfp and shDock3, respectively] [Fig. 2H(a)]. Similarly, the HAUS7 signal intensity at the growth cone was markedly reduced by shDock3 (a.u., 336.8 ± 43.9 versus 37.6 ± 4.6 for shGfp and shDock3, respectively) [Fig. 2H(b)]. These findings suggest that DOCK3 is essential for the intracellular transport of HAUS7 along the axon to the growth cone.

Fig. 2. Regulation of HAUS7 transport to the growth cone by DOCK3.

Fig. 2.

Open in a new tab

(A) Interaction between DsRed-HAUS7 and DHR1. Cos-7 cells were transfected with myc-tagged HAUS7-DsRed and His-tagged DHR1, and a pull-down assay demonstrated the binding between DHR1 and HAUS7-DsRed. (B) Validation of the shRNA-mediated knockdown of DOCK3 in primary cultured cortical neurons. (C) DOCK3 mediates the transport of HAUS7 to the growth cone. Neurons transfected with Haus7-DsRed showed prominent red fluorescence signals along the axon (arrowheads) when treated with shGfp-ZsGreen (left panels). The neurons electroporated with shDock3-ZsGreen exhibited a marked reduction in DsRed signal along the axon, although the signals remained present in the cell body (asterisk). An adjacent neuron not transfected with shDock3-ZsGreen displayed clear DsRed signals along with the axon (arrows). Images were captured 3 days postelectroporation. Scale bar, 50 μm. (D) Quantitative analysis of somatic DsRed levels shown in (C). The DsRed fluorescence intensity at the cell body was not affected by the transfection of shDock3 (n = 51 to 81 cells per group, Mann Whitney U test). (E) Plot graph illustrating DsRed intensity at various points along the axons. (F) Quantitative analysis of data from (C). The DsRed intensity was significantly reduced by shDock3 (n = 64 to 81 axons per group; ***P < 0.001 and *P < 0.05, two-way ANOVA followed by the Bonferroni test). (G and H) DOCK3 and HAUS7 levels at the growth cone. Representative images of DOCK3 and HAUS7 treated with shGfp- or shDock3-ZsGreen are shown in (G). Scale bars, 2 μm. Quantitative data of signal intensity for (a) DOCK3 and (b) HAUS7 are presented in (H). Compared with the shGfp-treated group, the shDock3-treated group exhibited significant reduction in DOCK3 and HAUS7 at the growth cone (n = 30 growth cones per group; ***P < 0.001, Mann Whitney U test). Data are presented as the means ± SEM.

HAUS7 is a downstream molecular target of BDNF-DOCK3 signaling

DOCK3 is known to be translocated to the plasma membrane of the growth cone after BDNF treatment (13). At the membrane, DOCK3 interacts with TrkB, WAVE1, and Rac1. Phosphorylation of DOCK3 dissociates WAVE1/Rac1, thereby triggering actin rearrangement and axon outgrowth. We sought to determine whether and how HAUS7 is involved in this process. First, we transfected Cos-7 cells with HAUS7 together with DOCK3-DHR1, farnesylated DOCK3-DHR1 (F-DHR1), or F-DHR1 with the farnesylated intracellular domain of TrkB (F-iTrkB) to trigger TrkB signaling without BDNF (9). The pull-down assay revealed that the membrane translocation of DHR1 by farnesylation (i.e., F-DHR1) induced the dissociation of HAUS7 from DHR1 (Fig. 3, A and B). F-iTrkB enhanced the dissociation of HAUS7 from DHR1. Given that DOCK3 phosphorylation is induced by the activation of TrkB, we investigated which type of tyrosine residue is required for its dissociation from HAUS7. We substituted four types of tyrosine residues (Y546, Y548, Y562, and Y572) of DOCK3 with phenylalanine. The reduced binding between HAUS7 and DOCK3-DHR1 by F-iTrkB was partially reversed in DOCK3Y548F, DOCK3Y562F, and DOCK3Y572F (Fig. 3, C and D). Among them, DOCK3Y562F showed the most significant restoration of DOCK3-HAUS7 binding. Structural analysis using the AlphaFold Protein Structure Database revealed that the Y562 residue on the DHR1 domain of DOCK3 is exposed to the surface (fig. S3). These findings indicate that phosphorylation of DOCK3 at Y562 plays a pivotal role in regulating the dissociation of DOCK3 from HAUS7. To specifically detect Y562 phosphorylation (Y562p) in DOCK3, we generated a Y562p-specific antibody. While translocation of DHR1 to the membrane via F-DHR1 did not trigger Y562p, a robust induction of Y562p occurred in the presence of F-iTrkB (Fig. 3E). This effect was abolished by the Y562F substitution in DOCK3-DHR1. These data suggest that under normal conditions, HAUS7 binds to nonphosphorylated DOCK3, but upon BDNF-TrkB signaling activation, phosphorylation of Y562 on DOCK3 triggers the dissociation of HAUS7.

Fig. 3. Regulation of DOCK3Y562 phosphorylation and its interaction with HAUS7.

Fig. 3.

Open in a new tab

(A) Cos-7 cell lysates, transfected as shown at the top of the images, were subjected to a His-tag pull-down assay with antibodies against the myc-tag and HAUS7. (B) Quantitative results for (A) (n = 4; *P < 0.05 and ***P < 0.0001, one-way ANOVA followed by the Bonferroni test). (C and D) Dissociation between HAUS7 and DHR1 is regulated by tyrosine phosphorylation of DOCK3. Substituting Y548 and Y572 with phenylalanine partially reversed the dissociation, with the Y562 substitution being the most effective (n = 4; *P < 0.05 and ***P < 0.0001, one-way ANOVA followed by the Bonferroni test). NS, not significant. (E) The Y562p of DHR1 is induced by F-iTrkB but not myc-DHR1 or myc-F-DHR1. F-iTrkB induced the Y562p of DHR1, which was attenuated by the substitution of Y562 for phenylalanine. (F) Interaction and BDNF-induced dissociation of endogenous DOCK3 and HAUS7. Upon BDNF stimulation, HAUS7 dissociated from DOCK3, a process associated with phosphorylation of DOCK3 at Y562. No actin was detected in the precipitated samples. (G) Regulation of HAUS7 membrane translocation by the Y562p of DHR1. Expression of HAUS7-His, myc-F-DHR1, and myc-F-iTrkB led to a significant increase in the Y562p signal at the plasma membrane in Neuro-2a cells. Scale bars, 5 μm. (H) Phosphorylation at Y562 of DOCK3 in the growth cone of primary cultured neurons. Under basal condition, the Y562p signals were hardly detected but were significantly increased after BDNF treatment (50 ng/ml). Scale bar, 2.5 μm. (I) Interaction between HAUS7 and DOCK3 at the growth cone of primary cultured cortical neurons. DOCK3 partially colocalized with HAUS7 (arrowheads in the inset), but their colocalization decreased after BDNF treatment. Scale bars, 1.0 and 2.5 μm for insets and main panels. (J) Quantification of DOCK3+/HAUS7+ colocalized puncta per growth cone shown in (I) (n = 20; ***P < 0.0001, Mann Whitney U test).

