Wnt/β -catenin signalling shapes macrophage behaviour during injury and repair in the mouse submandibular gland

Abstract

Xerostomia, the subjective sensation of dry mouth, is a debilitating consequence of head and neck radiotherapy and autoimmune disorders such as Sjögren’s syndrome. It severely impairs oral health and quality of life by promoting mucosal ulceration, infection, malnutrition, and speech difficulties, yet effective regenerative treatments remain limited. Macrophages have recently emerged as critical regulators of salivary gland repair through their roles in coordinating inflammation, fibrosis, and epithelial regeneration; however, the molecular mechanisms governing macrophage activation and function in the injured salivary gland remain poorly defined. The Wnt/β-catenin signalling pathway is a key regulator of inflammation and tissue homeostasis across multiple organs, but its role in salivary gland macrophages has not been well characterised. Here, we investigated canonical Wnt/β-catenin signalling following murine submandibular gland injury induced by main excretory duct ligation, with deligation used in selected experiments to model repair. Using Axin2^CreERT2/+;R26^mTmG/+ reporter mice, we observed an increase in Axin2⁺ cells and substantial recruitment of F4/80⁺ macrophages exhibiting active Wnt/β-catenin signalling within the injured, ligated gland. qPCR-based gene expression analysis revealed increased expression of Axin2, F4/80, and several Wnt genes, including Wnt2 and Wnt2b, at days 3 and 6 post-injury, and identified Wnt2 and Wnt2b as macrophage-secreted ligands. Notably, despite the injury-associated increase in Wnt/β-catenin signalling, Axin2⁺ cells did not give rise to acinar cells following deligation. Finally, conditional depletion of Wntless (Wls) using pCAG^CreERT2/+;Wls^fl/fl. and Axin2^CreERT2/+;Wls^fl/fl mice increased the number of CD206⁺ macrophages and reduced fibrosis, indicating a potential association between Wnt signalling, macrophage polarisation, and fibrotic repair. Together, these findings identify Wnt/β-catenin signalling as a regulator of macrophage phenotype and tissue repair in the injured salivary gland, suggesting that targeted modulation of Wnt activity may promote regeneration and enhance functional recovery.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-38873-1.

Introduction

Salivary glands are exocrine organs responsible for saliva secretion, and their function depends on coordinated interactions between epithelial, stromal, and immune cells^1,2. Injury to the salivary glands, whether caused by ductal obstruction, inflammation, or irradiation, results in acinar cell loss, fibrosis, and reduced secretory function^3–6. In response to such injury, the inflammatory process plays a critical role in determining the outcome of repair and regeneration^1,7. Despite the gland’s limited regenerative capacity, inflammatory and stromal signals are essential for coordinating the transition from injury to healing^8,9.

The Wnt/β-catenin signalling pathway has been shown to regulate inflammation and tissue regeneration by coordinating immune and stromal cell behaviour in several organs, including the heart, kidney, and liver^10–12. However, the role of Wnt signalling in injury and inflammation within the salivary glands remains poorly understood. Wnt/β-catenin signalling has been shown to be activated within the ductal compartment of injured murine submandibular glands using Wnt reporter mice¹³, and Wnt-responding ductal cells exhibit epithelial plasticity and can generate acinar cells under severe injury conditions such as irradiation⁸. Notably, these studies have focused primarily on epithelial and ductal compartments rather than the stromal or immune components.

Among the immune cells involved in the salivary gland’s response to injury, macrophages have been shown to contribute to inflammation, tissue repair, and fibrosis¹⁴. They are highly plastic cells that adopt distinct phenotypic and functional states depending on local environmental cues, which change dynamically throughout injury and repair¹⁵; accordingly, defining how macrophage populations shift after injury is critical for identifying therapeutic strategies. In radiation-induced injury, macrophages have been shown to drive chronic fibrosis and, thus, represent promising therapeutic targets^16,17. Consistently, macrophage-focused therapeutic modulation (e.g., macrophage-enriched E-MNC therapy) has been reported to improve salivary secretion and reduce fibrosis in irradiated salivary glands¹⁸. Additionally, transient activation of Hedgehog–Gli signalling after irradiation restores salivary gland function by recovering resident macrophages and promoting epithelial and endothelial repair¹⁹. More recently, CSF1R-dependent macrophages were shown to be essential for epithelial regeneration after radiation injury⁷, which highlights macrophage maintenance as a critical determinant of gland recovery.

Although these studies demonstrate the importance of macrophages in glandular repair, the molecular mechanisms regulating their activation state and function in non-radiation injuries remain unclear. Here, we investigate how canonical Wnt/β-catenin signalling regulates macrophage behaviour following ductal ligation injury in the adult murine submandibular gland (SMG) and whether modulation of this pathway alters macrophage-mediated repair and fibrosis.

