Engineering Complex Breast Tumor-Stroma Models: TMPyP4-Photodynamic Therapy Is More Effective at the Metastatic Site in Breast Tumors.

1
Background
Breast
cancer is the most widespread malignancy and the primary
contributor to cancer-related deaths in women across the globe. It is a heterogeneous disease that can be attributed
to the underlying genetics and the tumor microenvironment (TME). The TME is a dynamic environment that changes
over time as disease progresses and comprises cellular components
such as tumor cells and nontumor cells including fibroblasts, immune
cells, endothelial cells, adipocytes, and mesenchymal stromal cells
(MSCs). The TME also contributes a unique biophysical and biochemical
microenvironment which includes the oxygenation state, nutrient state,
pH gradients, signaling molecules, and extracellular matrix (ECM)
components. Stromal cells significantly influence the behavior of
cancer cells through direct and indirect means. Stromal cells can
remodel the tissue microenvironment directly through the deposition
of ECM proteins and matrix-degrading enzymes, but stromal cells also
secrete chemokines, cytokines, angiogenic and growth factors. Interactions between stroma and tumor cells, alongside
genetic aberrations in tumor cells, determine the growth patterns,
phenotype, and metastatic potential of the tumor. Breast cancer subtypes, including luminal, HER2-enriched,
Claudin-Low (CL), and basal-like/triple-negative breast cancer (TNBC),
are defined by gene expression of estrogen and progesterone receptor
(ER, PR), human epidermal growth factor 2 (HER2), and a cluster of
basal genes. The clinical course and treatment
response differ significantly among these subtypes. For instance, Luminal A tumors generally have the most
favorable prognosis, whereas TNBC (ER
‑
, PR
‑
,
and HER2
‑
) is often linked to lower
survival rates. CL tumors exhibit diminished
expression of cell-cell adhesion genes, including claudins 3, 4, and
7, occludin, and E-cadherin, as well
as stem cell-like characteristics.

While the existing therapeutic modalities for breast cancer have
successfully lowered mortality rates, they still face significant
challenges, including systemic side effects due to a lack of tumor-specific
accumulation, and the growing problem of drug resistance which complicates
treatment outcomes.
,
Drug resistance can be driven
by the interplay of several factors, among which are TME components.
They play a role in educating cancer cells to escape drug effects
through extensive molecular crosstalk. For instance, cancer-associated
fibroblasts (CAFs) enhance the aggressiveness and survival of cancer
cells through the secretion of growth factors and cytokines, as well
as the establishment of a “protective niche” that shields
against drugs. Likewise, immune cells contribute to immunosuppression
and metastasis. Adipocytes and MSCs participate
in the secretion of factors related to matrix remodeling and tumor
survival. Thus, the TME is now considered
to be a novel prognostic and predictive biomarker for breast cancer.
Metastases may also display heterogeneous responses to treatment and
common sites for metastasis for breast cancer are the brain, bone,
and lung.
There is a need for innovative and effective treatment
modalities
to enhance patient outcomes with one notable alternative being photodynamic
therapy (PDT), which has emerged as a minimally invasive therapy for
the treatment of diverse cancer types and nonmalignant lesions. This therapy requires molecular oxygen and a
photosensitizer that is activated by specific wavelengths of light,
including blue, red, or near-infrared (NIR) light, which triggers
the formation of reactive oxygen species (ROS). PDT has received approval
for the treatment of tumors, including prostate and esophagus and provides significant benefits for patients
by reducing the necessity for major surgical interventions. PDT carries
a low risk of both local and systemic morbidity and is characterized
by its tumor selectivity, ease of use, cost-effectiveness, and suitability
for repeated applications. A few clinical trials have assessed the
efficacy of PDT in breast cancer. The first phase I/lla clinical trial
study on PDT treatment of primary breast cancer was conducted by Banerjee
et al. (2020) and evaluated verteporfin (BPD)-PDT in a cohort of 12
patients. MRI data correlated with histopathology findings, which
revealed that tumor necrosis was augmented with incremental increases
in light dose. Other studies showed that
PDT treatment using Photofrin in breast cancer patients with chest
wall invasion/recurrences resulted in ∼65% of patients achieving
complete responses, while PDT treatment
using m-THPC led to complete responses in all patients.

