Skip to main content
NCBI home page
Clinical & Translational Immunology logoLink to Clinical & Translational Immunology
. 2024 Nov 29;13(12):e70018. doi: 10.1002/cti2.70018

Squamous cell carcinoma is associated with reduced IL34 expression, alterations in the Langerhans cell antigen‐processing‐presentation machinery and poor patient survival

Thi Viet Trinh Dang 1, Kevin R Gillinder 1, Quan Nguyen 2,3, Onkar Mulay 2,3, Tuan Vo 2,3,4,5, Ahmed M Mehdi 1,6, Chenhao Zhou 1, Andrew J Brooks 1, Graham R Leggatt 1, David A Hume 7, Ian H Frazer 1, a , Janin Chandra 1, a ,✉
PMCID: PMC11605367  PMID: 39619016

Abstract

Objectives

Langerhans cells (LCs) are epithelial antigen‐presenting cells (APC) contributing to immune surveillance. LCs depend on interleukin 34 (IL34) production by epithelial cells. This study aimed to uncover mechanisms of alteration of IL34 and LC function in squamous cell carcinoma (SCC).

Methods

Cancer cohort data were used to identify associations between SCC and IL34. ATAC‐seq of keratinocytes (KCs) and LCs from a murine model of epithelial hyperplasia, driven by HPV16 E7 oncoprotein (K14E7), was analysed. Transcriptomic data were used to validate findings. RNAscope, RT‐qPCR, ELISA and confocal imaging was used to analyse IL34 expression and LCs in a spatial context.

Results

IL34 mRNA is downregulated in human SCCs of the head and neck, the cervix, the lung and the oesophagus, and low IL34 expression is associated with poor survival. We demonstrate that KCs of K14E7 mice have reduced Il34 gene accessibility, mRNA and protein, as well as broad changes in promotor accessibility associated with cell adhesion and immune responses. Chromatin accessibility was substantially changed in LCs, including increased accessibility of the Csf1r gene, and changes in promotors associated with cytoskeleton arrangement and antigen processing and presentation. We discovered altered spatial LC dendrite organisation in hyperproliferative epithelium.

Conclusion

Squamous cell carcinoma of the cervix, head and neck, oesophagus and lung demonstrate downregulation of IL34, which is associated with poor survival, and with alterations in LC spatial organisation and function. These findings suggest that reduced IL34 expression in SCC may contribute to impaired local immunity through LC dysregulation.

Keywords: epithelial hyperplasia, IL34, Langerhans cells, squamous cell carcinoma

IL34 is critical for the development and maintenance of Langerhans cells in squamous epithelium. This study shows that low expression of IL34 determines poor survival in squamous cell carcinomas. We demonstrate that the downregulation of IL34 transcript and protein in hyperproliferative squamous epithelial cells is because of the reduction in chromatin accessibility at the enhancer region of the Il34 gene. This epigenetic alteration is associated with substantial chromatin changes at genes pivotal to the antigen processing and presentation machinery of Langerhans cells.

graphic file with name CTI2-13-e70018-g006.jpg

Introduction

Epithelial hyperplasia is a common response to chronic epithelial inflammation, injury or infection 1 and may predispose to the development of cutaneous and mucosal squamous cell carcinoma (SCCs). 2 Hyperplastic epithelium is associated with alterations to epithelial cell gene expression, 3 , 4 through transcriptional regulation. 5 Understanding the molecular basis of epithelial hyperplasia may facilitate the development of targeted therapeutic strategies for epithelial cancer.

A transgenic mouse expressing the E7 oncoprotein of the high‐risk human papillomavirus (HPV) type 16 under the control of the keratin 14 promoter (K14E7) exhibits epithelial hyperplasia 6 reminiscent of HPV‐infected human epithelium 7 and of actinic keratosis. 8 The K14E7 epithelium contains an immune‐suppressive immune infiltrate, including regulatory T cells, mast cells and NKT cells. 9 , 10 , 11 , 12 , 13 , 14 The K14E7 mouse thus enables investigation of immune regulation in epithelial proliferative diseases. 15 Langerhans cells (LCs) are the only professional antigen‐presenting cells (APCs) in the cutaneous epithelium in steady state. 16 Our team reported that steady‐state murine LCs exist in different cell states, including phagocytic, proliferative, fully mature stimulatory and semi‐mature inhibitory cell states. 4 However, in K14E7 hyperproliferative epithelium, fully mature LC cell states are absent, while semi‐mature LCs are enriched. 4 This aberration in K14E7 LC transcriptional profiles corresponds with reduction in the expression of surface markers, including Langerin (CD207), CD11b, MHCII and EpCAM, and with elevated expression of immune‐modulatory enzymes and cytokines, including Ido1, Arg1 and Il6. 17 These cells exhibit a diminished capacity for antigen uptake both in vitro 17 and in vivo, 18 as well as in antigen processing, when compared to WT LCs. 19 Interestingly, culturing K14E7 LCs ex vivo restores their capacity to present antigens and prime T cells. 17 , 19 This suggests that the hyperplastic epithelium's microenvironment continuously provides signals that maintain LC dysfunction.

Identified as an alternative ligand for the colony‐stimulating factor 1 (CSF1), IL34 has been shown to be essential for the maintenance and development of LCs in mouse skin. 20 , 21 , 22 LC presence in the murine epidermis is strictly dependent on IL34 and CSF1R. In Il34 knockout mice, LCs are absent 20 , 22 and, similarly, they are absent in Csf1r‐deficient mice. 23 However, LCs are present in mice with an inactivating mutation in the Csf1 gene, indicating that LC presence is CSF1‐independent. 23 We have previously demonstrated the alteration of keratinocytes (KCs) in K14E7 skin at transcriptomic levels, 3 and predicted an absence of IL34‐CSF1R ligand receptor interaction. 4 To deepen our understanding of genetic regulation in epithelial hyperplasia, in this study, we assessed the chromatin accessibility landscape broadly and within the Il34 and Csf1r gene loci in KCs and LCs between normal and hyperplastic epithelium.

Results

Il34 mRNA expression is reduced in SCC, and is associated with poor survival

We previously showed that IL34 mRNA expression is downregulated in cutaneous and mucosal squamous epithelial hyperplasia, including high‐grade cervical intraepithelial neoplasia (CIN3), psoriasis and eczema. 4 To expand on these findings, we compared IL34 mRNA expression in normal and tumor tissues from patients with mucosal SCCs. Our analysis demonstrated a consistent pattern of reduced IL34 mRNA expression in SCC of the head and neck, cervix, lung and oesophagus compared to normal tissues (Figure 1a). Furthermore, we observed significantly reduced survival in patients exhibiting low IL34 expression in cervical SCC, in head and neck SCC (HNSCC) and in lung SCC (Figure 1b–d). Noteworthy, a similar survival trend was also observed in patients with oesophageal SCC, although the statistical power in this data set was limited (Figure 1e). In HNSCC, we further observed higher IL34 levels in HPV+ HNSCC than in HPV‐ HNSCC (Figure 1f), aligning with previous findings of increased immune activity in HPV+ HNSCC. 24 Overall, these findings are consistent with a role of IL34 in determining outcome of SCC.

Figure 1.