Using Axin2^CreERT2/+;R26^mTmG/+ mice, we visualised Wnt-responsive cells following ligation and deligation, defined their identity using macrophage markers, quantified time-dependent Wnt pathway gene expression by qPCR, and assessed the effects of Wls depletion after injury both globally and within Axin2⁺ cells using pCAG^CreERT2/+;Wls^fl/fl and Axin2^CreERT2/+;Wls^fl/fl mice.

Material & methods

Animals

pCagCre^ERT2/+20, Axin2^CreERT2/+21, R26^mTmG/+22, and Wntless^{flox−flox (fl/fl)23} mice were obtained from The Jackson Laboratory. Wild-type (CD1) mice were also used in this study. Animals were maintained under standard conditions on a 12 h light/dark cycle with ad libitum access to food and water. Experiments were conducted using 12-week-old female mice, with at least three biologically independent animals (biological replicates) per experimental group, unless stated otherwise. PCR protocols and primer details for transgenic mouse lines are provided in Supplementary Table 1.

Tamoxifen induction, duct ligation and deligation

Cre recombination was induced by intraperitoneal tamoxifen administration (0.075 mg g⁻¹ body weight, dissolved in corn oil). Submandibular gland (SMG) duct ligation and deligation were performed under Hypnorm/Hypnovel anaesthesia (Hypnorm^®:sterile H₂O: Hypnovel^® at 1:2:1; 100 µL/10 g body weight, administered intraperitoneally). Anaesthetic depth was confirmed after 5 min by loss of toe or tail withdrawal reflex. Post-operative analgesia was provided by buprenorphine (Vetergesic^®, 0.3 mg mL⁻¹; 10 µL/10 g body weight, administered intraperitoneally), and mice recovered in a 27 °C incubator with ad libitum access to food and water. For ligation, a midline cervical incision (~ 10–15 mm) was made and the main excretory duct was exposed by blunt dissection and occluded with a haemostatic clip; the wound was closed with 6 − 0 silk sutures. For deligation, the incision was re-opened and the clip was carefully removed (by inserting a thin blade between the clip jaws to minimise duct damage), and the wound was closed with 6 − 0 silk sutures. The clip was removed after the specified ligation period (as indicated for each experiment), and glands were allowed to regenerate for a further 2 weeks (or longer, as indicated).

Euthanasia and tissue collection

Mice were euthanised by exposure to carbon dioxide gas in a dedicated chamber using a rising concentration. Animals were placed in the chamber containing room air (not pre-filled), and 100% carbon dioxide was introduced gradually at ~ 20% of the chamber volume per minute. Death was confirmed by cessation of breathing and absence of reflexes, and euthanasia was completed by cervical dislocation, taking care to avoid damage to the salivary glands prior to tissue collection.

RNA extraction and quantitative PCR

Total RNA was isolated from submandibular glands (SMG) using TRIzol reagent, followed by DNase treatment to eliminate genomic DNA. Complementary DNA (cDNA) was synthesised from 1 µg of RNA using random primers and reverse transcriptase. Quantitative real-time PCR was performed using SYBR Green chemistry on a Bio-Rad CFX384 thermocycler. Gapdh served as the internal control, and relative expression levels were calculated using the 2^–ΔΔCt method. The PCR cycling protocol and melt curve conditions are shown in Supplementary Tables 2 and 3. Primer sequences used for qPCR are listed in Supplementary Table 4.

Immunostaining

Tissues were fixed in 4% paraformaldehyde, cryo- or paraffin-embedded, and sectioned at 10 μm and 8 μm, respectively. Sections were permeabilised, blocked in serum-containing buffer, and incubated with primary antibodies overnight at 4 °C, followed by fluorescent secondary antibodies. Nuclei were counterstained with Hoechst 33,342, and slides were mounted in antifade medium. Imaging was performed using a Leica TCS SP5 confocal microscope. Details of antibodies and reagents used are provided in Supplementary Tables 5 and 8. Quantification of GFP⁺ mesenchymal cells and histological measurements of ductal dilation and capsule thickness are described in the Supplementary Methods.

Flow cytometry

Freshly dissected salivary glands were enzymatically and mechanically dissociated into single-cell suspensions. Cells were stained with fluorophore-conjugated antibodies against macrophage and immune cell markers (including F4/80, CD45, and CD206) and with a viability dye to exclude dead cells. Data were acquired on a BD Fortessa cytometer and analysed using FlowJo v10. Each experiment was performed in biological triplicate. Conjugated antibodies used for flow cytometry are listed in Supplementary Table 7. Key equipment used in this study is summarised in Supplementary Table 6. Flow cytometry gating strategy for identification of CD45⁺ immune cells is described in the Supplementary Methods.

