A mouse model for vitamin D-induced human cathelicidin antimicrobial peptide gene expression

Abstract

In humans and other primates, 1,25(OH)₂vitamin D₃ regulates the expression of the cathelicidin antimicrobial peptide (CAMP) gene via toll-like receptor (TLR) signaling that activates the vitamin D pathway. Mice and other mammals lack the vitamin D response element (VDRE) in their CAMP promoters. To elucidate the biological importance of this pathway, we generated transgenic mice that carry a genomic DNA fragment encompassing the entire human CAMP gene and crossed them with Camp knockout (KO) mice. We observed expression of the human transgene in various tissues and innate immune cells. However, in mouse CAMP transgenic macrophages, TLR activation in the presence of 25(OH)D₃ did not induce expression of either CAMP or CYP27B1 as would normally occur in human macrophages, reinforcing important species differences in the actions of vitamin D. Transgenic mice did show increased resistance to colonization by Salmonella typhimurium in the gut. Furthermore, the human CAMP gene restored wound healing in the skin of Camp KO mice. Topical application of 1,25(OH)₂vitamin D₃ to the skin of CAMP transgenic mice induced CAMP expression and increased killing of Staphylococcus aureus in a wound infection model. Our model can help elucidate the biological importance of the vitamin D-cathelicidin pathway in both pathogenic and non-pathogenic states.

Keywords: Cathelicidin; Cyp27b1; Innate immunity; Macrophage; TLR; Vitamin D.

Figure 1.. Generation of a CAMP transgenic mouse and expression in different tissues.

(A) This schematic generated from the UCSC Genome Browser December 2013 (GRCh38/hg38) Assembly shows that the human 6.02 kb Kpn1-EcoR1 genomic DNA fragment contains the CAMP gene located on chromosome 3 (base pairs 48,223,347 −48,225,491). It also shows the conserved 5’- and 3’-regions that possess potentially important transcriptional regulatory elements and the VDRE located in a primate-specific AluSx SINE conserved in only apes (chimp, gorilla, orangutan and gibbon), Old World primates (rhesus macaque, baboon and green monkey) and New World primates (Ma’s night monkey and marmoset), but not in prosimians (tarsier, black lemur and bushbaby) or other mammals (mouse, rat, dog and opossum). (B) The human CAMP mRNA is expressed in tissues of Tg mice. Total RNA was isolated from tissues of Tg/WT mice and expression of CAMP mRNA was measured by qRT-PCR. The ratio of CAMP to the reference gene Ywhaz was log10 transformed and graphed as the mean (+S.D.) with the minimum and maximum (Box plots, n=4–6 mice; non-box plots, n=3 mice).

Figure 2.. Tg/KO mice express hCAP18 in several lineages of immune cells.

Immune cells from three mice were pooled from spleen, peritoneal cavity, blood, and bone marrow and dual stained with lineage markers and either control rabbit IgG (left panels A, B, D, E) or anti-hCAP18 antibody (left panel C and all right panels) and analyzed by flow cytometry. (A) Splenocytes stained with the T-cell marker CD3. (B) Splenocytes stained with the B-cell marker CD19. (C) Peritoneal cells stained with CD19 (left panel) or F4/80 (right panel). (D) Blood cells stained with the granulocyte marker GR1. (E) Bone marrow cells stained with GR1. Representative plots from three different experiments are shown.

Figure 3.. Human CAMP gene does not alter phagocytosis or superoxide production in murine phagocytes.

(A) BM-derived macrophages from WT, Tg/KO, and KO genotypes were incubated with fluorescent pHrodo E. coli for 60 min and then analyzed by flow cytometry to measure the percentage of phagocytes. Data represent mean (+SD) of four experiments. (B) BM-derived macrophages from WT and Tg/WT mice were incubated with a luminol substrate and stimulated with PMA to induce superoxide as measured by relative light unit (RLU) emission over a time course. Each time point represents the mean (±SD) of three experiments. (C) BM neutrophils from WT and Tg/WT mice were incubated as described in panel B. Each time point represents the mean (±SD) of six biological replicates from two experiments for each genotype. No significant differences were observed between the genotypes.

Figure 4.. Human CAMP gene increases resistance to enteric infection in Camp Tg mice.

(A) Mice were infected by oral gavage with 8×10⁸ S. typhimurium. Five days post infection organs were harvested and plated to measure bacterial CFU shown on a log10 scale. The Tg/WT (n=14) group of mice had a significant decrease in cecum burden compared with both the WT (n=13) and KO (n=14) groups (Asterisk indicates p value <.0001, one-way ANOVA,) (B) Tg/KO and Camp KO mice were infected with a lower dose (8×10⁷) of S. typhimurium for 5 days. The CFU burden in Tg/KO mice (n=30) was significantly lower compared with the KO mice (n=33). The asterisk indicates a p value = 0.007 (two-tailed T-test). Data were plotted as the mean CFU (horizontal line) ±SD (box). The bars show the range of the data.

Figure 5.. Induction of CAMP gene and hCAP18/LL-37 expression with 1,25(OH)₂D₃ treatment.

