Characterising plasmacytoid and myeloid AXL+ SIGLEC-6+ dendritic cell functions and their interactions with HIV

Abstract

AXL+ Siglec-6+ dendritic cells (ASDC) are novel myeloid DCs which can be subdivided into CD11c+ and CD123+ expressing subsets. We showed for the first time that these two ASDC subsets are present in inflamed human anogenital tissues where HIV transmission occurs. Their presence in inflamed tissues was supported by single cell RNA analysis of public databases of such tissues including psoriasis diseased skin and colorectal cancer. Almost all previous studies have examined ASDCs as a combined population. Our data revealed that the two ASDC subsets differ markedly in their functions when compared with each other and to pDCs. Relative to their cell functions, both subsets of blood ASDCs but not pDCs expressed co-stimulatory and maturation markers which were more prevalent on CD11c+ ASDCs, thus inducing more T cell proliferation and activation than their CD123+ counterparts. There was also a significant polarisation of naïve T cells by both ASDC subsets toward Th2, Th9, Th22, Th17 and Treg but less toward a Th1 phenotype. Furthermore, we investigated the expression of chemokine receptors that facilitate ASDCs and pDCs migration from blood to inflamed tissues, their HIV binding receptors, and their interactions with HIV and CD4 T cells. For HIV infection, within 2 hours of HIV exposure, CD11c+ ASDCs showed a trend in more viral transfer to T cells than CD123+ ASDCs and pDCs for first phase transfer. However, for second phase transfer, CD123+ ASDCs showed a trend in transferring more HIV than CD11c+ ASDCs and there was no viral transfer from pDCs. As anogenital inflammation is a prerequisite for HIV transmission, strategies to inhibit ASDC recruitment into inflamed tissues and their ability to transmit HIV to CD4 T cells should be considered.

Fig 1. Identification of pDC and ASDCs in PBMCs using pDC Isolation Kit and DC Enrichment Kit.

(a) Representative dot plots showing the gating strategy to identify pDCs (AXL^- Siglec-6^- BDCA2⁺ CD123⁺), CD123⁺ ASDCs (AXL⁺ Siglec-6⁺ CD123⁺ CD11c^-/lo) and CD11c⁺ ASDCs (AXL⁺ Siglec-6⁺ CD11c⁺ CD123^-/lo). Top row demonstrates isolation from PBMCs, middle row uses Human pDC Isolation Kit II and bottom row uses Human Pan DC Enrichment Kit. (b) Mean proportion of Lin1^- HLA-DR⁺ cells and (c) mean proportion of pDCs, CD11c⁺ and CD123⁺ ASDCs using the Human Pan DC enrichment Kit (±SD, n = 5). (d) A representative distribution of pDCs (blue), CD11c (orange) and CD123 ASDCs (purple) on a t-SNE plot to verify the phenotypes of blood pDCs and ASDCs isolated via the Human Pan DC Enrichment Kit. (e) Heat map visualisations of the median fluorescence intensity of surface AXL, Siglec-6, CD123, BDCA2, CD11c and CD141 expression for the populations shown on t-SNE plot. (f) Representative contour plots showing the phenotypic characteristics of pDCs, CD11c⁺ and CD123⁺ ASDCs.

Fig 2. Transcriptomic and proteomic profile of pDCs and ASDCs.

(a) Workflow for isolating pDCs and ASDCs using the Pan DC enrichment kit followed by FACS cell sorting to separate pDCs, CD123⁺ ASDCS and CD11c⁺ ASDCs. RNA was extracted and processed via NanoString. (b) Heat map for genes delineating pDCs from ASDCs (n = 4). (c) Chemokine gene expression by NanoString in pDCs and ASDCs. Data are presented as ±SD (n = 4). Statistical analysis was performed using one-way ANOVA with Tukey’s multiple comparisons test, *p< 0.05; **p< 0.01. (d) Gene expression of HIV binding and uptake receptors on blood derived pDCs and ASDCs. Blood PBMCs (GSE94820) [4] were transcriptionally profiled by scRNAseq. pDC, CD11c⁺ ASDC and CD123⁺ ASDC annotations were determined using metadata provided by the authors. All genes pertaining to HIV binding lectin receptors, HIV entry receptors and SAMHD1 were compared between pDCs, CD11c⁺ ASDCs and CD123⁺ ASDCs and calculated as a scaled value. The scaled values were plotted on a heat map using Prism. (e) The expression of Siglec-1 (CD169), CLEC4A, CLEC5A, MR (CD206), Langerin (CD207), DC-SIGN (CD209), CD4, CXC4, CCR5 and SAMHD1 were determined on each cell type immediately after their isolation from blood. The percentage expression of each subset was calculated for each marker, and gMFI (grey dotted box) calculated when percent expression exceeded 80% for multiple subsets (n = 3–4). *p < 0.05, **p < 0.01 by one-way ANOVA with Tukey’s multiple comparisons test.

Fig 3. Transcriptomic profile of pDCs and ASDCs via scRNASeq.

UMAP plots and expression of genes delineating pDCs from ASDCs in data sets of cells derived from inflamed tissues of psoriasis diseased skin (E-MTAB-8142[15]) is shown in (a-b) and from colorectal cancer (GSE178341[16]) in (c-d).

