Page 27 - Read Online

Page 4 of 9 Felton et al. J Cancer Metastasis Treat 2018;4:51 I http://dx.doi.org/10.20517/2394-4722.2018.39

we cleansed the abdomen with povidine-iodine solution and alcohol and applied sterile drapes. We made a
small midline laparotomy and inserted a self-retaining retractor in the upper abdomen.

Following injection of human colon cancer cells and replacing the intestines, we approximated fascial edges
with 5-0 vicryl running sutures and closed the skin primarily with 4-0 nylon interrupted sutures. After ap-
plying skin glue (3M Vetbond tissue adhesive, St. Paul, MN) to the suture line, we awakened mice slowly
from anesthesia and placed them in a clean cage for recovery with close monitoring. After completion of the
operation, the mice were administered analgesia for at least 72 h post-operatively and monitored closely.

Surgical technique - cecal injection
To explore the predilection of human colon cancer cells to form tumors in different regions of the mouse, we
6
first injected 2 to 5 × 10 (40-100 µL) HT-29 human colon cancer cells into the cecum of nude or NSG mice.
We chose these cell numbers based on previous reports describing successful metastatic models of colon
cancer in mice [1,12,22,23] . We described pre-operative steps above. The cecum was located using moist sterile
cotton tip applicators and brought outside the abdomen onto a moist 2 × 2 sterile gauze. In all mice, we in-
6
jected 2-5 × 10 cells (40-100 µL) into the wall of the cecum using a 27-guage needle. After injection, we ap-
plied light pressure at the injection site for approximately 30 s with a moist sterile tip applicator and inspect-
ed the area for leakage. We irrigated the cecum and abdominal cavity with warm DPBS, and then returned
the cecum to its normal anatomic position within the abdomen. Closure of the abdomen was performed as
describe above.

Surgical technique - flank injection
In mice failing to form cecal tumors, we confirmed the ability of the HT-29 cells to form xenografts and me-
tastases following subcutaneous and splenic injection, respectively. For subcutaneous injections, we briefly
6
anesthetized mice with vaporized isoflurane, disinfected their flanks with alcohol, and injected 2 × 10 cells
(40 µL) in each flank. We recovered mice from anesthesia in their cages.

Surgical technique - splenic injection
We described pre-operative steps above. The spleen was located using moist sterile cotton tip applicators
and brought forward within the abdomen. In all mice, we injected 5 × 10 cells (100 µL) into the wall of the
6
spleen using a 27-guage needle. After injection, we applied light pressure at the injection site for approxi-
mately 30 s with a moist sterile tip applicator and inspected the area for leakage and bleeding. We irrigated
the spleen and abdominal cavity with warm DPBS, and then returned the spleen to its normal anatomic
position within the abdomen. After 1 h, we removed the spleen and irrigated the abdomen again with DPBS.
We closed the abdomen as described above.

Surgical technique - distal colon injection
6
To induce colon cancer growth and metastasis, we injected 5 × 10 (100 µL) HT-29 human colon cancer cells
into the wall of the distal colon of nude or NSG mice. We described pre-operative steps above. The distal
colon was located using moist sterile cotton tip applicators [Figure 1A]. In all mice, we injected 5 × 10 cells
6
(100 µL) into the wall of the distal colon using a 27-guage needle [Figure 1B]. After injection, we applied
light pressure at the injection site for approximately 30 s with a moist sterile tip applicator [Figure 1C] and
inspected the area for leakage. We irrigated the distal colon and abdominal cavity with warm DPBS. We
closed the abdomen as described above.

Statistical analysis
We used the unpaired Student’s t test (assuming unequal variance) to compare continuous variables between
two independent groups. For multi-group comparisons, we applied two-way ANOVA with one between-
subject factor (WT vs. FGF15-deficient) and one within-subject factor (normal tissue vs. tumor tissue) fol-
lowed by post hoc tests with Tukey-Kramer’s adjustment for P values. We used Fisher’s exact test to compare