The invasion of bladder cancer into the sub-urothelial muscle and vasculature are key determinants leading to lethal metastatic progression. However, the molecular basis is poorly understood, partly because of the lack of uncomplicated and reliable models that recapitulate the biology of locally invasive disease. We developed a surgical grafting technique, characterized by a simple, rapid, reproducible and high-efficiency approach, to recapitulate the pathobiological events of human bladder cancer invasion in mice. This technique consists of a small laparotomy and direct implantation of human cancer cells into the bladder lumen. Unlike other protocols, it does not require debriding of the urothelial lining, injection into the bladder wall, specialized imaging equipment, bladder catheterization or costly surgical equipment. With minimal practice, the procedure can be executed in <10 min. Tumors develop with a high take rate, and most cell lines exhibit local invasion within 4 weeks of implantation.


Invasive bladder cancer is deadly and is the sixth most common malignancy in the United States1 and one of the most widespread carcinomas around the globe2. Patients with tumors exhibiting local muscle invasion (MI) and lymphovascular invasion (LVI) carry the worst prognosis for metastasis and death3,4,5,6,7. Advanced bladder cancer may initially respond to chemotherapy; however, remissions are not durable and cures are rare3,8. Despite contemporary therapies, the median survival of 1 year for patients with metastatic bladder cancer following chemotherapy, has not been substantially improved for >3 decades. These facts illustrate the critical need to develop therapeutic strategies designed to inhibit bladder cancer in the early stages of invasion and prevent lethal metastasis. This requires biologically relevant, high-throughput models for preclinical evaluation of new therapies or approaches at various stages of tumor progression.

Although some very useful protocols exist for the establishment of orthotopic bladder cancer xenografts in mice9,10, many of them suffer from obstacles that challenge their feasibility and reproducibility. First, they are time consuming, requiring specialized surgical equipment and a high level of technical skill. Researchers must be experienced with catheterization and ultrasound technology, and have a strong understanding of bladder anatomy to successfully execute the protocols. Second, most of the available protocols focus primarily on tumor growth and have not examined in-depth pathobiological development of invasive human bladder cancer. To address these obstacles, our group developed a novel technique that exhibits local MI and LVI of human bladder cancer in the murine bladder. This protocol was recently used by Han et al.11 to investigate the molecular mechanisms of muscle-invasive bladder cancer. This protocol is based on a simple and rapid implantation of bladder cancer cells through the bladder wall directly into the bladder lumen of mice (Fig. 1). We describe how to implement this model here.

Fig. 1: Schematic illustration comparing the development of normal human bladder cancer to the development of bladder cancer in mice after this protocol is performed.
Fig. 1

a, Development of bladder cancer and spontaneous muscle invasion, as observed in paraffin-embedded human bladder tumor samples provided by A.M.U. These samples were identified as muscle- and lymphovascular-invasive urothelial carcinoma. The patient’s written informed consent was obtained; in this document, the patient was assured of confidentiality, as well as compliance with national and institutional guidelines. These procedures were approved by the University of Michigan Medical School Institutional Review Board. b, Xenografts mimicking human bladder cancer growth and spontaneous muscle invasion in days to weeks after inoculation into the lumen of the mouse’s bladder. a,b, H&E images represent muscle-invasive human (a, right) and human-in-mouse (b, right) bladder cancer. Scale bars, 200 μm. All mouse experiments were approved by the Institutional Animal Care and Use Committee of the University of Michigan. The use of human bladder specimens was approved by the University of Michigan Medical School Institutional Review Board. T, tumor; M, muscle.

Comparison with other mouse models of bladder cancer

Our orthotopic technique of direct bladder lumen injection is most similar to the intramural bladder wall xenograft protocol12,13. Both techniques use laparotomies, expose the bladder and use needles to inject human bladder cancer cells. The most obvious difference is that the intramural technique requires precise placement of the cells in the bladder wall without puncturing the bladder mucosa. By contrast, our protocol first removes urine from the bladder to create consistent working conditions and to allow room for the cell suspension. This is followed by an injection through the bladder wall and mucosa directly into the luminal space of the mouse bladder. The luminal injection protocol is most useful when large numbers of mice are required and when a user wishes to look at tumor progression from the mucosal lining.

The presence of tumor cells within the bladder can be visualized on the day the tumor cells are implanted via fluorescence imaging, whereas the development of intra-bladder tumors can be followed in real time via bioluminescence in live animals over several weeks. Most of the human bladder cancer cell lines tested by our group developed invasive tumors in the bladder within 4 weeks while reproducing the key biological features of invasive human bladder cancer (Fig. 1).

Even when exquisite alternative protocols for orthotopic xenotransplantation of bladder cancer cells exist, the technique described here offers several distinct advantages over existing comparable models (Table 1). Our protocol is defined by a series of key features and steps that allow us to model the early invasive progression of human bladder cancer. Because this technique can be applied to both male and female adult mice and is not limited to female mice9,10,13,14,15,16,17, it better reflects disease that occurs in males and females. This contrasts with most previous models, which have reproduced tumor growth mostly in female mice, which is problematic because bladder cancer incidence is threefold higher in men than in women1.

Table 1 Comparison of alternative orthotopic xenografting models

This protocol accurately reproduces the spontaneous invasive potential of individual cellular models. Sporadic spontaneous metastasis is also observed with this technique after 4 weeks of engraftment. Third, other models9,10,12,13,18 have been validated using a few bladder cancer cell lines, with the KU-7 cells being the most commonly used, whereas we successfully reproduced our orthotopic xenotransplantation technique with 13 different bladder cancer cell lines. Unlike alternative protocols that depend on scarification of the luminal urothelium9,10,14,15,16,17 or injection into the bladder wall10,12,13,19,20,21, our protocol recapitulates bladder cancer progression without chemical treatment of the luminal urothelium or major injury of the organ before implantation. Our method is also quicker to establish, as some alternative techniques require up to an hour for completion of surgery on one animal9,10,12,13. Procedures based on catheterization and intravesical instillation also display high rates of procedure-related complications and animal mortality9,18. By contrast, our model is rapid and easily mastered, making it amenable to larger high-throughput studies.

Potential applications of the protocol and future directions

The main purpose of this technique is to replicate the biology of human bladder cancer progression and the critical steps leading to local invasion into both the bladder musculature and the lymphovascular system. It can be used to examine general tumor progression, basic and translational tumor biology, and preclinical drug testing.

This technique is of special value to a broad array of scientists interested in tumor invasion and metastasis, or novel drug development, as well as those studying the molecular pathogenesis of bladder cancer. We and other molecular tumor biologists have utilized this model to interrogate the molecular mechanisms of human bladder cancer invasion11. This technique has been successfully used by several laboratories at the University of Michigan. Currently, colleagues in medical oncology and pharmacology are using this model to evaluate drug formulation, delivery and efficacy. In addition, we are utilizing this protocol in collaboration with several biopharmaceutical companies interested in this technique for the evaluation of in vivo efficacy of experimental compounds and in the investigation of novel molecular targets. This protocol will be important in pharmaceutical research and academic research, as well as research in education and practical training in animal modeling.

The outlined technique is very versatile; beyond its applicability to human bladder cancer pathobiology studies, it can be adapted for the evaluation of drug efficacy or new bladder cancer therapies by direct instillation of drug into the bladder. For this purpose, survival experiments can also be designed using lower cell counts as inoculum and allowing the tumor to develop for extended periods of time. In addition, our research group is working on the optimization of a similar experimental design for the visualization and subsequent quantification of spontaneous metastasis. Early time points can also be used for the study of initial submucosal invasion within the mouse bladder visualized by standard immunohistochemical analysis or whole mounts.

Expertise needed to implement the protocol

Because of its simplicity, extensive technical ability and previous training are not necessary to execute this protocol. It is expected that any competent undergraduate student, graduate student, postdoc or technician could successfully follow the outlined protocol and rapidly become proficient.