Statistical analysis

Data are presented as mean ± SD, with n indicating the number of biological replicates (individual mice) as stated in the figure legends. Key findings were reproduced in at least three independent experiments using separate biological samples. Statistical comparisons between two groups were performed using an unpaired two-tailed Student’s t-test. A P value < 0.05 was considered statistically significant. Analyses were performed using GraphPad Prism 10. No a priori power calculations were performed; sample sizes were chosen based on prior studies and practical/ethical considerations and are reported for transparency.

Results

SMG injury induces canonical Wnt/β-catenin activation in stromal cells

To assess Wnt/β-catenin signalling activity following SMG injury, Axin2^CreERT2/+;R26^mTmG/+ reporter mice (12 weeks old, n = 3 females) were used. Tamoxifen was administered on days 1, 3, and 6 after ligation of the Main Excretory Duct (MED), and glands were collected on day 8 (Fig. 1E, F). Histological analysis of ligated SMGs confirmed characteristic injury-associated changes, including acinar cell loss, ductal dilation, capsular thickening, and increased fibrotic tissue deposition (Fig. 1), consistent with previous studies^24–30. Corn oil–treated reporter controls showed no detectable GFP signal, confirming absence of basal Cre activity without induction (Supplementary Fig. 1).

Fig. 1.

Histological changes in the SMG 8 days after duct ligation. Sirius Red trichrome staining of Axin2^CreERT2/+;R26^mTmG/+ SMGs shows extensive acinar cell loss, ductal dilation, and capsule thickening with increased collagen deposition (B, D) compared with non-operated controls (A, C). Alcian Blue labels acinar cells (white arrows), while ducts appear pale yellow (black arrows). Double-headed arrows in (C) and (D) indicate SMG capsule thickness. (E) Schematic of the ligation procedure of the MED, with the sublingual duct left intact. (F, G) Experimental timeline indicating ligation, non-operated controls, and tamoxifen administration. TAM: tamoxifen; CO: Corn Oil; MED: Main Excretory Duct. Scale bar: 100 μm.

Building on a previous study¹³, we observed that ductal ligation induced a marked increase in Axin2⁺ (Wnt-responding) cells, but this response was confined to the SMG stroma, particularly in the capsule, septa, and interstitial spaces between acini and ducts relative to non-operated controls (Fig. 2). Furthermore, GFP⁺ mesenchymal cell density was significantly increased in the gland capsule/septa and interstitial spaces during ligation compared with the non-operated controls (Supplementary Fig. 2). These results indicate that ductal ligation activates canonical Wnt/β-catenin signalling within stromal compartments of the SMG.

Fig. 2.

Increased Axin2⁺ cells within the stromal compartment of the SMG after injury. GFP immunostaining of Axin2^CreERT2/+;R26^mTmG/+ SMGs shows a marked expansion of Axin2⁺ (GFP⁺) cells within the capsule, septa, and interstitial regions between acini and ducts 8 days after ligation (D–F) compared with non-operated controls (48 h) (A–C). Double immunostaining for GFP and the ductal marker K7 reveals few K7⁺ ductal cells with active Wnt/β-catenin signalling (C, white arrows). At baseline, only occasional elongated Axin2⁺ cells were detected in septa (B, inset). Nuclei were counterstained with Hoechst (blue). GFP: Green; Tomato/K7: Red. ED: Excretory duct. Scale bars: 250 μm (A, B, D–F); 25 μm (C).

CD45⁺ macrophages represent the major Axin2⁺ cell population after injury

Given the increase in stromal cells exhibiting active Wnt/β-catenin signalling 8 days after injury in Axin2^CreERT2/+;R26^mTmG/+mice, we next sought to identify the specific cell populations involved. Double immunofluorescence for GFP and either CD45 (a pan-immune cell marker) or F4/80 (a mature macrophage marker) was performed. Numerous CD45⁺ cells, located primarily within the gland capsule and septa, exhibited active Wnt/β-catenin signalling. A large proportion of these CD45⁺ cells also expressed F4/80, confirming their macrophage identity (Fig. 3A, B). To corroborate these observations, flow-cytometric analysis of ligated SMGs was conducted according to the experimental timeline shown in Fig. 1E, F. Approximately 18% of all cells were CD45⁺, of which 37% were F4/80⁺ (Fig. 3C, D). Among these immune populations, 29% of CD45⁺ cells and 34.3% of F4/80⁺ macrophages exhibited active Wnt/β-catenin signalling (Fig. 3E, F). Collectively, these data demonstrate that following injury, more than one quarter of CD45⁺ immune cells display canonical Wnt activity, and within this population, approximately one third of F4/80⁺ macrophages are Wnt-responding.

