Mouse Intratongue Injection With Cancer Cells—An Orthotopic Model for Head and Neck Squamous Cell Carcinoma

Background

This protocol was inspired by previous work [1–3] demonstrating the preclinical relevance of tongue orthotopic injection in the study of head and neck cancers. Myers et al. [1] implanted human oral cancer cell lines directly into the tongue, where tumors formed more efficiently and more closely reproduced the histology and metastatic behavior of human disease than more classic subcutaneous models. Of note, this observation was confirmed and further studied recently from an immune landscape perspective [4]. Myers’ seminal work accelerated the development of metastatic progression studies, either focusing on the local tumor microenvironment characterization [5–7] or the factors influencing the formation of lung metastases. As such, Benaich et al. [2] focused on the role of miR-203, and Pascual et al. [3] studied how lipid metabolism affects survival and metastatic initiation.

Our protocol was developed and optimized for the completion of two studies: Punovuori et al. [8] focused on the contribution of extracellular matrix composition and cancer-associated fibroblasts in tumor progression, and Kaivola et al. [9] focused on the role of tissue stiffening in vocal fold cancer progression. We employed mouse intratongue injection as an in vivo model system to compare the growth of advanced-stage T3 head and neck squamous cell carcinoma (HNSCC) cells and their therapeutic response to treatment with a clinically relevant YAP-TEAD inhibitor. We also recently used this protocol in Punovuori et al. [8], where we reprogrammed cancer cells through cancer-stroma (HNSCC cell–cancer-associated fibroblast) crosstalk in vitro using spheroid co-cultures and compared the growth of two distinct HNSCC models injected in mice tongues with or without cancer-associated fibroblasts.

Most previously reported applications of intratongue (also called intralingual) injection involve injecting cell-free solutions of chemicals (like sodium selenite [10] or choleratoxin B-saporin [11]), prions [12], or virus particles (AAV), often in the context of optogenetic studies [13–18]. Further, it has also been used for acute pain studies using large needles and volumes under isoflurane anesthesia [19,20].

In contrast, there are only a few carcinogenesis [21] and tumor xenograft studies published for this orthotopic injection site, first proposed by the Fidler lab [1] (2002). Technically, downstream applications involved histology (tissue staining, immunohistochemistry, and immunostaining) and whole-animal non-invasive imaging (bioluminescence- or fluorescence-based). Intravital imaging [22] was also applied to the tongue using a non-invasive holder and 2-photon illumination post-intratongue injection to study tumor invasion and intravasation [23,24]. Of note, the experimental endpoint post-injection varies from two weeks to several months, depending on cell lines and animal models, as well as the variable cell numbers used. Lastly, similar to the orthotopic injection protocol described here, intramucosal injection in the mouse cheek has been described [25]. However, to the best of our knowledge, a detailed protocol for intratongue injection is still lacking. We believe that our refined protocol is compatible with high reproducibility and aligns extremely well with current ethics on animal welfare worldwide.

As a brief description of the protocol (Figure 1A), mice are first treated with analgesic and anti-inflammatory drugs. The injection mix for the intratongue injection is prepared. The mice are anesthetized with a ketamine/xylazine mix that will keep them unconscious for approximately 40 min. When ready, the injection needle is prepared, and the mice are suitably positioned to enable smooth intratongue injections (illustrated in Figure 1C). The injection procedure itself takes less than 1 min per animal. Mice are monitored until the anesthesia subsides. Analgesic and anti-inflammatory treatment is administered 6 h post-intratongue injections and continued for the following three days. Fresh, softened food is provided daily throughout the experiment. Cancer progression is closely monitored until the experimental objective is achieved and/or the animal reaches the humane endpoint. As this protocol includes injection inside the mouth, we typically observe a modest, transient decrease in body weight shortly after injection. Mice usually recover the lost weight in the first week post-injection. Then, when approaching the endpoint, animals lose weight rapidly and irreversibly (Figure 1B). After sacrifice, the tongues and potential metastatic organs are harvested and processed. As a control, in addition to the experimental controls, two non-injected animals are included to provide a reference to normal tissue.

