Inhibition of dipeptidyl-peptidase 4 induces upregulation of the late cornified envelope cluster in keratinocytes

Abstract

Dipeptidyl-peptidase 4 (DPP4) is a multifunctional type II transmembrane glycoprotein that is expressed on various cell surfaces. While DPP4 inhibitors have a therapeutic role in the treatment of diabetes mellitus, they are an independent risk factor in the development of bullous pemphigoid. Contrarily, there are reports of improvement in psoriasis with DPP4 inhibition. We investigated the effect of DPP4 inhibition on primary human keratinocytes to determine whether DPP4 modulates keratinocyte inflammatory signaling and keratinocyte homeostasis. We performed RNA sequencing of primary adult human keratinocytes treated with DPP4 inhibitor, identifying 424 differentially expressed genes. Gene ontology analysis revealed significant enrichment of epidermal differentiation and cornified envelope genes. Using three-dimensional organotypic cultures and a pan-late cornified envelope 2 (LCE2) antibody, we demonstrate a dose dependent relationship between DPP4 inhibition and increased expression of LCE2 during epidermal development. The late cornified envelope gene clusters are expressed at the late stages of epithelial development, responding to stimuli such as calcium and ultraviolet light. While its biologic function is not fully understood, mutations in LCE3B/LCE3C confer a 40% increased risk in the development of plaque psoriasis. While we did not identify significant modulation of keratinocyte inflammatory markers, DPP4 inhibition increased expression of the late cornified envelope may offer a potential alternative therapeutic mechanism in psoriasis.

Background

Dipeptidyl-peptidase 4 (DPP4) is a multifunctional type II transmembrane glycoprotein. It is expressed on the cell surface of many cell types. The extracellular domain of DPP4 is a serine exopeptidase that cleaves X-proline or X-alanine dipeptide residues in the penultimate position from the NH2-terminus of its substrate. The cysteine rich domain of DPP4 permits interactions with the extracellular matrix. DPP4 and its inhibition has numerous dermatologic implications [15]. Perhaps most notable, is the increased risk of developing bullous pemphigoid in patients taking DPP4 inhibitors. Yet, DPP4 inhibition additionally has a therapeutic role in wound healing, and a questionable efficacy in treating psoriasis [15].

While DPP4 is most commonly described as a T-cell marker, it is additionally expressed in keratinocytes in an array of clinical entities, with notable dysregulation on keratinocytes in psoriasis and bullous pemphigoid [10, 14, 16, 19, 20]. Keratinocyte DPP4 enzymatic activity in basal cell culture conditions exceeds that of resting lymphocytes and approximates mitogen-stimulated peripheral blood mononuclear cells [6]. As such, it is evident that modulation of DPP4 may have a direct impact on keratinocytes, aside from its role in T-cells.

Despite the observed differential expression of DPP4 on keratinocytes in various dermatologic conditions, little is known regarding its physiologic function. In vitro studies on primary keratinocytes and HaCaT cells demonstrated a dose-dependent decrease in DNA synthesis with DPP4 inhibitors [21]. This finding does not provide insight into potential mechanisms by which DPP4 inhibition may exacerbate bullous pemphigoid or improve psoriasis. Given the numerous dermatologic sequelae of DPP4 inhibition, we sought to better understand the effect of DPP4 inhibitors on keratinocyte physiology.

Materials and methods

Two-dimensional cell culture

Primary adult human keratinocytes (PHKs) (Thermo Fisher Scientific, Waltham, MA) were cultured in EpiLife Medium with Human Keratinocyte Growth Supplement in a CO₂ incubator. Fresh culture media were changed every 48 h. At 80–90% confluence, cells were treated with 3-N-[(2S,3S)-2-amino-3-methylpentanoyl]-1,3-thiazolidine, a potent and selective dipeptidyl-peptidase 4 inhibitor (10 μM) (BioVision, Inc., Milpitas, CA), or vehicle for 24 h. Then cells were washed twice with PBS and were stored at − 80 °C until RNA extraction.

RNA isolation and qPCR

Total RNA from PHKs was isolated using miRNeasy Mini Kit following the manufacturer’s instructions (Qiagen Inc., Valencia, CA). On-Column DNase I Digestion was performed to remove possible contaminated genomic DNA. 1 μg of total RNA was reverse-transcribed to 100 μl of cDNA. The PCR was carried out using a Stratagene Mx 3000 real-time PCR machine (Santa Clara, CA). The cycling condition was as follows: 95 °C, 10 min followed by 40 cycles of (95 °C, 15 s; 55 °C, 60 s). GAPDH was used as the internal reference. Primers used are listed in Table 1.

Table 1.

