University of Southern California Dissertations and Theses

Bone marrow-derived mesenchymal stromal cells in the tumor microenvironment of neuroblastoma

(USC Thesis Other)

Bone marrow-derived mesenchymal stromal cells in the tumor microenvironment of neuroblastoma

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

BONE MARROW-DERIVED MESENCHYMAL STROMAL CELLS IN THE
TUMOR MICROENVIRONMENT OF NEUROBLASTOMA
by
Christine Marie Hogan

A Thesis Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the Requirements for the Degree
MASTER OF SCIENCE
(EXPERIMENTAL & MOLECULAR PATHOLOGY)

May 2015

Committee Members:
Dr. Yves DeClerck
Dr. Cheng-Ming Chuong
Dr. Alan Epstein
Dr. Andre Ouellette

Copyright 2015 Christine Marie Hogan
2

Table of Contents

Acknowledgements 3
List of Figures 4
Abstract 6
Introduction 8
Results 10
Discussion 16
Materials and Methods 21
Future Directions 26
Figures 28
Bibliography 46

3

Acknowledgements
There have been so many who have helped me along my journey and among the many who
deserve my thanks and appreciation is my mentor and advisor, Dr. Yves DeClerck. His support, guidance
and encouragement throughout this project have benefited me in many ways. Dr. DeClerck’s passions
for teaching, science and medicine have inspired me to work hard, always strive for continuous success
and to make myself better. I am forever grateful.
With my limited experience working in the lab, I would like to acknowledge our lab manager, Dr.
Laurence Sarte. She has helped me tremendously throughout this project and provided guidance on
every aspect along the journey. I am incredibly appreciative of her help. In addition, I’d like to thank Dr.
Scott Bergfeld for introducing me to the project, techniques and helping me lay the foundation.
My colleagues in the DeClerck lab, Dr. Lucia Borriello, Dr. Marta Kubala, Dr. Veronica Placencio,
Rie Nakata and Jackie Rosenberg for their assistance and insight throughout my time working on the
project. These individuals have taught me numerous invaluable skills and I thank each of them for this
wonderful opportunity and experience.
I would not have been successful in the pursuit of my master’s degree and this project without
the members of my committee, Dr. Cheng-Ming Chuong, Dr. Alan Epstein, and Dr. Andre Ouellette.
Their feedback, recommendations and ensuring I was on track has been critical in shaping and refining
my research.
Last, and certainly not least, family and friends, in particular, Dr. Travis Williams, who have given
me the encouragement I needed to get through the hard days. I would not have made it without their
continuous love and support. Thank you!
4

List of Figures
Figure 1: Bone marrow mesenchymal stromal cells (BMMSC) migrate from the bone marrow into the
circulating blood stream where they differentiate over time. 28
Figure 2: Migration of human BMMSC increases in the presence of neuroblastoma cell lines. 29
Figure 3: Multiplex “Luminex” immunosorbent cytokine assay reveals increased expression of IL-6 and
IL-8 in co-culture. 31
Figure 4: ELISA assays reveal increased expression levels of MCP-1, IL-8, and IL-6 when human BMMSC
are co-cultured with CHLA-255. 32
Figure 5: Orthotopic model of pediatric solid tumors (neuroblastoma). 34
Figure 6: CHLA-255 tumor growth in the left kidney monitored by MRI. 35
Figure 7: CHLA-255 tumor growth in the left kidney monitored by bioluminescence. 36
Figure 8: Labeling of BMMSC cell membrane with PKH67 green fluorescent membrane dye in vitro. 37
Figure 9: Tracking of PKH67 green fluorescent membrane labeled BMMSC in vivo. 38
Figure 10: Mouse BMMSC injected intravenously monitored by bioluminescence. 39
Figure 11: Mouse BMMSC injected in the femur monitored by bioluminescence. 40
Figure 12: Kidney tumors shown in comparison to normal “control” kidneys. 42
Figure 13: H&E stains of tumor kidney show necrotic and normal morphology. 43
Figure 14: Natural tissue fluorescence has potential to reveal misleading signals. 44
5

Figure 15: Sensitivity of bioluminescence reveals misleading signals. 45

6

Abstract
Bone marrow mesenchymal stromal cells (BMMSC) migrate from the bone marrow into the
circulating blood stream where overtime they differentiate to adiposities, chondrocytes, myocyte,
fibroblasts, osteoblasts or other cell derivatives. We hypothesized that BMMSC migrate from the bone
marrow into the primary tumor and participate in the tumor microenvironment. Cells expressing
characteristics of BMMSC were obtained from human and mouse bone marrow. Using a migration
assay, human BMMSC were assessed for their ability to migrate toward five neuroblastoma cell lines.
Higher migration rates were observed in co-culture conditions of CHLA-255 (P < 0.001), SK-N-BE(2) (P <
0.001), SK-N-SH (P < 0.001), SH-5YSY (P < 0.005) and NB-19 (P < 0.001), concluding a chemoattraction
(Figure 2). To analyze the presence of small molecules, cytokines and chemokines with a potential role
in chemoattraction, Luminex was used as a screening panel (Figure 3) and ELISA assays (Figure 4)
validated increased expression levels of IL-6 (462 pg/mL), IL-8 (208 pg/mL) and MCP-1 (1,584 pg/mL),
when human BMMSC were co-cultured with CHLA-255. Further analysis of MCP-1 as a recombinant
protein did not show increased migration (P > 0.5) concluding an insignificant role in chemoattraction.
In vivo, mouse BMMSC were placed in the femur (Figure 11) or intravenously (Figure 10) and tracked via
bioluminescence for their ability to migrate towards neuroblastoma CHLA-255 tumors orthotopically
injected into the kidney of immuno-deficient (nude) mice over a period of six weeks. On day 24, mouse
BMMSC injected into the femur appeared to have migrated to the tumor kidney; however image may
have presented a misleading signal based on additional analysis of the lysed tumor kidney levels of
luminescence (P < 0.001). Preliminary data may suggest that some BMMSC may have migrated toward
the tumor, but experiments to validate this observation are needed. NIH-3T3 murine skin fibroblasts
injected via femur or intravenously were used as a control and did not migrate to the tumor kidney, as
confirmed by ex vivo luminescence of lysed tumor kidney (P < 0.001). We demonstrated that BMMSC
7

are attracted to neuroblastoma cells in vitro, and could migrate from the bone marrow to primary
tumors in vivo while human fibroblasts and NIH-3T3 did not migrate to the tumor kidney. Overall, this
data illustrates that BMMSC may be able to migrate from the bone marrow into the primary tumor.
Future studies should assess BMMSC as being important mediatord of the tumor microenvironment and
understanding the chemoattraction pathways.

