Three-Dimensional Analysis of the Effect of Osteosarcoma on Sensory Nerves Innervating the Femur in a Murine Model of Osteosarcoma-Induced Bone Pain

Abstract

Background: The ways in which peripheral sensory nerves interact with osteosarcomas are important to understand because it could lead to development of new approaches to treat bone cancer pain. This study aimed to determine how cancer affects sensory nerve density and distribution in a murine model of osteosarcoma-induced bone pain.

Methods: The femoral marrow cavities of male C3H/HeNHsd mice were injected with either NCTC 2472 primary osteosarcoma (cancer) cells or phosphate buffered saline (control). Pain behavior was assessed using limb use score and static weight bearing assays. At the experimental endpoint, femurs were collected, decalcified, immunolabeled, cleared and imaged using light sheet microscopy (Ultramicroscope Blaze, Miltenyi Biotec). The distribution of sensory nerves was traced through the marrow cavity of the proximal femur and the periosteum overlying the third trochanter (Imaris, Bitplane).

Results: Weight bearing on the injected limb was decreased in osteosarcoma-injected but not saline-injected mice. Filament tracing revealed a reduced density of neurofilament 200 kDa-positive (NF200+; myelinated nerve marker) but not calcitonin gene-related peptide-positive (CGRP+; peptidergic nerve marker) sensory nerves in the marrow cavity of osteosarcoma-injected relative to saline-injected mice. There was increased density of CGRP+ but not NF200+ nerves in the periosteum of osteosarcoma-injected relative to saline-injected mice.

Conclusions: Osteosarcoma differentially affects the density and distribution of different subtypes of peripheral sensory nerves in bone. Understanding how osteosarcomas affect different populations of sensory nerves could lead to more targeted mechanism-based treatments for bone cancer-induced pain.

Keywords: CGRP; NF200; bone cancer pain; bone pain; osteosarcoma; pain; sensory nerves.

Figure 1

Pain behavior and tumor growth. (A) Limb use score following intrafemoral injection of phosphate buffered saline (PBS; control animals) or NCTC 2472 cells (osteosarcoma bearing animals). Data are presented as the median ± range. (B) Percentage of weight bearing on the injected hindlimb, at baseline and experimental endpoint, in control vs. osteosarcoma bearing animals. Data presented as the mean ± standard error of the mean, two-way ANOVA with Fisher’s LSD. (C) Low power (4×) image of a 40 µm z-projection of the whole femur (left) and marrow cavity (inset) from a control animal. The marrow cavity had a normal appearance. (D) Low power (4×) image of a 40 µm z-projection of the whole femur (left) and marrow cavity (inset) from an osteosarcoma bearing animal. By the endpoint, osteosarcoma cells had proliferated and replaced hematopoietic cells in the entire marrow cavity and had begun to impact on cortical bone but did not break through to the periosteum. (E) Radiographs of hindlimbs from control and osteosarcoma bearing animals. Arrowhead indicates an area of reduced bone density. (F) Relative bone density of the femurs from control and osteosarcoma bearing animals at baseline and endpoint. Data presented as the mean ± standard error of the mean, two-way ANOVA with Fisher’s LSD. Red circles represent control animals, orange triangles represent osteosarcoma bearing animals.

Figure 2

Neurofilament 200 kDa (NF200)-immunolabeled nerve profiles in the marrow cavity of control and osteosarcoma bearing animals. (A) Low power (4×) image of a 200 µm z-projection of the femur showing NF200+ nerves in the marrow cavity of a control animal, and the region of interest that analysis was applied to (inset, (B)). (B) High power (12×) image of a 1500 µm z-projection, through the full thickness of the marrow cavity, showing NF200+ nerves in the region of interest. A 3D projection of these nerve profiles is available as a video (Supplementary Data, nf200-control.mp4). Arrow indicates primary nerve entry point in region of interest. (C,D) Insets show 50 µm z-projections of NF200+ nerve fibers indicated in (B). (E) Full thickness z-projection of NF200+ nerve filaments traced from (B). (F) Low power (4×) image of a 200 µm z-projection of the femur showing NF200+ nerves in the marrow cavity of an osteosarcoma bearing animal, and the region of interest that analysis was applied to (inset, (G)). (G) High power (12×) image of a 1500 µm z-projection through the full thickness of the marrow cavity, showing NF200+ nerves in the region of interest. A 3D projection of these nerve profiles is available as a video (Supplementary Data, nf200-cancer.mp4). (H,I) Insets show 50 µm z-projections of NF200+ nerve fibers indicated in (G). (J) Full thickness z-projection of NF200+ nerve filaments traced from (G). There was a reduction in (K) filament length, (L) number of branch points, and (M) number of terminal points in osteosarcoma bearing (n = 4) relative to control (n = 4) animals. Data presented as the mean ± standard error of the mean, unpaired t-test.

Figure 3

Calcitonin gene-related peptide (CGRP) immunolabeled nerve profiles in the marrow cavity of control and osteosarcoma bearing animals. (A) Low power (4×) image of a 200 µm z-projection of the femur showing CGRP+ nerves in the marrow cavity of a control animal and the region of interest that analysis was applied to (inset, (B)). (B) High power (12×) image of a 1500 µm z-projection, through the full thickness of the marrow cavity, showing CGRP+ nerve profiles in the region of interest. A 3D projection of these nerve profiles is available as a video (Supplementary Data, cgrp-control.mp4). (C,D) Insets show 50 µm z-projections of CGRP+ nerve fibers indicated in (B). (E) Full thickness z-projection of CGRP+ nerve filaments traced from (B). (F) Low power (4×) image of a 200 µm z-projection of the femur showing CGRP+ nerves in the marrow cavity of an osteosarcoma bearing animal, and the region of interest that analysis was applied to (inset, (G)). (G) High power (12×) image of a 1500 µm z-projection through the full thickness of the marrow cavity, showing CGRP+ nerves in the region of interest. A 3D projection of these nerve profiles is available as a video (Supplementary Data, cgrp-cancer.mp4). (H,I) Insets show 50 µm z-projections of CGRP+ nerve fibers indicated in (G). (J) Full thickness z-projection of CGRP+ nerve filaments traced from (G). There was no change in (K) filament length, (L) number of branch points and (M) number of terminal points in osteosarcoma bearing (n = 4) relative to control (n = 4) animals. Data presented as the mean ± standard error of the mean, unpaired t-test.

Figure 4

Sprouting of calcitonin gene-related peptide (CGRP) but not neurofilament 200 kDa (NF200) nerves in the periosteum of osteosarcoma bearing animals. (A,C,H,J) Full thickness z-projections of high power (12×) images of nerves in the periosteum overlying the posterior aspect of the third trochanter. (B,D,I,K) Full thickness z-projections of nerve filaments traced from (A,C,H,J). There was no difference in (E) filament length, (F) number of branch points and (G) number of terminal points of NF200+ nerve profiles in osteosarcoma bearing (n = 4) relative to control (n = 4) animals. There was an increase in (L) filament length, (M) number of branch points and (N) number of terminal points of CGRP+ nerve profiles in osteosarcoma bearing (n = 4) relative to control (n = 4) animals. Data presented as the mean ± standard error of the mean, unpaired t-test.