Fig. 3.

Inflammatory and macrophage populations exhibit active Wnt/β-catenin signalling after SMG injury. (A, B) Double immunostaining of ligated Axin2^CreERT2/+;R26^mTmG/+ SMGs shows colocalisation of GFP⁺ cells with CD45⁺ immune cells (A) and F4/80⁺ macrophages (B) within the capsule and septa 8 days after injury, indicating activation of canonical Wnt signalling. Nuclei were counterstained with Hoechst (blue). GFP: Green; CD45/F4/80: Red. (C–F) Flow cytometry analysis reveals that 18% of total cells were CD45⁺, of which 37% were F4/80⁺; 29% of CD45⁺ cells and 34.3% of macrophages exhibited GFP-associated Wnt/β-catenin activity. Scale bar: 250 μm.

Concomitant upregulation of F4/80 and Axin2 following SMG injury

Having identified that approximately one third of F4/80⁺ macrophages exhibit active Wnt/β-catenin signalling, we next examined the temporal relationship between macrophage recruitment and Wnt activation using wild-type (WT) mice. SMG ligation was performed and terminated at days 1, 2, 3, 4, 5, 6, and 7 post-surgery (n = 6 adult females per time point), with three mice serving as sham-operated controls at each time point. Quantitative PCR analysis of SMG tissue revealed significant upregulation of Axin2, a canonical Wnt/β-catenin signalling readout, at days 3 and 6 after injury, whereas other time points showed minimal changes (Fig. 4A). Similarly, F4/80 expression exhibited a parallel trend, with marked peaks at days 3 and 6 (Fig. 4B), suggesting a strong correlation between macrophage recruitment and Wnt activation after injury.

Fig. 4.

Concomitant upregulation of Axin2 and F4/80 expression following SMG injury. Quantitative PCR analysis of wild-type SMGs shows dynamic changes in gene expression after duct ligation. (A) Axin2 expression peaks significantly at days 3 and 6 post-injury. (B) F4/80 expression follows a similar pattern, with significant upregulation at the same time points, indicating a temporal correlation between macrophage recruitment and Wnt/β-catenin activation. Data are normalised to Gapdh expression. *P < 0.05, **P < 0.01 (Student’s t-test).

To determine which of these peaks contributed to the increase in F4/80⁺Axin2⁺ cells observed at day 8 post-injury, Axin2^CreERT2/+;R26^mTmG/+ mice were used. Tamoxifen was administered either at days 2–3 or days 5–6 following ligation, and SMGs were collected at the end of each respective period (Fig. 5E, F). Consistent with this early timepoint, the GFP⁺ population labelled by tamoxifen on days 2–3 was predominantly CD45⁺ immune cells, including a small subset of F4/80⁺ macrophages. Immunofluorescence for GFP showed substantially more labelled cells when tamoxifen was administered at days 5–6 compared with days 2–3 (Fig. 5A–D). Quantification confirmed a significant increase in GFP⁺ mesenchymal cell density at day 6 versus day 3 following ligation (Supplementary Fig. 3), indicating that the majority of Axin2⁺ cells arise around day 6 post-injury. This timing corresponds with the appearance of pro-reparative macrophages described in other organs, such as the liver, where F4/80⁺ macrophages display active Wnt/β-catenin signalling¹¹. These findings suggest that Axin2⁺ macrophages in the SMG are likely pro-reparative and activate canonical Wnt signalling during the later stages of injury.

Fig. 5.

Axin2⁺ cells are predominantly recruited to the SMG 6 days after injury. Immunofluorescence staining for GFP in ligated Axin2^CreERT2/+;R26^mTmG/+ SMGs reveals few Axin2⁺ cells within the capsule and septa at day 3 post-injury (A, B), but a marked increase by day 6 (C, D). Schematics illustrate tamoxifen administration and collection timelines for 3-day (E) and 6-day (F) injury protocols. Nuclei were counterstained with Hoechst (blue). GFP: Green; Tomato: Red. TAM: tamoxifen. Scale bar: 250 μm.