8

Introduction
The tumor microenvironment consists of many of cell types that perform intricate cross-talk by
sending signals and pro-inflammatory stimuli (1). Among those cells are bone marrow derived
mesenchymal stromal cells (BMMSC), which are multi-potent stromal cells that can differentiate into a
variety of cell types: osteoblasts, chondrocytes, myocyte, fibroblast and adipocytes (2) (Figure 1A). In
recent years, investigators have shown that BMMSC contribute to the tumor microenvironment in
colonizing tumors, promoting tumor survival and progression, particularly in adult cancers (3).
Tumors have many origins, including genetic and environmental factors. In this study is tumor
growth by neuroblastoma, a sporadic disease that arises in children from the developing central nervous
system of the neural crest (4). Neuroblastoma is the most common type of extra-cranial ‘solid’
childhood cancer and is diagnosed primarily in children under five and rarely in adults (5). Because of
vague symptoms early in the disease, neuroblastoma is often diagnosed once metastases have prevailed
in other parts of the body (6). This disease frequently originates in the adrenal gland, although it can
develop in the pelvis, neck, chest or abdomen (5). A heterogeneous and complex disease with few
genetic drivers (MYCN and ALK) and markers, very low mutation; neuroblastoma is one of the few
human malignancies known to demonstrate spontaneous regression from an undifferentiated state to a
completely benign cellular appearance (8).
In 70% of patients with advanced disease, there are metastases in the bone marrow (6) where
BMMSC reside (Figure 1B). There, the BMMSC provide an alternate pathway of osteoclast activation and
therefore promote bone destruction by cancer cells (9). BMMSC are unique as they have the ability to
self-renew, migrate and/or differentiate. Upon BMMSC migration, these cells participate in primary
tumors (e.g. breast, pancreatic and ovarian cancer) and upon recruitment they differentiate into tumor
associated fibroblasts (TAF) which can produce mitogenic and angiogenic factors, display extracellular
9

matrix degrading enzymes and secrete survival factors, cytokines, chemokines to protect tumor cells
from apoptosis by providing a suitable environment (6).
Although scientific investigation has gained much ground in understanding the molecular,
cellular and genetic mechanisms (1, 3, 9), much has yet to be discovered in regard to non-genetically
driven events (including the tumor microenvironment) contributing to neuroblastoma progression.
Further studies, later discussed, will allow improvement in patient outcomes and advanced treatment
therapies. We hypothesize that native human bone-marrow derived mesenchymal stromal cells possess
the ability to migrate from the bone marrow to primary neuroblastoma tumors where they contribute
to a reactive pro-tumorigenic stroma.

10

Results
Neuroblastoma cells attract BMMSC in vitro
To determine if chemoattraction was occurring between neuroblastoma and BMMSC, the
migration of BMMSC toward five neuroblastoma cell lines was analyzed via a series of transwell
migration assays (Figure 2A, 2B). The approach consisted of plating BMMSC in the upper compartment
with neuroblastoma in the lower compartment of a co-culture system and counting the BMMSC that
migrated through the semi-permeable (8 uM) pores membrane (Figure 2C, 2D, 2E). In one of two
independent experiments, data revealed an increase in migration of BMMSC when co-cultured, CHLA-
255 (P < 0.001), SK-N-BE(2) (P < 0.001), SK-N-SH (P < 0.001), SH-5YSY (P < 0.005) and NB-19 (P < 0.001),
although spontaneous migration does occur (Figure 2A, 2B). Among the neuroblastoma cell lines, the
co-cultured condition with CHLA-255 shows the highest increase (Figure 2A and 2B) with 2,278 cells per
filter (Figure 2A) and in another experiment, 5,814 cells per filter. From the data, we conclude that
neuroblastoma cells promote the migration of BMMSC.
To analyze what factors may help optimize these transwell migration assays, we looked at two
variables, time (Figure 2F) and ratio (Figure 2G) of cells. Time was looked at to determine whether
proliferation could be responsible for the higher number of cells having crossed. In this assessment, two
neuroblastoma cell lines, CHLA-255 and SK-N-BE(2), were used. Among the members of the three
groups, there is no statistically significant difference among the 16, 24 and 36 hr experiments. BMMSC
at 16 hr migrated at a rate of 1,536 per well, BMMSC at 24 hr migrated 1,622 per well (P = 0.272), and
BMMSC at 36 hours migrated at 1,995 per well (P < 0.001). BMMSC co-cultured with CHLA-255 also did
not reveal a statistically significant difference among the three time points. BMMSC migrated towards
CHLA-255 at a rate of 2,656 cells per well at 16 hr; 2,850 cells per well for 24 hour time point (P = 0.310);
and 2,995.3 cells per well over 36 hr (P = 0.175). Likewise, BMMSC co-cultured with SK-N-BE(2) did not
11

show statistically significant difference among time points. BMMSC migrated towards SK-N-BE(2) at
rates per well of 2,874, 2,815 and 3,040 at 16, 24, 36 hr, respectively. However, findings did reveal a
statically significant difference between the control of BMMSC plated alone and the co-cultured
conditions as seen in other experiments discussed above. This time experiment was important to
optimize and correctly report results because these cells may have started to proliferate after crossing
the barrier and thus not provided an accurate count of migration when analyzing.
Additionally, ratios of BMMSC to neuroblastoma play a key factor (Figure 2G). Changing the
number of neuroblastoma cells was done to determine the specificity for the observation. The more
neuroblastoma present, the more migration we see. In this assay, BMMSC were consistently plated at
15,000 per well on top, and spontaneously migrated at 1,619 per well. The co-culture condition with
CHLA-255 plated at 15k, 30k, 45k and 60k revealed migration rates of 1,800 (P = 0.135), 2,255 (P =
0.003), 2,772 (P = 0.0006) and 2,607 (P = 0.002) when compared to the control group of BMMSC plated
alone. Co-culture conditions with SK-N-BE(2) also showed increased migration rates at higher
concentrations of neuroblastoma. SK-N-BE(2) plated at 15k, 30k, 45k and 60k revealed migration rates
of 2,081 (P = 0.032), 2,143 (P = 0.013), 2,984 (P < 0.0001), and 2,951 (P = 0.004) when compared to the
control of BMMSC plated alone. In conclusion, we found neuroblastoma cells to attract BMMSC in vitro.
Co-Culture of BMMSC and neuroblastoma cells increases production of chemokines/cytokines
After concluding an increase in cell migration was occurring when co-cultured, we explored the
chemokines and/or cytokines that affect these signaling pathways. The serum free contents of the
transwell migration media were profiled for these factors by a multiplex “Luminex” Immunosorbent 10-
plex cytokine assay (Figure 3) and a series of enzyme-linked immunosorbent assays (ELISA) (Figure 4).
Both assays have similar protocol of using a capture antibody for the antigen(s) or sample(s), and a
secondary antibody conjugated to an enzyme allowed quantitative measurement of an Ag-Ab complex
12

compared against a standard. The results of the multiplex immunosorbent cytokine assay show
significant increase in IL-6 (Figure 3B) with human BMMSC and SK-N-BE(2) at 942 pg/mL, IL-6 is a
cytokine linking innate and adaptive immunity which has primary effects on antibody production and
anti-inflammation. Studies within the DeClerck lab reveal IL-6, secreted by BMMSC, as an important
player in the tumor microenvironment. Additionally, this assay revealed increased levels of IL-8 (Figure
3C) at 1,631 pg/mL when human BMMSC were co-cultured with SK-N-BE(2). IL-8 is a cytokine that
attracts neutrophils, increases adhesion of neutrophils to endothelial cells and causes pro-angiogenic
effects. In addition, the following cytokines were averaged and analyzed in triplicate; however were not
detected in the assay: IL-1b, IL-10, GM-CSF, IL-5, INF-g, TNF-a, IL-2 and IL-4 (Figure 3A). ELISA was used
to confirm the observations made with the Luminex assay and to identify other chemokines possibly
involved in the chemoattraction of BMMSC.
Results of the enzyme-linked immunosorbent assays (ELISA) revealed an increase in expression
of IL-6 (462 pg/mL), IL-8 (208 pg/mL) and MCP-1 (1,584 pg/mL) when human BMMSC were co-cultured
with CHLA-255 (Figure 4). MCP-1, is an important chemoattractant for monocytes, T-cells and NK cells
(Figure 4A). Additionally, high amounts (225 pg/mL) of SDF-1, a protein that activates leukocytes and
induced by pro-inflammatory stimuli, were seen in human BMMSC cultured alone, whereby data
showed a 9-fold decrease when co-cultured (Figure 4C). IL-8 was found to be significant with levels
greater than 208 pg/mL in the presence of BMMSC co-cultured with CHLA-255 (Figure 4B), and increases
of IL-6 higher than 450 pg/mL have a 9-fold increase of the co-cultured conditions from human BMMSC
and CHLA-255 (Figure 4D). From the data, we see increases in few chemokines / cytokines that could
have a chemoattractive function, such as IL-8, but further investigation, including the source of
production is needed.
13