To further validate this mechanism, we examined the interaction between endogenous DOCK3 and HAUS7 and their response to BDNF stimulation. Wild-type (WT) TrkB and GST-Elmo were expressed in primary cultured neurons by electroporation, and cell lysates were subjected to GST pull-down followed by immunoblotting with antibodies against DOCK3 or HAUS7 (Fig. 3F). Endogenous DOCK3 and HAUS7 were coimmunoprecipitated, confirming a direct interaction between DOCK3 and HAUS7 under unstimulated conditions. Following BDNF stimulation, HAUS7 was no longer coimmunoprecipitated with DOCK3, indicating the dissociation of the two proteins associated with Y562p. The absence of actin in the precipitated samples ruled out the possibility of nonspecific interactions.

We also demonstrate representative immunocytochemical images using Neuro2a neuroblastoma cells, which corroborate the Western blotting quantification data (Fig. 3G). In the control group, both HAUS7-His and myc-DHR1 were localized in the cytoplasm, with no detectable Y562p signals. When DHR1 was translocated to the plasma membrane using myc-F-DHR1, HAUS7 followed suit, relocating to the plasma membrane, although Y562p signals remained largely undetectable. In the presence of F-iTrkB, Y562p signals emerged and were prominently observed at the plasma membrane. Under these conditions, a portion of the membrane-translocated HAUS7 shifted back to the cytoplasm. This effect was reversed by the DHR1Y562F mutation, which eliminated both Y562p signals and cytoplasmic HAUS7 localization.

In addition, we assessed endogenous Y562p within the DHR1 domain of DOCK3 at the growth cones of primary neurons in response to BDNF. In the absence of BDNF, few Y562p signals were detected (Fig. 3H, upper panel). However, following BDNF stimulation (50 ng/ml for 20 min), a significantly higher number of Y562p signals were observed (Fig. 3H, lower panel). We further explored the endogenous DOCK3-HAUS7 interaction and dissociation at the growth cone of primary neurons. Under unstimulated conditions, a subset of DOCK3 colocalized with HAUS7, indicating their close association (Fig. 3I, left panel and arrowheads in the inset). After BDNF stimulation, while DOCK3 and HAUS7 remained in proximity, their colocalization diminished, suggesting that HAUS7 dissociates from DOCK3 upon stimulation (Fig. 3I, right panel). Quantitative analysis showed that BDNF stimulation significantly reduced the number of DOCK3+/HAUS7+ colocalized puncta per growth cone (Fig. 3J).

Knockdown of Haus7 and tubulin impairs axon regeneration

If HAUS7 acts downstream of the BDNF-DOCK3 pathway at the growth cone, it should play a critical role in axon regeneration. To test this hypothesis, we generated knockdown plasmids for Haus7 (shHaus7) and successfully knocked down Haus7 in vitro (Fig. 4A). This knockdown was further confirmed in the primary cultured neurons and RGCs in the retina of WT mouse (fig. S4, A and B). We then created adeno-associated virus type 2 (AAV2)-shHaus7 and investigated its effect on RGC axon regeneration in vivo. We hypothesized that AAV2-shHaus7 will inhibit axon regeneration, and thus, we intravitreally injected AAV2-K-RasV12, a constitutively active form of K-Ras that activates Akt and ERK1/2 (extracellular signal–regulated kinase 1/2) signaling to promote RGC axon regeneration (25), and examined the inhibitory effects of AAV2-shHaus7 in an ONC model. We estimated the RGC number because if the RGC number was markedly changed, it would be unclear whether the changes in the number of regenerating axons are due to altered axon elongation or RGC degeneration/protection. Knocking down Haus7 did not demonstrate any protective or detrimental effects on RGC soma [121.8 ± 1.5 (AAV2-shHaus7) and 113.0 ± 5.9 (control) Rbpms-positive cells per retina; P = 0.211] (Fig. 4B). We also found that knocking down Haus7 significantly suppressed AAV2-K-RasV12–induced axon regeneration (Fig. 4, C and D). These findings highlight the significance of HAUS7 in axon regeneration after the ONC.

Fig. 4. Essential role of HAUS7 and tubulin in optic nerve regeneration.

Fig. 4.

Open in a new tab

(A) Validation of shRNA-mediated knockdown of HAUS7. Cos-7 cell lysates, transfected as depicted at the top of the images, were analyzed by immunoblotting using antibodies against myc-tag and actin. (B) Representative image of Rbpms-positive cells in the retina 1 week after the ONC with prior treatment with AAV2-shGfp (upper panel) or AAV2-shHaus7 (lower panel) administered 1 week before the ONC. Scale bar, 500 μm. (C) Representative image of an optic nerve section illustrating regenerated axons in mice 1 week after the ONC, following treatment with AAV2-shGfp (upper panel) or AAV2-shHaus7 (lower panel) in addition to AAV2-K-RasV12 administered 1 week before the ONC. The asterisk indicates the crush site. Scale bar, 200 μm. (D) Quantitative assessment of regenerating axons in (C). n = 4 mice per group. **P < 0.01. A two-way ANOVA followed by the Bonferroni test was used for multiple group comparisons. (E) Verification of shRNA-mediated knockdown efficiency for tubulin proteins. Cos-7 cells expressing myc-tagged tubulin β3 (Tubb3) or tubulin γ1 or γ2 (Tubg1 or Tubg2) were transfected with shRNA specific to each subunit. (F) Evaluation of AAV2-shRNAs for tubulins on RGCs showed no protective or detrimental effects. The mice were treated with AAV2-K-RasV12 followed by ONC. Scale bar, 500 μm. (G) Representative optic nerve sections illustrating regenerated axons in mice 1 week after the ONC following treatment with AAV2-shGfp (upper panel), AAV2-shTubb3 (middle panel), or AAV2-shTubg1/2 (lower panel) in conjunction with AAV2-K-RasV12 administered 1 week before the ONC. Scale bar, 200 μm. (H) Quantitative assessment of regenerating axons in (G). n = 4 mice per group. **P < 0.01 and *P < 0.05. A two-way ANOVA followed by the Bonferroni test was used for multiple comparisons. Data are expressed as the means ± SEM.

HAUS7 is well known as an indispensable regulator of tubulin and microtubules (17), and the geometry, density, and directionality of microtubules are crucial for axon regeneration. Microtubules consist of dimers of tubulins, including α- and β-tubulin, while γ-tubulin forms the ring complex necessary as a template for de novo microtubule formation. To assess their roles, we initially conducted knockdown experiments targeting βIII-tubulin, which is highly expressed in neurons and plays an essential role in axon regeneration (26, 27). The shRNA plasmid for βIII-tubulin (shTubb3) effectively reduced Tubb3 expression in myc-Tubb3–transfected Cos-7 cells (Fig. 4E). These shRNAs also demonstrated efficacy in primary cultured neurons and in the RGCs of the WT retina (fig. S4, C and S4D). Similarly, shRNA for γ-tubulins (shTubg1/2) efficiently decreased the expression of γ1-tubulin (Tubg1) and γ2-tubulin (Tubg2) (Fig. 4E). We then investigated the contributions of Tubb3 and Tubg1/2 to axon regeneration. Neither AAV2-shTubb3 nor AAV2-shTubg1/2 exhibited any protective or detrimental effects on RGC numbers after the ONC in the presence of AAV2-K-RasV12 [105.8 ± 11.4 (AAV2-shTubb3), 107.8 ± 3.0 (AAV2-shTubg1/2), and 117.0 ± 9.6 (control) Rbpms-positive cells per retina; P = 0.641] (Fig. 4F). However, both AAV2-shTubb3 and AAV2-shTubg1/2 significantly reduced axon regeneration induced by AAV2-K-RasV12 (Fig. 4, G and H), phenocopying the results of AAV2-shHaus7. These findings support the idea that HAUS7-mediated regulation of tubulin might be involved in axon regeneration.

