HMGA1 and HMGA2 expression and comparative analyses of HMGA2, Lin28and let-7 miRNAs in oral squamous cell carcinoma

This study included human squamous cell carcinoma, non neoplastic controls, and tumour derived cell line samples which were obtained from 12 patients (9 male, 3 female, age 20–71 years) who underwent surgery at the Department of Oral and Maxillofacial Surgery, Hannover Medical School. Ethical approval and informed patient consent was obtained for all patients. This study was approved by the local ethics committee at the Hannover Medical School (Ref No. 984–2011). No patients had received preoperative chemotherapy or radiotherapy. The tumours (patients 2–12) were staged according to TNM staging system and were classified as follows: patient 2- pT4apN1, patient 3- pT1pN0, patient 4- rpT0 rpN2b R2 M1, patient 5- pT3pN0, patient 6- pT4apN0, patient 7- pT4apN1, patient 8- pT4apN2b, patient 9- pT4apN2c, patient 10- pT4apN2b, patient 11- pT2pN0, and patient 12-pT4 pN0.

Seven canine tumour and two healthy control samples (five female, four male) were used covering seven breeds: Boxer, Fox Terrier, Irish Terrier, Landseer, Retriever, Sheltie (n = 1 respectively), and three Mixed-breeds. Age ranged between a half year and eleven years. Samples derived from the maxilla (4), tongue (2), mandible (1), palate (1), and pharynx (1). All tumours were analysed immunohistologically. All diagnoses were cytologically and histologically confirmed according to the WHO Nomenclature. The tumours were staged and graded as follows: patient 3- grade IV (poor) stage T3bN1bM0, patient 4- grade I (well) stage T2aN0M0, patient 5- grade I (well) stage T3bN1aM0, patient 6- grade I (well) stage T3bN1bM0, patient 7- grade III (moderate) stage T3bN1bM0, patient 8- grade I (well) stage T1aN1bM0, patient 9- grade I (well) stage T2aN0M0. The non neoplastic control samples were collected from clinically unaltered tongue and palate tissues and the dogs were euthanized due to oral squamous cell carcinoma unrelated diseases. All samples were taken and provided by the Small Animal Clinic, University of Veterinary Medicine, Hannover, Germany according to the legislation of the state of Lower Saxony, Germany.

Due to the possibility to access fresh neoplastic material of both species we decided to aim at an establishment of OSCC cell lines as tools for further experimental approaches. The successful establishment of new cell lines allowed us to compare the gene expression patterns of the neoplastic primary tissues and the cell lines of both species. The respective human and canine tumour samples were verified to be squamous cell carcinomas by routine histopathologic characterisation. The samples were analysed by either a human or veterinary pathologist respectively. Two human cell lines were generated from freshly isolated squamous cell carcinoma biopsies derived from patient 4 and patient 12 (tumour staging see above). Single cell suspensions were prepared with a gentleMACS™ tissue dissociator (Miltenyi Biotec GmbH, Bergisch Gladbach, Germany). Samples were cut into small pieces of approximately 5 mm, transferred to a C Tube (Miltenyi Biotec GmbH, Bergisch Gladbach, Germany) containing 5 ml Dulbecco’s modified eagle medium (DMEM PAA, Pasching, Austria) and subjected to the first homogenisation step. After the addition of 1500 Units collagenase I (Cell Systems, St. Katharinen, Germany) and 0,5 mg neutral peptidase (Cell Systems, St. Katharinen, Germany), samples were incubated for 40 min at 37°C. Digested samples were subjected to a second homogenisation step followed by removal of tissue debris using a 70 μM cell strainer (BD Biosciences, Heidelberg, Germany). The cells were washed twice with culture medium (DMEM, 10% fetal calf serum, 20 mM Hepes, 1000 IU/ml penicillin and 0.1 mg/ml streptomycin; all PAA, Pasching, Austria) and plated on 100 mm cell culture dishes (Greiner, Frickenhausen, Germany) with DMEM and incubated at 37 C and 5% CO₂ until confluent.

The canine cell line was generated from a freshly isolated oral squamous cell carcinoma biopsy. Due to the limited amount of bioptic material this sample was not used in the primary tissue screenings. The tumour tissue sample was cut into small pieces with a sterile scalpel and treated with collagenase (0.26%) for 2 hours at 37°C. The dissociated cells were transferred into a sterile 10 ml tube and centrifuged for 10 min at 1000*g. After centrifugation the supernatant was discarded and the resuspended cell pellet transferred into a sterile flask and incubated in 5 ml culture medium (Medium 199 (Invitrogen, Frankfurt, Germany), 10% fetal calf serum (PAA, Pasching, Austria) 200 U/ml penicillin and 200 ng/ml streptomycin (Biochrom, Berlin, Germany) and incubated at 37°C and 5% CO₂ until confluent.

Tissue samples were homogenised using the stainless steel-beads and Qiagen-TissueLyser II homogeniser method accordingly to the manufacturer’s instructions (Qiagen, Hilden, Germany). Lysates of cultured cells were homogenised with QIAshredder columns accordingly to the manufacturer’s protocol (Qiagen, Hilden, Germany).

RNA from tissue samples and cultured cells was isolated using the RNeasy Mini Kit according to the manufacturer’s instructions (Qiagen, Hilden, Germany). On-column DNase digestion was performed with the RNase-Free DNase set (Qiagen, Hilden, Germany). cDNA syntheses was performed using 250 ng RNA and the QuantiTect Reverse Transcription Kit following the manufacturer’s protocol (Qiagen, Hilden, Germany).

