Robust expression of LINE-1 retrotransposon encoded proteins in oral squamous cell carcinoma

Abstract

Background: Oral Squamous Cell Carcinoma (OSCC) results from a series of genetic alteration in squamous cells. This particular type of cancer considers one of the most aggressive malignancies to control because of its frequent local invasions to the regional lymph node. Although several biomarkers have been reported, the key marker used to predict the behavior of the disease is largely unknown. Here we report Long INterpersed Element-1 (LINE1 or L1) retrotransposon activity in post-operative oral cancer samples. L1 is the only active retrotransposon occupying around 17% of the human genome with an estimated 500,000 copies. An active L1 encodes two proteins (L1ORF1p and L1ORF2p); both of which are critical in the process of retrotransposition. Several studies report that the L1 retrotransposon is highly active in many cancers. L1 activity is generally determined by assaying L1ORF1p because of its high expression and availability of the antibody. However, due to its lower expression and unavailability of a robust antibody, detection of L1ORF2p has been limited. L1ORF2p is the crucial protein in the process of retrotransposition as it provides endonuclease and reverse transcriptase (RT) activity.

Methods: Immunohistochemistry and Western blotting were performed on the post-operative oral cancer samples and murine tissues.

Results: Using in house novel antibodies against both the L1 proteins (L1ORF1p and L1ORF2p), we found L1 retrotransposon is extremely active in post-operative oral cancer tissues. Here, we report a novel human L1ORF2p antibody generated using an 80-amino-acid stretch from the RT domain, which is highly conserved among different species. The antibody detects significant L1ORF2p expression in human oral squamous cell carcinoma (OSCC) samples and murine germ tissues.

Conclusions: We report exceptionally high L1ORF1p and L1ORF2p expression in post-operative oral cancer samples. The novel L1ORF2p antibody reported in this study will serve as a useful tool to understand why L1 activity is deregulated in OSCC and how it contributes to the progression of this particular cancer. Cross-species reactivity of L1ORF2p antibody due to the conserved epitope will be useful to study the retrotransposon biology in mice and rat germ tissues.

Keywords: Cancer and L1 retrotransposon; L1 retrotransposon; L1ORF1p antibody; L1ORF2p antibody; OSCC and L1 retrotransposon.

Fig. 1

Generation of the antigen for the production of Human L1 ORF2 specific antibody: a Schematic diagram of full length active human L1 retrotransposon with two encoded proteins (L1ORF1p and L1ORF2p). L1ORF2p (1275 amino acids in length with predicted MW 150 kDa) has three partially characterized domains: endonuclease (EN) (AA:1–239), reverse transcriptase (RT) (AA: 453–883) and a cysteine-histidine-rich domain (CCHC) (AA: 1096–1275). The EcoRI-HindIII restriction fragment (AA: 479–558 fragment name hL1RT_EH) from the RT region was used as an antigen to make antibody against hL1ORF2p (shaded by yellow). b Alignment of the selected eighty amino acids stretches of the RT domain (hL1RT_EH) among human, mouse and rat L1. The selected RT stretch showed 76.25 and 73.75% identity at the protein level with the same stretch present in mice and rat L1ORF2, respectively. c Predicted structure of hL1RT_EH fragment and complete human RT domain (generated using PyMOL) [38]. d Sub-cloning scheme of human hL1RT_EH fragment in pET-28a bacterial expression vector. Bacterial expressed hL1RT_EH fragment encodes 130 amino acid polypeptides (from N terminal to C terminal AA: 1–42 vector, 43–124 RT fragment, and 125–130 vector) with predicted MW around 14 kDa. e Whole-cell lysate SDS-PAGE of E.coli expressed pET-hL1RT_EH. Induced protein with a molecular weight around 14 kDa is shown by the arrow. f Purification of hL1RT_EH from inclusion bodies by dissolving the pellet fraction in 8 M urea buffer (details in materials and methods). The purified protein in elution 1, 2 and 3 is shown by an arrow. g Dialysed and concentrated hL1RT_EH fragment protein (antigen) injected to mice for antibody generation. h Silver staining of purified hL1RT_EH (antigen) show the purity of antigen used to generate the antibody. i Western blotting of hL1RT_EH using an anti-His antibody. MW: Molecular Weight, FT: Flow-through