HAUS7 is essential for microtubule assembly in regenerating axons

To corroborate the role of HAUS7 in microtubule regulation, we generated Haus7 knockout (KO) mice (fig. S5, A and B). Immunoblot and immunohistochemical analyses confirmed the absence of HAUS7 protein in the retinas of 10-week-old KO mice (fig. S5, C and D). Notably, no apparent abnormalities were observed in the KO mice (fig. S5E), and their visual functions, as assessed through the optokinetic response and the visual evoked potential, remained normal (fig. S5, F and G). RGC numbers in the KO mice were comparable to those in WT mice (fig. S5H). Under baseline conditions, the distribution of retinal neuron markers, such as calretinin, calbindin, recoverin, and Rbpms, exhibited no apparent anomalies in KO mice (fig. S5I). In the KO retina, the expression levels of other Haus complex members (i.e., Haus1 to Haus6 and Haus8) remained unchanged (fig. S5J). We next examined whether HAUS7 is necessary for microtubule assembly and axon regeneration after the ONC. We estimated the RGC number because if the number was markedly changed, it would affect the number of regenerating axons. The RGC numbers in KO mice after the ONC did not differ from those in WT mice (Fig. 5, A and B). However, KO mice exhibited a significant reduction in RGC axon regeneration evoked by AAV2-K-RasV12 (Fig. 5, C and D). Further investigation into the impact of HAUS7 loss on microtubule assembly involved an ultrastructural analysis of regenerating RGC axons using electron microscopy (EM). In the EM images of WT axons, microtubules appeared as dark, circular structures with an approximate diameter of ~25 nm, consistent with a previous report (18). Despite no observable deficiencies in neuronal cell numbers, the EM images revealed that KO axons displayed a slight but significant (~20%) reduction in the number of microtubules, even under baseline conditions, without any alterations in axon size (Fig. 5, E and F). In contrast to healthy, myelinated axons, the regenerating axons were identifiable as unmyelinated axons without any abnormal intracellular organelles (Fig. 5G). While the axon sizes were similar, the microtubule density in regenerating KO axons was notably lower than that in WT axons (Fig. 5H).

Fig. 5. Suppressed optic nerve regeneration in Haus7 KO mice.

Fig. 5.

Open in a new tab

(A) Representative images of Rbpms-positive cells in the retinas of WT (upper panel) and KO mice (lower panel) at 2 weeks after the ONC. Scale bar, 500 μm. (B) Quantitative analysis of Rbpms-positive cells in (A). n = 4 mice per group. Data are presented as the means ± SEM. (C) Representative images of optic nerve sections 2 weeks after the ONC, illustrating regenerated axons in WT (upper) and KO mice (lower) treated with AAV2-K-RasV12 administered 2 weeks before the ONC. Asterisks indicate the crush site. Scale bar, 200 μm. (D) Quantitative analysis of regenerating axons in (C). n = 4 mice per group. ***P < 0.001, **P < 0.01, and *P < 0.05. A two-way ANOVA followed by the Bonferroni test was used for multiple group comparisons. Data are expressed as the means ± SEM. (E) Representative EM images of axons in WT (left) and KO mice (right). Intact microtubules are indicated by red arrowheads in the inset. Scale bars, 100 (inset) and 200 nm. (F) Quantitative analysis of (a) axon area and (b) calculated microtubule density at various distances from the eyeball (n = 90 axons from three mice per group; ***P < 0.001 and **P < 0.01, two-way ANOVA followed by the Bonferroni test). (G) Representative electron micrographs of regenerating axons in WT (left) and KO mice (right) 2 weeks after the ONC, treated with AAV2-K-RasV12 2 weeks before the ONC. Scale bar, 200 nm. (H) Quantitative analysis of (a) axon area and (b) calculated microtubule density (n = 58 to 209 axons from three mice per group; ***P < 0.001 and **P < 0.01, two-way ANOVA followed by the Bonferroni test). Data are expressed as the means ± SEM.

To support our data showing that HAUS7 is crucial for microtubule assembly, we performed time-lapse imaging of neuronal microtubule dynamics using fluorescent dye. In the absence of stimulation, polymerized tubulin near the growth cone was observed as thick fibers with relatively thin fiber structures extending from the tips. The thinner fibers at the tips exhibited continuous extension and retraction [movie S1 and fig. S6A(a)]. As this fluorescent dye did not visualize the structure of the growth cone, fluorescence images were overlaid with differential interference contrast (DIC) images. Even without stimulation, the growth cone observed in the DIC images showed highly dynamic structural changes [movie S2 and fig. S6B(a)]. The relatively thin polymerized tubulin fibers were observed within the interior of the growth cone. Upon BDNF stimulation (50 ng/ml), the polymerized tubulin near the extending tips exhibited more dynamic structural changes, and in some frames, growth cone-like structures were observed [movies S3 and S4 and fig. S6, A(b) and B(b)]. These results suggest that dynamic tubulin polymerization and depolymerization occur at the growth cone in response to BDNF stimulation. When neuronal HAUS7 was knocked down by shHaus7-ZsGreen, the dynamic behavior of polymerized tubulin at the extending tip disappeared, even though the thick fibers distant from the growth cone appeared relatively normal [movie S5 and fig. S6A(c)]. Because of numerous debris from shRNA electroporation, DIC imaging did not clearly reveal the growth cone; therefore, we used ZsGreen to visualize it. Consistent with the tubulin fluorescence images, the growth cone observed with ZsGreen displayed much less dynamic activity compared to nonstimulated or BDNF-stimulated conditions [movie S6 and fig. S6B(c)]. These data suggest that HAUS7 is essential for regulating the structural dynamics of polymerized tubulin within the neuronal growth cone.

Transcriptome analysis of HAUS7 in RGCs using public databases

To gain deeper insights, we conducted a bioinformatics analysis using publicly available bulk and single-cell RNA sequencing (scRNA-seq) databases. In murine retinal scRNA-seq data, the Haus7 gene was relatively enriched in RGCs among retinal neurons (Fig. 6A) (28). Human HAUS7 was also enriched in the RGCs in the human retinal scRNA-seq dataset (fig. S7A) (29). Similarly, the Dock3/DOCK3 gene displayed relatively higher expression in RGCs in the mouse and human retinas (Fig. 6A and fig. S7B). Subsequently, we explored whether the expression of the Haus7 gene in mice is down-regulated under pathological conditions such as glaucoma. To investigate this, we used the Glaucoma Discovery database, which collects data from the retina and the optic nerve head (ONH) of the DBA/2J mouse, an inherited model of primary open-angle glaucoma (30, 31). Our findings revealed a significant down-regulation of Haus7 expression in the retina at stage 4 (Fig. 6B), which corresponds to a phase when 10 to 50% of RGC axons are damaged (30). Retinal Dock3 also exhibited down-regulation at stage 4. In the ONH, no changes were observed in Haus7 expression, but Dock3 expression was significantly decreased as early as stage 2, an initial stage of glaucoma (Fig. 6C). These results suggest that HAUS7 and DOCK3 are expressed in RGCs and their expression levels are reduced in glaucoma.

Fig. 6. Bioinformatic analysis of HAUS7 using public databases.

Fig. 6.