Figure 1

Open in a new tab

Mucosal SCC and K14E7 KCs express reduced levels of IL34. (a) TNM data. 50 Boxplots of IL34 gene expression in HNSCC (normal = 44, tumor = 500), cervical SCC and endocervical adenocarcinoma (normal = 3, tumor = 304), lung SCC (normal = 476, tumor = 501) and esophageal carcinoma (normal = 418, tumor = 161). Unpaired two‐tailed Mann–Whitney U‐test, *P < 0.05; ****P < 0.0001. (b–e) Kaplan–Meier 68 plots based on IL34 expression of HNSCC (b), cervical SCC (c), lung SCC (d) and oesophageal SCC (e) samples. (f) The comparison between IL34 expression and HPV status in HNSCC.

Epithelial hyperplasia is associated with reduced Il34 expression

Employing single‐cell RNA sequencing (scRNA‐seq) data from sorted KCs from a murine K14E7 transgenic model of epithelial hyperplasia, 3 we identified significantly lower Il34 mRNA (adjusted P = 2.84E−174) in K14E7 KCs than non‐transgenic wild‐type (WT) KCs (Figure 2a). Basal and differentiated WT‐derived KCs primarily produced Il34 transcripts, while proliferative KCs and all K14E7‐derived KC cell states exhibited lower contributions (Figure 2b and c). Only few KCs expressed Csf1 transcripts, with no significant difference between WT and K14E7 (Figure 2d), suggesting Csf1 does not compensate for the lack of Il34 in epithelial hyperplasia. We investigated other macrophage survival factors, such as GM‐CSF (CSF2), FLT3L or TGF‐β, with Csf2 and Flt3lg not detected, and Tgfb1 expressed at low, non‐differential levels in our existing data, which includes bulk RNA‐seq data 8 and scRNA‐seq data of epithelial cells from K14E7 and WT mice. 3 Reduction in Il34 mRNA was further validated using RT‐qPCR (Figure 2e). Correspondingly, decreased IL34 protein was observed using ELISA (Figure 2f). In situ hybridisation revealed higher expression of Il34 mRNA in the epidermis than in dermis, and reduced Il34 expression per cell in K14E7‐derived KCs (Figure 2g). Collectively, the results demonstrate that squamous epithelial hyperplasia is associated with decreased expression of epithelial Il34 mRNA and IL34 protein.

Figure 2.

Figure 2

Open in a new tab

K14E7 KCs express reduced levels of Il34. (a) scRNA‐seq data 3 was used to visualise (a) sample groups (left) and Il34 expression (right) using Nebulosa, (b) cell states as t‐SNE plot, (c) Il34 expression across K14E7 and WT cell states as violin plot and (d) Csf1 expression across K14E7 and WT cell states as violin plot. (e, f) Comparison of Il34 mRNA expression (normalised by Rps6, Rpl5, eIF3a) (e) and IL34 protein expression (f) between K14E7 and WT epithelial skin cell lysates (n = 6 combined from 2 independent experiments). Unpaired t‐test with Welch's correction. ****P < 0.0001. (g) Representative of Il34‐targeted in situ hybridisation using cross‐sections of ears. The dashed line indicates the epithelium. Scale bars represent 25 μm for the top images and 5 μm for the bottom images. Il34 expression levels per cell. Unpaired t‐test with Welch's correction. ***P < 0.001. All error bars represent standard deviation (SD).

The chromatin accessibility landscape in hyperproliferative KCs

The gene expression profile of K14E7 KCs is vastly different from WT KCs. 3 To investigate whether altered gene expression correlates with changes in chromatin accessibility, we performed bulk ATAC‐seq on sorted KCs from K14E7 and WT epithelium (Supplementary figure 1). After quality assessments, we obtained approximately 50 million reads per sample (Supplementary figure 2a). We established a consensus peak set comprising 71 998 peaks, where 52 432 were common to both WT and K14E7, 8782 were unique to WT, and 10 784 were exclusive to K14E7 (Figure 3a). Differential analysis revealed 34 018 differentially enriched sites (DES) between WT and K14E7 KCs, with 20 192 displaying reduced accessibility and 13 826 showing greater accessibility in K14E7 KCs (Figure 3b). To increase biological significance, we applied an absolute logFC cut‐off of 0.5, resulting in a reduction to 9242 DES. Approximately 50% of DES overlapped with gene bodies, while 40% were in intergenic regions (Supplementary figure 2b). The top five more accessible DES in K14E7 KCs were associated with Nppb, Kirrel3, Kcna4, Lrrc38 and Plpp4 gene loci, while the top five less accessible DES were linked to Kif6, 4930452A19Rik, Kcnt2, Kcnip4 and C6 gene loci (Supplementary figure 2c). Analysing the regulatory feature association revealed that K14E7 KCs had 767 DES with decreased promoter access and 631 DES with increased promoter access, and 195 DES with decreased enhancer access and 608 with increased enhancer access (Figure 3c). Gene ontology analysis suggested that less accessible enhancer and promoter regions were linked to immune regulation and cell adhesion processes (Figure 3d), while more accessible regions were associated with immune responses to biotic stimuli and antigen presentation (Figure 3e). Overall, these findings indicate significant changes in chromatin accessibility in KCs from epithelial hyperplasia with extensive alterations in access to promoter and enhancer regions, suggesting alterations in cell‐to‐cell organisation and an activated immune response.

Figure 3.

Figure 3

Open in a new tab

Chromatin accessibility is significantly changed in hyperplastic KCs and reduced at the Il34 gene locus. (a, b) Venn diagrams, generated in DiffBind, show the overlap of the enriched sites identified after defining the consensus peakset (a) and after differential enrichment analysis (b). Epithelial samples from four independent experiments, each pooling cells from 6 WT or 3 K14E7 mice. (c) The association of DES with enhancers and promoters. (d, e) Gene ontology analysis of less accessible DES (d) and more accessible DES (e) that were associated with enhancer and promoter regions. (f) IGV visualisation of ATAC‐seq peaks at the Il34 gene and DES within its locus spanning 50 kb upstream and downstream. (g) IGV visualisation of Il34_DES1 (logFC = 0.53122, Adj.P = 0.00235) and Il34_DES5 (logFC = −0.69091, Adj.P = 0.00003). The red bar highlights Il34_DES5 overlapping with an upstream enhancer of Il34 highlighted by the green bar.

The Il34 gene locus in hyperproliferative KCs is less accessible

The Il34 gene in mice has two transcription start sites (TSS) that are mainly active in the brain and skin (FANTOM 5 data 25 ). The brain‐specific TSS is at the first coding exon, while the skin‐specific TSS is further upstream at the first non‐coding exon, separated by a large 50 kb intron (Figure 3f). Conducting a DES analysis at the Il34 gene locus spanning 50 kb upstream and downstream, we identified nine DES, with seven significantly less accessible in K14E7 KCs compared to WT KCs (Table 1 and Figure 3f). From the distribution of these sites, we identified that Il34_DES1 and Il34_DES5 exhibited close proximity to the skin‐specific TSS. Notably, Il34_DES5 partially overlapped with the upstream enhancer of the Il34 gene (Figure 3g), implying perturbations in regulatory elements governing Il34 gene expression.

Table 1.