(A) BM-derived macrophages from Tg/KO mice were treated 48 h in vitro with increasing concentrations (10⁻⁹, 10⁻⁸ and 10⁻⁷ M) of 1,25(OH)₂D₃. hCAP18 and F4/80 staining cells (macrophages) were detected by flow cytometry. Mean fluorescent intensity (MFI) of hCAP18 is noted in the upper right quadrant. Graphs are representative of four experiments. (B-F) Dorsal skin of Tg/KO mice was treated with either vehicle control or 1 nmole 1,25(OH)₂D₃ for 24 h (mRNA) or 48 h (protein). Expression of CAMP (B) and Cyp24a1 (C) mRNAs (n=3 mice) determined by qRT-PCR was normalized to the housekeeping gene Ywhaz. The ratio was log10 transformed and plotted as the mean (+SD). D) hCAP18 protein levels were determined by ELISA (n=6 mice per treatment) and plotted as the mean (ng hCAP18/mg of total protein, ±SD). (E) hCAP18 expression in tissues from untreated Tg/KO and WT mice was analyzed by Western blot using 50 g protein per lane and an anti-hCAP18 polyclonal antibody. The 18-kDa pro-protein was detected in BM, skin, lung and spleen of the Tg/KO mice. The 14-kDa processed cathelin-domain was detected in the BM and skin. No hCAP18 was detected in the WT skin. GAPDH was used as a loading control, except this particular antibody does not identify GAPDH in bone marrow. (F) BM (untreated) and skin (from treated mice, panel C) samples from WT and Tg/KO mice were analyzed by Western blot with an anti-LL-37 monoclonal antibody. Synthetic LL-37 peptide (4-kDa) was included as a positive control. Both the 18- and 4-kDa forms were detected in the BM. In skin, primarily the 18-kDa was detected with very faint 4-kDa bands. Each Western blot lane represents an individual mouse. Asterisks denote statistical significance (p<0.05) determined by a two-tailed T-test.

Figure 6.. Topical application of 1,25(OH)₂D₃ induces hCAP18 in Tg/KO mouse skin.

Tg/KO mice were topically treated for 48 h with either 50% ethanol/glycerin vehicle (control) or 1,25(OH)₂D₃ in vehicle. Skin samples were fixed and paraffin embedded. Sections were stained with anti-hCAP18 polyclonal antibody and then detected with a secondary antibody conjugated to Alexafluor 488 (green) and counterstained with DAPI (blue). The dashed yellow line indicates the border between the epidermis (E), dermis (D) of the skin, and hair follicles (HF), white scale bar = 50 μm. Top panels are multiple sections from vehicle control mice, while the bottom panels are multiple sections from 1,25(OH)₂D₃-treated mice.

Figure 7.. The human CAMP gene restores wound healing and topical treatment with 1,25(OH)₂D₃ lowers S. aureus burden in Tg/KO mice.

On Day 0, horizontal dermal wounds were generated on the backs of mice using a 5 mm dermal punch (n=4 per genotype). (A) Percent wound closure was determined by measuring the wound size on days 3, 5, 7, and 9. On Day 5, both WT and Tg/KO mice showed similar wound healing and contraction rates, as measured by total wound area, while a significant lack of wound closure was observed for KO mice (Asterisk indicates p < 0.05, Student’s t-test). (B) Illustrative examples are shown for each genotype at initial wounding day 0 (top panels) and 5 days post wounding (lower panels). (C) Single 5mm dermal wounds were formed on Tg/KO mice and 1 day later mice were treated with either vehicle control (n=5) or 1,25(OH)₂D₃ (n=7). The following day 8×10⁵ CFU of S. aureus were applied to the wound. At 2 days post infection, the mice were treated again with either vehicle or 1,25(OH)₂D₃ at the wound site. At 5 days post infection, mice were euthanized and a 10 mm circular biopsy of the wound site was used to enumerate surviving S. aureus bacteria expressed as log10 CFU. The dashed line across the graph indicates the starting infectious CFU, while the center line and brackets indicate the mean and S.D. per group of mice. (Asterisk indicates p < 0.05, two tailed t-test)

Figure 8.. Expression of Cyp27b1 in mouse macrophages and tissues after TLR ligand stimulation.

TLR ligands do not induce Cyp27b1 expression in mouse macrophages. (A) BM-derived Tg/WT mouse macrophages and human PBMC-derived macrophages were stimulated with TLR ligands for 24 h and Cyp27b1 gene expression was measured by qRT-PCR, normalized to β-actin and graphed on a log10 scale. Graphs are a single experiment representative of three independent experiments (Asterisks denote significance from control; p<0.05, two-tailed T-test). (B) Tg/WT mice were injected IP with either 10 μg/g body weight LPS or vehicle control for 24 h. Expression of Cyp27b1 was normalized to Ywhaz by qRT-PCR and graphed on a log10 scale. Asterisk denotes a significant increase as compared with the control (p<0.05, T-test with Welch’s correction). (C) WT mice were injected IP with either 0.5 g/g body weight LPS or vehicle control. Kidneys were collected 3 h post injection. Cyp27b1 expression was normalized to Gapdh and graphed on a log10 scale. D) Cyp24a1 expression was normalized to Gapdh and graphed on a log10 scale. Data were graphed as the mean (center line) ±SD (brackets). Each symbol represents an individual mouse for B, C and D. Asterisks denote significance from control (p<0.05, two-tailed T-test).