Additional analysis in using MCP-1 as a recombinant protein show no statistically significant
difference in chemoattraction with BMMSC as the control having migration rates of 312 cells per filter
when compared to BMMSC with MCP-1 at 1 ng/mL to have 373 cells per filter (P = 0.761), at 5 ng/mL
with 336 cells per filter (P = 0.56) and at 10 ng/mL with 322 cells per filter (P = 0.770). These data reveal
that MCP-1 is unlikely to be the chemoattractant (Figure 4E).
Neuroblastoma cells may attract BMMSC in vivo
After in vitro analyses, a series of experiments were performed in mouse models to test the
hypothesis that neuroblastoma attracts BMMSC from the bone marrow to the primary tumor. To
answer this question, our approach consisted of three in vivo experiments in which mice were
orthotopically injected with CHLA-255 cells into the left kidney (Figure 5). In the first experiment, CHLA-
255 cells stably transfected with a virus that carried an expression vector that included the cDNA of
firefly luciferase were monitored by MRI (Figure 6) and bioluminescence (Figure 7) for tumor growth.
After tumor growth of 21 days, human BMMSC (Figure 8C, 8D) and human fibroblasts (Figure 8A, 8B) as
the control, both labeled with green fluorescence membrane dye (Figure 8), were injected intravenously
and migration patterns analyzed after 24 hr and 48 hr. Findings reveal both cell types at the 24 hour
time point to have migrated into the lungs (Figure 9A, 9D). However, neither human BMMSC (Figure 9B,
9C) nor human fibroblasts (Figure 9E, 9F) were detected in the tumor kidney, non-tumor kidney, liver or
lungs at the 48 hour time points as analyzed by frozen sections, where nuclei were stained in blue with
Dapi (Figure 9).
In order to follow the cell migration patterns for a longer period of time, we attempted to
transfect the human BMMSC with a virus carrying an expression vector that included the cDNA of firefly
luciferase. Unfortunately this transduction was unsuccessful and although the control human fibroblast
transduction was successful, the experimental design was changed to a mouse / human system. Mouse
14

BMMSC that had previously been labeled would allow us to follow cell migration patterns via
bioluminescence. In this second in vivo experiment, similar to the third, mouse BMMSC and NIH-3T3
mouse fibroblasts, as a control, expressing the firefly luciferase gene were monitored throughout a six
week time interval (Figure 10A, 11A) in tumor-bearing mice. One week prior, mice were given one
million neuroblastoma CHLA-255 cells, not expressing luciferase into the left kidney.
On day 0 of the intravenous injection of mouse BMMSC or NIH-3T3, mice were imaged and
bioluminescence scans revealed that cells in both groups have migrated into the lungs and away from
the lungs by the next scan on day 6 (Figure 10A). Both mouse BMMSC and NIH-3T3 remained in the
retro-orbital area several weeks following injection. No visible detection of either cell type was found in
the tumor kidney or other organs from the bioluminescence scans. Upon sacrificing these mice, the
Xenogen instrumentation (a device that captures the light emitted of luciferin and luciferase with a
sensitive camera, see materials and methods) showed no detectable luminescence in the tumor kidney
(Figure 10B), (normal) kidney, liver or lungs. In addition, the tumor kidney was imaged ex vivo (Figure
10C) and lysates of tumor fragments were obtained and examined for luciferase activity but revealed
negligible detectable levels of luminescence (M#1 - 58, M#2 - 50, M#4 - 72, M #5 - 78 luminescence per
10mg (RLU) in mice with BMMSC and M#4 - 90 and M#5 - 62 luminescence per 10 mg (RLU) in mice with
NIH-3T3) compared to a control of 500 luminescence per 10mg (RLU) (Figure 10B).
In the last of the in vivo experiments, the same cells (mouse BMMSC and NIH-3T3 mouse
fibroblasts, expressing the firefly luciferase gene) were injected into the right femur of tumor-bearing
mice and monitored throughout the same time period (Figure 11A). Results showed BMMSC and NIH-
3T3 migrate to the lungs on day 0 (Figure 11B) and away from the lungs as illustrated in the following
scan on day 6. Both groups show cell lines remain in the femur following injection (Figure 11A). On day
24, the BMMSC group appeared to have migrated to the tumor kidney in mouse #3 (Figure 11C). Images
15

on day 41 of the BMMSC group show mice #1, #2, and #5 have signals in the abdomen (Figure 11D).
From these data, we conclude, as expected, the control group of NIH-3T3 mouse fibroblasts do not
migrate into the tumor kidney. We determined the results inconclusive because of the limitations and
sensitivity of the bioluminescent tracking method since the Xenogen instrumentation shows background
signals which are misleading (Figure 15). To counteract this uncertainty, additional attempts were made
to further analyze, which included ex vivo imaging of the tumor kidney (Figure 11E) and analysis of
lysates of the tumor kidney for detectable levels of bioluminescence; however non-detectable amounts
were found in the alternate methods (M#1 - 54, M#2 - 56, M#3 - 66, M#4 - 66, M #5 - 60 luminescence
per 10 mg (RLU) in mice with BMMSC and M#1 - 64, M#2 - 46, M#3 - 74, M#4 - 68 and M#5 - 58
luminescence per 10 mg (RLU) in mice with NIH-3T3) when compared to a control of 500 luminescence
per 10 mg (RLU) (Figure 11F).

16

Discussion
This work is important to further understand the role BMMSC play in the progression of the
tumor microenvironment. Our studies revealed three important findings: (A) neuroblastoma cells
attract BMMSC in vitro, (B) co-culture of BMMSC and neuroblastoma cells increases production of
chemokines/cytokines, and (C) neuroblastoma cells may attract BMMSC in vivo.
In our attempts to understand if migration of BMMSC increased toward neuroblastoma, we
analyzed five neuroblastoma cell lines (Figure 2A, 2B) and modified factors such as time (Figure 2F) and
cell ratio (Figure 2G) to optimize these assays. Adjusting the time variable was essential for correctly
reporting data because if the cells had been plated for too long, they may have started proliferating
after crossing the barrier and therefore data counts would have been inaccurate. Additionally, changing
the number of neuroblastoma cells was performed to determine the specificity for the observation to
see if increasing amounts of neuroblastoma caused BMMSC to migrate at higher rates, which we
concluded to be true. After assessing these factors, we modified the assay and concluded that
neuroblastoma cells promote the migration of BMMSC in vitro.
Interestingly, when human BMMSC and neuroblastoma are co-cultured in serum free media,
expression levels of cytokines / chemokines increases; this was our second major finding. We looked at
MCP-1 in more detail in an attempt to see if it plays a major factor in the chemoattraction of BMMSC
towards neuroblastoma. MCP-1 was used as a recombinant protein in a transwell migration assay which
showed no statistically significant difference in the rates of BMMSC migration toward the protein (Figure
4E). From these findings, we concluded that MCP-1 is unlikely to be the chemoattractant. Another
potential candidate of this chemoattraction is SDF-1; however, data show this protein is secreted by
BMMSC (Figure 4C) thus would be the reason for neuroblastoma to migrate towards BMMSC (not the
other direction). This knowledge is important in metastases analysis of the bone marrow whereby
17