DES at Il34 gene locus in KCs extended by 50 kb upstream to 50 kb downstream

DES Coordinates Width LogFC P‐value Adj.P
Il34_DES1 a chr8:110 803 073–110 803 473 401 0.53122 0.00070 0.00235
Il34_DES2 chr8:110 780 123–110 780 523 401 0.40227 0.02109 0.04483
Il34_DES3 chr8:110 775 709–110 776 109 401 −0.81453 0.00272 0.00770
Il34_DES4 chr8:110 731 362–110 731 762 401 −0.77457 0.00003 0.00016
Il34_DES5 b chr8:110 808 291–110 808 691 401 −0.69091 0.00001 0.00003
Il34_DES6 chr8:110 777 822–110 778 222 401 −0.64581 0.00011 0.00048
Il34_DES7 chr8:110 846 865–110 847 265 401 −0.61925 0.00003 0.00016
Il34_DES8 chr8:110 721 241–110 721 641 401 −0.55986 0.00006 0.00026
Il34_DES9 chr8:110 713 090–110 713 490 401 −0.38404 0.00766 0.01880
Open in a new tab

Positive logFC means increased chromatin accessibility in K14E7 KCs, while negative logFC indicates decreased chromatin accessibility in K14E7 KCs.

a

Proximity to the skin‐specific Il34 TSS.

b

Proximity to the skin‐specific Il34 TSS and overlap with Il34 upstream enhancer.

The LC chromatin accessibility landscape

As introduced, we previously established the transcriptional, phenotypic and functional alterations of K14E7 LCs. 4 , 17 , 18 , 19 To investigate whether chromatin accessibility changes are associated with these alterations, we conducted bulk ATAC‐seq on sorted K14E7 and WT LCs (Supplementary figure 1). Post‐alignment and quality control, each sample yielded approximately 50 million reads (Supplementary figure 2a). We established a consensus peak set comprising 70 496 peaks, where 48 573 were common to both WT and K14E7, 10 526 were unique to WT, and 11 397 were exclusive to K14E7 (Figure 4a). Differential analysis revealed 26 290 DES between WT and K14E7 LCs, with 7168 displaying reduced accessibility and 19 122 showing greater accessibility in K14E7 LCs (Figure 4b). To focus on changes more likely to have biological consequences, we applied an absolute logFC cut‐off of 0.5, resulting in 7236 DES. Approximately 50% of DES intersected with gene body segments, 37.44% were in distal intergenic regions, and 11% with promoter regions (Supplementary figure 2d). The top five more accessible DES in K14E7 LCs were associated with Olfr810, Slc13a1, Bcl10, Olfr808 and Cmtm6 gene loci, while the top five less accessible DES were linked to Tnfaip2, Cd83, Zbtb10, Mir592 and Iqcn gene loci (Supplementary figure 2e). Analysing the distribution of these DES at regulatory regions, we observed a large number of K14E7‐derived LC‐enriched DES overlapped with enhancer and promoter regions (Figure 4c). Gene ontology analysis revealed that less accessible enhancer regions were dominantly involved with metabolic processes, while less accessible promoter regions were involved in cell adhesion, organisation of actin filaments, cytoskeleton structures and stress fibre assembly (Figure 4d). Conversely, more accessible enhancer and promoter regions in K14E7 LCs were associated with immune responses (Figure 4e). Overall, these findings demonstrate a proportionally increased chromatin accessibility of LCs in hyperproliferative epithelium, particularly notable in the pronounced enrichment at promoter and enhancer regions.

Figure 4.

Figure 4

Open in a new tab

Chromatin accessibility is significantly changed in LCs within hyperplastic epithelium and enhanced at Csf1r locus. (a, b) Venn diagrams, generated in DiffBind, show the overlap of the enriched sites identified after defining the consensus peakset (a) and after differential enrichment analysis (b). Sorted LC samples from four independent experiments, each pooling cells from six WT or three K14E7 mice. (c) The association of DES with enhancers and promoters. (d, e) Gene ontology analysis of less accessible DES (d) and more accessible DES (e) that were associated with enhancer and promoter regions. (f) IGV visualisation of Csf1r gene and DES within its locus spanning 50 kb upstream and downstream. (g) IGV visualisation of Csf1r_DES3 (logFC = 0.43334, Adj.P = 0.001611) and Csf1r_DES6 (logFC = 0.33054, Adj.P = 0.02786). The red bar highlights Csf1r_DES3 overlapping with FIRE enhancer highlighted by the blue bar. (h) scRNA‐seq data 4 was used to analyse the expression of Csf1r gene in K14E7 and WT LCs presented as a violin plot.

LCs resident to epithelial hyperplasia have increased chromatin accessibility and transcript of the Csf1r gene

We determined whether epithelial hyperplasia was associated with chromatin changes in the Csf1r gene locus. We identified six DES within the Csf1r locus spanning 50 kb upstream and downstream with all sites demonstrating increased accessibility in K14E7 LCs compared to WT (Table 2 and Figure 4f). Notably, the Csf1r_DES3 partially overlapped with a known Csf1r enhancer (Figure 4g), the fms‐intronic regulatory element (FIRE), which is critical for Csf1r transcription, and its deletion is associated with the loss of LCs. 26 Csf1r_DES6 overlapped with a macrophage‐specific TSS and the upstream promoter element (Figure 4g). Our scRNA‐seq data revealed a significant increase in Csf1r transcripts (adjusted P = 1.89E−96) in K14E7 LCs (Figure 4h). These findings demonstrate that the Csf1r gene in K14E7 LCs exhibits enhanced accessibility for transcription, resulting in increased Csf1r gene expression.

Table 2.

DES at Csf1r locus in LCs extended by 50 kb upstream to 50 kb downstream

DES Coordinates Width LogFC P‐value Adj.P
Csf1r_DES1 chr18:61 083 100–61 083 500 401 1.18853 1.569E−08 1.935E−07
Csf1r_DES2 chr18:61 085 308–61 085 708 401 1.21678 2.320E−07 2.284E−06
Csf1r_DES3 a chr18:61 108 443–61 108 843 401 0.43334 3.490E−04 1.611E−03
Csf1r_DES4 chr18:61 074 606–61 075 006 401 0.66097 5.574E−04 2.431E−03
Csf1r_DES5 chr18:61 069 779–61 070 179 401 0.35667 5.969E−03 1.885E−02
Csf1r_DES6 b chr18:61 105 563–61 105 963 401 0.33054 9.449E−03 2.786E−02
Open in a new tab

Positive logFC means increased chromatin accessibility in K14E7.

a

Overlap with FIRE enhancer.

b

Overlap with promotor region.

Spatial organisation of CSF1R‐expressing cells and of their dendrites in epithelial hyperplasia is disturbed

Promotor accessibility analysis suggested that K14E7 KCs had decreased promotor activity in cell adhesion, while K14E7 LCs showed decreased promotor activity in cytoskeleton organisation and actin filament assembly. This may alter the spatial LC organisation and dendrite formation in hyperproliferative epithelium. To study LC distribution and phenotype, we crossed K14E7 animals with CSF1R‐FRed reporter mice. 27 In whole mount images, CSF1R‐FRed+ cell numbers per area were similar in K14E7 and WT epithelium, but K14E7 CSF1R‐FRed+ cells appeared to have fewer horizontally oriented dendrites from a top view of the skin; however, an in‐focus image of LCs from the hyperplastic K14E7 epithelium was difficult to obtain (Figure 5a). Using cross‐sections and Cd207‐targeted in situ hybridisation, we observed that WT LCs were primarily situated in the basal layer of the epidermis, while K14E7 LCs were distributed throughout the hyperplastic epithelium, with mostly vertically oriented dendrites compared to the horizontal dendrites of WT LCs (Figure 5b). This was confirmed with 3D imaging of CSF1R‐FRed+ cells in K14E7 and WT epithelium (Figure 5c, Supplementary videos 1 and 2), indicating that LC dendrite formation is not impaired but is directed throughout the additional epithelial hyperplastic layers.