For the inoculation of human cancer cells into mice, immunocompromised mouse strains must be used to ensure that the tumor cells are not rejected by the recipient mouse. The relationship between cancer and the immune system has been very well documented22,23. Thus, the requirement for immunocompromised mice, with attenuated immune components, remains a limitation for disease modeling and immunotherapy investigation in bladder cancer24. Another limitation is the engraftment of human cancer cells in a mouse host, which does not fully represent all aspects of human pathophysiology24. However, we have been able to recapitulate many of the invasive features of human bladder cancer progression using this xenograft technique. Therefore, by modeling critical features of human disease, our human-in-mouse model can efficiently be used for the pathobiological and preclinical evaluation of bladder cancer. In addition, post-surgical complications such as infection and tumor obstruction of the urethra or ureters can arise before the end point, requiring early sacrifice of the animal and loss of data; however this is rare.

Experimental design

Our orthotopic xenografting method is characterized by its simplicity, swiftness and ability to spontaneously reproduce the early events of human bladder cancer progression within the bladder microenvironment (Fig. 1), including local MI and LVI. Relying on a small laparotomy and direct inoculation of the cancer cells into the bladder lumen of mice (Fig. 2, Supplementary Video 1), this process does not entail extensive damage to the urothelium or the bladder wall. In addition, the technique does not require special surgical or imaging equipment, special expertise, or critical anatomical knowledge.

Fig. 2: Orthotopic inoculation of bladder cancer cells in a recipient NOD-SCID mouse.
Fig. 2

a, Schematic representation of the anatomical location of the mouse bladder. bl, Step-by-step visual description of the orthotopic implantation procedure (see also Supplementary Figs. 1–4 and Supplementary Video 1). b, The surgical plane is cleared of hair with depilatory cream, and the skin is disinfected with iodine. c, A 1-cm incision is made through the skin and abdominopelvic wall, exposing the internal organs. d, The bladder is exteriorized. e, Urine is aspirated from the bladder, clearing the bladder lumen for inoculation. f, The diagram shows the needle placement during urine aspiration by piercing the dome of the bladder until the needle point is within the lumen. g, The base of the bladder is exposed by bringing the superior end of the bladder to the inferior end of the pelvic area. h, The needle is gently inserted into the bladder base until visualized in the vesical lumen. Then, the cancer cells are slowly injected. i, Needle placement during the intra-bladder implantation, by inserting the needle point into the region where the ureters drain into the organ. j, A cotton-tipped swab is used to absorb bodily fluids. k, The bladder is returned to the abdominopelvic cavity. The abdominal muscle and skin are sutured and closed. l, The skin is closed with surgical glue (as pictured) or wound clips. This procedure was approved by the Institutional Animal Care and Use Committee of the University of Michigan and followed the animal welfare recommendations of the National Institutes of Health.

Usually, the entire procedure, from anesthesia to the surgical end point, takes <10 min (Fig. 2). Although this protocol could be performed by one individual, we prefer a team of two to three, depending on available space and the number of animals included in the experiment. When we use a team, we distribute the work as follows. One person is responsible for the preparation of human cancer cells. While the cells are prepared, another person prepares the surgical space (Supplementary Fig. 1). When performing the surgeries, one person is responsible for monitoring the animals, recording important notes and preparing the syringes with cells, and the other individual(s) perform(s) the laparotomy and implantation. Distributing the tasks reduces the time required for the procedure, as well as the variability within the experiments and animal recovery.

During the procedure (Fig. 2, Supplementary Video 1), the mouse is placed in dorsal recumbency. Then, a laparotomy with an incision size of 1 cm is performed to open the abdominal cavity to exteriorize the bladder. Minimizing the laparotomy incision size limits inflammation at the surgical site, and decreases the time needed to close the incision and the risk of infection. Once the bladder is located and exteriorized, the urine is aspirated from the apex with a syringe and the cellular suspension is injected into the empty lumen. These steps are the most critical and unique characteristics of our protocol, which does not require catheterization or assistance of special ultrasound equipment, denuding of the mucosa or extensive damage to the ventral bladder during implantation. Visualization of the injection process is also key to avoiding leakage and tumor seeding into the abdominopelvic cavity. After the cells are injected, the bladder is returned to the abdominal cavity and the laparotomy incision is closed using standard surgical methods. The mice are then allowed to wake from anesthesia, with the only restriction being that their water is withheld for 1 h, which reduces post-surgical urination, giving the cells time to adequately interact with the urothelium and seed into the bladder.

Adaptations to the experimental design are required if different cellular models and murine strains are used. If a researcher is interested in using a cell line other than those mentioned here, we recommend running a pilot study. Although we discuss only the cell lines used by our group, we have successfully used many bladder cancer cell lines with this protocol. For a pilot study, the researcher must determine the optimal number of cells to inoculate and the time required for the tumors to develop. We recommend starting with a cell count of 1 × 106 cells per animal, with an experimental end point at 4 weeks. As the experiment nears completion, special attention should be given to the animal’s overall health and behavior to ensure that the tumor is not causing the animal pain or distress. Once the experiment is completed and the histopathology has been analyzed, the optimal cell count and time of tumor development can be calculated. Besides broad utilization of non-obese diabetic/severe combined immunodeficient (NOD-SCID) mice, we have gained recent experience implementing this technique with NOD-SCID gamma (NSG) and nude mice, showing that these strains represent suitable alternative hosts. Therefore, this suggests that this model could be successfully adapted to other immunocompromised murine strains. If this protocol is standardized with a different strain, a pilot study should be performed first. This technique could also be adapted for utilization with humanized mice.


For this protocol, proper controls include age, gender and strain-matched mice to minimize biological variations. It is best to use mice from the same litter with the same date of birth whenever possible. Mice for our experiments generally ranged in age from 6 to 12 weeks old. Most commonly, bladder cancer develops in adults; thus, it is important to use mature adult mice for bladder cancer experiments. Although most of our work has been completed with NOD-SCID mice, alternative immunocompromised strains such as NSG and nude mice can be successfully used for this protocol. Finally, because bladder cancer affects men and women, sex as a biological variable should be addressed by using both male and female mice to have the best representation of patients. In addition, experimental controls must be included within the study. As this protocol is a surgical procedure, surgical controls, or sham mice, should be included in the experiment. The sham mice undergo the surgical procedure, but instead of injecting cells, saline or culture medium is injected into the bladder. These controls are especially important for comparing histological differences between mice with tumors and mice without tumors. When imaging the mice, it is important to include a negative control mouse with no bioluminescent or fluorescent signal. This will determine the background threshold. If this protocol is used for drug efficacy testing, a vehicle control group should also be included, as well as an untreated control group. The number of control animals should be kept to the lowest number necessary to achieve the experimental observations with the proper statistical power (P < 0.05).


Biological materials

  • Laboratory animals. We use NOD-SCID mice (NOD.CB17-Prkdcscid/J) (Jackson Laboratory, cat. no. 001303), NSG mice (NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ) (Jackson Laboratory, cat. no. 005557) and nude mice (Charles River, cat. no. 490). We purchased breeders from the originating laboratory, and breeding was carried out by the Unit for Laboratory Animal Medicine (ULAM) Breeding core facility at the University of Michigan


    All animal experiments should be performed in accordance with federal and local guidelines and regulations. Our experiments performed with animals were approved by the Animal Welfare Committee of the University of Michigan. Animals were housed in the ULAM housing facility in Ann Arbor, Michigan.


    The animals in our experiments range in age from 6 to 12 weeks old, and both male and female adult mice are used. The presence of thymic lymphoma is commonly observed in the NOD-SCID strain, and it should be monitored in all adult mice.

  • Human bladder cancer cells. We have established orthotopic bladder xenografts with 13 human bladder cancer cell lines. In this protocol, we describe results from using the human transitional cell carcinoma lines UM-UC-6 (refs. 25,26), UM-UC-9 (ref. 26) and UM-UC-13 (ref. 27), which were generously obtained from the originators (H.B. Grossman, University of Texas MD Anderson Cancer Center; and M. Liebert, University of Michigan) and banked in our laboratory. Human transitional cell carcinoma lines SW 780 (cat. no. CRL-2169) and T24 (cat. no. HTB-4) can be used and purchased from the American Type Culture Collection (ATCC). All cells should be grown in DMEM with high glucose, enriched with 10% (vol/vol) FBS, in a humidified incubator with an atmosphere of 5% CO2 at 37 °C


    The identity of the cell lines should be genetically verified by restriction fragment length polymorphism. We tested the identity of our cells through the Research Animal Diagnostic Laboratory, Columbia, MO.


