Allele-specific recognition by LILRB3 and LILRA6 of a cytokeratin 8 - associated ligand on necrotic glandular epithelial cells

The LILRs are a family of receptors that regulate the activities of myelomonocytic cells. We found that specific allelic variants of two related members of the LILR family, LILRB3 and LILRA6, interact with a ligand exposed on necrotic glandular epithelial cells. The extracellular domains of LILRB3 and LILRA6 are very similar and their genes are highly polymorphic. A commonly occurring allele, LILRB3*12, displayed particularly strong binding of these necrotic cells and further screening of the products of LILRB3 alleles identified motifs that correlated with binding. Immunoprecipitation of the ligand from epithelial cell lysates using recombinant LILRB3*12, identified cytokeratins 8, 18 and 19. Purified proteins obtained from epithelial cell lysates, using anti-cytokeratin 8 antibodies, were able to activate LILRB3*12 reporter cells. Knock-down of cytokeratin 8 in epithelial cells abrogated expression of the LILRB3 ligand, while staining with recombinant LILRB3*12 showed co-localisation with cytokeratin 8 and 18 in permeabilised breast cancer cells. Necrosis is a common feature of tumours. The finding of a necrosis-associated ligand for these two receptors raises the possibility of a novel interaction that alters immune responses within the tumour microenvironment. Since LILRB3 and LILRA6 genes are highly polymorphic the interaction may influence an individual's immune response to tumours.


LILR-CD3ζ constructs and NFAT-GFP reporter cells
The extracellular regions of LILRB3 and LILRA6 were amplified from previously cloned sequences using the primers NK1320 (ATGCAGATCTCCCTTCCCCAAACCCAC) and NK1321 (ATGCGTCGACAGGTGTGGAGGGCGGCC). LILR PCR products were ligated into the pDISPLAY vector (Life Technologies) such that they were fused to the vector N terminal of the haemagglutinin A (HA) epitope and the transmembrane domain of platelet-derived growth factor receptor (PDGFR). The extracellular region of LILRB1 was cloned into pDISPLAY using the primers NK1306 (ATGCAGATCTCACCTCCCCAAGCCCACC) and NK1309 (ATGCGTCGACACTCTGGGGATCCGACC). HA-tagged, LILR-PDGFR transmembrane constructs were then subcloned into the vector pMx puro which encoded the sequence of the cytoplasmic tail of the human CD3ζ chain (a gift from Dr Alex Barrow, St Louis USA) [4] using the primers NK1313 (ATGCTTAATTAATCCACCATGGAGACAGAC) and NK1314 (CTTCTCGAGCCAAAGCATGATGAGGATG).
Following transient transfection of PLAT-E retroviral packaging cells with the pMx HA-LILR-PDGFR-CD3ζ fusion constructs, recombinant retrovirus was used to transduce 2B4 T cell hybridoma cells that had previously been stably transfected with a NFAT-GFP reporter construct (a kind gift from Lewis Lanier, UCSF, San Francisco, California, USA) [5]. Hereafter these cells are referred to as 2B4 reporter cells. Following transduction, 2B4 reporter cells were sorted using a MoFlo cytometer (Beckman Coulter, Brea, CA, USA) for similar levels of expression of each LILR based on the level of expression of the LILR-fused N terminal HA epitope stained with an anti-HA-PE monoclonal antibody (Miltenyi Biotec, Bergisch Gladbach, Germany).