Figure 5.

Figure 5

Open in a new tab

Spatial organisation of CSF1R‐expressing cells and of their dendrites in the hyperproliferative epithelium. (a) Confocal imaging of CSF1R‐Fred K14E7 and CSF1R‐Fred WT skin. Arrowheads show positions of dendrites. Scale bars represent 20 μm. Epithelial CSF1R‐Fred+ cells were enumerated following confocal imaging; unpaired t‐test with Welch's correction; ns, nonsignificant; error bars represent SD. (b) Representative target Cd207 RNA expression at a single‐cell level using RNAscope on fresh frozen ear pinnae of K14E7 and WT mice. The dashed line indicates the epithelium. Arrowheads show positions of LCs. Scale bars represent 20 μm. (c) 3D images of CSF1R‐Fred K14E7 and CSF1R‐Fred WT skin. The xy surface (indicated by red and green arrows, respectively) indicates the outermost layers of the skin. Scale bars represent 20 μm.

The chromatin landscape of the antigen processing and presentation machinery is dysregulated in LCs within hyperplastic epithelium

The impairment of K14E7 LCs in antigen uptake, processing and presentation described previously 17 , 18 , 19 prompted us to investigate chromatin accessibility changes in genes involved in the antigen processing and presentation machinery (APPM). We established a comprehensive APPM gene list based on three existing sources (Supplementary table 1). 4 , 28 , 29 In antigen uptake, there were diverse changes with decreased accessibility at key genes involved in phagocytosis, including clathrin (Clt), scavenger receptor (Scarb2), Fc Receptor‐Like Protein 1 (Fcrl1) and Langerin (Cd207) (Figure 6a). Both increased and decreased accessibility were observed at MHCI‐related genes and non‐classical MHCII genes (Figure 6b). Intriguingly, substantially decreased accessibility was identified at classical MHCII genes, including H2‐Aa, H2‐Ab1, and H2‐Eb1 and Cd74 (Figure 6b–d). Notable numbers of co‐stimulatory genes (Cd86, Pvr, Timd4, Icam1), as well as pro‐inflammatory genes (Il1a, Il1b, Il23a), exhibited reduced accessibility (Figure 7a and b). Conversely, the anti‐inflammatory gene Il10 and a number of co‐inhibition genes, including Cd247 (PD‐L1) and Pdcd1lg2 (PD‐L2), showed increased accessibility, suggesting a more tolerogenic phenotype (Figure 7b and c). Interestingly, substantially increased accessibility was observed at genes involved in antigen processing and migration (Figure 7d and e), correlating with elevated mRNA levels (Supplementary table 1). Diverse changes were also observed at some peptide‐loading genes (Figure 7f).

Figure 6.

Figure 6

Open in a new tab

Chromatin accessibility is significantly decreased at genes involved in antigen presentation in K14E7 LCs. (a, b) DES of genes involved in the antigen uptake (a) and the antigen presentation pathway (b) were plotted based on logFC, with corresponding adjusted P‐values provided on the right. Negative logFC values indicate reduced accessibility in K14E7 LCs, and positive values signify increased accessibility. (c, d) IGV visualisation of ATAC‐seq peaks at classical MHCII gene loci (c) and Cd74 (d). Green bars indicate identified DES.

Figure 7.

Figure 7

Open in a new tab

Chromatin accessibility alterations at genes involved in the antigen processing and presentation machinery. IGV visualisation of ATAC‐seq peaks at genes involved in co‐stimulation (a), cytokine production (b), co‐inhibition (c), processing (d), migration (e) and peptide loading (f).

Collectively, these data indicate that APPM function is dysregulated in LCs of hyperproliferative epithelium and that the previously observed dysfunction in antigen processing and presentation originates at a dysregulation at chromatin level, and is consistent with alterations in gene expression.

Discussion

Here, we show that reduced IL34 gene expression is a hallmark of multiple SCC, including cervical, head and neck, lung and oesophageal SCC, and is linked to poor survival. In a mouse model of epithelial hyperplasia, we show that reduced IL34 expression correlated with alterations in chromatin accessibility at the Il34 locus, and loss of biological processes referring to cell adhesion. This was further associated with substantial chromatin changes in LCs, affecting morphology and antigen processing and presentation.

IL34 is abundantly expressed in skin and brain tissues, 30 playing pivotal roles in LC and microglia development, 20 , 22 as well as in cellular processes including proliferation, differentiation, migration and cytokine/chemokine production. 31 However, its involvement in diseases ranges widely from inflammation and autoimmunity to cancer, 32 with varying impacts depending on the disease context. Elevated IL34 levels have been observed in several solid cancers, including breast, 33 colorectal, 34 lung, 35 , 36 ovarian 37 and hepatocellular carcinoma, 38 often correlating with poor prognosis and tumor progression. 38 , 39 , 40 Contrary to these findings, our study showed a consistent pattern of decreased IL34 expression in mucosal SCC such as cervical, head and neck, lung and oesophageal SCC, suggesting that low IL34 levels are a common characteristic in these conditions. This is aligned with our previous observations with reduced IL34 levels in epithelial hyperplasia conditions such as CIN3, psoriasis and eczema. 4 Furthermore, in a spatial context, we have recently described that IL34‐CSF1R co‐expression is specifically absent from the neoplastic region in CIN3 samples, while the adjacent normoplastic tissue regions present normal IL34‐CSF1R co‐expression. 41 Importantly, we here highlight the clinical significance of low IL34 expression, as it correlates with poor survival in cervical cancer patients, 4 in HNSCC and in lung SCC. Availability of large‐scale gene expression and clinical data from other SCC patient cohorts including cutaneous, esophaegeal, anal, vulval, penile and vaginal SCCs will in future enable to establish whether IL34 plays a similar role across these different disease entities.

Human papillomavirus is widely recognised as the primary cause of cervical SCC, and between 38% and 80% of HNSCC cases test positive for HPV. 42 , 43 , 44 Here, we showed that HPV+ HNSCC exhibits higher IL34 expression levels than HPV− HNSCC. Our recent work revealed that HPV+ HNSCC exhibits a heightened immune activity score, 24 and research has consistently shown that HPV+ HNSCC typically features a ‘hot’ tumor immune microenvironment, resulting in improved treatment responses and clinical outcomes. 44 , 45 , 46 Collectively, this suggests that IL34 may play a beneficial role in mucosal squamous epithelial cancers, and skin lesions associated with epithelial hyperplasia.