3D models better replicate the physiological
3D tissue architecture
of the TME reflected in biomimetic cell morphological characteristics,
cell proliferation, tumor heterogeneity, and drug response compared
to monolayer 2D culture. Herein, we utilized
breast tumoroids using dense collagen l as a scaffold for 3D in vitro modeling of breast cancer. Spatially compartmentalized
dense 3D collagen l gels were engineered to mimic the TME by embedding
a tumor mass within a stromal compartment. Tumoroids offer a structural
basis for constructing matrix-relevant models, enabling meticulous
regulation of the density and composition of the ECM alongside a distinctly
defined boundary between tumor and stroma. These models include physical,
chemical, and cellular barriers closely replicating the in
vivo interactions between cancer cells and stromal components. The stiffness of these scaffolds falls within
the range of solid breast tumors, and tumoroids are compatible with
monitoring in real-time the formation of hypoxia gradient. The present
study aims to investigate the effect of PDT and combined PDT/Chemotherapy
in a biomimetic 3D breast tumor model using advanced real-time noninvasive
assessment modes.

2
Methods
2.1
Cell Lines
MDA-MB231 and MCF-7 human
breast adenocarcinoma cells were obtained from the European Collection
of Authenticated Cell Cultures (ECACC). MCF-7 cells represent luminal
A subtype, and MDA-MB-231 cells represent CL tumors. They were cultured
using Dulbecco’s Modified Eagle’s Medium/Nutrient Mixture
F12 Ham (DMEM/F12) growth medium (Sigma-Aldrich, Dorset, U.K.). Adipose
tissue-derived mesenchymal stem cells (MSCs) were acquired from ATCC
(American Type Culture Collection, Virginia, United States) and human
normal lung fibroblast (HNLF) were purchased from Lonza, Bioscience.
Both cell types were cultured in DMEM supplemented with 1g/L d-glucose (Gibco through Thermo Fisher Scientific, Loughborough, U.K.),
10% FBS (Gibco through Thermo Fisher Scientific, Loughborough, U.K.),
and 1% penicillin (5000 units/ml)/streptomycin (5000 μg/ml)
(Gibco through Thermo Fisher Scientific, Loughborough, U.K.). Human
umbilical vein endothelial cells (HUVECs) were acquired from Promocell,
Heidelberg, Germany, and were cultured in endothelial growth media
supplemented with 10% FBS and 1% penicillin/streptomycin obtained
from Promocell, Heidelberg, Germany. All cells were cultured under
5% carbon dioxide (CO2) and 95% atmospheric air at 37 °C, with
routine passaging carried out in 2D monolayers. Passage numbers for
cells were: HUVECs < 6 passage number, HNLF < 10 passage number,
MSCs < 20 passage number, MCF-7 and MDA-MB-231 cells < 50 passage
number.

2.2
Engineering of 3D Breast Tumor Model
3D tumoroids were constructed using the RAFT protocol (Lonza, Basel,
Switzerland), which uses the plastic compression method (Figure
) to generate physiologically
relevant collagen densities.

2.2.1
Simple Tumoroids (Tumor Mass)
Tumor
mass (TM) of either MDA-MB-231 or MCF-7 at 2 × 104 seeding density was formed by mixing 10xMEM (Gibco through Thermo
Fisher Scientific, Loughborough, U.K.), type I collagen 2 mg/ml (First
Link, Birmingham, U.K.), 10 μg/ml laminin (Sigma-Aldrich, Dorset,
U.K.), neutralizing solution (17% 10 M NaOH (Sigma-Aldrich, Dorset,
U.K.) in 1 M HEPES buffer (Gibco through Thermo Fisher Scientific,
Loughborough, U.K.) and cells suspended in culture medium. A volume
of 240 μL of cell-containing gel mixture was added in 96-well
plates (Corning through Thermo Fisher Scientific, Loughborough, U.K.)
and incubated at 37 °C for 15 min. Following polymerization,
absorbers were placed on top of each gel for 15 min. Following compression,
100 μL of cell culture media was added to each TM. The seeding
density was determined by culturing different cell densities within
the range of previously published cultures for colorectal cancer (5
× 104 cells per tumor mass) (Pape et al. 2020). However, following assessment and monitoring
of invasion using the breast cancer cell lines, equivalent invasion
was measured using 2 × 104 cells, and this was deemed
a more sustainable option.