neuroblastoma migrate towards BMMSC. Additionally, our data showed a statistically significant
increase in IL-6 from co-cultured conditions (Figure 3B, 4D) although the DeClerck lab has found IL-6 to
originate from BMMSC. Other scientists have shown human BMMSC but not murine BMMSC stimulate
tumor growth primarily through the paracrine production of secreted IL-6 (6). Specifically for this
project, we are analyzing factors which the neuroblastoma secretes in the presence of BMMSC, not
what BMMSC are producing. Lastly, promising data show IL-8 with increased expression levels in co-
cultured conditions (Figure 3C, 4B), which may aid in our understanding of the mechanistic ways in
which these cells interact. Further studies must be performed to look at migration with IL-8
recombinant proteins and which cell is producing this protein.
Other studies to continue this work should use additional screening panels to reveal and analyze
what other factors exist that could be increased in the co-culture of BMMSC and neuroblastoma cells as
we only addressed 12 factors within the scope of this project. Once other factors are revealed to
increase in the presence of co-culture conditions, analyzing gene expression through differential display
of the RNA when compared to c-DNA libraries, antibody blocking experiments (receptor or ligand)
and/or recombinant proteins in migration assays should be performed.
Next, the third finding showed that neuroblastoma cells may attract BMMSC in vivo, and here
we briefly analyze the orthotopic model used for all in vivo work. Critiques of this method argue that
because the kidney is highly vascularized and upon injecting this organ it may become damaged (Figure
13) and the tumor will metastases to other areas of the body or grow external to the kidney (Figure 6),
which presents challenges when analyzing migration of BMMSC into the primary tumor. Although
alternative methods are available, we determined this method sufficient because having a primary
tumor is the important aspect for this project.
18

In the first in vivo model where human BMMSC and human fibroblasts were stained with PKH67
green fluorescence membrane dye and injected intravenously in the retro-orbital, we found our cells of
interest migrated into the lungs after 24 hours (Figure 9A), away from the lungs at 48 hours and were
not found in the liver, (tumor) kidney or (normal / control) kidney at any time points. This experiment
was difficult to analyze because of background and the natural fluorescence of tissue (Figure 14).
However, to overcome this limitation, photographs were overlaid with Dapi to stain the nuclei blue
which allowed measurement of green “spots” for comparison to the average nucleus size (signals which
were smaller than blue stained nuclei cannot be the cells of interest). This dye method was used for the
ability to label the cell membrane quickly and due to the ease of detection. However, we were limited
in the length of time these cells could be tracked in vivo because of the brightness decreasing by half
with every cell division.
In the second in vivo experiment to continue analyzing our findings that neuroblastoma may
attract BMMSC in vivo, we attempted to transfect human BMMSC with a virus that carries an expression
vector which includes the cDNA of firefly luciferase. However, after two attempts, these cells presented
with an altered membrane and were not proliferating, thus concluding it to be unsuccessful. BMMSC
are difficult to transfect since they are primary, not established cells. Additionally, each batch of human
BMMSC behave differently and because these BMMSC are mortal, they begin to show signs of slower
growth around passages 6-8 but typically survive to passages of no greater than 10. As a control to
analyze how much of the culture had picked up the virus, and to ensure the design was working as
expected, puromycin was used to select for cells that have been effectively infected and in which the
vector is effectively expressing since it contains a puromycin resistant cDNA thus it kills those cells which
were not expressing the gene. In varying levels of puromycin concentration, it was unclear if the cells
were exposed to toxic concentrations that resulted in necrosis even as low as 0.2 μg/mL - 0.5 μg/mL
when the recommended dose in cell cultures is within a range of 1-10 μg/mL (per manufacture’s
19

protocol). The virus transfection of the human fibroblast control appeared to have behaved as
expected.
Although transfection of the human BMMSC was unsuccessful, alternate cell lines previously
transfected with the firefly luciferase gene were used, mouse BMMSC and NIH-3T3 mouse fibroblasts.
Using these cells allowed us to continue analysis if BMMSC are attracted to neuroblastoma primary
tumors in vivo. The cell change resulted in a mouse/human system because the tumors were CHLA-255,
a human cell line from patients at Children’s Hospital Los Angeles. However, a migration assay showed
that mouse BMMSC in the presence of CHLA-255 (P < 0.001) and SK-N-BE(2) (P < 0.001) does have a
statistically significant difference in migrating rates when compared to spontaneous migration in vitro.
In vivo experiment #2 followed mouse BMMSC and NIH-3T3 injected intravenously into tumor
bearing mice; however bioluminescence scans did not reveal signals near the kidney which contained
the tumor. However, in vivo experiment #3 illustrated in three separate scans that BMMSC migrate out
of the femur and into the body (Figure 11A). On day 0, BMMSC migrated into the lungs, on day 24 a
signal is found in the left kidney area and on day 41, three mice show signals in the abdominal cavity.
Although we see signals, there are limitations to this method.
The limitations in the interpretation of the data lie primarily in the instrumentation of the
bioluminescence Xenogen machine. Although increasingly more accurate in recent years, both
endogenous and exogenous factors can impact the various components of the luciferase reaction which
may lead to erroneous readouts. Examples include signal quantification, background, half-life of the
enzyme and stability, cellular environment (oxygen, hypoxia and oxidative stress), substrate availability,
luciferin efflux, administration route of the substrate, light quenching and scattering, spatial resolution
and more (12). The sensitivity of the scans, primarily in background, can make it difficult to analyze
(Figure 15). Therefore this has the potential to create signals which may not be real. To confirm or deny
20

the presence of luminescence in the tumor kidneys, ex vivo scans were performed in the Xenogen
machine of the entire tumor kidney (Figure 10C, 11E), and again by a homogenate mixture with plate
reader instrumentation (Figure 10B, 11F). Neither group tested positive for luminescence in the non-
tumor kidney, lungs, and liver, although varying time points had positive signals in those locations.
Additional analysis has yet to be performed by immunohistochemistry using an anti-firefly luciferase
antibody where one single cell could be detected. If there are very few cells that have migrated, it may
not be detectable in the bioluminescent scans or homogenate mixture luminescence analysis thus we
conclude that BMMSC may migrate to the primary tumor in the kidney; however additional processing is
required.
In analyzing and comparing published literature on this scientific niche, similar studies
performed with mouse BMMSC illustrate that these cells migrate into the primary tumor located
subcutaneously; however in one mouse study using mouse BMMSC intravenously injected and tumors
of 4T1 mammary carcinoma injected subcutaneously where n=5, only one mouse showed luminescence
of 3,000 per 10 mg, another had 1,250 per 10 mg, while the other three mice showed negligible
amounts (3). This study used an aggressive cancer, with tumors just below the skin, and were able to
illustrate these mice BMMSC migrate to the tumor just after two weeks. Our study differs in the type of
tumor (neuroblastoma), tumor location (in the kidney) and the amount of BMMSC injected (2 million in
the Bergfeld study versus the 300,000 cells used in this study).
To conclude, we found neuroblastoma to attract BMMSC in vitro, co-cultures of BMMSC and
neuroblastoma cells increase production of chemokines/cytokines, and neuroblastoma cells may have
the ability to attract BMMSC in vivo; however additional analysis is needed.