Fig 4. Identification of pDCs and ASDCs in human tissues.

(a) Representative dot plots identifying pDCs (AXL^- CD123⁺) and ASDCs (AXL⁺ Siglec-6⁺) in a rectum of a patient with ulcerative colitis. (b) Back gating to show the surface expression of pDCs and ASDC and to confirm the phenotype detected in a. (c-d) Representative data showing pDCs and ASDCs in the rectum of patient with chronic ulcerative colitis in c and in mesenteric lymph nodes of inflamed rectum in d. All cells were first gated as live HLA-DR⁺ CD45⁺ CD19^- CD3^- Autofluorescent^- CD14^- CD16^- cells. (e) Absence of pDCs and ASDCs in non-inflamed rectal tissue (top) after adopting the gating strategy listed in a. Siglec-6 FMO and AXL FMO (bottom) confirmed real expression. (f) ASDCs AXL⁺ CD303⁺ (pink arrows), ASDCs AXL⁺ CD11c⁺ (orange arrows) and pDCs (blue arrows) were identified by microscopy at the periphery of a submucosal lymphoid aggregate in an inflamed human colon. ASDCs were located proximal to a blood vessel based on CD31 expression. (g) Representative gating strategy as demonstrated by inner foreskin lamina propria to identify pDCs, ASDCs and cDC2s. All cells were first gated as live CD45⁺ HLA-DR⁺ CD3^- CD19^-; pDCs (blue) were defined as CD14^lo CD5^- CD123⁺, CD123⁺ ASDCs (pink) as CD14^lo CD5⁺ CD1c⁺ Siglec-6⁺ CD123⁺, CD11c⁺ ASDC (orange) as CD14^lo CD5⁺ CD1c⁺ Siglec-6⁺ CD11c⁺ and cDC2 (green) as CD14^- CD5⁺ CD1c⁺ CD11c⁺ CD163^-. This gating strategy differed to the gating in a-e as a different panel design was used to allow for identification of cDC2s.

Fig 5. Expression of key HIV binding receptors on ASDCs, pDCs and cDC2s from human tissues and blood.

(a) Immune cells liberated from abdominal epidermis (abdo epi) (n = 1), labia epidermis (n = 2), inner foreskin epidermis (n = 1), labia dermis (derm) (n = 3), outer foreskin dermis (n = 1) and inner foreskin lamina propria (LP) (n = 1) were stained for flow cytometry. The percentage expression of Siglec-1, DC-SIGN, MR, Langerin, and CD4 on pDCs, CD11c⁺ ASDCs, CD123⁺ ASDCs and cDC2s for each tissue was plotted on a heat map and compared to blood-derived pDCs, ASDCs and cDC2s in (b).

Fig 6. Functional phenotyping of blood-derived pDCs and ASDCs.

(a) Cell surface expression of CD80, CD83, CD86, HLA-DRM and ICAM-1 was determined on each cell type immediately after their isolation from blood. The percentage of expression and gMFI (when % of expression was above 80%, outlined in grey dotted box) of each subset was calculated. (b) Workflow for functional phenotyping. (c-d) FACS sorted pDCs, CD123 and CD11c ASDC were cultured at 37°C with Cell trace Violet-stained naïve T cells at a ratio of 1 ASDC or pDC: 10 T cells. On day 6, cultures were analysed by flow cytometry to assess CD4 T cell (c) proliferation via CTV stain and activation via CD25 expression. (d) on day 6, all conditions were treated with anti-CD3/CD28 for 24 hours. Supernatants were collected to assess cytokine production. Data is presented as ±SD (n = 3–4). Statistical analysis was performed using one-way ANOVA with Tukey’s multiple comparisons test. *p < 0.05; **p < 0.01; ***p < 0.001.

Fig 7. Chemokine and cytokine production after HIV interaction with pDCs and ASDCs.

LEGENDplex assay was carried out to determine the concentration of cytokines in the supernatants of mock and HIV-1 BaL treated pDCs and ASDCs at 18 hours post HIV treatment. Data presented as ±SD (n = 4). Statistical analysis was performed using one-way ANOVA with Tukey’s multiple comparisons test. *p< 0.05; **p< 0.01; ***p< 0.001; ****p< 0.0001.

Fig 8. HIV interactions with blood-derived pDCs and ASDCs and viral transfer to T cells.

(a) HIV transfer to JLTR cells was determined by measuring JLTR GFP fluorescence intensity by flow cytometry 2h and 96h post co-culture for first phase and second phase transfer respectively. Representative flow data and graph (data presented as ±SD, n = 3) show transfer of HIV from HIV-treated cells to JLTR (GFP⁺ JLTR). (b) Number of infected TZMBL cells after inoculation with supernatants collected at 96 hpi of pDCs, CD123⁺ and CD11c⁺ ASDCs. Data presented as ±SD (n = 3). (c) HIV transfer to JLTR cells in the presence and absence of Maraviroc. Data presented as ±SD (n = 3). Statistical analysis was performed using Wilcoxon Rank Sum Test.