To determine whether these Wnt-responding cells were derived from resident or circulating populations, Axin2^CreERT2/+;R26^mTmG/+ mice were injected with tamoxifen either before or after injury. In the first scenario, Axin2⁺ cells were labelled during homeostasis, followed by 8 days of ductal ligation. In the second, mice underwent ligation first and received tamoxifen injections at days 1, 3, and 6 post-surgery, coinciding with known peaks in macrophage recruitment. The number of GFP⁺ cells was markedly lower in mice labelled before injury (Fig. 6B, E, H) compared with those labelled after injury (Fig. 6C, F, I). Moreover, GFP⁺ mesenchymal cell density was significantly higher when tamoxifen was administered during ligation than with pre-injury labelling, and remained minimal in non-operated controls (48 h) (Supplementary Figs. 2 and 4). This suggests that canonical Wnt signalling is activated in immune cells in response to injury. These findings indicate that the majority of Wnt active cells are derived from circulating Axin2⁻ monocytes that infiltrate the gland and switch on Wnt/β catenin signalling following injury, although a minor contribution from resident macrophages cannot be excluded. This aligns with previous findings showing that salivary gland macrophage recovery involves early local proliferation followed by sustained recruitment from circulating monocytes⁷. Collectively, these data suggest that macrophages dynamically engage Wnt/β-catenin signalling during SMG injury, with peak activity around day 6, suggesting an association with a more pro-reparative state.

Fig. 6.

Axin2⁺ cells arise predominantly from recruited Axin2⁻ cells following SMG injury. Immunofluorescence staining for GFP in Axin2^CreERT2/+;R26^mTmG/+ SMGs shows few Axin2⁺ cells when tamoxifen was administered before ligation (B, E, H) compared with a marked increase when given after injury (C, F, I). Recruited Axin2⁺ cells were abundant in the interstitial space, gland capsule, and around excretory ducts relative to non-operated controls (A, D, G). Schematics illustrate tamoxifen administration and tissue collection timelines for the non-operated control (J), tamoxifen-before (K), and tamoxifen-after (L) protocols. Nuclei were counterstained with Hoechst (blue). GFP: Green; Tomato: Red. Scale bar: 250 μm.

Several Wnt genes are upregulated following SMG injury

Having identified a significant increase in Axin2⁺ cells after SMG injury, we next examined which Wnt ligands might activate canonical Wnt/β-catenin signalling in these cells and determined their potential cellular sources. Given the two peaks of Axin2 upregulation at days 3 and 6 after injury, coinciding with F4/80 expression peaks, we screened the entire Wnt gene family by quantitative PCR at these time points. At day 3, several Wnt ligands, including Wnt1, Wnt5b, and Wnt9a, were upregulated, alongside novel candidates such as Wnt2, Wnt2b, Wnt7b, and Wnt11 (Fig. 7A). By day 6, Wnt2 and Wnt2b remained significantly upregulated, with additional increases in Wnt8a and Wnt8b (Fig. 7B). Notably, many of these Wnt ligands have been reported to be secreted by macrophages in other tissues; for example, Wnt7b is released by macrophages during kidney injury¹². To determine the cellular source of these ligands in the SMG, immunostaining was performed for Wnt2 and Wnt2b, the most highly upregulated canonical Wnt components. Both ligands were detected within F4/80⁺ macrophages (Fig. 7E, F, arrowheads), suggesting that macrophages express Wnt proteins after injury and may contribute to local Wnt signalling. Together, these findings indicate that macrophages participate in a dual mode of Wnt regulation in the SMG injury response, both producing Wnt ligands and responding to them through autocrine or paracrine activation of canonical Wnt/β-catenin signalling.

Fig. 7.

Multiple Wnt ligands are upregulated following SMG injury, and Wnt2/Wnt2b are expressed by macrophages. qPCR analysis of wild-type SMGs shows significant upregulation of several Wnt genes 3 days (A) and 6 days (B) after duct ligation compared with sham-operated controls. Wnt2 and Wnt2b were among the most highly induced ligands at both time points. Immunofluorescence staining 3 days after injury confirms expression of Wnt2 and Wnt2b within F4/80⁺ macrophages (E, F; arrowheads). Schematics illustrate the ligation experiment timeline and tissue collection time points for 3D ligation (C) and 6D ligation (D). Data were normalised to Gapdh. *P < 0.05, **P < 0.01, ***P < 0.001 (Student’s t-test). Lig/L: Ligation; D: Day. W: Week. SMG: Submandibular gland. WT: Wild Type. Scale bar: 300 μm.

Acinar cell regeneration is independent of Axin2⁺ cells after injury

The salivary gland is composed primarily of acinar cells responsible for secretion. Following ductal ligation, these cells undergo widespread degeneration, and regeneration occurs after ductal patency is restored. To determine whether Axin2⁺ cells contribute to acinar regeneration during the recovery phase, Axin2^CreERT2/+;R26^mTmG/+ mice (12 weeks old, n = 5) underwent right SMG ligation followed by tamoxifen administration at days 1, 3, and 6 after injury. 8 days later, ligation was reversed, and glands were allowed to regenerate for two weeks or longer before analysis.