21

Materials and Methods
Cells:
Human bone marrow mesenchymal stromal cells (BMMSC) were harvested from patients at
Children’s Hospital Los Angeles (patient #3 and NO450). These cells along with mouse BMMSC were
cultured in either Iscove’s Modified Dulbecco Medium (IMDM) with 20% fetal calf serum (FCS) and 1%
penicillin-streptomycin or Dulbecco’s Modified Eagle Medium (DMEM) which included 20% fetal calf
serum (FCS) and 1% penicillin-streptomycin.
Human skin fibroblasts and NIH-3T3 mouse fibroblasts were used as a control in both the in vitro
and in vivo experiments. These cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM)
which included 20% fetal calf serum (FCS) and 1% penicillin-streptomycin.
Five neuroblastoma cell lines were used in this project, consisting of CHLA-255, NB-19, SK-N-
BE(2), SK-N-SH, and SH-5YSY. All neuroblastoma cells were cultured in Dulbecco’s Modified Eagle
Medium (DMEM) containing 10% fetal calf serum (FCS) and 1% penicillin-streptomycin.
Transwell Migration Assay:
The transwell migration assays were completed in a multi-step process. After multiple protocol
changes to optimize conditions, experiments had neuroblastoma cell lines, 45,000 cells per well, plated
in DMEM with 2% FCS for 24 hours. At the 24 hr time mark, the media and floating/dead cells were
removed and additional media (DMEM, serum free) was added to the remaining neuroblastoma cells.
Immediately adding new media, BMMSC, human fibroblasts or NIH-3T3 (all at 15,000 cells per well)
were plated in a 24-well, 8.0 um poor size transwell inserts and exposed to the various neuroblastoma
cell lines below (or not co-cultured as a control to view spontaneous migration). Once migration
22

completed, inserts were removed and cleaned by swab on the upper side to displace the cells which did
not migrate. Filters were then stained by using a Hema3 Stat Pack dye kit; manufacture protocol was
strictly followed. The filters were removed and mounted on slides for analysis. Each filter was counted
in 9 separate fields on 20x, those values were multiplied by the surface area and averaged with standard
deviations.
Luminex Assay:
To analyze the quantification of the expression levels from the cytokines and chemokines in the
supernatant from the transwell migration assays, a human cytokine, 10-Plex Panel Luminex platform
from Life Technologies was used. The multiple panels provided data from each sample well for averages
in triplicate. Protocol was strictly followed as per the manufacturer and a concept identical to that of
ELISA (with the only difference in that multiple capture antibodies “beads” were plated on the bottom).
ELISA Assay:
Enzyme-linked immunosorbent assay (ELISA) is a technique that uses antibodies and color
change to identify a substance. The supernatant from the transwell migration assays was used in
triplicate for the ELISA assays to measure concentrations of MCP-1, IL-8, IL-6 and SDF-1 and to confirm
the findings of the Luminex assay. The protocol consisted of (1) the plate being coated with a capture
antibody and sat for a minimum of 24 hours (2) samples were added in triplicate and antigens present
bound to the capture antibody (3) detection antibody was added that bound the antigen (4) enzyme-
linked secondary antibody was added that bound to the detecting antibody and lastly (5) substrate was
added to be converted by enzyme into a detectable form.

23

Cell Staining:
Human BMMSC and human fibroblasts were labeled on the same day as injection with PKH67
green fluorescence dye in which the fluorescence is emission of a lower energy (green) light upon
exposure to high energy (UV) light when analyzed by microscopy that give the ability to observe where
these cells migrate by looking at tissue sections of organs of interest.
Virus Transduction:
Mouse BMMSC, NIH-3T3 and CHLA-255 cells were given a virus that carries an expression vector that
includes the cDNA of firefly luciferase for the ability to follow via bioluminescence. This is the emission
of a light through a chemical reaction (luciferin, ATP, luciferase) that produce visible wavelengths at 560
nm and are captured with the Xenogen machine and analyzed among time points.
Animal In-Vivo Experimental Design:
The first in vivo experiment neuroblastoma cell line, CHLA-255 labeled with firefly luciferase was
injected into the kidneys of immune-depressed mice (Figure 4). In addition, the mice were injected with
“green dye” labeled human BMMSC or human fibroblasts as a control (Figure 7) via intravenous
injection for in vivo experiment one. The migration of these cell lines was analyzed after 24 hr and 48
hr time points to see if their homing in orthotopic tumors could be confirmed by immune fluorescence.
In a second in vivo experiment, mouse cell lines BMMSC and NIH-3T3, which had undergone
transfection of the firefly luciferase virus, were injected (300,000 per mice) via intravenous injection to
immuno-depressed mouse to see if they migrate toward the orthotopic tumors in the left kidney (Figure
10). One week prior, mice had been given 1 million CHLA-255 neuroblastoma cell tumors.
In the third in vivo experiment, same cells from the second in vivo experiment (mouse BMMSC
and NIH-3T3 labeled with the firefly luciferase gene) were used at an amount of 300,000 per mouse and
24

introduced into immuno-compromised (nude) mice via right femur injection and followed by
bioluminescence (10) (Figure 11). One week prior, mice had been given 1 million CHLA-255
neuroblastoma cell tumors via orthotopic injection into the left kidney.
Orthotopic Tumor Surgery:
Each mouse is put under anesthesia via isofluorine and the surgical site is sterilized with iodine.
A cut is then made medial to left lateral with a blade. The fat pad is used to grab the kidney and expose
the kidney, resting it outside the body cavity with a cotton swab for easy access. A needle is then
injected inferior to superior and cells delivered are in the superior portion of the kidney. The kidney is
placed back inside the body cavity and the cut is closed with one stitch followed by a staple applied on
top of stitch. Heat pads may be used to comfort healing animals and it is highly recommended to
provide animals with pain killers and antibiotics five days post surgery (11) (Figure 5).
Magnetic Resonance Imaging:
Magnetic resonance imaging (MRI) was used to aid in our investigation of the anatomy and
physiology of the mouse bodies to track tumor progression (disease) and monitor health of animal. The
animals were injected with 30 uL of Megnevist, to reduce background, and then put under with
anesthesia (isofluorine). Mice were placed one at a time, in an open pipe and inserted into the scanner
for three minutes where magnetic fields and radio waves formed images of the body (Figure 6).
Statistics:
Statistics for the transwell migration assays used the t-test in which the test statistic follows a
Student's t distribution if the null hypothesis is supported. The purpose was to determine if two data
sets of data were significantly correlated or different from each other. Studies performed in Microsoft
25

Excel used two tails with an unequal two sample variance between the control group and analyzed
group.