    All cells should be maintained for a limited number of passages and evaluated for Mycoplasma sp. contamination. We regularly test the cell lines using PlasmoTest (InvivoGen, cat. no. rep-pt1) before commencing with experiments. In addition, all cell lines were shown to be free of infectious agents by Charles River Research Animal Diagnostic Services.


Cell culture

  • DMEM with l-glutamine (4 mM), sodium pyruvate (1 mM), phenol red and glucose (25 mM) (Thermo Fisher Scientific, cat. no. SH30243.02)

  • FBS (Thermo Fisher Scientific, cat. no. SH300070.03)

  • TrypLE (Thermo Fisher Scientific, cat. no. 12604-013)

  • PBS (Thermo Fisher Scientific, cat. no. SH30256.02)

  • 0.4% (wt/vol) Trypan blue (Thermo Fisher Scientific, cat. no. 15250061)

(Optional) Cell labeling


See ‘Reagent setup’ for instructions on how to use the following reagents.

  • pLentipuro3/TO/V5-GW/EGFP-firefly luciferase plasmid (deposited in Addgene: ID 119816)

  • EGFP (from pEGFP-C1, Clontech Laboratories, cat. no. 6084-1)

  • Firefly luciferase (from pGL3-Basic, Promega)

  • Phusion High-Fidelity DNA polymerase (New England BioLabs, cat. no. M0530S)

  • pENTR/D-TOPO (Thermo Fisher Scientific, cat. no. K240020)

  • Gibson Assembly Kit (New England BioLabs, cat. no. E5510S)

  • pLentipuro3/TO/V5-DEST (a gift from A. Aplin, Thomas Jefferson University)

  • LR Clonase II enzyme mix (Thermo Fisher Scientific, cat. no. 11791100)

  • Qtracker Cell Labeling Kit (Qtracker 655 (red), Thermo Fisher Scientific, cat. no. Q25021MP; or Qtracker 525 (green), Thermo Fisher Scientific, cat. no. Q25041MP)


  • Isoflurane (Fluriso; Vet One, cat. no. 501017)

  • Medical oxygen (21.5% (vol/vol) O2, 78.5% (vol/vol) N2)

  • Povidone iodine scrub solution (7.5% (wt/vol); Medline, cat. no. MDS093908)

  • Saline (Thermo Fisher Scientific, cat. no. 12608697)

  • Carprofen (Rimadyl; Zoetis)


  • Conical 15-mL tubes (Fisher Scientific, cat. no. 0644321)

  • Conical 50-mL tubes (Falcon, cat. no. 352196)

  • 5-mL Sterile tubes (Eppendorf, cat. no. 0030119460)

  • Inverted microscope with bright field (Nikon, model no. DIAPHOT 200)

  • Automated cell counter (TC20; Bio-Rad Laboratories, cat. no. 1450102)

  • Cell-counting slides for TC20 cell counter, dual chamber (Bio-Rad Laboratories, cat. no. 1450015)

  • CO2 incubator (5% (vol/vol) CO2, 37 °C)

  • 1.5-mL Microcentrifuge tubes (Eppendorf, cat. no. 0030120086)

  • Refrigerated centrifuge with swinging-bucket rotor (Eppendorf, cat. no. 5810 R)

  • 100-mm Dishes (tissue culture treated; Corning, cat. no. 353003)

  • Bucket with ice

  • Instrument sterilization cassette (World Precision Instruments, cat. no. WP-3019 or lab-prepared surgical wrap)

  • Stevens needle holder (World Precision Instruments, cat. no. 14133)

  • Iris scissors (World Precision Instruments, cat. no. 14225-G)

  • Ceramic-coated cupped forceps (World Precision Instruments, cat. no. 501984)

  • Adson forceps (World Precision Instruments, cat. no. 14226)

  • Tissue forceps (World Precision Instruments, cat. no. 15918)

  • Micro-dissecting angled forceps (Roboz Surgical Store, cat. no. RS-5359)

  • Pointed forceps (Roboz Surgical Store, cat. no. RS-5043)

  • Absorbable sterile synthetic sutures (5-0 coated Vicryl (polyglactin 910), 18-inch C-3 reverse cutting; Ethicon US, cat. no. J385H)

  • Sterile towel drape (Dynarex, cat. no. 4410)

  • Veterinary ophthalmic ointment (Paralube; Dechra Veterinary Products)

  • Tissue adhesive (Vetbond; 3M, cat. no. 1469SB)

  • Syringes (1-mL tuberculin, slip tip; BD Medical, cat. no. 309659)

  • 30 1/2-gauge needles (Terumo, cat. no. 0197)

  • Sterile alcohol prep pads (Fisherbrand, cat. no. 22-363-750)

  • Sterile half drape (Kimberly-Clark, cat. no. 89101)

  • Tool sterilizer (Germinator 500; Braintree Scientific, cat. no. GER 5287-120V)

  • Traditional isoflurane vaporizer (Kent Scientific, cat. no. VetFlo-1210S)

  • Disposable feather scalpel (no. 21; Thermo Fisher Scientific, cat. no. NC0159443)

  • Depilatory lotion (Nair; Church & Dwight)

  • Sterilized gauze

  • Cotton-tipped swabs

  • Warming pads

  • Scale

  • Labeling tape, pens and markers

Reagent setup

Complete DMEM

Mix 500 mL of DMEM with 50 mL of FBS in sterile flow hood. Store at 4 °C for up to 1 year.

(Optional) Luciferase labeling

Amplify the pLentipuro3/TO/V5-GW/EGFP-firefly luciferase plasmid and package into lentivirus. Transduce the lentivirus into bladder cancer cells as we previously described28,29. The pLentipuro3/TO/V5-GW/EGFP-firefly luciferase lentivirus plasmid is generated as follows: EGFP (from pEGFP-C1) and firefly luciferase (from pGL3-Basic) are amplified with Phusion High-Fidelity DNA polymerase and cloned into pENTR/D-TOPO using the Gibson Assembly Kit. The EGFP-luciferase coding region is recombined into pLentipuro3/TO/V5-DEST using LR Clonase II enzyme mix. The resultant pLentipuro3/TO/V5-GW/EGFP-firefly luciferase lentivirus plasmid is packaged (in our case, by the University of Michigan Vector Core) and transduced into bladder cancer cells as reported28,29.

(Optional) Quantum dot labeling

Fluorescently label bladder cancer cells using red or green quantum dots and following the Qtracker Cell Labeling Kit manufacturer’s recommendations.

Animal welfare

Carprofen (Rimadyl; 5 mg/kg) can be stored under refrigeration (2–8 °C for up to 6 months). Carprofen should be diluted to 0.05% (wt/vol) with saline solution (0.9% (wt/vol) NaCl) before use.


Always prepare working solutions fresh before each procedure.

Equipment setup

Surgical room

The Guide for the Care and Use of Laboratory Animals sets forth that any survival surgery performed on rodents or other small species should be executed in a space dedicated to only surgery or related activities. The area must be easily sanitized and not used for any other purpose than surgery for the full duration of the procedure. The surgery must be performed in an aseptic environment, with access to an isoflurane anesthesia machine and ventilator (Supplementary Fig. 1). Ensure that enough isoflurane and medical oxygen are present to complete all surgeries.