We utilised the K14E7 transgenic mouse model, which demonstrates epithelial hyperplasia resembling a thickened epithelium and gene expression similar to HPV‐driven human neoplastic tissue, 7 to investigate the genetic mechanisms underlying IL34 downregulation in hyperplastic epithelium. Our study uncovered a corresponding decrease in Il34 mRNA levels and protein expression. ATAC‐seq analysis of KCs revealed significantly reduced chromatin accessibility at multiple sites within the Il34 locus, including regions near the skin‐specific TSS and overlap with an upstream enhancer of Il34, indicating reduced transcriptional accessibility of Il34 in hyperplastic epithelium. Although reduction in Il34 chromatin accessibility was notable in K14E7 KCs, they were less pronounced compared with the reductions observed in mRNA and protein expression, suggesting that factors beyond chromatin accessibility may contribute to suppressing IL34 expression in hyperproliferative epithelium. We explored whether other macrophage survival factors and growth factors such as CSF2, FLT3L, TGF‐β may compensate for the lack of IL34, but gene transcripts were either not present or not differentially expressed in hyperproliferative epithelium. Intriguingly, we observed significantly increased chromatin accessibility at the Csf1r locus in LCs within hyperplastic epithelium, correlating with elevated transcript levels. CSF1R, the primary receptor of IL34, is expressed exclusively on mononuclear phagocyte cells, 47 including LCs. In response to constant ligand stimuli, the receptor is typically downregulated to prevent excessive cellular reactions to the ligand. 48 This implies that LCs in hyperplastic epithelium may have an increased CSF1R expression because of the reduced availability of IL34, hence not triggering ligand‐mediated downregulation of the receptor. Interestingly though, while genetic depletion of IL34 leads to a complete absence of LCs, 20 , 22 reduced IL34 in hyperplastic epithelium did not reduce LC numbers, suggesting that either sufficient IL34 remains in the skin to retain LCs or other factors contribute to LC survival in this environment.

Previously, we demonstrated reduced antigen uptake 17 , 18 and impaired antigen processing capability in LCs within hyperplastic epithelium, 19 associated with decreased expression of classical MHCII genes including H2‐Aa, H2‐Ab1, H2‐Eb1 and Cd74. 4 In line with this, we observed reduced chromatin accessibility at classical MHCII and critical phagocytosis genes. Importantly though, we observed both increased and decreased chromatin accessibilities across different antigen processing and presentation pathways, suggesting that LCs resident to epithelial hyperplasia are not simply shut down in the overall antigen processing and presentation machinery, but dysregulated in a complex way. Additionally, consistent with our previous findings, K14E7 LCs exhibited reduced horizontal dendritic extension and a rounded shape from a top view of the skin, but we discovered that dendrite formation was not impaired, but rather oriented vertically through the additional epithelial cell layers. 17

In conclusion, our research highlights that IL34 downregulation is a common occurrence in diseases associated with epithelial dysplasia, including inflammatory skin diseases and various SCCs such as lung, head and neck, cervical and oesophageal SCC. This aligns with our previous findings of reduced IL34 transcripts in CIN3, psoriasis and eczema. Moreover, decreased IL34 levels correlate with poorer survival rates in patients with cervical, lung and head and neck SCC, underscoring its clinical significance. This study provides a comprehensive molecular portrait of the impact of epithelial hyperplasia on LCs, highlighting significant shifts in antigen processing and presentation pathways. These observations contribute to understanding the complex immune dysregulation within the microenvironment of hyperplastic epithelium. Future research is required to provide a cause‐and‐effect link between reduces IL34 production and dysregulated LC function, and whether this signalling axis can be therapeutically manipulated to increase APC function and T cell priming in SCC.

Methods

Mice

Eight‐ to twelve‐week‐old WT, K14E7, 49 CSF1R‐Fred/WT 27 and CSF1R‐Fred/K14E7 transgenic female mice were used for all experiments. Used littermates were bred either between K14E7 male mice with C57BL/6 females or between CSF1R‐Fred/WT mice with K14E7 mice. All animal experiments were approved by the University of Queensland Animal Ethics Committee (2021/AE000460 and 2023/AE000617). They were maintained under specific pathogen‐free conditions at the Translational Research Institute's Biological Research Facility.

RNA expression of IL34 in human tumors

Gene expression of IL34 in tumor and normal tissues of each cancer set was obtained from the TNMplot database. 50 Plots were generated using Graphpad Prism 10.

Kaplan–Meier survival analysis

Survival plots were generated using the Kaplan–Meier Plotter 51 (kmplot.com) and the following settings: survival (OS), auto selection for best cut‐off (all), follow‐up threshold (60 months). ‘All’ was selected for all other setting options.

HPV status vs Il34 expression

Gene expression data from 466 primary tumors of the HNSCC cohort were sourced from Xenabrowser. 52 The relationship between IL34 expression and HPV status was examined in 111 subjects where HPV status was available. The P‐value was calculated using the Wilcoxon rank‐sum test because of non‐normal data distribution.

scRNA‐seq analysis

Previously acquired scRNA‐seq data of sorted KCs and LCs from WT and K14E7 mice were used. 3 , 4 Cell clusters were manually reannotated based on genetic markers. Differential gene expression analysis was performed using the Seurat package 53 with significant genes (adjusted P < 0.001) reviewed for their differentiation cell states and biological relevance. tSNE and gene expression plots were generated using Seurat.

Preparation of epithelial single‐cell suspensions

Mice were euthanised by CO2 asphyxiation, and ear pinnae were harvested and split into dorsal and ventral halves. These halves were digested in PBS with Dispase II (Cat. #04942078001, Roche, Basel, Switzerland) at 37°C for 1 h, and the epidermis was separated from the dermis. The epidermis was chopped into small pieces, further digested with collagenase D (Cat. #11088858001, Roche, Basel, Switzerland) and DNase (Cat. #11284932001, Roche, Basel, Switzerland) in PBS at 37°C for 30 min and filtered through a 70‐μm cell strainer. The resulting single‐cell suspension was used for RNA and protein extraction, sorting and flow cytometry.

RT‐qPCR

RNA from epithelial single‐cell suspensions was extracted using the RNeasy Mini kit (Cat. #74104, QIAGEN, Germantown, USA) and DNase‐treated with the TURBO DNA‐free™ Kit (Cat. #AM1907, Thermo Fisher Scientific, Waltham, USA). RNA was reverse‐transcribed into cDNA using Superscript III RT (Cat. #18080093, Thermo Fisher Scientific, Waltham, USA) with the following conditions: 25°C for 5 min, 50°C for 90 min and 70°C for 15 min. Quantitative PCR (qPCR) was performed using the QuantStudio™ 7 Flex Real‐Time PCR System (Thermo Fisher Scientific, Waltham, USA) with specific primers (Supplementary table 2).

ELISA

IL34 expression in the epithelium was measured from total protein from lysed epithelial cells using the LEGEND MAX™ Mouse IL34 ELISA Kit (Cat. #439107, BioLegend, San Diego, USA) following the manufacturer's instructions. Absorbances were measured using a Multiskan GO (Thermo Fisher Scientific, Waltham, USA).

RNA in situ hybridisation

OCT‐embedded fresh frozen skin tissues were sectioned at 10 μm thickness. Target probes (Cat. #ADV1222201T3 and #ADV452521T2, Advanced Cell Diagnostics, Newark, USA) were used with the RNAscope™ HiPlex12 Reagent Kit v2 Standard Assay Manual. Slides were fixed in 4% paraformaldehyde for 1 h at room temperature, washed with PBS and dehydrated with ethanol (50%, 70% and 100% for 5 min each). Protease IV (Cat. #322340, Advanced Cell Diagnostics, Newark, USA) was applied for 30 min at room temperature, followed by incubation with the probe mix at 40°C for 2 h. Probes were amplified and fluorescently labelled using the Detection‐RNAscope HiPlex 12 Reagents Kit (Cat. #324409, Advanced Cell Diagnostics, Newark, USA). Slides were washed, stained with DAPI and mounted with ProLong Gold Antifade Mountant (Cat. #P36930, Thermo Fisher Scientific, Waltham, USA). Imaging was performed on the LSM900 Airyscan confocal microscope, and images were stitched and adjusted using the ZEN software (version 3.2). Signals from different rounds were merged using the HiPlex v2—Image Registration Software. The cell segmentation was performed using the StarDist 54 in QuPath 55 and the average pixel intensity of each targeted mRNA probe signal per cell was quantified.