Furthermore, total RNA including small RNAs like miRNAs was isolated using the mirVana miRNA Isolation Kit according to the manufacturer’s instructions (Ambion, Applied Biosystems, Darmstadt, Germany). The respective cDNA syntheses were performed using 100 ng total RNA of each sample and the TaqMan MicroRNA Reverse Transcription Kit following the manufacturer’s protocol (Applied Biosystems, Darmstadt, Germany).

Relative quantification real time PCRs for both species were carried out using the Eppendorf Mastercycler ep realplex real-time PCR System (Eppendorf AG, Hamburg, Germany).

For analysis of the human target genes, 2 μl of each cDNA was amplified in a total volume of 25 μl using universal PCR Mastermix and commercially purchased TaqMan gene Expression Assays (HMGA1– Assay ID: Hs00600784_g1; HMGA2– Assay ID: Hs00971724_m1; Lin28A- Assay ID: Hs04189307_g1; HPRT- Assay ID: Hs02800695_m1; GUSB- Assay ID: Hs99999808_m1; (Applied Biosystems, Darmstadt, Germany)).

For analysis of canine target genes, 2 μl of each cDNA was amplified using universal PCR Mastermix, self-designed TaqMan based Assays ([9, 22] (canine HMGA1 (NM_001003387)- forward primer: 5′ ACCCAGTGAAGTGCCAACACCTAA 3′, reverse primer: 5′ CCTCCTTCTCCAGTTTTTTGGGTCT 3′, probe: 5′ 6-FAM-AGGGTGCTGCCAAGACCCGGAAAACTACCA-TAMRA 3′; canine HMGA2 (DQ316099)- forward primer: 5′ AGTCCCTCCAAAGCAGCTCAAAAG 3′, reverse primer: 5′ GCCATTTCCTAGGTCTGCCTC 3′, probe: 5′ 6-FAM-CGCCCACTACTATGCCATCGTGTG-TAMRA 3′; canine HPRT (NM_001003357)- forward primer: 5′CCTTCTGCAGGAGAACCT 3′, reverse primer: 5′TCATCACTAATCACGACGCT 3′, probe: 5′6-FAM-CCTCCTGTTCAGGCTGCCGTCA-TAMRA 3′; canine GUSB (NM_001003191)- forward primer: 5′ TGGTGCTGAGGATTGGCA 3′, reverse primer: 5′ CTGCCACATGGACCCCATTC 3′, probe: 5′ 6-FAM-CGCCCACTACTATGCCATCGTGTG-TAMRA 3′) and commercially purchased TaqMan gene Expression Assays (Lin28- Assay ID: Cf02725509_g1 (Applied Biosystems, Darmstadt, Germany)).

The canine and human HMGA1 qPCR assays detected both splicing variants (HMGA1a and HMGA1b) simultaneously. PCR conditions were as follows: 10 min at 95°C, followed by 45 cycles with 15 s at 95°C and 1 min at 60°C. All human and canine samples were measured in triplicate and for each run non-template controls and non-reverse transcriptase control reactions were included.

Relative quantification of the human and canine let-7a, mir-98 and RNU6B micro RNA transcript levels were carried out using 1.33 μl of each cDNA amplified in a total volume of 20 μl using TaqMan Universal PCR Master Mix, No AmpErase UNG and TaqMan MicroRNA assays for each gene (Let-7a- Assay ID: 000377; mir-98- Assay ID: 000577; RNU6B- Assay ID: 001093 (Applied Biosystems, Darmstadt, Germany)).

PCR conditions were as follows: 10 min at 95°C, followed by 45 cycles with 15 s at 95°C and 1 min at 60°C. All samples were measured in triplicate and for each run non-template controls and non-reverse transcriptase control reactions were included.

A precedent efficiency analysis of the microRNA PCR assays which were used in this study was performed by applying the same template and dilution steps.

The formalin-fixed specimens were embedded in paraffin, sectioned at 4 μm and routinely stained with haematoxylin and eosin (H&E). Thereafter, polyclonal goat-anti human HMGA2 (R&D Systems, Minneapolis, MN, USA) (1:400) or mouse-anti human Ki-67 antibody (Dianova, Hamburg, Germany) (1:100) was applied and allowed to incubate for approximately 16–18 h. Sections were incubated with biotin-conjugated horse antibody to goat IgG or goat anti-mouse IgG (both Vector Laboratories, Burlingame, CA, USA) (1:200) followed by ABC solution (Vectastain Elite ABC kit, Vector). Tyramine amplification reaction was performed according to the method of Adams (1992) [23] for 15 min (only HMGA2). The chromogen 3-amino-9-ethyl-carbazol (AEC) (Biologo, Kronshagen, Germany) was used for visualization followed by counterstaining with Mayer’s haematoxylin. Negative control sections were prepared by substituting the primary antibody with PBS. For the scoring, the percentage of carcinoma cells with intense red positive nuclear labelling for HGMA2 was estimated by examining the centre of the tumour and the invasive front (0: no expression; <25% weak expression; 25-50% moderate expression; >50%: strong expression).

Statistical analysis of the relative real time PCR results applying the Hypothesis Test was performed with the Relative expression software tool REST 2008, version 2.0.7 [24]. A p-value of <0.05 was considered statistically significant.