Surgical preparation

Before initiating surgery, the following tools must be sterilized using a sterilization cassette (Supplementary Fig. 1): ceramic-coated cupped forceps, Adson forceps, tissue forceps, Stevens needle holder, micro-dissecting angled forceps, pointed forceps, iris scissors and cotton-tipped swabs. Ensure that there are enough autoclaved tool packs so that a new sterilized tool pack can be used for each new cage of mice. Thoroughly wash your hands and lower arms with disinfecting soap. Dress with all personal protective equipment, including gown, hair bonnet, face mask, sterile gloves, shoe covers and eye protection, before commencing further activities in the surgical room. Wipe the surgical station well with disinfectant before placing a sterile drape over the table. Place the sterile tools, sterile gauze and cotton tips, individually wrapped alcohol pads and sutures on the dominant side of the surgeon. Place a warming pad at the nose cone, and place a sterile towel drape over the warming pad. Be sure that there is a sharps container and a container for garbage nearby for waste. Turn on the hot-bead sterilizer to sterilize tools between animals so that it reaches proper temperature for sterilization. Tools must be sterilized between animals of the same cage.

Animal preparation

Before initiation of the experiment, the lowest number of mice necessary for statistical power should be calculated. Only healthy mice that have been fully evaluated before surgery should be used. The day before the surgeries are to occur, the animals must be weighed, and they must receive an identifier (tattoo, ear tag and tail mark) that will remain for the duration of the experiment. The hair of NOD-SCID mice, and other furry strains such as NSG, must be removed at the incision site. We prefer to use depilatory lotion because it can fully remove the coat of animal fur (Supplementary Fig. 2). Using a dime size (18 mm) amount of lotion, remove the hair by working into the fur, using a quick but gentle, circular motion. Rinse and remove any excess lotion left on the animal, as prolonged contact can irritate the animal’s skin.


The protocol must be approved by the institute’s Animal Care and Use Committee. All animals must receive humane care according to the criteria outlined in the United States National Institutes of Health Guide for the Care and Use of Laboratory Animals.


Do not remove fur on the day of surgery, as wet fur can lead to hypothermia while the animals are under anesthesia. Clipping is the only acceptable alternative to a depilatory lotion for removal of the mouse hair. Do not shave the animals; small lacerations can lead to infections at the incision site.

Animal housing

House animals under standard housing conditions in ventilated cages. Immediately following surgery, place mice in a cage on a warming pad to assist in recovery and withhold water for 1 h. Following surgery and successful recovery from anesthesia, mice that were previously housed together can remain together. Do not place mice that have incisions in the same cage as mice that did not undergo surgery, as this could lead to fighting and the recovering animals within the cage could be harmed.


Preparation of human bladder cancer cells

Timing ~15 min

  1. 1

    Before commencing work with cells, pre-warm PBS, supplemented culture medium and TrypLE to room temperature (23 °C). Prepare a bucket of ice.

  2. 2

    Aspirate culture medium from plates of healthy cells at 60–80% confluence.


    Cancer cells are a biohazard and must be handled with care. Wear gloves, lab coat and eye protection when handling. This procedure must be performed only under a laminar flow hood.


    To maintain sterility, all cell work must be performed under a laminar flow hood.

  3. 3

    Detach cells from the plate with nonenzymatic cell dissociation solution, TrypLE. Ensure that the TrypLE solution covers the full area of the dish. Place the dish in a 37 °C incubator.

    Critical step

    Our lab has worked with trypsin and TrypLE when preparing cells for animal inoculation and have repeatedly found a better tumor take when TrypLE is used as the dissociation reagent. It is gentler on cells and does not denude the surface proteins critical for adhesion to the bladder mucosa. After 3–5 min of incubation, monitor cell dissociation with an inverted bright-field microscope.

    Critical step

    Incubating the cells for too long could lead to decreased viability and, unsuccessful or reduced xenograft development. Incubation time with TrypLE will vary among cell lines, and close monitoring under a microscope should be performed to determine the optimal time for each line.


  4. 4

    Once cells have detached, add culture medium to the dish to neutralize the TrypLE. Be sure to thoroughly mix the suspension. Pipette the cell suspension up and down rapidly to break up clumps, and wash the dish sides to remove any remaining cells.

  5. 5

    Pipette the cell suspension into a 15- or 50-mL sterile centrifuge tube and mix well until a uniform suspension is achieved.

    Critical step

    Great care must be taken to accurately determine the cell count and maintain uniformity between intra-bladder injections.

  6. 6

    Mix well 10 μL of homogeneous cell suspension with 10 μL of Trypan blue in a 1.5-mL microcentrifuge tube. Place the remaining cellular suspension on ice. Count the live cells with a cell-counting chamber or automatic counter. Base the amount of inoculum on the live cell count.

    Critical step

    The number of cells to graft should be determined on the basis of cell line and mouse strain. In general, we inoculate between 0.5 and 1 × 106 cells per mouse in 50 µL of DMEM (SW 780: 0.5 × 106; UM-UC-6 and UM-UC-9: 1.0 × 106); however, a higher number is desired when less aggressive cells are grafted (T24: 1.5 × 106, or UM-UC-13: 5 × 106 cells). Alternatively, a lower cell number may be desired for studies in which you plan to maintain mice for longer than 4 weeks.


  7. 7

    After calculating the volume of cell suspension needed for the grafting, vortex the suspension for 30 s to assure uniformity. Pipette the volume needed into a new 15- or 50-mL sterile centrifuge tube.

  8. 8

    If cells are not tagged with the luciferase construct and you wish to follow them using fluorescence, stain them now with quantum dots by following the manufacturer’s staining protocol.

  9. 9

    Centrifuge the new cell suspension tube at 200g at 4 °C for 5 min. A cellular pellet will form at the bottom of the tube.

  10. 10

    Aspirate the supernatant and discard it.

  11. 11

    Add enough complete DMEM to the cellular pellet to reach the desired concentration of cells per injection. The volume of solution to inject per animal is 50 μL. Pipette the final cell suspension into a sterile microcentrifuge tube of appropriate size.

    Critical step

    Break up cell clumps well by pipetting up and down rapidly or vortexing for 5 s to ensure that the cell suspension is uniform before injection.

  12. 12

    Cells are ready for grafting. Place the cell suspension on ice until needed.


Pause point

Cells can be kept on ice for up to 3 h.

Preparation for surgery

Timing ~3 min per mouse


All animal experiments must conform to relevant local and national guidelines and regulations. All surgical procedures should be performed in a designated surgical space. There should be one autoclaved tool pack per cage of mice.

  1. 13

    Record mouse weights and provide pre-emptive analgesic carprofen (5 mg/kg) 1 h before administration of anesthesia.

  2. 14

    Place the mouse in the induction chamber with ~5.0% (vol/vol) isoflurane until fully anesthetized.


    Isoflurane can be hazardous if overexposure occurs. Ensure that charcoal filters are in place to collect waste gas fumes.

  3. 15

    Reduce isoflurane flow to ~2.0% (vol/vol) for maintenance, apply ophthalmic ointment and transfer the mouse to a nose cone on a surgical plane.

    Critical step

    Adjust the isoflurane concentration to between 1.5 and 2.5% (vol/vol) according to the size of the mouse. Larger males will generally require a higher concentration of isoflurane than smaller female mice. Ensure that the mouse is fully asleep by testing that reflexes are lost via toe-pinch reflex.

  4. 16

    Clean and disinfect the surgical site on the animal with povidone–iodine solution. Allow the solution to sit on the skin for 5–10 s and then wipe away with alcohol wipes. Repeat this step three times.


Surgical orthotopic implantation

Timing ~4 min per mouse


Tools should be sterilized with a Germinator between animals from the same cage.


A full description of the surgical technique described in this section is provided in Supplementary Video 1.

  1. 17

    Make a 1-cm midline incision through the skin of the abdominopelvic region (Fig. 2a,b) by grasping only the skin layer with the Adson forceps and cutting with the iris scissors. Separate the skin and muscle layers surrounding the incision with the Adson forceps and the pointed forceps.

  2. 18

    Section the exposed abdominal wall by making a 1-cm incision with the iris scissors and following the white midline known as linea alba (Fig. 2c). Use the Adson forceps to hold the tissue while cutting. This will generate a small opening of the abdominopelvic cavity. Place sterile gauze around the incision to maintain the sterility of the mouse’s surgical plane.

    Critical step

    When entering the abdominopelvic cavity to approach the urinary bladder, minimize blood loss and nerve damage by making a midline incision through the linea alba. This is a fibrous structure with poor vascularization.