Cell sorting

Epithelial single‐cell suspensions were stained with 300 μL of anti‐mouse CD16/CD32 antibody at a 1/100 dilution (Cat. #553141, BD Biosciences, New Jersey, USA) and Fixable Aqua Dead Cell Stain at a 1/200 dilution (Cat. #L34965, Thermo Fisher Scientific, Waltham, USA) for 10 min on ice in the dark. Samples were washed with PBS and further incubated for 20 min with 300 μL antibody cocktail including anti‐mouse CD45‐PerCP‐Cy5.5 at a 1/400 dilution (Cat. #103131, Biolegend, San Diego, USA), EpCAM‐PE at a 1/400 dilution (Cat. 118205, Biolegend, San Diego, USA) and MHCII‐APC‐Cy7 at a 1/400 dilution (Cat. #107627, Biolegend, San Diego, USA). Samples were washed with PBS and resuspended in PBS with 2% FBS and 5 mM EDTA. For sorting, live CD45− cells (KCs) and live CD45+ MHCII+ EpCam+ cells (LCs) were sorted into 100% FBS using the MoFlo Astrios Cell Sorter (Beckman Coulter, Brea, USA).

ATAC‐seq library preparation and sequencing

Bulk ATAC‐seq was performed on 16 samples, comprising four cell types: K14E7_KC, WT_KC, K14E7_LC and WT_LC, sorted from four independent experiments. Each experiment used pools of epidermal single‐cell suspensions from six WT or three K14E7 mice. 50 000 LCs and 50 000 KCs from WT and K14E7 mice were processed for ATAC‐seq. Cell pellets were washed with ice‐cold PBS, lysed in cold lysis buffer (1 M Tris–HCl pH 7.4, 5 M NaCl, 1 M MgCl2, 10% NP‐40, 10% Tween‐20, nuclease‐free water), centrifuged and washed again. Nuclei were resuspended in transposition buffer (25 μL Tagment DNA (Cat #FC‐121‐1030, Illumina, San Diego, USA), 2.5 μL Tn5 Transposase (Cat #FC‐121‐1030, Illumina, San Diego, USA), 22.5 μL nuclease‐free water) and fragmented at 37°C for 30 min. Fragmented chromatin was purified and amplified using PCR, then further purified and eluted in nuclease‐free water. Library quality was assessed using the High Sensitivity D1000 ScreenTape Assay, with successful libraries having quantities above 10 ng μL‐1. SPRI select was performed to exclude fragments outside the 100 bp to 1 kb range. Pair‐end sequencing was done using NGS, aiming for 50+ base pairs per read and approximately 50 million reads per sample.

ATAC‐seq analysis

Raw sequences were qualified using FastQC. 56 Low‐quality and adapter bases were removed using fastp 57 and mapped to the mouse genome GRCm38 (mm10) using Bowtie2. 58 BAM files were generated by using SAMtools 59 and Picard. 60 Peaksets were derived by using MACS3 peak caller. 61 Consensus peaksets were established by restricting analysis to peaks identified consistently in all replicates. Differential analysis was executed by default using DESeq2 (ref. 59) interrogated in DiffBind package. 62 BEDtools 63 was used to define the association of DES with known mouse promoters and enhancers from the Fantom database. 25 , 64 , 65 Annotation of DES was performed using ChIPseeker package. 66 Genomic Regions Enrichment of Annotations Tool (GREAT) 67 was used to identify biological processes.

Confocal microscopy

Direct imaging of whole ears was conducted with minimal disturbance through the ear surface. Images were captured using the Olympus FV3000 microscope and FLUOVIEW software. Epithelial CSF1R‐Fred+ cells were enumerated using CellProfiler. 68 3D images were processed using Imaris (Bitplane, Zurich, Switzerland).

Statistics

Statistical analyses were performed using GraphPad Prism 11. Differences between two groups were assessed with an unpaired two‐tailed t‐test or a Mann–Whitney U‐test. Results were deemed significant at *P < 0.05, **P < 0.01, ***P < 0.001 and****P < 0.0001.

Author contributions

Thi Viet Trinh Dang: Conceptualization; data curation; formal analysis; investigation; visualization; writing – original draft; writing – review and editing. Kevin R Gillinder: Conceptualization; data curation; formal analysis; investigation; supervision; writing – review and editing. Quan Nguyen: Conceptualization; writing – review and editing. Onkar Mulay: Formal analysis; writing – review and editing. Tuan Vo: Investigation; writing – review and editing. Ahmed M Mehdi: Formal analysis; writing – review and editing. Chenhao Zhou: Formal analysis; writing – review and editing. Andrew J Brooks: Supervision; writing – review and editing. Graham R Leggatt: Supervision; writing – review and editing. David A Hume: Conceptualization; supervision; writing – review and editing. Ian H Frazer: Conceptualization; funding acquisition; supervision; writing – original draft; writing – review and editing. Janin Chandra: Conceptualization; funding acquisition; investigation; supervision; writing – original draft; writing – review and editing.

Conflict of interest

The authors declare no conflict of interest.

Supporting information

Supplementary figure 1

Supplementary figure 2

Supplementary table 1

Supplementary table 2

CTI2-13-e70018-s002.pdf (1.3MB, pdf)

Video S1

Download video file (100.4MB, mp4)

Video S2

Download video file (112.4MB, mp4)

Acknowledgments

This study was carried out at the Translational Research Institute (TRI), Woolloongabba, Queensland 4102, Australia, The Translational Research Institute is supported by a grant by the Australian Government. We acknowledge TRI and the University of Queensland's School of Biomedical Sciences for providing cutting‐edge Core Facilities including Flow Cytometry, Microscopy, Histology and Biological Resource Facility (BRF) that enabled this research. We particularly thank Lynn Tolley and the staff of TRI BRF for the excellent care of animals and assistance with animal experimentation; Natalie Molotkov for support with histology and the staff of the Flow Cytometry, Microscopy, Histology facilities for their technical assistance. IHF is supported by the National Health and Medical Research Council of Australia (APP1173927), the Merchant Foundation and the Metal Manufacturing Pty Ltd. JC is supported by a Mid‐Career Fellowship by the Passe & Williams Foundation and by the Zelman Cowan Academic Initiative. Open access publishing facilitated by The University of Queensland, as part of the Wiley ‐ The University of Queensland agreement via the Council of Australian University Librarians.

Data availability statement

The ATAC‐seq data generated in this study have been submitted to the Gene Expression Omnibus (GEO) databases under accession number GSE268544. The URL is as follows: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE268544.