Open in a new tab

(A) Expression of Haus7 and Dock3 genes in the mouse retina. Haus7 and Dock3 genes were relatively enriched in the RGCs among all murine retinal neurons. (B and C) Changes in Haus7 and Dock3 in the retina and ONH in a glaucoma model (i.e., DBA/2J mice). **P < 0.01 and ***P < 0.001 compared to the control group. (D and E) Haus7 and Dock3 expression in RGC subclusters in immature (D) and adult (E) retinas. (F) Relationship between axon regeneration–promoting gene therapy and Haus7 expression in RGCs. (a) Regenerating RGCs were classified into around 40 subclasses. RGCs expressing the Dock3 [red signals, (b)] or Haus7 [blue signals, (c)] genes in each cluster corresponded well to RGCs that underwent PTEN knockdown with CNTF overexpression [green signals, (d)].

Recent scRNA-seq techniques have provided us with the capability to identify subclasses of RGCs (10, 12). In the mouse retina, there are ~40 subclasses of RGCs. In immature RGCs, the Haus7 gene displayed ubiquitous expression across all subtypes of RGCs (Fig. 6D). However, its expression became limited to a small subset of mature RGC subclasses (Fig. 6E) (10). Dock3 genes in neonatal RGCs were also expressed in nearly all of the clusters (Fig. 6D). The age-associated reduction in Dock3 expression was relatively mild compared to that in Haus7 (Fig. 6E). The scRNA-seq analysis using the ONC model revealed that RGCs exhibit sensitivity or resilience against ONC (12), and the intrinsic protective mechanism overlaps with the regenerative mechanism (11). To investigate the potential role of HAUS7 in axon regeneration, we used scRNA-seq data from regenerating RGCs (i.e., 3, 7, and 21 days after the ONC) treated with gene therapy [involving the knockdown of Pten and overexpression of Cntf genes (PTEN-KD/CNTF-OE)], which significantly enhances axon regeneration (11). We observed Dock3 and Haus7 gene expression in a subset of cells in each subcluster [Fig. 6F(a to c)]. These populations of Dock3- and Haus7-expressing RGCs overlapped with the PTEN-KD/CNTF-OE RGCs [Fig. 6F(d) and fig. S8C]. These findings suggest that Dock3 and Haus7 gene expression significantly diminishes during maturation, glaucoma, or injury and that gene therapy promoting axon regeneration is associated with an up-regulation of the Dock3 and Haus7 genes.

DISCUSSION

Our findings demonstrate the crucial role of HAUS7 in facilitating axon regeneration following optic nerve injury. We have identified HAUS7 as the binding partner of DOCK3, a molecule that acts downstream of BDNF signaling and promotes axon regeneration in RGCs. In the absence of HAUS7, we observed impaired microtubule formation in axons under both normal and regenerating conditions. Bioinformatic analyses indicated the down-regulation of Haus7 expression in cases of glaucoma, whereas it was up-regulated in regenerating RGCs.

HAUS7 is a subunit of the augmin complex, which plays an essential role in microtubule assembly. It forms a hetero-octamer complex with HAUS1-8 (19, 32). A previous study demonstrated that augmins, including HAUS7, regulate polarized axon elongation in postmitotic neurons under normal conditions (21). However, the regulatory mechanisms and their roles in pathological conditions remained unclear. In this study, we have demonstrated that HAUS7 binds to the C terminus of DOCK3-DHR1 independently of WAVE1, indicating that HAUS7 has a distinct role from WAVE1 in DOCK3-mediated axon regeneration. WAVE1 regulates actin polymerization at the growth cone during axon regrowth (13). As HAUS7 is essential for microtubule assembly, we hypothesized that it is important in regenerating axons. Supporting this notion, both knockdown or deletion of Haus7 (Haus7 KD/KO) and knockdown of microtubules (Tubb3 and Tubg1/2) significantly impaired RGC axon regeneration after the ONC. EM analysis demonstrated that Haus7 KO led to a reduction in the number of microtubules in both naive and regenerating axons. The reason why Haus7 KO mice exhibited normal eye development and visual function might be attributed to the higher expression of all HAUS subunit genes in the immature RGCs, as observed in the public database (10). It is possible that other subunits may compensate for the absence of the Haus7 gene. In adult tissues, Haus7 gene expression is limited to only a small subset of RGCs and further reduced after the ONC (12), which could lead to more challenging conditions for RGC axon regeneration.

In Cos-7 cells, HAUS7 is bound with DOCK3 in the absence of external stimuli, and the HAUS7-DOCK3 complex is localized in the cytoplasm. When DOCK3 was farnesylated, HAUS7 was translocated to the plasma membrane and dissociated in response to DOCK3 phosphorylation by F-iTrkB. The reason why almost half of HAUS7 dissociated from DOCK3 when it was farnesylated but without any exogenous stimulation is that the endogenous activity of DOCK3 may be changed by its farnesylation and translocation to the plasma membrane. In primary cultured neurons, most of HAUS7 colocalized with nonphosphorylated DOCK3 at the growth cone. Upon BDNF stimulation, the amount of phosphorylated DOCK3 increased, and the dissociation of HAUS7 from DOCK3 was enhanced. These findings suggest that HAUS7 binds with nonphosphorylated DOCK3 under normal conditions, dissociates from phosphorylated DOCK3 (e.g., at Y562), and plays an essential role in regulating axon regeneration by promoting microtubule assembly at the growth cone.

The neuronal microtubule cytoskeleton serves as a transport track for molecular cargo and organelles, performing essential functions in maintaining the integrity of RGCs and the optic nerve. One of the most critical characteristics of RGC axons in glaucoma is the disruption of their transport system, which is observed in rodents (33), nonhuman primate models (34), and human patients (35). Given that Haus7 KO reduced the number of microtubules in the axons of naive mice, it is likely to be crucial for maintaining their functions. The age-associated decline in the expression of the Haus7 gene in RGCs was more significant than that of the Dock3 gene, suggesting that the age-associated reduction in regenerative capacity may be more affected by expression changes in the Haus7 gene. Notably, analysis using a public database revealed that Haus7 gene expression in the retina was significantly decreased in the moderate stage of glaucoma (in a DBA/2J mouse model) (30, 31), and Dock3 gene expression was also down-regulated at the ONH from an early stage of glaucoma. Although these transcriptome data are obtained from the whole retina and ONH rather than specifically from RGCs, the DBA/2J mouse shows a relatively selective reduction of RGCs with disease progression. Thus, the reduction in the Dock3 and Haus7 expression may be a result of RGC degeneration, and the reduction of DOCK3 at the ONH might impair HAUS7 transport to the axon and hinder microtubule maintenance.

Does the up-regulation of HAUS7 promote axon regeneration? The accelerated RGC axon regeneration induced by PTEN-KD/CNTF-OE was associated with higher expression of the Haus7 gene in the feature plot. Given that PTEN-KD/CNTF-OE induces significant axon regeneration, the larger amount of HAUS7 would likely contribute to enhanced microtubule assembly and axon elongation. In our recent study, we observed that constitutively active TrkB signaling in RGCs induced by F-iTrkB markedly accelerated axon regeneration (9). Although there was a tendency for an increase in Haus7 expression, it was not statistically significant (1.81-fold, P = 0.22), while other Haus genes were up-regulated (Haus2, 2.27-fold, P = 0.0017; Haus3, 5.81-fold, P = 0.006) in our previous data (9). The functional acceleration of HAUS7 by F-iTrkB might be sufficient, or the up-regulation of other HAUS subunits may facilitate axon regeneration.