  3. 19

    Locate the urinary bladder. It is found near the distal end of the incision (or hypogastric region), behind fatty tissue and the large intestine. Gentle pressure lateral to the incision may cause it to rise up through incision. Fully distended bladders will be easier to locate (Fig. 2d). However, if the animal has recently urinated, locating the bladder can be difficult for inexpert personnel because of its small size. Once located, use pointed forceps to gently grasp and exteriorize the bladder. Typically, pointed forceps can grasp the serosal peritoneum of the bladder, although ceramic-coated cupped forceps can also be used to grasp the body of the bladder.

    Critical step

    Be extremely gentle with internal organs and tissues. Excessive or aggressive handling can increase inflammation and pain at the surgical site.

  4. 20

    Keep hold of the bladder with either the pointed or ceramic-coated cupped forceps. Insert a 30 1/2-gauge needle with a syringe attached into the dome of the bladder to aspirate the urine (Fig. 2e,f).


  5. 21

    While still keeping hold of the bladder with the pointed or ceramic-coated cupped forceps, expose the inoculation site by bringing the superior end of the bladder to the inferior end of the abdominal cavity (Fig. 2g). Using a 30 1/2-gauge needle attached to a 1-mL syringe filled with the cellular suspension, pierce the base of the bladder (Fig. 2h).

    Critical step

    To prevent lesions, pierce the bladder in a different area from that used for urine removal (Step 21) and avoid ureters. We have tried diverse areas of inoculation within the bladder body and have observed equal results. We prefer to pierce the bladder base, where the muscle wall is thicker, to help retain the injected cellular suspension and avoid bladder leakage.


  6. 22

    Before pressing the plunger of the syringe, be sure to visualize the needle point within the distal lumen of the bladder. Once visualized, inject 50 μL of cellular suspension into bladder lumen (Fig. 2h,i) and use a sterile cotton swab to absorb any fluid at the injection site.

Critical step

When preparing syringes with cellular suspension, mix the suspension well to ensure that the same number of cells are injected into each animal. Ensure that little to no dead space remains in the syringe and needles with cells.

  1. 23

    Place the bladder back inside the body cavity and make sure that all other organs are in their proper places. Again, use a sterile cotton-tipped swab to absorb excess fluid and blood in the abdominopelvic cavity (Fig. 2j). Pull up the abdominal tissue layers on each side (including the abdominal muscle, connective tissue and peritoneum; this is observed as a thick muscular layer by the surgeon).

  2. 24

    Close the abdominopelvic cavity by suturing the muscle and associated layers. Our lab utilizes a continuous, absorbable suture to close the abdominal wall (Fig. 2k, Supplementary Fig. 3).


  3. 25

    Close the skin layer using surgical adhesive or autoclips and ensure that the incision is fully sealed (Fig. 2l).


  4. 26

    Remove the animal from the isoflurane nose cone, return the animal to its cage with a warming pad underneath and allow the animal to recover.


Post-surgical recovery and follow-up

Timing ~10 min for recovery from anesthesia and 10 min following the surgical procedure


Surgical records monitoring animal recovery should be filled out according to national and institutional regulations.

  1. 27

    Monitor the animals as they wake up from anesthesia, but do not allow them to have water.

    Critical step

    Ensure that all animals are awake in each cage before ending the post-surgical monitoring period. Mice will attack cage mates that are unconscious for too long.


  2. 28

    1 h following surgery, check on the mice again and return water to the cage containing mice.

    Critical step

    Limiting the post-surgical water intake to a minimum is critical to prevent urination of the recently injected cancer cells and allow maximum interaction with the bladder mucosa while promoting the recovery of the mice. However, if the water consumption is restricted for longer periods of time, it could severely affect the animals during post-surgical recovery, causing dehydration and shock.

  3. 29

    Monitor wound healing (Supplementary Fig. 4) and recovery.


Troubleshooting advice can be found in Table 2 and Supplementary Fig. 5.

Table 2 Troubleshooting table


  • Steps 1–12, preparation of human bladder cancer cells: ~15 min

  • Steps 13–16, preparation for surgery: ~3 min per mouse

  • Steps 17–26, surgical orthotopic implantation: ~4 min per mouse

  • Steps 27–29, post-surgical recovery and follow-up: ~10 min for recovery from anesthesia and 10 min following the surgical procedure

Anticipated results

Post-surgical monitoring of mice is critical for overall experimental success. Our institute requires daily monitoring of mice for 10 d following the surgical procedure. We recommend continuing this monitoring schedule for the duration of the experiment to ensure the health and recovery of the animals. Check the surgical incision daily to ensure that it is healing and there are no signs of infection (Supplementary Fig. 4). Weigh the animals at least weekly and note fluctuations in the animal’s weight. If significant discomfort or weight loss is observed, the animals should be allowed supportive care, such as fluid administration and/or diet gel. The development of the tumor in the bladder must also be monitored. The monitoring schedule can be weekly at first; however, as the experiment progresses, we recommend monitoring the tumor development daily. To monitor the tumor’s growth in the animal, perform gentle palpations of the bladder by pressing down on the abdominopelvic cavity. ~1 week following implantation, the bladder tumor should be palpable. In addition, while palpating the bladder, ensure that the animal is able to urinate by gently pressing on the bladder and observing the release of urine. If the animal is not able to urinate or significant hematuria is observed, euthanize the animal.

The evolving needs of the experiment will dictate whether to perform bioluminescence or fluorescence in vivo imaging to check the success of the inoculation and follow the development of the intra-bladder tumors (Fig. 3a,b). We utilized the IVIS Spectrum Optical Imaging System for all in vivo imaging because this system has the capability to capture both bioluminescent and fluorescent signals. In general, we recommend using cells labeled with quantum dots for experiments requiring early time points (<1 week) and cells tagged with luciferase for experiments requiring longer time points (>1 week). However, we have seen a fluorescent quantum dot signal as early as the day of inoculation and a bioluminescent signal <1 week following inoculation.

Fig. 3: End-stage bladder cancer xenograft growth as observed with different methods.
Fig. 3

ac, UM-UC-9 (a), UM-UC-6 (b) and SW 780 (c) bladder cancer xenografts in NOD-SCID mice are shown as examples. a, In vivo bioluminescence imaging with the IVIS Spectrum Optical Imaging System, displaying the absence (left panel) or presence (right panel) of tumor at the anatomical bladder localization. Scale bar indicates radiance (photons). b, Similar experiment with quantum dot–labeled cancer cells. The tumor presence was evaluated by epifluorescence via the IVIS system. These composite images are the result of spectral unmixing, which was used to distinguish between quantum dot (Qtracker 655) fluorescence and the autofluorescence of the food. The autofluorescence caused by the food in the intestines is red, whereas the tumor-specific signal is yellow. c, Representative anatomic and histological observations at the experimental end point. Top left, a bladder before excision from the mouse abdominopelvic cavity. The tumor is easily observed within the bladder. Scale bar, 5 mm. Top right, ex vivo fluorescence scanning by the IVIS system, showing tumor growth within the bladder lumen (right panel). Yellow areas indicate high fluorescence signal originating from the cancer cells labeled with quantum dots (Qtracker 525). No signal is observed in the control bladder (left panel). Scale bars, 5 mm. Red–yellow scale indicates epifluorescence units. Bottom left, excised bladders can be photographed using a dissecting stereomicroscope to record gross anatomopathological changes. Bottom right, representative 5-µm tissue section of a bladder tumor xenograft stained with H&E. Control images represent mice that underwent control sham surgeries. Scale bars for bottom images, 1 mm. The procedures were approved by the Institutional Animal Care and Use Committee of the University of Michigan and followed the animal welfare recommendations of the National Institutes of Health.

The number of inoculated cells must be determined on the basis of the aggressiveness and in vivo proliferation of each cell line. We suggest starting with 0.5–1 × 106 cells and increasing the number if the tumors do not develop after 4 weeks. When inoculating 0.5–5 × 106 cells per mouse, we have found that extending the experiments longer than 4 weeks often leads to mice with obstruction of the urinary bladder due to aggressive tumor development. The most important symptoms of an obstructed bladder in the mouse include distended abdomen, inability to urinate or the ability to urinate only small drops, and hematuria. Other associated signs, such as shakiness, lethargy and low body temperature, and pale eyes, ears and gums may also be present. In such cases, the mice should be euthanized immediately.