26

Future Directions
The findings from this research illustrate potential for important discoveries. Future directions
should aim to eliminate the controversial role of BMMSC in the tumor microenvironment and disease
progression. Since BMMSC are multi-potent stromal cells that differentiate, their ability and reasons for
doing so should be examined by studies focusing on the factors, signals and stimuli that are causing this
intricate cross-talk between BMMSC and tumor cells, not solely on what is attracting BMMSC to
neuroblastoma, but what BMMSC may be secreting that attracts neuroblastoma to the bone marrow.
Results will aid in further knowledge of migration patterns of both BMMSC and tumor cells in
metastasis. Once a grasp on the mechanistic action is understood, therapeutic studies to block,
promote or modify behavior of these cells will be key in treatment of patients.
This project specifically, should focus on additional experiments that use an in vivo human /
human system (human BMMSC and human neuroblastoma), to provide a realistic view on the action of
these cells, opposed to the mouse / human system used in a majority of in vivo work for this project.
Next steps will perform screens of additional molecules on their chemoattraction ability using multiplex
panel platforms with samples from supernant of co-cultured BMMSC and neuroblastoma. The proteins
that have increased expression levels in this assessment must be further analyzed to understand which
cell is secreting them (BMMSC or neuroblastoma) by using recombinant proteins in migration assays to
see if a change in migration is noted when only this protein is present. Other ways to analysis up
expression levels is to look at the gene expression of these cells and determine which have amplified
regions. Another approach is to perform antibody blocking experiments (based on findings from the
multiplex screening panel) on either receptor or ligand to look for changes in migration.
In vivo studies should follow-up these findings by using human BMMSC expressing the firefly
luciferase gene, injected into their place of origin, the femur, and monitored for a lengthy amount of
27

time or until the experimental end point (tumor size) has been reached for each animal. For optimal
results, a larger amount of cells (2 million) should be injected for easier detection and monitoring. Once
migration has been understood in vivo, the additional work should be used in mouse models to block
those molecules that affect chemoattraction and look what changes are taking place in cell migration.
Future studies with reproducible, reliable data will allow us to have a better understanding of
the mechanisms in the tumor microenvironment and therefore, ability to detect disease early, or
potentially create specific cancer therapeutics. If BMMSC are pro-tumorigenic, one could target them
with a monoclonal antibody in the tumor microenvironment. Another way could also target these
BMMSC by using chimeric antigen receptors (CARs). Investigators developed CARs for expression on T
cells, and when it is derived from an antibody, the resultant cell should combine the desirable targeting
features of an antibody (e.g. lack of requirement for major histocompatibility complex recognition and
ability to recognize non-protein antigens) with the persistence, trafficking, and effector functions of a T
cell (13). Additionally, if BMMSC migrate to the tumor, one could use these cells to deliver drugs once
they arrive in the tumor microenvironment by using as vector vehicle to transfer genetic material to a
tumor cell. Knowledge and application of these mechanisms will improve the quality and length of life
for patients with this disease. This is an exciting space to further investigate.

28

Figures
Figure 1: Bone marrow mesenchymal stromal cells (BMMSC) migrate from the bone marrow into the
circulating blood stream where they differentiate over time.

A. Bone marrow mesenchymal stromal cells (BMMSC) reside in the bone marrow where they regenerate or
migrate out to differentiate into chondrocytes, osteoblasts, neurons, adipocyte, myocyte, or fibroblasts. B. Cross-
talk occurs between BMMSC and neuroblastoma as seen by BMMSC found in the tumor microenvironment, while
neuroblastoma metastasizes in the bone marrow where BMMSC reside.
A
B
29
Figure 2: Migration of human BMMSC increases in the presence of neuroblastoma cell lines.

0
500
1000
1500
2000
2500
3000
Human
BMMSC
BMMSC +
CHLA-255
BMMSC + NB-
19
BMMSC + SH-
5YSY
BMMSC + SK-
N-BE(2)
BMMSC + SK-
N-SH
Cells per Filter
0
1000
2000
3000
4000
5000
6000
7000
Human
BMMSC
BMMSC +
CHLA-255
BMMSC + NB-
19
BMMSC + SH-
5YSY
BMMSC + SK-
N-BE(2)
BMMSC + SK-
N-SH
Cells per Filter
A
B
C D
E
P < 0.001
P < 0.001
P < 0.001
P < 0.001 P < 0.001
P < 0.001
P < 0.001
P < 0.005
P < 0.001
P < 0.001
30

A. Data represent mean number of human BMMSC per 9 fields with standard deviation. Results are from 1 of 2 experiments
performed in triplicate, with 2% FCS of a transwell migration assay. BMMSC were plated on top of semi-permeable barrier
with five neuroblastoma (NB) cells lines analyzed: CHLA-255, NB-19, SK-N-BE-2, SK-N-SH, and SH-5YSY. B. Results from 2 of 2
experiments as noted in panel A. C. Photographs shows stained filter of spontaneous migration of BMMSC plated on top of
2% FCS, 10x. D. Filter with increased number of BMMSC migrated towards CHLA-255, 10x. E. Filter with increased migration
of BMMSC towards SK-N-BE(2), 10x. F. Data represent the mean number of BMMSC that migrated in 1 experiment,
performed in triplicate for times of 16, 24 and 36 hours. G. Data show mean number of human BMMSC to have migrated
with varying amounts of neuroblastoma cell lines, in ratios of BMMSC to NB in 1:1, 1:2, 1:3, and 1:4, respectively. H. Data
show increased levels of mouse BMMSC migrating towards human neuroblastoma cell lines CHLA-255 and SK-N-BE(2).
0
500
1000
1500
2000
2500
3000
3500
4000
Cells per Filter
16 h
24 h
36 h
0
500
1000
1500
2000
2500
3000
3500
Cells per Filter
15 k
30 k
45 k
60 k
0
1000
2000
3000
4000
5000
6000
7000
Mouse BMMSC Mouse BMMSC + CHLA-255 Mouse BMMSC + SK-N-BE(2)
Cells per Filter
F
G
P = .272
P < 0.001
P = .310
P = 0.173
P = .546
P = 0.117
16hr P < 0.001
24hr P < 0.005
36hr P = 0.012

16hr P < 0.001
24hr P < 0.001
36hr P < 0.001

P = 0.009
P = 0.002
P = 0.004
P = 0.629
P = 0.003
P = 0.012
15k P = 0.135
30k P = 0.003
45k P = 0.0006
60k P = 0.002

15k P = 0.032
30k P = 0.013
45k P < 0.0001
60k P = 0.004

P < 0.001
P < 0.001
H
31

Figure 3: Multiplex “Luminex” immunosorbent cytokine assay reveals increased expression of IL-6 and IL-8 in co-
culture.
IL-1b IL-10 IL-6
GM-
CSF IL-5 INF-g TNF-a IL-2 IL-4 IL-8
Human BMMSC
alone
0
pg/mL
0
pg/mL
57
pg/mL
0
pg/mL
0
pg/mL
0
pg/mL
0
pg/mL
0
pg/mL
0
pg/mL
0
pg/mL
CHLA-255 alone 0 0 0 0 0 0 0 0 0 0
SK-N-BE(2) alone 0 0 0 0 0 0 0 0 0 46
Human BMMSC +
CHLA-255
0 0 239 0 0 0 0 0 0 204
Human BMMSC +
SK-N-BE(2)
0 0 942 0 0 0 0 0 0 1631
Human fibroblast
alone
0 0 2371 0 0 0 0 0 0 3553

A. Data shown from a one-time multiplex “Luminex” immunosorbent 10-plex cytokine assay with serum-free samples
performed in triplicate with standard deviation. Cytokines IL-1b, IL-10, GM-CSF, IL-5, INF-g, TNF-a, IL-2, and IL-4 had
negligible detectable amounts (in pg/mL). Results from IL-6 show human BMMSC alone at 57 pg/mL, negligible amounts of
CHLA-255 alone and SK-N-BE(2) alone, human BMMSC co-cultured with CHLA-255 at 239 pg/mL, human BMMSC co-cultured
with SK-N-BE(2) at 942 pg/mL and human fibroblasts with the highest at 2,371 pg/mL B. Data from the one-time multiplex
assay show in graph form of IL-6. C. Data from the one-time multiplex assay show in graph form of IL-8.
0
500
1000
1500
2000
2500
IL-6 (pg/ml)
IL-6
0
500
1000
1500
2000
2500
IL-8 (pg/ml)
IL-8
MSC
A
B C
32
Figure 4: ELISA assays reveal increased expression levels of MCP-1, IL-8, and IL-6 when human BMMSC
are co-cultured with CHLA-255.