In summary, our data demonstrated that HAUS7 and WAVE1 bind to DOCK3 and are transported to the growth cone of elongating axons (Fig. 7). BDNF/TrkB signaling induces membrane translocation of the HAUS7-WAVE1-DOCK3 complex. Phosphorylation of DOCK3 dissociates WAVE1 and HAUS7, leading to actin polymerization and microtubule assembly, respectively. This coordinated cytoskeletal regulation may promote effective axon regeneration in RGCs. Furthermore, here, we identified the phosphorylation site (Y562) in DOCK3 and generated an antibody to detect phosphorylated DOCK3 at Y562. This specific antibody was useful to elucidate the function of phosphorylated DOCK3. Deletion of the Haus7 gene (Haus7 KO) or conditions like glaucoma (characterized by Haus7 gene expression reduction and/or functional deficits) result in impaired microtubule assembly and maintenance. In such scenarios, axon regeneration following injury is compromised. Even without injury, the loss of microtubule maintenance can lead to RGC axon damage and the onset of glaucoma. Given the critical role of microtubule integrity at the ONH in glaucoma pathogenesis, HAUS7-targeted gene therapy may represent an attractive treatment approach for glaucoma.

Fig. 7. HAUS7 mediates RGC axon regeneration.

Fig. 7.

Open in a new tab

(A) Role of HAUS7 under regenerative conditions. HAUS7 is transported through RGC axons to the growth cone via a DOCK3-dependent mechanism without external stimuli. Upon BDNF stimulation, the HAUS7-DOCK3 complex is recruited to the plasma membrane. Upon DOCK3 phosphorylation, HAUS7 dissociates, promoting microtubule assembly, while WAVE1 promotes actin polymerization at the growth cone. (B) KO of Haus7 results in a reduced number of microtubules in RGC axons and significantly impairs axon regeneration after the ONC. In glaucoma, both HAUS7 and DOCK3 levels are down-regulated. The loss of microtubule maintenance can lead to axonal injury and trigger glaucomatous pathologies.

MATERIALS AND METHODS

Sex as a biological variable

Our study used male and female animals, except for the Haus7 KO mice, and similar findings are reported for both sexes. Haus7−/− female mice could not be generated and thus were not used in this study, possibly due to the poor reproductive capacity of the corresponding Haus7−/Y male parents (36).

Animals

A list of mouse strains used in this study is provided in table S1. All animal experiments were conducted in strict compliance with the ARVO Statement for the Ethical Use of Animals in Ophthalmic and Vision Research. The experimental protocols involving animals were reviewed and approved by the Institutional Animal Care and Use Committee of the Tokyo Metropolitan Institute of Medical Science (approval numbers: 21-008, 22-034, 23-068, and 24-009). C57BL/6J and Jcl:ICR mice were procured from Japan SLC (Shizuoka, Japan). No specific statistical methods were used for determining the sample size. The sample size used in this study is consistent with the standards of the research field. All mice were housed in a pathogen-free environment under controlled temperature conditions (23°C) with humidity maintained at 55%. Lighting was regulated on a 12-hour cycle (8:00 a.m. to 8:00 p.m.), and the mice had ad libitum access to both food and water.

Haus7 KO mice

Haus7+/− female mice were generated via homologous recombination in C57B6 mouse embryonic stem cells at UNITECH Co., Ltd. (Chiba, Japan). Haus7+/− female mice were mated with WT Jcl:ICR male mice to generate a Haus7-deficient mouse line. In summary, a targeting vector was introduced between exons 1 and 5 of the HauS7 gene allele (fig. S5A). Haus7−/Y male mice (i.e., with the Haus7 gene deleted from the X chromosome) were successfully obtained through the mating of founder Haus7+/− female mice with male WT Jcl:ICR mice. Genotyping of transgenic mice was performed via polymerase chain reaction (PCR) analysis using mouse tail genomic DNA and a specific primer set (fig. S5B and table S2).

Plasmid construction

We PCR amplified the full-length fragments of DOCK3, DOCK3-DHR1, WAVE1, Tubb3, Tubg1, Tubg2, and HAUS1-8 from mouse cDNAs and subcloned them into the pCMV expression vector (635689, Takara Bio, Shiga, Japan). These proteins were expressed with myc, HA, FLAG, GST, or His tags, as indicated in the figures. Site-directed mutagenesis was performed using primers for four DOCK3-DHR1 mutants (Y546F, Y548F, Y562F, and Y572F; table S3). To induce the translocation of iTrkB and DOCK3-DHR1 to the plasma membrane, we added the farnesylation signal (CAAX) to them (9). Short hairpin plasmids were constructed following the manufacturer’s instructions for the AAVpro Helper Free System (6658, Takara Bio). The shRNA sequence was modified to have the same hairpin sense and antisense sequences (table S4). A fusion protein of HAUS7 and DsRed was incorporated into the plasmid vector provided in the AAVpro Tet-One Inducible Expression System (634310, Takara Bio).

Pull-down assay

We conducted transient transfection of His- or GST-tagged proteins into Cos-7 cells using polyethyleneimine MAX (24765, Polysciences, Warrington, PA). After 24 hours of transfection, the cells were lysed with a lysis buffer [25 mM tris (pH 7.4), 150 mM NaCl, and 1% Triton X-100] and subsequently centrifuged at 20,000g for 30 min. The resulting supernatant was then incubated with TALON resin (635501, Takara Bio) or glutathione sepharose 4B resin (GE17-0756-01, GE Healthcare, Chicago, IL) for 30 min at 4°C with gentle agitation. Following the incubation, coprecipitated samples underwent immunoblot analysis using the appropriate antibodies (table S1). Quantitative analysis of the immunoblot bands in Fig. 3 (A and C) was performed using ImageJ version 2.0.0.11 (National Institutes of Health, Bethesda, MD).

Primary cultured cortical neurons

Primary cortical neurons were obtained from E16 mice following established procedures (13). Cerebral cortices from E16 mice were dissected and then digested for 30 min at 37°C in phosphate-buffered saline (PBS) containing trypsin (0.25%), glucose (0.5%), and deoxyribonuclease I (1%). Following a PBS wash, the neurons were dissociated by pipetting. For electroporation, we coelectroporated 5 μg of the Tet-on HAUS7-DsRed plasmid and 5 μg of the shGfp-U6-CMV-ZsGreen or shDock3-U6-CMV-ZsGreen, shHaus7-U6-CMV-ZsGreen, shβIIIfTubulin-U6-CMV-ZsGreen, CAG-WT full length TrkB, and CMV-GST-Elmo plasmids into 1 × 106 primary cortical neurons using the CUY21 EDIT II electroporator (BEX Co., Ltd., Tokyo, Japan) with the following settings: poring pulse (Pp), 275 V; Pp on, 1 ms; Pp off, 50 ms; driving pulse (Pd), 20 V; Pd on, 50 ms; Pd off, 50 ms; Pd N, 5; C, 940 μF. The cells were then seeded in poly-d-lysine–coated 24-well plates at a density of 1 × 105 cells per well and cultured in Dulbecco’s modified Eagle’s medium (08458-16, Nacalai Tesque, Kyoto, Japan) containing 10% fetal bovine serum (10082-147, Thermo Fisher Scientific, Waltham, MA). After 24 hours of electroporation, neurons were maintained in Dulbecco’s modified Eagle’s medium with a 2% B27 supplement (17504044, Thermo Fisher Scientific). Three days after electroporation, we introduced doxycycline (100 ng/μl; D5207, Sigma-Aldrich, St. Louis, MO) to induce the expression of the HAUS7-DsRed fusion protein. Following 8 hours of doxycycline exposure, the cells were fixed with 4% paraformaldehyde (PFA).