After 4 weeks of tumor development in vivo, the mice should be euthanized according to guidelines, and tissues of interest should be collected at necropsy. Ex vivo bioluminescence or fluorescence imaging can be performed, as well as photography of excised bladders with a dissection microscope for gross examination of the specimen (Fig. 3). Bladders should be fixed in formalin and paraffin-embedded until ready for immunohistochemical and histopathological analysis (Fig. 3c).

Following the technique described in this protocol, human bladder cancer cell lines develop locally invasive intra-bladder tumors 4 weeks after transplantation. The establishment of orthotopic bladder cancer xenografts is summarized in Table 3. We followed the presence of bladder tumors before the experimental end point using different methods. UM-UC-9 xenografts were detected by bioluminescence (Fig. 3a) and, UM-UC-6 and SW 780 xenografts were detected by bioluminescence or quantum dot fluorescence (Fig. 3b). We also corroborated the incidence of intravesical tumors by ex vivo imaging and histological assessment (Fig. 3c). Overall, the xenografts modeled with the orthotopic surgical technique yielded high implantation rates. We successfully established SW 780, UM-UC-6 and UM-UC-9 tumors in 100% of the NOD-SCID mice after intravesical injection of 1 million cells (Table 3). Interestingly, in preliminary pilot studies, we observed a reduction in the incidence of UM-UC-9 tumors according to the number of cells injected. Among these mice, the grafting success rate was 61–93% when 0.2–0.6 million cells per mouse were inoculated (Table 3). Additional pilot studies with NSG and nude mice were also performed. A total of 11 NSG mice and five nude mice were inoculated intravesically with 1 million UM-UC-9 cells (Supplementary Fig. 6, Table 3). Despite these strains not yet being established models, we have successfully grafted tumors in 100% of the mice, thus indicating that NSG and nude strains have solid potential to serve as alternative modeling hosts.

Table 3 Surgical orthotopic implantation rate

At the histological level, the xenografted cell lines showed an aggressive growth pattern characterized by solitary intra-bladder tumors invading the bladder submucosa and exhibiting local muscle and LVI. These key microscopic characteristics are represented in Fig. 4. Of particular interest, the microscopic examination of UM-UC-9, UM-UC-6 and SW 780 tumors displayed a deep extent of invasion into the bladder wall (Fig. 5a). Cytokeratin 8 (CK8) tumor cells invaded beyond the lamina propria with clear infiltrates into the muscular layer. These observations represent muscle-invasive T2, and occasionally T3, staged tumors. In all, the tumors exhibited marked cellular atypia with evident architectural irregularities, high mitotic activity and pleomorphism. The pathological sections also showed frequent tumor emboli within the lymphovascular spaces when MI was present (Fig. 5b). These distinctive histological features of the implanted tumors resemble the key pathobiological characteristics of human tumors, including MI and LVI (Fig. 5c). Typical results from our previously established UM-UC-3 and T24 models11 are shown in Supplementary Fig. 7; these cell lines show less aggressive traits. The extent of MI should be assessed by recording the maximum anatomic depth of tumor invasion in a manner analogous to staging of human bladder cancers: (i) invasion of the inner half of the muscularis propria; (ii) invasion of the outer half of the muscularis propria; (iii) microscopic invasion of the perivesicular adipose tissue and (iv) gross invasion of the perivesicular adipose tissue. The extent of LVI should be assessed by counting the total number of hCK8+ tumor emboli in mCD31 and mLYVE1+ lymphovascular spaces in five representative sections of tumor.

Fig. 4: Xenograft models and their key histological features.
Fig. 4

The figure represents a cross-section of the bladder wall. The tumor developed by invading through the mucosa and submucosa, exhibiting local invasion into the muscle and the vasculature.

Fig. 5: Main histological features of the xenografted tumors.
Fig. 5

a,b, Representative immunohistochemical results after implementation of the surgical orthotopic technique in UM-UC-9, SW 780 and UM-UC-6 NOD-SCID models. Tumor cells, blood vessels and lymphatic vessels present, respectively, anti-cytokeratin 8 (CK8, clone EP17, 1:100, Epitomics, cat. no. AC-0007), anti-CD31 (clone SZ31, 1:50, Dianova, cat. no. DIA-310) and anti-LYVE1 (1:1,000, Abcam, cat. no. ab14917) immunopositivity. To assess the presence of invasive tumor growth into muscularis propria and/or lymphovascular spaces, all H&E sections and immunostains were reviewed concurrently by A.M.U. to verify muscle- (a) and/or lymphovascular-invasive (b) tumors. Cytokeratin 8 reactivity highlights tumor cells, and CD31 and LYVE1 outline the mouse blood vessels and lymphatics, respectively. a, Characteristic histological observations of muscle-invasive T2-implanted tumors. Tracks of cancer cells display extensive invasion within the muscularis propria. b, In addition to muscle invasion, lymphovascular invasion is observed. CK8+ tumor cells are detected within CD31 and LYVE1+ vessels. c, Comparative histological findings in human tumors with muscle and lymphovascular invasion. Distinctive CK8+ tumor cells penetrate the muscle and the CD31+ vessels (black arrows) of the human bladder. Scale bars, 50 μm (a,b) and 200 μm (c). The use of human bladder specimens was approved by the University of Michigan Medical School Institutional Review Board and followed the criteria of the National Institutes of Health.

Although not the focus of this protocol, other sporadic observations have been made using this technique, such as noninvasive intraepithelial lesions of the bladder (Supplementary Fig. 8a,b) and spontaneous metastasis when allowing tumor development through the 4th week (Supplementary Fig. 8c). Most frequently, we have observed distant metastasis in the liver and lungs.

In all, our orthotopic bladder cancer xenografts in mice allow us to model the early key features of spontaneous human bladder cancer invasion. We have shown that human bladder cancer cells are able to seed in their natural bladder niche without further requirements and spontaneously recapitulate MI and LVI, which are well-known histological features of aggressive human disease. Thus, this reproducible and simple technique efficiently models human bladder cancer invasion and represents a valuable tool for dissecting the cellular and molecular mechanisms that regulate the invasive progression of human bladder cancer.

Reporting Summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Data availability

The datasets generated during the current study are included in this published article and its SupplementaryInformation files.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Key references using this protocol

Han, A. L. et al. Oncogene 36, 5243–5251 (2017): https://www.nature.com/articles/onc2017149


  1. 1.

    Siegel, R. L., Miller, K. D. & Jemal, A. Cancer statistics, 2018. CA Cancer J. Clin. 68, 7–30 (2018).

  2. 2.

    Torre, L. A. et al. Global cancer statistics, 2012. CA Cancer J. Clin. 65, 87–108 (2015).

  3. 3.

    Agarwal, N. & Hussain, M. Management of bladder cancer: current and emerging strategies. Drugs 69, 1173–1187 (2009).

  4. 4.

    Bolenz, C. et al. Lymphovascular invasion is an independent predictor of oncological outcomes in patients with lymph node-negative urothelial bladder cancer treated by radical cystectomy: a multicentre validation trial. BJU Int. 106, 493–499 (2010).

  5. 5.

    Brunocilla, E., Pernetti, R. & Martorana, G. The prognostic role of lymphovascular invasion in urothelial-cell carcinoma of upper and lower urinary tract. Anticancer Res. 31, 3503–3506 (2011).

  6. 6.

    Smith, A. B. et al. Muscle-invasive bladder cancer: evaluating treatment and survival in the National Cancer Data Base. BJU Int. 114, 719–726 (2014).

  7. 7.

    Sylvester, R. J. et al. Predicting recurrence and progression in individual patients with stage Ta T1 bladder cancer using EORTC risk tables: a combined analysis of 2596 patients from seven EORTC trials. Eur. Urol. 49, 466–477 (2006).

  8. 8.