Figure 1. Experimental design for intratongue injection and subsequent monitoring. (A) Experimental timeline and summary of key steps. (B) Typical weight curve of the animals during the protocol. (C) Schematic of the position of the animal on the holder for intratongue injection. (D) Schematic of the tongue injection spot from three angles. The injection needle comes nearly parallel to the tongue, on the side, and the needle does not enter more than 2 mm into the tissue to stay in the submucosal compartment. (E) Representative pictures of the mouse tongue, directly post-injection (C57BL/6 mouse). (F) Representative pictures of the mouse tongue during the monitoring of the tongue (NSG mouse): directly post-injection, after 1 week, and after 2 weeks (endpoint).

Based on our expertise and existing literature, intratongue injection in mice is a relevant experimental system in research contexts ranging from fundamental to pre-clinical research. The tools, methods, and technical tips presented here are applicable to any tongue injection, with or without cells. In cancer research, it should be used as an orthotopic site of injection for upper respiratory tract cancer models. As previously reported [26], growing HNSCC lines in classical subcutaneous assays in the flank of mice is challenging. In our hands, all tested HNSCC cell lines grew a tumor in the tongue [8,9]. However, it has, a priori, no added value for other cancers, in which other orthotopic sites or more classical subcutaneous injection are more relevant and/or safer for the animals.

Our protocol is compatible with any kind of cellular mixes (cancer, immune, or stromal cells, separately or together). We believe that the addition of Matrigel is instrumental for the early engraftment of the tumor cells post-injection, but one can work without it if the cellular model allows. If well-positioned, liquid solutions do not leak out of the tongue when you pull the needle out. This model is suitable for distinct systemic pre-treatment or treatment during the experiment, and therapeutic agents can be administered in different ways (drinking water, gavage, injections, etc.). Although we did not perform live imaging, this approach is possible after the injection, as mentioned earlier (using bioluminescence, intravital imaging, etc.). Furthermore, tumor growth can be observed visually and with caliper measurements of animals under low isoflurane anesthesia. Visual inspection enables excellent quality control and evaluation of the progression of the experiment. After injection, the administered Matrigel–cell mixture appears like a small translucent lentil. After a week, a small whiter spot appears on the tongue, which grows over the weeks. Post-sacrifice, tongue dissection is relatively simple if well executed, and sample processing complexity is dependent on the experimental goal. Lastly, although depending on the aggressiveness of the cancer cell line and the cell number injected, the experiment duration will vary [9]; animals in our study reached the clinical endpoint in two weeks (precisely reproduced in independent experiments). We believe this is the shortest time to be expected.

This protocol requires training in animal experimentation. It uses intraperitoneal, subcutaneous, and intratongue injections. Experimenters must be familiar with animal anesthesia procedures and safety. It also requires a good appreciation of animal signs of discomfort, and significant dedicated time for animal monitoring (regular weighing and daily softened food change). Toward the endpoint of the experiment, mouse health will deteriorate very quickly; thus, careful and thorough monitoring is necessary. We recommend following the OBSERVE consensus guidelines to monitor the animals [27]. Based on our experience, this protocol is non-trivial and should be performed by advanced animal experimenters with potential support from animal facility personnel. Of note, the injection protocol presented here does not strictly require the presence of a second experimenter, although it is generally advisable to work in pairs during animal experimentation, as it can help during certain steps.

This protocol is not compatible with resection of the primary tumor for long-term relapse studies. Tumor growth impairs the animal’s capacity to eat and, ultimately, to breathe. Animal pain signs are visible soon after the injection (prompting treatment with analgesics) and are later replaced by low-level animal stress (hyperactive animals, sometimes more aggressive than non-injected counterparts). These are ethical, as well as experimental, concerns that must be considered before starting the experiment, with a clear definition of endpoints.

Last, since this procedure includes animal experimentation, all experiments should be performed in accordance with relevant guidelines and regulations. Our protocol was ethically assessed and authorized by the National Animal Experiment Board and in accordance with the Finnish Act on Animal Experimentation (animal license numbers ESAVI/6253/2024).

In this article, we present standard histology, classical fluorescence imaging modalities (widefield and confocal), and atomic force microscopy (AFM) measurements as examples. Indeed, this protocol does not detail the processing of the sample after dissection, including paraffin embedding, sectioning, and staining, as these are standard and are out of the scope of this article. We used classic histology approaches that are standard for many organs, standard immunofluorescence staining approaches, and published AFM on the fresh tissue slice protocol [28]. Finally, it also does not detail standard drug administration procedures via intraperitoneal injection, subcutaneous injection, or oral gavage.