Immunofluorescence staining

Retinal sections or fixed cells were incubated for 1 to 2 hours in a blocking solution (PBS containing 5% horse serum and 1% Triton X-100) at room temperature. The samples were then exposed to primary antibodies for 24 hours at 4°C in the blocking solution. After washing the tissue three times with PBS containing 0.1% Triton X-100, the samples were incubated with the secondary antibodies (table S1) for 2 hours at room temperature. Fluorescence images were captured using an FV3000 confocal microscope (Olympus, Tokyo, Japan). For RGC counting, the number of Rbpms-positive cells in the GCL of four retinal slices was manually enumerated, as previously described (9, 25).

Optic nerve crush (ONC)

Male adult mice (8 to 10 weeks old) were anesthetized by intraperitoneal injection of a mixture of midazolam (4 mg/kg), medetomidine (0.3 mg/kg), and butorphanol tartrate (5 mg/kg) before the ONC procedure. The optic nerve was intraorbitally exposed and crushed ~0.5 to 1.0 mm behind the eye globe using fine surgical forceps for 5 s, following previously established protocols (9, 13).

Assessment of axon regeneration

To assess axon regeneration, RGC axons were labeled with Alexa Fluor 647–conjugated cholera toxin β-subunit (CTB647, C34778, Thermo Fisher Scientific) as previously reported (9, 25). Five days after the ONC, 2 μl of CTB647 was injected intravitreally, and 2 days later, the mice were transcardially perfused with ice-cold 1× PBS, followed by 4% PFA. The tissues were then cryoprotected in 30% sucrose. Optic nerve cryosections were prepared at a thickness of 14 μm, and the number of CTB647-positive regenerating axons that crossed vertical lines drawn parallel to the lesion site at distances of 250, 500, 1000, 1500, and 2000 μm distal to the lesion was counted manually. Because injured RGC axons typically do not regenerate spontaneously, we used AAV2-K-RasV12 to enhance their regeneration (25) and examined the effects of AAV2-shRNAs targeting HAUS7 and tubulins. shGfp was used as a control. AAV2-K-RasV12 was injected intravitreally 1 week before the ONC. The investigators were blinded to group allocations during experiments and when counting the number of regenerating axons as well as when assessing outcomes.

Production of AAVs

The production and purification of AAVs followed previously established protocols (25). Briefly, subconfluent HEK293 cells were transiently transfected with the vector plasmid and two helper plasmids, pRC2-mi342 and pHelper (6652; Takara), using polyethyleneimine MAX (24765, Polysciences). At 72 hours after transfection, the cells were detached in PBS (10 mM sodium phosphate, pH 7.4, 150 mM NaCl) by scraping and then disrupted through three cycles of freezing and thawing to release the AAV vector. Lysates were subjected to Benzonase treatment (200 U/ml; Merck, Rahway, NJ) in the presence of 5 mM MgCl2 at 37°C for 1 hour. The Benzonase-treated viral solution was loaded onto an iodixanol step gradient and ultracentrifuged at 104,000g using an Optima L-90K (Beckman Coulter, Brea, CA) equipped with a swing rotor (SW28; Beckman) for 18 hours. The iodixanol solution containing the AAV vector was collected and diluted with a 15-ml Hanks’ balanced salt solution (pH 7.4, 1.25 mM CaCl2·2H2O, 0.8 mM MgSO4, 4.2 mM NaHCO3, 5 mM KCl, 0.4 mM KH2PO4, 0.3 mM K2HPO4, and 136 mM NaCl) and then concentrated using a VIVASPIN (VS2002, Sartorius Stedim Biotech GmbH, Göttingen, Germany). Viral stock titers were determined by quantitative PCR. The mice were anesthetized with isoflurane (Intervet Inc., Madison, NJ), and AAVs were intravitreally injected with a 34-gauge needle (TN*3404 M, Nanopass, Terumo, Tokyo, Japan) at a viral titer adjusted to 1.0 × 1013 genome copies/ml for intravitreal injection, with a volume of 2 μl.

EM procedures

EM observation of intact retinal RGC axons and regenerating axons was conducted following established procedures (25). Briefly, AAV2-K-RasV12 was intravitreally injected 2 weeks before the ONC. Two weeks after the ONC, the mice were transcardially perfused with a fixation buffer (4% PFA and 2.5% glutaraldehyde in 0.1 M phosphate buffer; pH 7.4). The eyeballs, together with their optic nerves, were enucleated and postfixed overnight with the fixation buffer at 4°C. After multiple washes with 4.5% sucrose in 0.1 M cacodylate buffer, the tissues were postfixed with 2% osmium tetroxide in 0.1 M cacodylate buffer for 2 hours. Subsequently, they were dehydrated through a series of graded ethanol, followed by propylene oxide, and then polymerized in epoxy resin (EPON 812, TAAB Laboratories, Berkshire, UK) at 60°C for 48 hours. Embedded nerves were sectioned at ~500, 1000, and 2500 μm from the ONC site and were adhered to another epoxy resin block. Embedded samples were sectioned into 1-μm semithin sections using an ultramicrotome (PowerTomeX; RMC Boeckeler, Tucson, AZ) and stained with toluidine blue before EM evaluation. Ultrathin sections of 50 to 80 nm in thickness were cut and placed on formvar-coated single-slot grids. After staining with uranyl acetate and lead citrate, the ultrathin sections were observed using a transmission electron microscope (JEM-1400, JEOL, Tokyo, Japan) equipped with a bottom-mount charge-coupled device camera (EM-14321DCAM, Hamamatsu Photonics, Shizuoka, Japan). On the basis of the characteristics of regenerating axons described in our previous paper (25), we imaged regenerating axons throughout the ultrathin sections for each site (n > 50 in all sites). Using ImageJ, we measured the area of each regenerating axon and counted the number of microtubules inside each axon. Microtubule density was defined as the microtubule count divided by the area of the regenerating axons.

Bioinformatic analysis of the HAUS7/Haus7 gene using public databases

To investigate the cell type–specific expression of the HAUS7 gene in the human retina, we used scRNA-seq data published by Lukowski et al. (37). The expression of the Haus7 gene at each disease stage in the retina or ONH of the DBA/2J mice was analyzed using the Glaucoma Discovery Platform Database (31). This platform allowed us to identify a statistically significant fold change in the gene of interest compared with a strain-matched control (D2-Gpnmb+) (30, 31). To explore the RGC subcluster–specific expression of the Haus7 gene in immature, mature, and injured adult RGCs, we used the RGC subtype gene browser (https://health.uconn.edu/neuroregeneration-lab/rgc-subtypes-gene-browser) (10–12). In addition, we investigated Haus7 expression in RGCs undergoing accelerated axon regeneration because of PTEN-KD/CNTF-OE. This analysis was conducted using the Single Cell Portal Platform provided by BROAD Institute (https://singlecell.broadinstitute.org/single_cell), using the scRNA-seq data published by Jacobi et al. (11).

Statistical analysis

A two-tailed Student’s t test was used for individual comparisons between pairs of groups. For differences among multiple groups, an analysis of variance (ANOVA) was conducted, followed by the Bonferroni test as a post hoc analysis. Error bars in all figures indicate the means ± SEM. Statistical analysis was carried out using JMP software version 16.0.0 (SAS Institute, Cary, NC). The investigators were blinded with regard to the treatment. A significance level of P < 0.05 was considered statistically significant.

Acknowledgments

We thank T. Hilts for editing this manuscript and Y. Azuchi, M. Kunitomo, T. Hara, and K. Endo for technical support and useful discussions.

Funding: This work was supported in part by Japan Society for the Promotion of Science (JSPS) KAKENHI Grants-in-Aid for Scientific Research (JP24H00583 to Y.S. and T.H.; JP25K02447 and JP25K02564 to Y.S., K.N., and T.H.; JP24K12795 to X.G.; JP21K20979 and JP22K16985 to E.N.; JP22K09804 and JP25K12864 to C.H.; JP23K06818 to K.N.; and JP19KK0229 to T.H.), the Takeda Science Foundation (to Y.S. and T.H.), a Shiseido Female Researcher Science Grant (to X.G.), and the Mitsubishi Foundation (to T.H.).

Author contributions: N.K.: methodology, formal analysis, investigation, data curation, writing—original draft, writing—review and editing, and visualization. Y.S.: methodology, software, validation, formal analysis, investigation, resources, data curation, writing—original draft, writing—review and editing, visualization, project administration, and funding acquisition. X.G.: investigation, writing—review and editing, methodology, funding acquisition, data curation, validation, and formal analysis. A.K.: writing—original draft, investigation, and writing—review and editing. K.K.: investigation, writing—review and editing, and formal analysis. E.N.: investigation and funding acquisition. S.H.: investigation. C.H.: investigation, writing—review and editing, and funding acquisition. T.N.: writing—review and editing and supervision. K.N.: writing—original draft, conceptualization, investigation, writing—review and editing, methodology, resources, funding acquisition, data curation, validation, supervision, formal analysis, project administration, and visualization. T.H.: writing—original draft, conceptualization, investigation, writing—review and editing, methodology, resources, funding acquisition, data curation, validation, supervision, and project administration.

Competing interests: The authors declare that they have no competing interests.

Data and materials availability: All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials. scRNA-seq data are accessible from the original publication and Broad Institute scRNA-seq portal (https://singlecell.broadinstitute.org/single_cell/study/SCP1846/overlapping-transcriptional-programs-promote-survival-and-axonal-regeneration-of-injured-retinal-ganglion-cells).

Supplementary Materials

The PDF file includes:

Supplementary Methods

Figs. S1 to S12

Tables S1 to S4

Legends for movies S1 to S6

sciadv.adq7105_sm.pdf (2.9MB, pdf)

Other Supplementary Material for this manuscript includes the following:

Movies S1 to S6

sciadv.adq7105_movies_s1_to_s6.zip (15.4MB, zip)