Sirius Red trichrome staining confirmed complete restoration of gland architecture, with abundant acinar structures and normal ductal morphology in deligated glands (Fig. 8A, B). Double immunofluorescence for GFP and either K7 (ductal marker) or Aqp5 (acinar marker) revealed that GFP⁺ Axin2⁺ cells were confined to intercalated and granular ducts (Fig. 8F, Supplementary Fig. 5). Importantly, no GFP⁺Aqp5⁺ cells were detected, indicating that Axin2⁺ cells, whether epithelial or stromal, did not give rise to regenerated acinar cells (Fig. 8C, D, Supplementary Fig. 5).

Fig. 8.

Axin2⁺ cells do not contribute to acinar cell regeneration following SMG injury. After duct ligation and subsequent deligation, Axin2CreERT2/+;R26mTmG/+ SMGs regained normal acinar architecture within two weeks, as shown by Sirius red trichrome staining (A, B). Double immunostaining for GFP with Aqp5 (acinar marker) or K7 (ductal marker) revealed no GFP⁺ acinar cells in the regenerated gland (C, D), whereas GFP⁺K7⁺ ductal cells persisted within intercalated ducts (F). Stromal GFP⁺ cells were also detected around excretory ducts (E). (G) Experimental timeline indicating ligation, deligation, and tamoxifen administration. Nuclei were counterstained with Hoechst (blue). GFP: Green; Tomato/K7/Aqp5: Red. ED: Excretory duct; ID: Intercalated Duct; GD: Granular Duct. Lig: Ligation; Delig: Deligation; TAM: tamoxifen. Scale bars: 100 μm (A–C, E), 300 μm (D), 250 μm (F).

These findings are consistent with previous reports showing that prelabelled Axin2⁺ cells fail to generate acinar cells after reversible SMG injury⁸. Together, these results suggest that Axin2⁺ ductal cells support recovery by replenishing the ductal compartments after ductal ligation injury, while showing limited capacity to generate acinar cells.

Depletion of Wls increases CD206⁺ macrophages and reduces fibrosis following SMG injury

Having established that macrophages both activate Wnt/β-catenin signalling and secrete Wnt2 and Wnt2b ligands, we next examined the effect of Wnt ligand secretion loss on salivary gland repair. To this end, we used pCAG^CreERT2;Wls^fl/fl (Wls) and Axin2^CreERT2/+;Wls^fl/fl (Axin2Wls) mice subjected to SMG ligation (12 weeks old, n = 12 females). In Wls mice, Wls was depleted one week before injury by three tamoxifen injections, while in Axin2Wls mice, tamoxifen was administered on days 1, 3, and 5 post-injury, and SMGs were collected on day 7 (Fig. 9M, N).

Fig. 9.

Wls depletion increases CD206⁺ macrophages and reduces fibrosis following SMG injury. Sirius red staining and immunofluorescence of ligated Axin2^CreERT2/+;Wls^fl/fl and pCAG^CreERT2/+;Wls^fl/fl SMGs show fewer dilated ducts and a thinner gland capsule following Wls depletion compared with controls (A–H). Immunostaining for CD206 (pro-reparative macrophages) and CD45 reveals a marked increase in CD206⁺ macrophages within the interstitial space of Wls-depleted glands (J, L) relative to controls (I, K). Schematics illustrate the timeline of ligation, tamoxifen administration, and tissue collection for the Wls (M) and Axin2Wls (N) protocols. Nuclei were counterstained with Hoechst (blue). TAM: tamoxifen; CO: Corn Oil. Scale bars: 100 μm (I, J), 250 μm (K, L).

Histological analysis revealed a general reduction in ductal dilation and capsule thickness in Wls and Axin2Wls glands compared with controls (Fig. 9A–H). Consistent with this, quantification showed significantly reduced capsule thickness and fewer dilated ducts in Wls glands relative to controls (Supplementary Fig. 6A, B), and similar reductions in Axin2Wls glands (Supplementary Fig. 7A, B). Collectively, these findings indicate a potential reduction in fibrosis following Wls depletion. Immunofluorescence for CD45 and CD206 showed a marked increase in CD206⁺ cells within the interstitial spaces of injured glands in Wls mice (Fig. 9J, I, white arrows), whereas in controls, CD206⁺ macrophages were largely confined to the capsule (Fig. 9I, white arrowhead). Similarly, Axin2Wls mice exhibited increased numbers of CD206⁺ macrophages within the injured gland (Fig. 9K, L). These findings suggest that loss of Wnt secretion enhances the recruitment or maintenance of CD206⁺ pro-reparative macrophages and may attenuate fibrotic remodelling. This aligns with previous observations that macrophages contribute at both early and late stages of salivary gland repair, with interstitial macrophage populations supporting epithelial regeneration after injury⁷. Comparable effects have been reported in the heart, where macrophage-specific Wls deletion following myocardial infarction increased reparative macrophages and reduced fibrosis^10,31.