0
200
400
600
800
1000
1200
1400
1600
1800
2000
MCP-1 pg/mL
0
50
100
150
200
250
IL-8 pg/mL
0
50
100
150
200
250
300
350
400
450
500
IL-6 pg/mL
0
50
100
150
200
250
SDF-1 pg/mL
A B
C
D
33

A. Data represent mean amount in triplicate of MCP-1 (in pg/mL)with standard deviations from serum-free media
conditions in varying cell cultures (alone and co-cultured). B. Data represent mean amount in triplicate of IL-8 (in
pg/mL) with standard deviations from serum-free media conditions in varying cell cultures (alone and co-cultured).
C. Data show mean amount in triplicate of SDF-1 (in pg/mL) with standard deviations from serum-free media
conditions in varying cell cultures (alone and co-cultured). D. Data represent mean amount in triplicate of IL-6 (in
pg/mL) with standard deviations from serum-free media conditions in varying cell cultures (alone and co-cultured).
E. Data represent mean amount in triplicate of migration of human BMMSC towards recombinant proteins of MCP-
1 in increasing concentrations of 1ng/mL, 5ng/mL and 10ng/mL with standard deviations. F. ELISA of human MCP-
1 antibody illustrate expression levels in varying conditions of mouse BMMSC or NIH-3T3 with CHLA-255, SK-N-
BE(2) or TH-MYCN.

0
100
200
300
400
500
600
Human
MSC
Human
MSC +
MCP-1 @
01 ng/mL
Human
MSC +
MCP-1 @
05 ng/mL
Human
MSC +
MCP-1 @
10 ng/mL
Cells per Filter
0
20
40
60
80
100
120
MCP-1 pg/mL
E
F
34

Figure 5: Orthotopic model of pediatric solid tumors (neuroblastoma).

A. Each mouse is put under anesthesia via isofluorine and the surgical site is sterilized with iodine. B. Cut is made
medial to left lateral. C. The fat pad is used to grab the left kidney and the kidney is exposed with a cotton swab.
D. Needle is injected inferior to superior and neuroblastoma cells delivered are in the superior portion of the
kidney (orthotopically). E. The kidney is then placed back inside the body cavity. F. The cut is closed with one
stitch. G. Staple is applied on top of the stitch. H. A heat pad is used to comfort healing animals.
A B
E
F
D C
G H
35

Figure 6: CHLA-255 tumor growth in the left kidney monitored by MRI.

A B
A. Post surgical image day 11, from an
orthotopic injection of 1.2 million CHLA-255
neuroblastoma cells labeled with luciferase and
injected into the superior portion of the left
kidney. B. MRI image collected 31 days
following the orthotopic injection of 1 million
CHLA-255 neuroblastoma cells into the left
superior kidney. Image shows superficial tumor
growth on left lateral. C. Image shows tumor
growth in superior portion of left kidney 11 days
after an orthotopic injection of 1.2 million
CHLA-255 neuroblastoma cells labeled with
luciferase.

C
Tumor
Tumor
Tumor
36

Figure 7: CHLA-255 tumor growth in the left kidney monitored by bioluminescence.

A. Post surgical image of 10 mice on day 11, following an orthotopic injection of 1.2 million CHLA-255
neuroblastoma cells labeled with luciferase and injected into the superior portion of the left kidney. B. Post
surgical image of 8 mice on day 18, following an orthotopic injection of 1.2 million CHLA-255 neuroblastoma cells
labeled with luciferase and injected into the superior portion of the left kidney.

A
B
37

Figure 8: Labeling of BMMSC cell membrane with PKH67 green fluorescent membrane dye in vitro.

PKH67 green fluorescent membrane dye is used to label the cell membrane: A. Human fibroblasts, shown at 10x.
B. Human fibroblasts, 40x. C. Human BMMSC, 10x D. Human BMMSC, 40x.

A B
C
D
38

Figure 9: Tracking of PKH67 green fluorescent membrane labeled BMMSC in vivo.

A. Human BMMSC, labeled with green fluorescence membrane dye, shown in lung frozen section after 24 hours,
10x. B. Human BMMSC not seen in lung after 48 hours, 10x. C. Human BMMSC not in tumor kidney after 48
hours, 10x. D. Human fibroblasts in lung after 24 hours, 10x. E. Human fibroblasts not in lung after 48 hours, 10x. F.
Human Fibroblasts not in tumor kidney of frozen sections after 48 hours, 10x.

39

Figure 10: Mouse BMMSC injected intravenously (retro-orbital) and monitored by bioluminescence.
Day Mouse BMMSC NIH-3T3 Day Mouse BMMSC NIH-3T3
0

24

6

28

10

35

14

41

17

44

21

0
100
200
300
400
500
Control M1 M2 M4 M5
Luminescence per 10mg (RLU)
Intravenous Injection
NIH-3T3
BMMSC
A
A. Image scans on day 0 of the group
injected with 300,000 labeled mouse
BMMSC show migration to the lungs via
bioluminescence. Both cell lines
(mouse BMMSC and NIH-3T3 mouse
fibroblasts) remain in the orbital weeks
following injection. No visible detection
found in the tumor kidney with
bioluminescence. B. Ex vivo analysis of
homogenized left tumor kidney after 44
days did not reveal detectable amounts
of luminescence from either cell type.
C. Ex vivo imaging of entire tumor
kidneys in bioluminescence Xenogen
machine, on day 44 revealed no
detectable levels of bioluminescence.
B
C

40
Figure 11: Mouse BMMSC injected in the femur monitored by bioluminescence.

0
6
10
14
17
21
24
28
35
41
44
Mouse BMMSC NIH-3T3 Mouse Fibroblast
A

41

A. Images show bioluminescence scans from day 0, when 300,000 mice BMMSC and 300,000 NIH-3T3, both
labeled with luciferase were injected into the femur (bone marrow) of 10 mice, through day 44. B. Scans on day 0
show both cells migrate from the femur into the lungs. C. On experimental day 24, signal of mouse #3 shows
BMMSC in the region of the left tumor kidney. D. On day 41, signal of mice #1, #2, #5 show BMMSC in the region
of the abdomen. E Ex vivo imaging of entire tumor kidneys in bioluminescence Xenogen machine, on day 44
revealed no detectable levels of bioluminescence. F. Ex vivo analysis of homogenized left tumor kidney of all mice
after 44 days did not reveal detectable amounts of luminescence from either cell type compared to control or
bone marrow (BM).
0
100
200
300
400
500
Control BM M1 M2 M3 M4 M5
Luminescence per 10mg (RLU)
Femur Injection
NIH-3T3
BMMSC
B
C
F
D
E
42

Figure 12: Kidney tumors shown in comparison to normal “control” kidneys.

A. Left kidney pictured on top that had one million CHLA-255 neuroblastoma cells injected seven weeks prior.
Bottom shows the normal “control” kidney. B. and C. Same as panel A, although illustrates differences in tumor
size from different mice.
Tumor
Kidney
Normal
Kidney
A B C
43

Figure 13: H&E stains of tumor kidney show necrotic and normal morphology.