Preparation of cell lysates
Adherent cells were detached using non-enzymatic cell dissociation buffer (NECDB, Sigma-Aldrich, St. Louis, MO, USA) and then washed with PBS pH 7.4 before lysis. For whole cell lysates, each 1x10 7 cells to be lysed were pelleted and resuspended in 100μl of 1% Nonidet P-40, 150 mM NaCl, 5 mM EDTA, 1mM MgCl 2 , 50 mM Tris HCl pH 7.5, and 1x Proteoblock protease inhibitor cocktail (Fermentas, Vilnius, Lithuania) for 10 minutes on ice. Cell debris and nuclei were removed by centrifugation at 17000g at 4°C for 30 min; lysate supernatants were used immediately or snap frozen and stored at -80°C.
For lysates enriched for membrane proteins, each 1x10 8 cells were resuspended in 1ml of cold PBS pH 7.4 containing 1mM MgCl 2 . Cells were frozen and thawed twice and membrane proteins were pelleted by centrifugation for 30 min at 17000g at 4°C. Pellets were then resuspended and dissolved in the same lysis buffer used for whole cells.

shRNA silencing of cytokeratin 8 expression by MCF-7 cells
A lentiviral vector was used to stably transduce MCF-7 cells with shRNA construct designed to target human cytokeratin 8 mRNA. Transduction with an shRNA sequence with no target in the human genome was used as a negative control. shRNA oligomer pairs were annealed and ligated into AgeI/EcoRI cut pLKO-1-puro lentiviral vector (Addgene). HEK293T cells were transfected with the pLKO-1 constructs and the packaging plasmids pCMV-Δ8.9 and pMDG (a gift from Louise Boyle, CIMR, Cambridge). Supernatants containing virus were harvested after 4 days culture, 0.2μm-filtered and incubated with MCF-7 cells for 18hrs. Cells were subsequently placed under selection with 1.5μg/ml puromycin. Cells surviving selection and displaying low levels of cytokeratin 8 expression, as determined by the anti-human cytokeratin 8 monoclonal antibody 1E8, were sorted and cloned using a MoFlo cytometer (Beckman Coulter). Cytokeratin 8 expressing, puromycin resistant cells transduced with the control shRNA virus were maintained and used as bulk cultures.

Tissue culture conditions
All cell lines were cultured in RPMI 1640 supplemented with 10% FBS, penicillin-streptomycin (50 U/ml) and Amphotericin B (2.5 μg/ml) at 37°C in a 5% CO 2 in air atmosphere. Cell lines used were as follows: the EBV transformed B cell 721.221 (HLA-Class I deficient) both untransfected and stably transfected with HLA-G [1]; the EBV positive and negative Burkitt's lymphoma B cell lines Daudi (β 2 microglobulin deficient) and BJAB; the epithelial cell lines MCF-7 (breast glandular), T47D (breast ductal) and HCT-116 (colon); the human embryonic kidney cell line thought to be of neuronal origin HEK293T and the 2B4 cell line, a mouse T cell hybridoma stably transfected with a NFAT-GFP reporter construct (a kind gift from Lewis Lanier, UCSF, San Francisco, California, USA). Human breast epithelial cells of non-tumour origin were cultured as previously described [6]. All primary human material was derived from reduction mammoplasties carried out at Addenbrooke's NHS Trust, Cambridge, UK, under full informed consent and in accordance with Ethics Committee approval (08/H0308/178) as part of the Adult Breast Stem Cell Study.  Table S2) deposited in Genbank (http://www.ncbi.nlm.nih.gov/genbank/) (total number of sequences n=50; alignment A). Genbank sequences that contained unique substitutions were excluded from the alignment to avoid confusion from possible sequence errors. The p values and normalised dN-dS values are given in the left side of the table. A significance threshold of p<0.05 was used in the case of MEME, while a threshold of p<0.1 was used for the more conservative SLAC, FEL and IFEL tests. Results from FUBAR were considered significant when a posterior probability (Post. Pr.) of >0.95 was achieved.

SUPPLEMENTARY REFERENCES
Three further sequence alignments were analysed, consisting of: B) LILRB3 alleles sequenced in [3] (n=14); C) LILRA6 alleles [3] (n=11); and D) the combined alleles of LILRB3 and LILRA6 [3] (n=25). The summaries of these analyses are provided in the right side of the table, where a statistically significant result is denoted by a letter (A-D) that corresponds to the alignment in which it was detected. Only Codons that were discovered to be under significant selection by two or more tests are shown.