Fig. 2

Characterization of human L1 ORF2p antibody (anti-hL1ORF2): a) Immunoblot analysis with anti-hL1ORF2RT_EH on induced bacterial lysate obtained from and pEThL1ORF1_RRM, pET28A and pET-hL1RT_EH. Human L1ORF1p RRM antigen used to make anti-hL1ORF1p didn’t show any cross-reaction with anti-hL1ORF2p [26, 27] (b) Immunoblot analysis with anti-hL1ORF1_RRM on total bacterial lysate obtained from pET-hL1RT_EH, and pEThL1ORF1_RRM. No cross-reaction of anti-hL1ORF1with hL1RT_EH fragment used as antigen to make ORF2p antibody [26]. c Detection of L1ORF2p (exogenous), L1ORF1p (endogenous and exogenous) and GAPDH (endogenous) [26, 27] by immunoblotting in HEK 293T cells after transfecting full-length L1 construct (pL1_RPEGFP) [39]. All the original immuno-blots are shown in supplementary file

Fig. 3

Detection of L1ORF1p and L1ORF2p in mouse and rat tissues: a Immunoblot analysis of somatic and germline tissues of the mouse (testis, ovary, liver and kidney) with anti-L1ORF2 (panel 1), anti-L1ORF1 (panel 2), and anti-GAPDH (panel 3) as loading control. b Immunoblot analysis of somatic and germline tissues of rat (testis, ovary, liver and kidney) with anti- L1ORF2 (panel 1), anti-L1ORF1 (panel 2), and GAPDH (panel 3) as loading control. Original Western blots are attached in supplementary file. c Immunohistochemical analysis of somatic and germline tissues of the mouse (testis, ovary, liver and kidney) with anti- L1ORF1 (panel 2) and anti- L1ORF2 (panel 3). Tissue sections stained with hematoxylin-eosin are shown in panel 1. d Immunohistochemical analysis of somatic and germline tissues of rat (testis, ovary, liver and kidney) with anti-L1ORF1 (panel 2) and anti-L1ORF2 (panel 3). Hematoxylin-eosin stained samples are shown in panel 1

Fig. 4

Immuno-peroxidase detection of L1ORF2p expression in post-operative OSCC samples: Immunohistochemistry with anti-L1ORF2p was performed in total 39 post-operatives OSCC. Samples exhibit high (C1 and C2), moderate (C8, C11) and no expression (C3 and C4) (two representatives from each group are shown); IHC staining of the rest of the samples are shown in the supplementary figure. Immuno-staining with anti-His and non-immune mice sera didn’t show any signal. Staining with anti-GAPDH served as a positive control. Images were taken at 40X magnification. The Pie diagram showed more than 50% post-operative OSCC samples expressed a significant amount of L1ORF2p

Fig. 5

Immuno-peroxidase detection of L1ORF1p, L1ORF2p, Pan-CK, Ki-67 and p53 in operated OSCC samples. IHC staining of five samples (C1, C2, C8, C18, and C19) using all five antibodies are shown. The samples were also stained with hematoxylin and eosin

Fig. 6

Expression analysis of L1ORF1p and L1ORF2p in OSCC(a) Bar diagram showing the percent expressing L1ORF1p, L1ORF2p, L1ORF1p + L1ORF2p, p53, Pan-CK and Ki67. b Immunoblot analysis of L1ORF1p and L1ORF2p expression in paired normal cancer samples (C1, C2, C8, C18, and C19). Anti-L1ORF2p detects a distinct band at around 150 kDa corresponds to the MW of L1ORF2p in all the cancer tissues but not in paired normal. The same blot was re-probed with anti-L1ORF1p and anti-GAPDH (original blots are attached in supplementary file)