A. H&E stain made from
paraffin section of kidney
after tumor had grown for 7
weeks shows necrotic tissue
and normal morphology. B.
Photograph shows tumor
invading normal tissue in the
kidney. C. Multiple nodules
“islands” of tumor
surrounded by stroma, a very
common neuroblastoma
characteristic.
A
B C
44

Figure 14: Natural tissue fluorescence has potential to reveal misleading signals.

A. Misleading signals after 24 hours in the non-tumor kidney of a mouse injected with green fluorescent human
BMMSC. Frozen section shown at 10x. B. Frozen section of tumor kidney of a mouse injected with human BMMSC
illustrates a misleading signal after 24 hours, 10x. C. Another misleading signal after 48 hours in the non-tumor
kidney of a mouse injected with human BMMSC shown from a frozen section at 10x. D. Misleading signals in
human BMMSC, tumor kidney, 48 hours. E. Misleading signal in human BMMSC, tumor kidney, 48 hours.

A B
C
D E
45

Figure 15: Misleading fluorescence due to the sensitivity of the bioluminescence instrumentation.

A. Bioluminescence Xenogen instrumentation reveals misleading signals on day 17 of post femur injection of
labeled mouse BMMSC. Photographs show side and belly scans. B. Bioluminescence Xenogen instrumentation
reveals misleading signals on day 17 of post femur injection of labeled NIH-3T3 mouse fibroblasts. Photographs
show side and belly scans.
A
B
46
Bibliography
1. Barcellos-de-Souzaa, Pedro, Valentina Gori, and Franco Bambi. "Tumor Microenvironment:
Bone Marrow-mesenchymal Stem Cells as Key Players." BiochimicaEtBiophysicaActa (BBA) - Reviews on
Cancer 1836.2 (2013): 321-55.
2. Prockop DJ. Repair of tissues by adult stem/progenitor cells (MSCs): controversies, myths, and
changing paradigms. MolTher 2009; 17:939-46.
3. Bergfeld, Scott A., Laurence Blavier, and Yves A. DeClerck. "Bone Marrow–Derived
Mesenchymal Stromal Cells Promote Survival and Drug Resistance in Tumor Cells.” Division of
Hematology-Oncology, Children's Hospital Los Angeles, 27 Jan. 2014.
4. Cheung, Nai-Kong V., and Michael A. Dyer. "Neuroblastoma: Developmental Biology, Cancer
Genomics and Immunotherapy." Nature Reviews Cancer. Nature, June 2013.
5. Maris, John M. "Recent Advances in Neuroblastoma." New England Journal of Medicine
362.23 (2010): 2202-211.
6. DuBois, SG, Y. Kalika, JN Lukens, and GM Brodeur. "Metastatic Sites in Stage IV and IVS
Neuroblastoma Correlate with Age, Tumor Biology, and Survival." Journal of Pediatric Hematology /
Oncology 21.3 (1999): 181-89.
8. Brodeur, Garrett M. "Neuroblastoma: Biological Insights into a Clinical Enigma." Nature
Reviews Cancer 3.3 (2003): 203-16.
9. Sohara, Y., H. Shimada, C. Minkin, A. Erdreich-Epstein, JA Nolta, and YA DeCleck. "Bone
Marrow Mesenchymal Stem Cells Provide an Alternate Pathway of Osteoclast Activation and Bone
Destruction by Cancer Cells." Cancer Reseach 65.4 (2005): 1129-135.
47

10. Yasuyoshi Sohara; Hiroyuki Shimada; Miriam Scadeng; Harvey Pollack; Shinya Yamada; Wei
Ye; Charles Patrick Reynolds; Yves A DeClerck. Lytic bone lesions in human neuroblastoma xenograft
involve osteoclast recruitment and are inhibited by bisphosphonate. Cancer research 2003;63 (12):3026-
31.
11. Patterson, Danielle M., Jason M. Shohet, and Eugene S. Kim. "Preclinical Models of Pediatric
Solid Tumors (Neuroblastoma) and Their Use in Drug Discovery."Cellular and Animal Models in Oncology
and Tumor Biology. Current Protocols in Pharmacology, Mar. 2011.
12. Badr, Christian E. "Chapter 1: Bioluminescence Imaging - Basics and Practical."
Bioluminescent Imaging: Methods and Protocols. New York: Humana, 2014. 1-17.
13. Dotti, G., S. Gottschalk, B. Savoldo, and MK Brenner. "Design and Development of Therapies
Using Chimeric Antigen Receptor-expressing T Cells." Immunological Reviews 257.1 (2013): 107-26.
Web. 19 Mar. 2015.

Abstract (if available)

Abstract Bone marrow mesenchymal stromal cells (BMMSC) migrate from the bone marrow into the circulating blood stream where overtime they differentiate to adiposities, chondrocytes, myocyte, fibroblasts, osteoblasts or other cell derivatives. We hypothesized that BMMSC migrate from the bone marrow into the primary tumor and participate in the tumor microenvironment. Cells expressing characteristics of BMMSC were obtained from human and mouse bone marrow. Using a migration assay, human BMMSC were assessed for their ability to migrate toward five neuroblastoma cell lines. Higher migration rates were observed in co-culture conditions of CHLA-255 (P < 0.001), SK-N-BE(2) (P < 0.001), SK-N-SH (P < 0.001), SH-5YSY (P < 0.005) and NB-19 (P < 0.001), concluding a chemoattraction (Figure 2). To analyze the presence of small molecules, cytokines and chemokines with a potential role in chemoattraction, Luminex was used as a screening panel (Figure 3) and ELISA assays (Figure 4) validated increased expression levels of IL-6 (462 pg/mL), IL-8 (208 pg/mL) and MCP-1 (1,584 pg/mL), when human BMMSC were co-cultured with CHLA-255. Further analysis of MCP-1 as a recombinant protein did not show increased migration (P > 0.5) concluding an insignificant role in chemoattraction. In vivo, mouse BMMSC were placed in the femur (Figure 11) or intravenously (Figure 10) and tracked via bioluminescence for their ability to migrate towards neuroblastoma CHLA-255 tumors orthotopically injected into the kidney of immuno-deficient (nude) mice over a period of six weeks. On day 24, mouse BMMSC injected into the femur appeared to have migrated to the tumor kidney

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Bone marrow derived mesenchymal stem cells promote survival and drug resistance in tumor cells

PDF

From mesenchymal stem cell therapy to discovery of drug therapy for systemic sclerosis

PDF

The calcium-sensing receptor in the specification of normal and malignant hematopoietic cell localization in the bone marrow microenvironment

Asset Metadata

Creator Hogan, Christine Marie (author)

Core Title Bone marrow-derived mesenchymal stromal cells in the tumor microenvironment of neuroblastoma

School Keck School of Medicine

Degree Master of Science

Degree Program Experimental and Molecular Pathology

Publication Date 04/21/2015

Defense Date 03/20/2015

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Bone Marrow,cancer,mesenchymal stromal cells,neuroblastoma,OAI-PMH Harvest,oncology,stem cells,tumor microenvironment

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor DeClerck, Yves A. (committee chair), Chuong, Cheng-Ming (committee member), Epstein, Alan L. (committee member), Ouellette, Andre J. (committee member)

Creator Email hogancm@usc.edu,norcal12@pacbell.net

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-554193

Unique identifier UC11298778

Identifier etd-HoganChris-3345.pdf (filename),usctheses-c3-554193 (legacy record id)

Legacy Identifier etd-HoganChris-3345.pdf

Dmrecord 554193

Document Type Thesis

Format application/pdf (imt)

Rights Hogan, Christine Marie

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA