HB-EGF Promotes Horizontal Basal Cell Proliferation and Olfactory Sensory Neuron Regeneration in the Zebrafish Olfactory Epithelium

Maintenance and regeneration of the zebrafish olfactory epithelium (OE) are supported by distinct progenitor cell populations that occupy discrete stem cell niches and respond to different tissue conditions. Globose basal cells (GBCs) reside at the inner and peripheral margins of the sensory OE and are constitutively active to replace sporadically dying olfactory sensory neurons (OSNs). In contrast, horizontal basal cells (HBCs) are more uniformly distributed across the tissue, including basal layers of the sensory region, and are selectively activated by acute injury conditions that affect the morphological integrity of the OE. Here we show that expression of the heparin-binding epidermal growth factor-like growth factor (HB-EGF) is strongly and transiently upregulated in response to OE injury and signals through the EGF receptor (EGFR), which is expressed by HBCs. Exogenous stimulation of the OE with recombinant HB-EGF promotes HBC expansion and OSN neurogenesis within the sensory OE, resembling the tissue response to injury. In contrast, pharmacological inhibition of HB-EGF shedding, HB-EGF availability, and EGFR signaling strongly attenuate or delay injury-induced HBC activity and OSN restoration without affecting maintenance neurogenesis by GBCs. Thus, HB-EGF/EGFR signaling appears to be a critical component of a complex signaling network that controls HBC activity and, consequently, repair neurogenesis in the zebrafish OE.


278
The transcriptome profile indicates that hbegfa is highly and transiently upregulated immediately after 279 OE injury, while the second zebrafish paralog, hbegfb, does not show any significant changes in 280 expression ( Figure 2B). Hbegfa transcript levels rise 13.6 ± 0.9-fold by 4 hpl but decline rapidly to 4.1 281 ± 0.3-fold higher expression by 12 hpl and revert to pre-injury levels between 24 and 72 hpl. However,

320
Tissue damage by 0.1% TrX increased expression of hbegfa in all major OE cell types at 4 hpl ( Figure   321 4G

413
To examine if the increased mitotic activity is neurogenic and results in the generation of a higher than

511
To further discriminate between different cell types, we analyzed the number of Krt5/EdU double-

519
To further test this possibility, we performed RNA in situ-hybridization against transcripts of the 520 proneural gene ascl1, which is expressed by GBCs. Stimulation of the OE with 200 ng/µl HB-EGF also 521 induced ascl1-positive cells within the sensory region ( Figure 6G), which suggests that HB-EGF 522 activates a neurogenic program in HBCs that is identical to injury-induced OSN regeneration.

525
In the zebrafish retina, HB-EGF is part of an autoregulatory gene expression network that includes      The spreadsheet contains the cell counts of tp63 and hbegfa single-and double-positive cells for 620 vehicle-treated and HB-EGF-stimulated OEs and statistics information (Student's t-tests) for the data 621 shown in Figure 6I.
activates EGFR (Higashiyama et al., 2008). We intercepted this signaling route at three different levels in the injured zebrafish OE (Figure 7). Cell proliferation and OSN neurogenesis in the intact OE occurs

634
To prevent HB-EGF ectodomain shedding, we utilized the broad-spectrum MMP inhibitor Marimastat 16.5 ± 1.9 cells/hemi OE; padj = 9.5x10 -6 ), the increase in mitotic activity was only 1.9 ± 0.2-fold higher 678 in comparison to unlesioned vehicle controls (padj = 0.011). This is most likely due to the effect of the 679 inhibitor on ongoing activity in the intact tissue, which was much lower in PD153035-injected animals.

680
Taken together, these observations strongly suggest that HB-EGF/EGFR signaling is required during 681 the early phase of the injury response to trigger HBC activity in the lesioned zebrafish OE at 24 hpl.

682
Inhibition of HB-EGF/EGFR signaling prevents the development of the typical spatial and temporal 683 pattern of mitotic activity that drives OSN neurogenesis and tissue regeneration. In contrast, inhibition has no major effect on GBC activity at the ILC and SNS, further strengthening the notion that HB-

807
We therefore suspected that OSN repair neurogenesis was severely delayed in CRM197-and   The spreadsheet provides the raw positional cell counts and statistics information (one-way ANOVA,  (Figure 7). This is most likely because the MMP inhibitors Marimastat and GM6001 have synergistically or redundantly to Notch, may play a more prominent role through direct induction of Wnt employed an experimental injury model that globally destroys OE cells and, accordingly, observed widespread induction of hbegfa expression (Figure 3) and activation of HBC proliferation (Figure 6).

1218
In situ-hybridization against hbegfa, egfr, and ascl1a mRNA transcripts were performed as described after the lesion until analysis of early effects at 24 hpl, or between 48 and 72 hpl for analysis of late sequencing data for untreated control OEs and OEs dissected at 4, 12, 24, 72, and 120 hpl was described previously (Kocagöz et al., 2022) and read counts were converted to fragments per kilobase (R Core Team 2022) and gene/genome information obtained from the Ensembl data base The authors are grateful to Umut Sahin for valuable suggestions and reagents.
The authors declare no competing financial and non-financial interests.
All data generated or analyzed during this study are included in the manuscript and supporting files;