    Iyer, G. et al. Prevalence and co-occurrence of actionable genomic alterations in high-grade bladder cancer. J. Clin. Oncol. 31, 3133–3140 (2013).

  9. 9.

    Chan, E., Patel, A., Heston, W. & Larchian, W. Mouse orthotopic models for bladder cancer research. BJU Int. 104, 1286–1291 (2009).

  10. 10.

    Jäger, W. et al. Orthotopic mouse models of urothelial cancer. in Urothelial Carcinoma: Methods and Protocols, Methods in Molecular Biology Vol. 1655 (eds. Schulz, W.A., Hoffmann, M. & Niegisch, G.) 177–197 (Springer Science+Business Media, Berlin, 2018).

  11. 11.

    Han, A. L. et al. Fibulin-3 promotes muscle-invasive bladder cancer. Oncogene 36, 5243–5251 (2017).

  12. 12.

    Dinney, C. P. et al. Isolation and characterization of metastatic variants from human transitional cell carcinoma passaged by orthotopic implantation in athymic nude mice. J. Urol. 154, 1532–1538 (1995).

  13. 13.

    Jäger, W. et al. Ultrasound-guided intramural inoculation of orthotopic bladder cancer xenografts: a novel high-precision approach. PLoS ONE 8, e59536 (2013).

  14. 14.

    Huebner, D. et al. An orthotopic xenograft model for high-risk non-muscle invasive bladder cancer in mice: influence of mouse strain, tumor cell count, dwell time and bladder pretreatment. BMC Cancer 17, 790 (2017).

  15. 15.

    Chong, L. et al. Characterization of a novel transplantable orthotopic murine xenograft model of a human bladder transitional cell tumor (BIU-87). Cancer Biol. Ther. 5, 394–398 (2006).

  16. 16.

    Watanabe, T. et al. An improved intravesical model using human bladder cancer cell lines to optimize gene and other therapies. Cancer Gene Ther. 7, 1575–1580 (2000).

  17. 17.

    Yang, X. H. et al. A new method of establishing orthotopic bladder transplantable tumor in mice. Cancer Biol. Med. 9, 261–265 (2012).

  18. 18.

    Ahmad, I., Sansom, O. J. & Leung, H. Y. Exploring molecular genetics of bladder cancer: lessons learned from mouse models. Dis. Model. Mech. 5, 323–332 (2012).

  19. 19.

    Gust, K. M. et al. Fibroblast growth factor receptor 3 is a rational therapeutic target in bladder cancer. Mol. Cancer Ther. 12, 1245–1254 (2013).

  20. 20.

    Hayashi, T. et al. Not all NOTCH is created equal: the oncogenic role of NOTCH2 in bladder cancer and its implications for targeted therapy. Clin. Cancer Res. 22, 2981–2992 (2016).

  21. 21.

    Hu, C. et al. Intravenous injections of the oncolytic virus M1 as a novel therapy for muscle-invasive bladder cancer. Cell Death Dis. 9, 274 (2018).

  22. 22.

    Calì, B., Molon, B. & Viola, A. Tuning cancer fate: the unremitting role of host immunity. Open Biol. 7, 170006 (2017).

  23. 23.

    Schreiber, R. D., Old, L. J. & Smyth, M. J. Cancer immunoediting: integrating immunity’s roles in cancer suppression and promotion. Science 331, 1565–1570 (2011).

  24. 24.

    Kobayashi, T., Owczarek, T. B., McKiernan, J. M. & Abate-Shen, C. Modelling bladder cancer in mice: opportunities and challenges. Nat. Rev. Cancer 15, 42–54 (2015).

  25. 25.

    Grossman, H. B., Wedemeyer, G., Ren, L., Wilson, G. N. & Cox, B. Improved growth of human urothelial carcinoma cell cultures. J. Urol. 136, 953–959 (1986).

  26. 26.

    Shinohara, N., Liebert, M., Wedemeyer, G., Chang, J. H. & Grossman, H. B. Evaluation of multiple drug resistance in human bladder cancer cell lines. J. Urol. 150, 505–509 (1993).

  27. 27.

    Sabichi, A. et al. Characterization of a panel of cell lines derived from urothelial neoplasms: genetic alterations, growth in vivo and the relationship of adenoviral mediated gene transfer to coxsackie adenovirus receptor expression. J. Urol. 175, 1133–1137 (2006).

  28. 28.

    Abel, E. V. et al. HNF1A is a novel oncogene that regulates human pancreatic cancer stem cell properties. Elife 7, e33947 (2018).

  29. 29.

    Day, K. C. et al. HER2 and EGFR overexpression support metastatic progression of prostate cancer to bone. Cancer Res. 77, 74–85 (2017).

Download references


This work was supported by a University of Michigan Rogel Cancer Center Research Grant to E.T.K.; Department of Urology Research funds to M.L.D.; NIH R01CA154252 to M.L.D.; the European Egyptian Pharmaceutical Industries (EEPI) research collaboration with M.L.D. and, the Genmab research collaboration with M.L.D. In addition, EEPI provided salary funds to L.E.-S., K.C.D., L.J.B. and M.L.D., as well as consulting fees to L.E.-S. and M.L.D. Genmab provided salary support for G.L.H. and A.L.C. G.L.H. was also supported by the Postdoctoral Translational Scholars Program of the Michigan Institute for Clinical and Health Research (UL1TR002240), NIH. We thank J. Escara-Wilke for her assistance with the banked samples utilized in this study, E. Breij for her conceptual advice, and M.C. Winkler, who aided with the surgical setup and preliminary stages of the research.

Author information

Author notes

    • Guadalupe Lorenzatti Hiles
    •  & Hannah L. Briggs

    Present address: Division of Head and Neck Surgery, Department of Otolaryngology, University of Michigan, Ann Arbor, MI, USA

    • Angelica L. Cates

    Present address: College of Veterinary Medicine, Michigan State University, East Lansing, MI, USA

    • Amy L. Han

    Present address: School of Medicine, Anschutz Medical Campus, University of Colorado Denver, Aurora, CO, USA

    • Amir Emamdjomeh

    Present address: College of Pharmacy and Health Sciences, Wayne State University, Detroit, MI, USA

    • Andrew Chou

    Present address: College of Medicine, University of Cincinnati, Cincinnati, OH, USA

  1. These authors contributed equally: Guadalupe Lorenzatti Hiles, Angelica L. Cates, Layla El-Sawy, Kathleen C. Day.


  1. Division of Urologic Oncology, Department of Urology, University of Michigan, Ann Arbor, MI, USA

    • Guadalupe Lorenzatti Hiles
    • , Angelica L. Cates
    • , Layla El-Sawy
    • , Kathleen C. Day
    • , Luke J. Broses
    • , Amy L. Han
    • , Hannah L. Briggs
    • , Amir Emamdjomeh
    • , Andrew Chou
    • , Monica Liebert
    • , Evan T. Keller
    •  & Mark L. Day
  2. University of Michigan Rogel Cancer Center, Ann Arbor, MI, USA

    • Guadalupe Lorenzatti Hiles
    • , Angelica L. Cates
    • , Layla El-Sawy
    • , Kathleen C. Day
    • , Luke J. Broses
    • , Amy L. Han
    • , Amir Emamdjomeh
    • , Andrew Chou
    • , Ethan V. Abel
    • , Monica Liebert
    • , Phillip L. Palmbos
    • , Evan T. Keller
    •  & Mark L. Day
  3. European Egyptian Pharmaceutical Industries, Alexandria, Egypt

    • Layla El-Sawy
  4. Department of Pathology, University of Michigan, Ann Arbor, MI, USA

    • Amy L. Han
    • , Aaron M. Udager
    •  & Evan T. Keller
  5. Department of Molecular and Integrative Physiology, University of Michigan, Ann Arbor, MI, USA

    • Ethan V. Abel
  6. Division of Haematology and Oncology, Department of Internal Medicine, University of Michigan, Ann Arbor, MI, USA