Together, these results suggest that Wnt signalling may contribute to macrophage phenotypic regulation, and that Wls depletion correlates with a shift toward a pro-reparative macrophage profile and a reduction in fibrotic changes after SMG injury.

Discussion

The mechanisms by which macrophages coordinate injury resolution and tissue regeneration in the salivary gland remain incompletely defined. Here, we show that submandibular gland (SMG) injury induces a robust increase in Wnt/β-catenin signalling within stromal and interstitial regions of the gland. These Wnt-responding cells are predominantly immune in origin, with a significant proportion identified as F4/80⁺ macrophages. qPCR-based temporal analysis revealed that Wnt activation peaks 6 days after injury, coinciding with macrophage accumulation and the early onset of fibrosis. These findings suggest that Wnt-active macrophages may contribute to the inflammatory-to-reparative transition following salivary gland injury. Thus, macrophage-Wnt crosstalk emerges as a potential regulatory node in restoring immune balance during glandular repair.

Previous studies have shown that macrophages are critical for epithelial recovery in the SMG, acting through distinct subsets across early and late repair phases and supporting regeneration after irradiation-induced injury⁷. Building on this concept, our data indicate that canonical Wnt/β-catenin signalling regulates macrophage activation state rather than their persistence or viability, thereby shaping their reparative potential and influencing fibrotic outcome. In our model, Wnt-responsive macrophages were predominantly circulating or infiltrating cells, although a minor population of resident macrophages also displayed Wnt activation and may contribute locally to regenerative signalling, consistent with prior observations¹⁹. By secreting Wnt ligands such as Wnt2 and Wnt2b, these cells establish an autocrine and paracrine feedback loop that sustains Wnt activity within the macrophage compartment. Evidence from cardiac and pulmonary fibrosis models supports a conserved role for macrophage-derived Wnt signalling in tissue remodelling^31,32. Collectively, these findings suggest that Wnt-dependent macrophage activation may represent a mechanism associated with immune regulation and epithelial repair in the salivary gland.

Persistent Wnt activation after ductal ligation appears to contribute to the maintenance of a fibrotic environment. In line with this, depletion of Wnt secretion through Wls loss increased the number of CD206⁺ macrophages and reduced collagen deposition, suggesting an association with a more pro-reparative phenotype. CD206⁺ macrophages were detected in both stromal and parenchymal compartments by immunostaining, although these regions were not quantified separately. Importantly, CD206 alone does not fully define functional M2 polarisation³³; additional markers (e.g., Arg1, Il10) would be required to confirm macrophage functional state, and this warrants further investigation. Similar outcomes have been observed in other organs, where macrophage-specific deletion of Wls enhanced reparative macrophage accumulation and attenuated fibrosis after myocardial infarction^10,31. More broadly, canonical Wnt/β-catenin signalling regulates inflammation, repair, and fibrotic remodelling across multiple organ systems^34–36. Taken together, these findings suggest that Wnt signalling may constrain macrophage polarisation toward a reparative phenotype, such that its inhibition could facilitate tissue recovery. Conversely, sustained Wnt activity may promote pathological remodelling and reinforce chronic fibrosis in the glandular stroma.

Although Wnt/β-catenin signalling was strongly activated in the stroma, Axin2⁺ epithelial cells did not contribute to acinar regeneration following deligation. This finding is consistent with previous work showing that Axin2⁺ ductal cells fail to generate new acinar cells after reversible injury, although they can acquire plasticity under conditions of severe damage such as irradiation⁸. These data indicate that Wnt-responding ductal cells contribute primarily to ductal maintenance rather than acinar regeneration under mild or reversible injury conditions. This distinction suggests that stromal Wnt signalling may play a prominent role in coordinating immune activity and tissue remodelling to support epithelial regeneration during salivary gland injury.