REFERENCES AND NOTES

  • 1.Burton M. J., Ramke J., Marques A. P., Bourne R. R. A., Congdon N., Jones I., Ah Tong B. A. M., Arunga S., Bachani D., Bascaran C., Bastawrous A., Blanchet K., Braithwaite T., Buchan J. C., Cairns J., Cama A., Chagunda M., Chuluunkhuu C., Cooper A., Crofts-Lawrence J., Dean W. H., Denniston A. K., Ehrlich J. R., Emerson P. M., Evans J. R., Frick K. D., Friedman D. S., Furtado J. M., Gichangi M. M., Gichuhi S., Gilbert S. S., Gurung R., Habtamu E., Holland P., Jonas J. B., Keane P. A., Keay L., Khanna R. C., Khaw P. T., Kuper H., Kyari F., Lansingh V. C., Mactaggart I., Mafwiri M. M., Mathenge W., McCormick I., Morjaria P., Mowatt L., Muirhead D., Murthy G. V. S., Mwangi N., Patel D. B., Peto T., Qureshi B. M., Salomao S. R., Sarah V., Shilio B. R., Solomon A. W., Swenor B. K., Taylor H. R., Wang N., Webson A., West S. K., Wong T. Y., Wormald R., Yasmin S., Yusufu M., Silva J. C., Resnikoff S., Ravilla T., Gilbert C. E., Foster A., Faal H. B., The lancet global health commission on global eye health: Vision beyond 2020. Lancet Glob. Health 9, e489–e551 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Laha B., Stafford B. K., Huberman A. D., Regenerating optic pathways from the eye to the brain. Science 356, 1031–1034 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Weinreb R. N., Leung C. K., Crowston J. G., Medeiros F. A., Friedman D. S., Wiggs J. L., Martin K. R., Primary open-angle glaucoma. Nat. Rev. Dis. Primers. 2, 16067 (2016). [DOI] [PubMed] [Google Scholar]
  • 4.Benowitz L. I., Yin Y., Optic nerve regeneration. Arch. Ophthalmol. 128, 1059–1064 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Silver J., Miller J. H., Regeneration beyond the glial scar. Nat. Rev. Neurosci. 5, 146–156 (2004). [DOI] [PubMed] [Google Scholar]
  • 6.Williams P. R., Benowitz L. I., Goldberg J. L., He Z., Axon regeneration in the mammalian optic nerve. Annu. Rev. Vis. Sci. 6, 195–213 (2020). [DOI] [PubMed] [Google Scholar]
  • 7.Sun F., Park K. K., Belin S., Wang D., Lu T., Chen G., Zhang K., Yeung C., Feng G., Yankner B. A., He Z., Sustained axon regeneration induced by co-deletion of PTEN and SOCS3. Nature 480, 372–375 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.de Lima S., Koriyama Y., Kurimoto T., Oliveira J. T., Yin Y., Li Y., Gilbert H. Y., Fagiolini M., Martinez A. M., Benowitz L., Full-length axon regeneration in the adult mouse optic nerve and partial recovery of simple visual behaviors. Proc. Natl. Acad. Sci. U.S.A. 109, 9149–9154 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Nishijima E., Honda S., Kitamura Y., Namekata K., Kimura A., Guo X., Azuchi Y., Harada C., Murakami A., Matsuda A., Nakano T., Parada L. F., Harada T., Vision protection and robust axon regeneration in glaucoma models by membrane-associated Trk receptors. Mol. Ther. 31, 810–824 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Rheaume B. A., Jereen A., Bolisetty M., Sajid M. S., Yang Y., Renna K., Sun L., Robson P., Trakhtenberg E. F., Single cell transcriptome profiling of retinal ganglion cells identifies cellular subtypes. Nat. Commun. 9, 2759 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Jacobi A., Tran N. M., Yan W., Benhar I., Tian F., Schaffer R., He Z., Sanes J. R., Overlapping transcriptional programs promote survival and axonal regeneration of injured retinal ganglion cells. Neuron 110, 2625–2645.e7 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Tran N. M., Shekhar K., Whitney I. E., Jacobi A., Benhar I., Hong G., Yan W., Adiconis X., Arnold M. E., Lee J. M., Levin J. Z., Lin D., Wang C., Lieber C. M., Regev A., He Z., Sanes J. R., Single-cell profiles of retinal ganglion cells differing in resilience to injury reveal neuroprotective genes. Neuron 104, 1039–1055.e12 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Namekata K., Harada C., Taya C., Guo X., Kimura H., Parada L. F., Harada T., Dock3 induces axonal outgrowth by stimulating membrane recruitment of the WAVE complex. Proc. Natl. Acad. Sci. U.S.A. 107, 7586–7591 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Namekata K., Tsuji N., Guo X., Nishijima E., Honda S., Kitamura Y., Yamasaki A., Kishida M., Takeyama J., Ishikawa H., Shinozaki Y., Kimura A., Harada C., Harada T., Neuroprotection and axon regeneration by novel low-molecular-weight compounds through the modification of DOCK3 conformation. Cell Death Discov. 9, 166 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Laurin M., Cote J. F., Insights into the biological functions of Dock family guanine nucleotide exchange factors. Genes Dev. 28, 533–547 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Cote J. F., Motoyama A. B., Bush J. A., Vuori K., A novel and evolutionarily conserved PtdIns(3,4,5)P3-binding domain is necessary for DOCK180 signalling. Nat. Cell Biol. 7, 797–807 (2005). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Akhmanova A., Kapitein L. C., Mechanisms of microtubule organization in differentiated animal cells. Nat. Rev. Mol. Cell Biol. 23, 541–558 (2022). [DOI] [PubMed] [Google Scholar]
  • 18.Kollman J. M., Merdes A., Mourey L., Agard D. A., Microtubule nucleation by γ-tubulin complexes. Nat. Rev. Mol. Cell Biol. 12, 709–721 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Goshima G., Mayer M., Zhang N., Stuurman N., Vale R. D., Augmin: A protein complex required for centrosome-independent microtubule generation within the spindle. J. Cell Biol. 181, 421–429 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Stiess M., Maghelli N., Kapitein L. C., Gomis-Ruth S., Wilsch-Brauninger M., Hoogenraad C. C., Tolic-Norrelykke I. M., Bradke F., Axon extension occurs independently of centrosomal microtubule nucleation. Science 327, 704–707 (2010). [DOI] [PubMed] [Google Scholar]
  • 21.Sanchez-Huertas C., Freixo F., Viais R., Lacasa C., Soriano E., Luders J., Non-centrosomal nucleation mediated by augmin organizes microtubules in post-mitotic neurons and controls axonal microtubule polarity. Nat. Commun. 7, 12187 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Namekata K., Enokido Y., Iwasawa K., Kimura H., MOCA induces membrane spreading by activating Rac1. J. Biol. Chem. 279, 14331–14337 (2004). [DOI] [PubMed] [Google Scholar]
  • 23.Cunha-Ferreira I., Chazeau A., Buijs R. R., Stucchi R., Will L., Pan X., Adolfs Y., van der Meer C., Wolthuis J. C., Kahn O. I., Schatzle P., Altelaar M., Pasterkamp R. J., Kapitein L. C., Hoogenraad C. C., The HAUS complex is a key regulator of non-centrosomal microtubule organization during neuronal development. Cell Rep. 24, 791–800 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Namekata K., Harada C., Guo X., Kimura A., Kittaka D., Watanabe H., Harada T., Dock3 stimulates axonal outgrowth via GSK-3beta-mediated microtubule assembly. J. Neurosci. 32, 264–274 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Kiyota N., Namekata K., Nishijima E., Guo X., Kimura A., Harada C., Nakazawa T., Harada T., Effects of constitutively active K-Ras on axon regeneration after optic nerve injury. Neurosci. Lett. 799, 137124 (2023). [DOI] [PubMed] [Google Scholar]
  • 26.Latremoliere A., Cheng L., DeLisle M., Wu C., Chew S., Hutchinson E. B., Sheridan A., Alexandre C., Latremoliere F., Sheu S. H., Golidy S., Omura T., Huebner E. A., Fan Y., Whitman M. C., Nguyen E., Hermawan C., Pierpaoli C., Tischfield M. A., Woolf C. J., Engle E. C., Neuronal-specific TUBB3 is not required for normal neuronal function but is essential for timely axon regeneration. Cell Rep. 24, 1865–1879.e9 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Moskowitz P. F., Smith R., Pickett J., Frankfurter A., Oblinger M. M., Expression of the class III β-tubulin gene during axonal regeneration of rat dorsal root ganglion neurons. J. Neurosci. Res. 34, 129–134 (1993). [DOI] [PubMed] [Google Scholar]
  • 28.Li J., Choi J., Cheng X., Ma J., Pema S., Sanes J. R., Mardon G., Frankfort B. J., Tran N. M., Li Y., Chen R., Comprehensive single-cell atlas of the mouse retina. iScience 27, 109916 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Lukowski S. W., Lo C. Y., Sharov A. A., Nguyen Q., Fang L., Hung S. S., Zhu L., Zhang T., Grunert U., Nguyen T., Senabouth A., Jabbari J. S., Welby E., Sowden J. C., Waugh H. S., Mackey A., Pollock G., Lamb T. D., Wang P. Y., Hewitt A. W., Gillies M. C., Powell J. E., Wong R. C., A single-cell transcriptome atlas of the adult human retina. EMBO J. 38, e100811 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Howell G. R., Macalinao D. G., Sousa G. L., Walden M., Soto I., Kneeland S. C., Barbay J. M., King B. L., Marchant J. K., Hibbs M., Stevens B., Barres B. A., Clark A. F., Libby R. T., John S. W., Molecular clustering identifies complement and endothelin induction as early events in a mouse model of glaucoma. J. Clin. Invest. 121, 1429–1444 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Howell G. R., Walton D. O., King B. L., Libby R. T., John S. W., Datgan, a reusable software system for facile interrogation and visualization of complex transcription profiling data. BMC Genomics 12, 429 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Manka S. W., Structural insights into how augmin augments the mitotic spindle. Nat. Commun. 14, 2073 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Chidlow G., Ebneter A., Wood J. P., Casson R. J., The optic nerve head is the site of axonal transport disruption, axonal cytoskeleton damage and putative axonal regeneration failure in a rat model of glaucoma. Acta Neuropathol. 121, 737–751 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Quigley H. A., Anderson D. R., Distribution of axonal transport blockade by acute intraocular pressure elevation in the primate optic nerve head. Invest. Ophthalmol. Vis. Sci. 16, 640–644 (1977). [PubMed] [Google Scholar]
  • 35.Hernandez M. R., Pena J. D., The optic nerve head in glaucomatous optic neuropathy. Arch. Ophthalmol. 115, 389–395 (1997). [DOI] [PubMed] [Google Scholar]
  • 36.Li L., Sha Y. W., Su Z. Y., Mei L. B., Ji Z. Y., Zhang Q., Lin S. B., Wang X., Qiu P. P., Li P., Yin C., A novel mutation in HAUS7 results in severe oligozoospermia in two brothers. Gene 639, 106–110 (2018). [DOI] [PubMed] [Google Scholar]
  • 37.Lukowski S. W., Patel J., Andersen S. B., Sim S. L., Wong H. Y., Tay J., Winkler I., Powell J. E., Khosrotehrani K., Single-cell transcriptional profiling of aortic endothelium identifies a hierarchy from endovascular progenitors to differentiated cells. Cell Rep. 27, 2748–2758.e3 (2019). [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Methods

Figs. S1 to S12

Tables S1 to S4

Legends for movies S1 to S6

sciadv.adq7105_sm.pdf (2.9MB, pdf)

Movies S1 to S6

sciadv.adq7105_movies_s1_to_s6.zip (15.4MB, zip)

Articles from Science Advances are provided here courtesy of American Association for the Advancement of Science

RESOURCES