Human primers used for real-time RT-PCR

Gene name	Forward primer 5′ → 3′	Reverse primer 5′ → 3′
LCE1C	TAG GTC CCA CTG CCA CAG AC	TTA GCA GCA GCC TCC AGA GT
LCE3B	AAG ACC TGG GTG CTG TGG T	TGC CTC TGT CAC AGG AGT TG
LCE3C	CCT CCA CCC TCT TCT GAC TG	ATT GAT GGG ACC TGA AGT GC
LCE3D	CCC AAA GAG CCC AGT ACA GT	CCT GTG GTG GTT CAG GAA G
LCE3E	CAG CAG AAC CAG AAG CAG TG	ACA GCC AGA GGA AGC TGG AG
SPRR2D	ATC AAC AGC AGC AGT GCA AG	TGT GGA CAC TTT GGT GAT GG
SPRR2G	ATG TCT TAC CAG CAG CAG CA	GAC AAG GAG GAG GCA GGT AA

RNA-seq library preparation

RNA quantity and quality were assessed using a Nanodrop 1000 spectrophotometer (Nanodrop, Wilmington, DE, USA), with RNA integrity number and concentration checked using an Agilent 2100 Bioanalyzer (Agilent Technologies, Inc., Santa Clara, CA, USA). Isolated RNA samples were subsequently prepared for construction of transcriptome libraries. Poly (A) mRNA was isolated and enriched from total RNA using oligo (dT)-attached magnetic beads according to the manufacturer’s (Illumina) instructions. Enriched and purified mRNA was subsequently broken into an approximately 200 nt short mRNA. Random hexamers were used as primers to synthesize first-strand cDNA while second-strand cDNA was synthesized in a buffer containing dNTPs, DNA polymerase I, and RNaseH. Suitable fragments were isolated and enriched by PCR amplification. Finally, the constructed cDNA libraries of the samples were sequenced using an Illumina HiSeq sequencing platform by CDGenomics (New York, NY).

Bioinformatic analysis

Filtered clear reads were mapped to the reference genome by HISAT2 [7]. Mapped reads were converted to Fragments Per Kilobase of transcript per Million mapped reads (FPKM), and subsequently quantified with Cuffquant and Cuffnorm [17]. DESeq was used to analyze differentially expressed genes (DEGs) [2]. Fold change ≥ 2 and FDR < 0.05 were set as screening criteria. Gene Ontology (GO) was assessed using the topGO package [1], utilizing the gene ontology database (http://geneontology.org/). Heat maps were generated using FPKM values using the iDEP server [4].

Three-dimensional human skin equivalent culture (3D HSE)

3D HSEs were initiated and propagated as described [3] using primary neonatal human keratinocytes. Three independent cell lines derived from three donors were used for 3D HSEs. Cultures were treated with DPP4 inhibitor (10 or 50 μM) or vehicle (0.05% or 0.2% dimethyl sulfoxide) at days 0, 2, and 5 after lifting to the air–liquid interface. Cultures were harvested at day 6.

Immunostaining

3D HSEs were embedded in optimal cutting temperature medium and sectioned at 4 μm. LCE2 antibody was kindly provided by Dr. Masashi Narita. Fresh cryosections of 3D HSE were incubated for 15 min with 5% goat serum/PBS. Tissue sections were then incubated at room temperature for 1 h with LCE2 antibody (1:500 diluted in 1%BSA/PBS), then subsequently washed with PBS for 5 min thrice. Sections were incubated with 1:200 diluted Alexa fluor 647 conjugated donkey anti rabbit IgG (Thermo Fisher, Carlsbad, California) for 1 h and washed with PBS. Slides were allowed to dry, then mounted with mounting buffer containing DAPI.

Fluorescent quantification

Immediately after staining, images were acquired using the Life Sciences EVOS M5000 using the same gain, lighting, and contrasting settings for LCE2. Quantification of fluorescent signal was performed using Image J as previously described by us with minor modifications [18]. Briefly, images were analyzed generating a histogram of intensity/color using a positive control to define the color range of the 647 nm secondary antibody. After defining cutoffs, the number of pixels falling in this color range were recorded and divided by the width of the tissue section by an investigator blinded to the experimental conditions. 3–4 sections of each culture were measured. The number of positive pixels was normalized and compared between the treated and untreated arms using one-way ANOVA followed by Tukey’s test with P < 0.05 using Graphpad Prism 6 (San Diego, California, USA).

Results

RNA-seq revealed 424 DEGs (167 up, 257 down) with adjusted P values (FDR) < 0.05 and at least a log2fold change of ± 1. DEGs between control and DPP4 inhibitor treated PHKs are shown in Online Resource 1, sTable 1 sorted by FDR. A heat map of the 20 most DEGs is shown in Fig. 1.

Fig 1.