This study also highlights the limitations of macrophage depletion strategies in the salivary gland. Attempts to ablate macrophages using clodronate liposomes resulted in high mortality and incomplete depletion, consistent with reports that this method removes fewer than 5% of macrophages and has only transient effects⁷. Functional modulation using genetic approaches, rather than depletion, therefore provides a more physiologically relevant strategy to study macrophage function in vivo. This suggests that full macrophage removal can obscure reparative subpopulation dynamics and distort interpretation of immune contributions to regeneration. In addition, the duct ligation/deligation model primarily reflects obstructive injury and does not fully capture the epithelial damage or chronic immune dysregulation typical of irradiation or autoimmune disease (e.g., Sjögren’s syndrome). Accordingly, macrophage behaviour and Wnt/β-catenin signalling dynamics may differ in these contexts, warranting further investigation.

The results further carry translational relevance for inflammatory and fibrotic diseases of the salivary gland. Aberrant macrophage activation and persistent inflammation are key drivers of glandular dysfunction in conditions such as radiation-induced xerostomia and Sjögren’s syndrome. In both contexts, macrophage-derived cytokines including TNF-α, IL-1β, and TGF-β1 promote fibroblast activation and extracellular matrix deposition, contributing to chronic glandular fibrosis and acinar atrophy^37–39. Dysregulated Wnt/β-catenin signalling has been detected in chronically inflamed salivary tissues, where it correlates with macrophage accumulation, myofibroblast expansion, and collagen deposition^40–42. Beyond macrophages, our results also indicate that other GFP⁺ CD45⁺ cells may represent additional leukocyte populations, and further studies will be required to define their specific roles during injury and repair. In addition to cytokine-mediated effects, chemokine-driven recruitment can further amplify immune cell infiltration; notably, CCR1^hi- CCL5^hi macrophages promote CCL5^hi T-cell chemotaxis and subsequent lymphocytic infiltration in Sjögren’s syndrome⁴³. Wnt/β-catenin signalling may intersect with these chemokine circuits by shaping macrophage activation and macrophage-lymphocyte crosstalk, although whether it directly modulates CCR1/CCL5-linked recruitment or downstream immune programming remains to be determined. Targeting macrophage Wnt/β-catenin signalling may therefore provide a means to dampen persistent inflammation and restore the balance between injury and repair in the salivary gland. Future studies integrating macrophage-specific Wnt modulation with anti-inflammatory therapies could reveal combinatorial strategies for restoring salivary gland function and limiting chronic fibrotic remodelling.

Collectively, these results identify canonical Wnt/β-catenin signalling as an important regulator of macrophage behaviour in the injured salivary gland. We propose a model in which Wnt-active macrophage sustain local signalling that may promote or sustain aspects of inflammation and fibrosis. Disruption of this signalling through Wls depletion promotes the accumulation of CD206⁺ macrophages consistent with a more pro-reparative profile and reduces fibrotic tissue deposition. These findings support Wnt signalling as a central pathway linking macrophage phenotype to fibrotic outcomes in the salivary gland and highlight selective modulation of macrophage Wnt activity as a potential strategy to enhance regeneration and restore glandular function. Such macrophage-centred interventions may ultimately inform reshape therapeutic approaches for chronic salivary inflammation and fibrotic repair.

Summary of key findings

• SMG injury triggered a stromal/interstitial increase in Wnt/β-catenin signalling, with a substantial fraction of Wnt-responsive cells identified as F4/80⁺ macrophages.

• qPCR revealed increased expression of several Wnt ligands (including Wnt2 and Wnt2b) at days 3 and 6 post-injury, consistent with macrophages contributing to Wnt ligand production.

• Wnt activation peaked at day 6 post-injury, coinciding with macrophage accumulation and early fibrosis, suggesting a potential link between Wnt-active macrophages and the injury-to-repair transition.

• Genetic disruption of macrophage Wnt secretion was associated with increased CD206⁺ macrophages and reduced collagen deposition, supporting the idea that macrophage–Wnt/β-catenin signalling may shape fibrotic outcome.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Araz Ahmed: Conceptualisation, methodology (design and performance of experiments), formal analysis (data analysis), writing in the original draft, writing in review and editing, and final approval of the manuscript. Isabelle Miletich: Supervision, scientific guidance, support with data analysis, and writing in review and editing. Simon Whawell: Scientific discussion and writing in review and editing. Suveer Sachdeva: Writing in review and editing. All authors contributed to the manuscript, with the final version reviewed and approved by Araz Ahmed.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethics approval

All animal procedures were approved by the King’s College London Ethical Review Process and conducted in accordance with the UK Home Office Animals (Scientific Procedures) Act 1986 (Project Licence 70/7866) and institutional welfare guidelines. All animal procedures and reporting were performed in accordance with the ARRIVE guidelines (https://arriveguidelines.org).