References

  • 1. Neville B, Damm D, Allen C, Chi A. Chapter 10 – Epithelial Pathology. Amsterdam: Elsevier; 2019. [Google Scholar]
  • 2. Speight PM. Update on oral epithelial dysplasia and progression to cancer. Head Neck Pathol 2007; 1: 61–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Lukowski SW, Tuong ZK, Noske K et al. Detection of HPV E7 transcription at single‐cell resolution in epidermis. J Invest Dermatol 2018; 138: 2558–2567. [DOI] [PubMed] [Google Scholar]
  • 4. Tuong ZK, Lukowski SW, Nguyen QH et al. A model of impaired Langerhans cell maturation associated with HPV induced epithelial hyperplasia. iScience 2021; 24: 103326. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Jin L, Chen Y, Muzaffar S et al. Epigenetic switch reshapes epithelial progenitor cell signatures and drives inflammatory pathogenesis in hidradenitis suppurativa. Proc Natl Acad Sci USA 2023; 120: e2315096120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Griep AE, Lambert PF. Role of papillomavirus oncogenes in human cervical cancer: Transgenic animal studies. Proc Soc Exp Biol Med 1994; 206: 24–34. [DOI] [PubMed] [Google Scholar]
  • 7. Tuong ZK, Noske K, Kuo P, Bashaw AA, Teoh SM, Frazer IH. Murine HPV16 E7‐expressing transgenic skin effectively emulates the cellular and molecular features of human high‐grade squamous intraepithelial lesions. Papillomavirus Res 2018; 5: 6–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Zhussupbekova S, Sinha R, Kuo P, Lambert PF, Frazer IH, Tuong ZK. A mouse model of hyperproliferative human epithelium validated by keratin profiling shows an aberrant cytoskeletal response to injury. EBioMedicine 2016; 9: 314–323. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Bergot AS, Ford N, Leggatt GR, Wells JW, Frazer IH, Grimbaldeston MA. HPV16‐E7 expression in squamous epithelium creates a local immune suppressive environment via CCL2‐ and CCL5‐mediated recruitment of mast cells. PLoS Pathog 2014; 10: e1004466. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Gosmann C, Frazer IH, Mattarollo SR, Blumenthal A. IL‐18, but not IL‐12, induces production of IFN‐gamma in the immunosuppressive environment of HPV16 E7 transgenic hyperplastic skin. J Invest Dermatol 2014; 134: 2562–2569. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Gosmann C, Mattarollo SR, Bridge JA, Frazer IH, Blumenthal A. IL‐17 suppresses immune effector functions in human papillomavirus‐associated epithelial hyperplasia. J Immunol 2014; 193: 2248–2257. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Mattarollo SR, Rahimpour A, Choyce A, Godfrey DI, Leggatt GR, Frazer IH. Invariant NKT cells in hyperplastic skin induce a local immune suppressive environment by IFN‐gamma production. J Immunol 2010; 184: 1242–1250. [DOI] [PubMed] [Google Scholar]
  • 13. Mittal D, Kassianos AJ, Tran LS et al. Indoleamine 2,3‐dioxygenase activity contributes to local immune suppression in the skin expressing human papillomavirus oncoprotein e7. J Invest Dermatol 2013; 133: 2686–2694. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14. Zhou C, Tuong ZK, Lukowski SW, Chandra J, Frazer IH. Antigen nonspecific induction of distinct regulatory T cell states in oncogene‐driven hyperproliferative skin. Immunohorizons 2021; 5: 102–116. [DOI] [PubMed] [Google Scholar]
  • 15. Chandra J, Dutton JL, Li B et al. DNA vaccine encoding HPV16 oncogenes E6 and E7 induces potent cell‐mediated and humoral immunity which protects in tumor challenge and drives E7‐expressing skin graft rejection. J Immunother 2017; 40: 62–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. Otsuka M, Egawa G, Kabashima K. Uncovering the mysteries of Langerhans cells, inflammatory dendritic epidermal cells, and monocyte‐derived Langerhans cell‐like cells in the epidermis. Front Immunol 2018; 9: 1768. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Abd Warif NM, Stoitzner P, Leggatt GR, Mattarollo SR, Frazer IH, Hibma MH. Langerhans cell homeostasis and activation is altered in hyperplastic human papillomavirus type 16 E7 expressing epidermis. PLoS One 2015; 10: e0127155. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Bashaw AA, Teoh SM, Tuong ZK, Leggatt GR, Frazer IH, Chandra J. HPV16 E7‐driven epithelial hyperplasia promotes impaired antigen presentation and regulatory T‐cell development. J Invest Dermatol 2019; 139: 2467–2476.e2463. [DOI] [PubMed] [Google Scholar]
  • 19. Chandra J, Miao Y, Romoff N, Frazer IH. Epithelium expressing the E7 oncoprotein of HPV16 attracts immune‐modulatory dendritic cells to the skin and suppresses their antigen‐processing capacity. PLoS One 2016; 11: e0152886. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20. Greter M, Lelios I, Pelczar P et al. Stroma‐derived interleukin‐34 controls the development and maintenance of langerhans cells and the maintenance of microglia. Immunity 2012; 37: 1050–1060. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21. Wang Y, Bugatti M, Ulland TK, Vermi W, Gilfillan S, Colonna M. Nonredundant roles of keratinocyte‐derived IL‐34 and neutrophil‐derived CSF1 in Langerhans cell renewal in the steady state and during inflammation. Eur J Immunol 2016; 46: 552–559. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Wang Y, Szretter KJ, Vermi W et al. IL‐34 is a tissue‐restricted ligand of CSF1R required for the development of Langerhans cells and microglia. Nat Immunol 2012; 13: 753–760. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Ginhoux F, Tacke F, Angeli V et al. Langerhans cells arise from monocytes in vivo. Nat Immunol 2006; 7: 265–273. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Mehdi AM, Zhou C, Turrell G et al. HPV status represents dominant trait driving delineation of survival‐associated gene co‐expression networks in head and neck cancer. Cancer Gene Ther 2023; 30: 629–640. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Severin J, Agrawal S, Ramilowski JA et al. ZENBU‐reports: A graphical web‐portal builder for interactive visualization and dissemination of genome‐scale data. NAR Genom Bioinform 2023; 5: lqad075. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Rojo R, Raper A, Ozdemir DD et al. Deletion of a Csf1r enhancer selectively impacts CSF1R expression and development of tissue macrophage populations. Nat Commun 2019; 10: 3215. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Grabert K, Sehgal A, Irvine KM et al. A transgenic line that reports CSF1R protein expression provides a definitive marker for the mouse mononuclear phagocyte system. J Immunol 2020; 205: 3154–3166. [DOI] [PubMed] [Google Scholar]
  • 28. Blees A, Januliene D, Hofmann T et al. Structure of the human MHC‐I peptide‐loading complex. Nature 2017; 551: 525–528. [DOI] [PubMed] [Google Scholar]
  • 29. Santambrogio L, Berendam SJ, Engelhard VH. The antigen processing and presentation machinery in lymphatic endothelial cells. Front Immunol 2019; 10: 1033. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30. Lin H, Lee E, Hestir K et al. Discovery of a cytokine and its receptor by functional screening of the extracellular proteome. Science 2008; 320: 807–811. [DOI] [PubMed] [Google Scholar]
  • 31. Baghdadi M, Umeyama Y, Hama N et al. Interleukin‐34, a comprehensive review. J Leukoc Biol 2018; 104: 931–951. [DOI] [PubMed] [Google Scholar]
  • 32. Lelios I, Cansever D, Utz SG, Mildenberger W, Stifter SA, Greter M. Emerging roles of IL‐34 in health and disease. J Exp Med 2020; 217: e20190290. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33. Kajihara N, Kitagawa F, Kobayashi T, Wada H, Otsuka R, Seino KI. Interleukin‐34 contributes to poor prognosis in triple‐negative breast cancer. Breast Cancer 2020; 27: 1198–1204. [DOI] [PubMed] [Google Scholar]