Heatmap of top 20 most differentially expressed genes defined by FDR in dipeptidyl dipeptidase 4 (DPP4) inhibitor versus vehicle treated primary keratinocytes (FPKM ≥ 500 shown)

DEGs were analyzed by GO and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway term enrichment. GO analysis organizes DEG to provide a controlled vocabulary to describe biological processes, molecular functions, cellular components, while KEGG analysis represents a curated list of signaling, metabolic, and regulatory pathways. GO analyses of the DEGs demonstrated that skin development, epidermal development, epidermal cell differentiation, and keratinocyte differentiation were the most enriched biological process terms, while extracellular matrix and cornified envelope were the most enriched cell components. Receptor ligand activity and cytokine receptor binding were the most enriched molecular function pathways. KEGG analysis demonstrated the enhancement of cytokine–cytokine receptor interaction, rheumatoid arthritis, and IL-17 signaling pathways. Enrichment analyses are shown in Online Resource 2, sFig. 1.

Given the significant enrichment of keratinocyte related genes, particularly cornification, we reviewed individual genes, identifying a large number of late cornified envelope and small proline-rich protein genes upregulated. We next validated 7 LCE genes by qPCR (LCE1C, LCE3B, LCE3C, LCE3D, LCE3E, SPRR2D, SPRR2G), noting correlation with our RNA-seq findings (Fig. 2).

Fig 2.

qPCR demonstrates upregulation of multiple components of the late cornified envelope gene cluster corresponding with next generation sequencing data. Values significant to P < 0.05 are shown with *.

Next, to determine whether these changes occurred at the protein level, as well as had an effect on keratinocyte differentiation in a stratified setting, we performed 3D HSE cultures of primary neonatal human keratinocytes. Cultures were treated with media containing DPP4 inhibitor (10 μM or 50 μM) or vehicle for 24 h. Cultures were harvested at 6 days and snap frozen. Sections of 3D HSE were stained with a pan-LCE2 antibody previously described [1]. Fluorescent intensity at the granular layer demonstrated a dose dependent increase in signal with DPP4 inhibition. (Fig. 3) Histology of 3D HSE did not demonstrate significant morphologic changes in keratinocytes (Online Resource 2, sFig. 2).

Fig 3.

A Dose dependent effect of DPP4 inhibition on 3D organotypic cultures demonstrates the absence of late cornified envelope at 6 days, but dose dependent increase with DPP4 inhibitor. B Fluorescent signal quantified from AlexaFluor 647 reveals significant dose dependent increase in LCE2 expression with DPP4 inhibition (n = 3/condition)

Discussion

We herein demonstrate that DPP4 inhibition induces keratinocyte expression of the late cornified envelope complex. The exact interplay of LCE2 and LCE3 proteins remains unclear. LCE3 are expressed at an earlier time point during epidermal differentiation, localizing to the upper stratum granulosum, while LCE2 are limited to the uppermost granular layer and stratum corneum [13]. The physiologic function of subtypes within the LCE families also appears to differ, with proteins demonstrating antimicrobial activity [12]. Deletion in the late cornified envelope genes LCE3B and LCE3C are associated with the development of psoriasis in individuals with the high-risk HLA-C*06 allele [12]. Notably, RNA-seq of keratinocytes treated with DPP4 inhibitor exhibited upregulation of most LCE genes except LCE3B/C, though our qPCR results demonstrated significant upregulation of LCE3C. This is likely a result of relatively low read counts for LCE genes.

The relationship between DPP4 inhibition and psoriasis remains unclear. Reports of glucose independent improvements countered by a failed phase II trial utilizing DPP4 inhibitors calls into question the therapeutic role of DPP4 inhibitors [11]. Likewise, mixed epidemiologic data indicating a potential increase of psoriasis in patients receiving DPP4 inhibitors would make it appear even less likely to be beneficial [7, 9]. Yet, changes to the cornified envelope are one of the best described and strongest genetic predispositions to the development of psoriasis. Whether increases in these late cornified envelope constituents due to DPP4 inhibition exert a sufficient protective effect in psoriasis patients with LCE3B/C deletions is yet to be further investigated. This could represent a patient subset who may benefit from therapies increasing late cornified envelope constituents.

While DPP4 inhibition is known to predispose towards the development of BP, the mechanism remains unclear. BP has a known complement-independent pathway, by which BP-IgG leads to internalization of collagen 17 as well as expression of numerous pro-inflammatory cytokines [8]. However, DPP4 inhibitors usually cause atypical noninflammatory BP where inhibition of plasmin by DPP4 inhibitors may contribute to development of the noninflammatory phenotype. [5] In line with this, we did not find upregulation of pro-inflammatory cytokines. It is also possible that DPP4 inhibition may affect T-cell regulation.