    • Phillip L. Palmbos


  1. Search for Guadalupe Lorenzatti Hiles in:

  2. Search for Angelica L. Cates in:

  3. Search for Layla El-Sawy in:

  4. Search for Kathleen C. Day in:

  5. Search for Luke J. Broses in:

  6. Search for Amy L. Han in:

  7. Search for Hannah L. Briggs in:

  8. Search for Amir Emamdjomeh in:

  9. Search for Andrew Chou in:

  10. Search for Ethan V. Abel in:

  11. Search for Monica Liebert in:

  12. Search for Phillip L. Palmbos in:

  13. Search for Aaron M. Udager in:

  14. Search for Evan T. Keller in:

  15. Search for Mark L. Day in:


G.L.H., A.L.C., L.E.-S. and K.C.D. conceived different aspects of the experimental design and implemented the surgical technique with contributions from E.T.K. and M.L.D. The luciferase construct was developed by E.V.A., and K.C.D. labeled the cancer cells with the luciferase tag. G.L.H., A.L.C., L.E.-S., K.C.D., L.J.B., A.L.H., A.E. and A.C. performed the surgical experiments. G.L.H., A.L.C., K.C.D., L.J.B., A.H., H.L.B., A.E. and A.C. collected, analyzed and interpreted data. G.L.H., A.L.C. and H.L.B. conducted the MI and LVI examinations. M.L., P.L.P., A.M.U., E.T.K. and M.L.D. provided conceptual advice. A.M.U. contributed to the histological analysis and provided the bladder cancer specimens. M.L., E.T.K. and M.L.D. provided reagents, materials and mice. G.L.H., A.L.C., K.C.D., L.J.B., H.L.B., E.T.K. and M.L.D. contributed to the manuscript preparation.

Competing interests

EEPI provided salary support to L.E.-S., K.C.D., L.J.B. and M.L.D., and consulting fees to M.L.D. and L.E.-S. M.L.D. has received research funds from EEPI as part of a sponsored research agreement. L.E.-S. is currently an employee of EEPI. Genmab provided research funds to M.L.D. and salary support to G.L.H. and A.L.C. Neither EEPI nor Genmab participated in the development, design, execution or analysis of the experiments, or the publication of the manuscript. The remaining authors declare no competing interests.

Corresponding authors

Correspondence to Evan T. Keller or Mark L. Day.

Integrated supplementary information

  1. Supplementary Figure 1 Required surgical supplies and setup.

    Main tools, equipment, and reagents necessary for successful orthotopic inoculations. a, Set of surgical tools to be autoclaved in a cassette for instruments (see also i). From left to right: ceramic-coated cupped forceps, Adson forceps, tissue forceps, Stevens needle holder, micro-dissecting angled forceps, pointed forceps, and iris scissors. b, Povidone iodine scrub solution, 7.5%. c, Isoflurane with anti-drip top. d, Germinator 500, bead sterilizer for surgical tools. e, Autoclaved cotton swabs, sterile gauze pads, alcohol swabs. f, Sterile towel drape. g, One milliliter tuberculin slip tip syringe and 30½ gauge needle. h, Ophthalmic ointment. i, Minimum required surgical setup.

  2. Supplementary Figure 2 Pre-surgical hair removal.

    Required hair removal procedure one day prior to surgery. Depilatory cream is utilized by our laboratory. A-NOD-SCID mouse is depicted immediately after the hair has been removed and the excess depilatory cream has been washed away. This procedure was approved by the Institutional Animal Care and Use Committee of the University of Michigan following the animal welfare recommendations by the National Institutes of Health.

  3. Supplementary Figure 3 Recommended wound closure tools and abdominal suture procedure.

    Incision closure information. a-c, Tools recommended for wound closure. a, Stevens needle holder. b, Adson forceps. c, Coated Vicryl absorbable suture. d, Tissue adhesive. e-l, Protocol utilized by our laboratory to suture the abdominal wall. The procedure is represented in a NOD-SCID mouse. e, Align the sharp end of the needle with the cranial end of the incision. The needle should be oriented such that it is perpendicular to the incision while the needle holders remain parallel to the incision. The sutures are applied from the cranial to the caudal end of the incision. f, Start to suture by inserting the needle from the outside of the muscle layer. Then, insert the needle on the opposite side of the incision from the inside of the muscle. Pull the needle and suture through the muscle layer so that ~2 cm of suture remains on the end of the suture. This small piece will be used to tie the suture. g, Tie three knots to secure the suture in place. The first knot will be double knot. h, The second and third knots will be single knots tied opposite of each other. The suture is now secured in place. i, Pierce the needle through the muscle, Then, grab the muscle on the opposite side and pierce the needle through it. Pull the suture tightly through. Repeat this step. Evenly separate the sutures until the incision is closed. j, Tie the suture closed as in (e) but, using the final loop of the suture. k, Tie one double knot followed by two single, opposite knots to end the suture. l, Gently tighten the suture and trim any excess away. The muscle layer is now fully closed. Sealed the skin closed with tissue adhesive or wound clips (not depicted). This procedure was approved by the Institutional Animal Care and Use Committee of the University of Michigan following the animal welfare recommendations by the National Institutes of Health.

  4. Supplementary Figure 4 Wound progression.

    Wound healing patterns observed following surgical xenotransplantation. While some fibrotic tissue is expected, the skin must remain sealed to prevent infection. The recovery period and wound healing process should be completed after 10 days. This procedure was performed in a NOD-SCID mouse and approved by the Institutional Animal Care and Use Committee of the University of Michigan following the animal welfare recommendations by the National Institutes of Health.

  5. Supplementary Figure 5 Troubleshooting and execution problems.

    Examples of experimental complications depicted in NOD-SCID mice. a, Improper wound healing in which the abdominal muscle was exposed due to improper closure and excessive retraction of the skin. Built-up of fibrotic tissue is observed at the wound site. b, Histological section showing the consequence of a failed inoculation. A tumor developed on the exterior wall of the bladder. Scale bars, 500 μm (left) and 200 μm (right).

  6. Supplementary Figure 6 Xenograft models in NSG and athymic nude mice.

    Representative tumor growth, followed by bioluminiscence imaging, at three weeks post-inoculation of 1 ×106 luciferase-tagged UM-UC-9 cells into the bladder of female NSG and male nude mice. This procedure was approved by the Institutional Animal Care and Use Committee of the University of Michigan following the animal welfare recommendations by the National Institutes of Health.

  7. Supplementary Figure 7 Key histological characteristics of UM-UC-13 and T24 xenograft models.

    The xenografted tumors display less aggressive pathobiological features and require a higher inoculum for their development. The figure shows representative histological results after implantation of 5 ×106 UM-UC-13 and 1.5 ×106 T24 human bladder tumor cells into the bladder lumen of NOD-SCID mice. The tumors developed for four weeks until end point. a, Characteristic histological analysis at the invasive edge of the tumor. Cytokeratin 8+ (CK8, clone EP17, Epitomics, 1:100, cat. no. AC-0007) positive cancer cells are observed invading the muscle of the bladder. In general, the mice exhibit muscle-invasive T2 bladder tumors at a solitary focal location. b, Lymphovascular invasion is also observed at lower frequency. CK8 positive tumor cells are detected within CD31 (Dianova, clone SZ31, 1:50, cat. no. DIA-310) and LYVE1 (Abcam, 1:1000, cat. no. ab14917) positive vessels. Scale bars, 50 μm.

  8. Supplementary Figure 8 Sporadic experimental observations.

    a, Normal bladder displaying a thin, triple-layered urothelium (insert). b, Non-invasive lesion with engrossment of the bladder urothelium (insert) c, Example of spontaneous sporadic metastasis. A liver metastatic tumor is observed by bioluminescence imaging. The figures display experiments with NOD-SCID mice. Scale bars, 500 μm (left) and 50 μm (right inserts).

Supplementary information

  1. Supplementary Text and Figures and Supplementary Methods

    Supplementary Figures 1–8, Processing of murine bladder samples for immunohistological analysis.

  2. Reporting Summary

  3. Supplementary Video 1

    Demonstration of the surgical orthotopic procedure to graft human tumor cells into the bladder of an immunodeficient mouse. Step-by-step protocol showing a NOD-SCID mouse undergoing surgery.

About this article

Publication history





By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.