Molecular evidence for natural killer-like cells in equine endometrial cups

Abstract

Objectives: To identify equine orthologs of major NK cell marker genes and utilize them to determine whether NK cells are present among the dense infiltration of lymphocytes that surround the endometrial cup structures of the horse placenta during early pregnancy.

Study design: PCR primers were developed to detect the equine orthologs of NKP46, CD16, CD56, and CD94; gene expression was detected in RNA isolated from lymphocytes using standard 2-step reverse transcriptase (RT) PCR and products were cloned and sequenced. Absolute real-time RT-PCR was used to quantitate gene expression in total, CD3+, and CD3- peripheral lymphocytes, and invasive trophoblast. Lymphocytes surrounding the endometrial cups (ECL) of five mares in early pregnancy were isolated and NK marker gene expression levels were assayed by quantitative RT-PCR.

Main outcome measures: Absolute mRNA transcript numbers were determined by performing quantitative RT-PCR and comparing values to plasmid standards of known quantities.

Results: NKP46 gene expression in peripheral CD3- lymphocytes was higher than in CD3+ lymphocytes, CD16 levels were higher in the CD3+ population, and no significant differences were detected for CD56 and CD94 between the two groups. Expression of all four NK cell markers was significantly higher in lymphocytes isolated from the endometrial cups of pregnant mares compared to PBMC isolated from the same animal on the same day (NKP46, 14-fold higher; CD94, 8-fold higher; CD16, 20-fold higher; CD56, 44-fold higher).

Conclusions: These data provide the first evidence for the expression of major NK cell markers by horse cells and an enrichment of NK-like cells in the equine endometrium during pregnancy.

Figure 1

Multi-species alignments of NK cell marker protein sequences. Amino acid sequences of equine NKp46 (A), CD16 (B), CD94 (C), and CD56 (D) orthologs were determined by translation of cDNA sequences determined from a combination of bioinformatic analysis of the equine WGS and amplification from horse PBMC mRNA by RT-PCR. Equus caballus (EC) sequences were aligned with sequences of Homo sapiens (HS), Bos taurus (BT), Sus scrofa (SS) and Mus musculus (MM) using Clustal W. Full- length clones were obtained for NKP46, CD16, and CD94 genes; a 247 bp partial clone was obtained from CD56, as indicated by brackets. Dots represent identities, dashes represent gaps. Dashed boxes represent transmembrane domains and grey bars represent conserved immunoglobulin-like domains. Asterisks highlight highly conserved transmembrane domain charged residues (arginine and aspartic acid) required for association with adaptor molecules responsible for intracellular signaling. Closed boxes represent cysteine residues known to stabilize intramolecular secondary structures. Genbank IDs for sequences used: NP_004820 (HS), NP_899209 (BT), NP_001116615.1 (SS), NP_034876 (MM), NKp46; AAH17865.1 (HS), AAI12757.1 (BT), NP_999556 (SS), NP_034318.2 (MM), CD16; NP_001107868 (HS), NP_001002890 (BT), EDK99936 (MM), CD94; NP_851996.2 (HS), NP_776824.1 (BT), NP_001074914.1 (MM), CD56.

Figure 2

Relation of equine and human NKP46 splice isoforms. (A) Schematic diagram of NKP46 genomic structure. (B) Agarose gel of PCR products generated from PBMC cDNA using primers designed to amplify the full-length CDS and partial 5’ and 3’ untranslated regions. Bands were extracted, cloned, and sequenced, yielding 3 transcripts with open reading frames (C). Equine transcript variant 1 is similar to full length human NKP46 isoform a (NM_004829.5). Variant 2 uses an alternate in-frame splice site at the exon 4–5 boundary, similar to human isoform b (NM_001145457.1). Variant 3 has a deletion of exon 4, corresponding to the loss of one Ig-like domain, similar to the human exon 3 deletion isoform d (NM_001242356.1). The band observed at 900bp is an intron-retention mutant of variant 3 with a premature stop codon (ψ, pseudogene).

Figure 3

Detection of NK cell marker expression in equine peripheral lymphocytes. Expression of equine NKP46 (A), CD16 (B), CD56 (C), and CD94 (D) was determined using quantitative RT-PCR performed on RNA isolated from PBL and chorionic girdle trophoblast (CG). CD3G (D) and GCM1 (E) expression levels were also measured as controls; n=5.

Figure 4

Expression of NK cell marker expression in CD3-depleted vs. CD3-enriched lymphocyte populations. PBL were magnetically sorted into CD3 depleted (CD3−) and enriched (CD3+) populations; isolated RNA was analyzed for expression of CD3G (A), NKP46 (B), CD16 (C), CD94 (D), and CD56(E) using quantitative RT-PCR; n=4.

Figure 5

Comparison of NK cell marker expression in peripheral and endometrial cup lymphocytes. Paired PBMC and ECL were isolated from five mares pregnant at days 43–46 of gestation. Expression of NKP46 (A), CD16 (B), CD56 (C), CD94 (D), CD3G (E), and GCM1 (F) were measured using quantitative RT-PCR.