  • 34. Kobayashi T, Baghdadi M, Han N et al. Prognostic value of IL‐34 in colorectal cancer patients. Immunol Med 2019; 42: 169–175. [DOI] [PubMed] [Google Scholar]
  • 35. Baghdadi M, Endo H, Takano A et al. High co‐expression of IL‐34 and M‐CSF correlates with tumor progression and poor survival in lung cancers. Sci Rep 2018; 8: 418. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. Baghdadi M, Wada H, Nakanishi S et al. Chemotherapy‐induced IL34 enhances immunosuppression by tumor‐associated macrophages and mediates survival of chemoresistant lung cancer cells. Cancer Res 2016; 76: 6030–6042. [DOI] [PubMed] [Google Scholar]
  • 37. Endo H, Hama N, Baghdadi M et al. Interleukin‐34 expression in ovarian cancer: A possible correlation with disease progression. Int Immunol 2020; 32: 175–186. [DOI] [PubMed] [Google Scholar]
  • 38. Zhou SL, Hu ZQ, Zhou ZJ et al. miR‐28‐5p‐IL‐34‐macrophage feedback loop modulates hepatocellular carcinoma metastasis. Hepatology 2016; 63: 1560–1575. [DOI] [PubMed] [Google Scholar]
  • 39. Chen T, Wang X, Guo L et al. Embryonic stem cells promoting macrophage survival and function are crucial for teratoma development. Front Immunol 2014; 5: 275. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40. Kobayashi H, Gieniec KA, Lannagan TRM et al. The origin and contribution of cancer‐associated fibroblasts in colorectal carcinogenesis. Gastroenterology 2022; 162: 890–906. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41. Pavilion G, Vu H, Xiong Z et al. Spatial analysis of HPV associated cervical intraepithelial neoplastic tissues demonstrate distinct immune signatures associated with cervical cancer progression. bioRxiv. 2024. 10.1101/2024.10.02.611623 [DOI]
  • 42. Boscolo‐Rizzo P, Furlan C, Lupato V, Polesel J, Fratta E. Novel insights into epigenetic drivers of oropharyngeal squamous cell carcinoma: Role of HPV and lifestyle factors. Clin Epigenetics 2017; 9: 124. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43. Dong H, Shu X, Xu Q et al. Current status of human papillomavirus‐related head and neck cancer: From viral genome to patient care. Virol Sin 2021; 36: 1284–1302. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44. Leemans CR, Snijders PJF, Brakenhoff RH. The molecular landscape of head and neck cancer. Nat Rev Cancer 2018; 18: 269–282. [DOI] [PubMed] [Google Scholar]
  • 45. Gameiro SF, Evans AM, Mymryk JS. The tumor immune microenvironments of HPV(+) and HPV(−) head and neck cancers. WIREs Mech Dis 2022; 14: e1539. [DOI] [PubMed] [Google Scholar]
  • 46. Kang H, Kiess A, Chung CH. Emerging biomarkers in head and neck cancer in the era of genomics. Nat Rev Clin Oncol 2015; 12: 11–26. [DOI] [PubMed] [Google Scholar]
  • 47. Hume DA, Gutowska‐Ding MW, Garcia‐Morales C et al. Functional evolution of the colony‐stimulating factor 1 receptor (CSF1R) and its ligands in birds. J Leukoc Biol 2020; 107: 237–250. [DOI] [PubMed] [Google Scholar]
  • 48. Shankaran H, Wiley HS, Resat H. Receptor downregulation and desensitization enhance the information processing ability of signalling receptors. BMC Syst Biol 2007; 1: 48. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49. Herber R, Liem A, Pitot H, Lambert PF. Squamous epithelial hyperplasia and carcinoma in mice transgenic for the human papillomavirus type 16 E7 oncogene. J Virol 1996; 70: 1873–1881. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50. Bartha A, Gyorffy B. TNMplot.com: A web tool for the comparison of gene expression in Normal, tumor and metastatic tissues. Int J Mol Sci 2021; 22: 2622. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51. Győrffy B. Integrated analysis of public datasets for the discovery and validation of survival‐associated genes in solid tumors. The Innovation 2024; 5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52. Goldman MJ, Craft B, Hastie M et al. Visualizing and interpreting cancer genomics data via the Xena platform. Nat Biotechnol 2020; 38: 675–678. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53. Hao Y, Hao S, Andersen‐Nissen E et al. Integrated analysis of multimodal single‐cell data. Cell 2021; 184: e3529. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54. Schmidt U, Weigert M, Broaddus C, Myers G. Cell detection with star‐convex polygons. Med Image Comput Comput Assist Interv 2018; 11071: 265–273. [Google Scholar]
  • 55. Bankhead P, Loughrey MB, Fernandez JA et al. QuPath: Open source software for digital pathology image analysis. Sci Rep 2017; 7: 16878. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56. Andrews S. FastQC. Available from: https://www.bioinformatics.babraham.ac.uk/projects/fastqc/
  • 57. Chen S, Zhou Y, Chen Y, Gu J. fastp: An ultra‐fast all‐in‐one FASTQ preprocessor. Bioinformatics 2018; 34: i884–i890. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58. Langmead B, Salzberg SL. Fast gapped‐read alignment with bowtie 2. Nat Methods 2012; 9: 357–359. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59. Li H, Handsaker B, Wysoker A et al. The sequence alignment/map format and SAMtools. Bioinformatics 2009; 25: 2078–2079. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60. Broad‐Institute . Picard. Available from: https://broadinstitute.github.io/picard/
  • 61. Zhang Y, Liu T, Meyer CA et al. Model‐based analysis of ChIP‐seq (MACS). Genome Biol 2008; 9: R137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62. Ross‐Innes CS, Stark R, Teschendorff AE et al. Differential oestrogen receptor binding is associated with clinical outcome in breast cancer. Nature 2012; 481: 389–393. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 63. Quinlan AR, Hall IM. BEDTools: A flexible suite of utilities for comparing genomic features. Bioinformatics 2010; 26: 841–842. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 64. FANTOM Consortium and the RIKEN PMI and CLST (DGT) , Forrest AR, Kawaji H et al. A promoter‐level mammalian expression atlas. Nature 2014; 507: 462–470. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 65. Severin J, Lizio M, Harshbarger J et al. Interactive visualization and analysis of large‐scale sequencing datasets using ZENBU. Nat Biotechnol 2014; 32: 217–219. [DOI] [PubMed] [Google Scholar]
  • 66. Yu G, Wang L‐G, He Q‐Y. ChIPseeker: An R/Bioconductor package for ChIP peak annotation, comparison and visualization. Bioinformatics 2015; 31: 2382–2383. [DOI] [PubMed] [Google Scholar]
  • 67. McLean CY, Bristor D, Hiller M et al. GREAT improves functional interpretation of cis‐regulatory regions. Nat Biotechnol 2010; 28: 495–501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68. Jones TR, Kang IH, Wheeler DB et al. CellProfiler analyst: Data exploration and analysis software for complex image‐based screens. BMC Bioinformatics 2008; 9: 482. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Supplementary figure 1

Supplementary figure 2

Supplementary table 1

Supplementary table 2

CTI2-13-e70018-s002.pdf (1.3MB, pdf)

Video S1

Download video file (100.4MB, mp4)

Video S2

Download video file (112.4MB, mp4)

Data Availability Statement

The ATAC‐seq data generated in this study have been submitted to the Gene Expression Omnibus (GEO) databases under accession number GSE268544. The URL is as follows: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE268544.

Articles from Clinical & Translational Immunology are provided here courtesy of Australasian Society of Immunology

RESOURCES