2.2.2
Compartmentalized Tumoroids
The
advanced structure of complex tumoroids encompasses a stromal compartment
that contains an embedded tumor mass. TM was manufactured as previously
described. To model the tumor stroma, a collagen solution was prepared
by mixing 10xMEM, type I collagen 2 mg/ml, neutralizing solution (17%
10 M NaOH in 1 M HEPES buffer) with 20 μg/ml of laminin and
2 mg/ml of fibrin (Sigma-Aldrich, Dorset, U.K.) and stromal cell populations.
650 μL of the collagen mix is placed in a 24-well plate (Corning
through Thermo Fisher Scientific, Loughborough, U.K.) and first polymerized
at 37̊C for 15 min. Then, the TMs were placed on top of the
set gels followed by adding a further 650 μL of the collagen
mix, and incubated for further 15 min at 37 °C. Once fully polymerized,
the tumoroids were plastic-compressed using the 24-well RAFT absorbers
at room temperature for 15 min, as shown in Figure
. Finally, 1 mL of cell culture media was
added to each gel. Tumoroids were cultured at 37 °C under 5%
CO2 and 95% atmospheric air for 21 days, with 50% of cell culture
media replaced every 48 h.
2.2.2.1
Engineering a Primary
Stroma Compartment
The stromal compartment is composed of
1 × 104 MSCs
and 1 × 105 HUVECs in addition to 20 μg/mL laminin
and 2 mg/ml fibrin.

2.2.2.2
Engineering a Metastatic
Stroma Compartment
The stromal compartment is composed of
1 × 104 HNLF
and 1 × 105 HUVECs in addition to 20 μg/mL laminin
and 2 mg/ml fibrin.
Initially, a 1:1 ratio of breast cancer
cell lines to stromal cells was attempted because this is a very common
ratio utilized by many in vitro 3D coculture studies.
However, on closer inspection, we observed that the presence of stromal
cells with this ratio caused the tumoroids to contract, which made
it difficult to conduct specific assessments such as fluorescence
imaging. We then tried a 2:1 ratio of breast cancer cell lines to
stromal cells (2 × 104 of breast cancer cells and
1 × 104 of stromal cells), which was considered a
more convenient option as it maintained both the tumoroids structure
stability, cancer invasion, and the development of vascular networks.

2.2.3
Characterization of Stromal Compartments
The primary stromal compartment was comprised of adipose-derived
MSCs, and the metastatic stromal compartment was comprised of human
normal lung fibroblast (HNLF). Both are cell lines and were acquired
from ATCC and Lonza, Bioscience, respectively.
As such, the
characterization conducted for these cells focused on how cells behaved in vitro. Characterization included observing and quantifying
the invasion of both tumor cell lines into compartments containing
either MSCs or lung fibroblasts, and showing that invasion is greater
in the presence of the lung fibroblast stroma (Supporting 1). This is expected as both tumor cell lines are
derived from lung pleural effusions. Furthermore, characterization
work demonstrated that MSCs support the aggregation of ECs to fuse
and form networks, which is not the case with the lung fibroblasts
(Supporting 1). Characterization of the
MSCs using flow cytometry analysis to check the “stemness”
of the adipose-derived MSCs found positive expression of cell surface
markers for CD90, CD73, and CD105 (excess of 93% for all groups) and
negative markers for CD19, CD34, CD11b, CD45, and HLA DR (Ho, 2024).

2.3
Cell Viability Assay
The cell viability
of 2D monolayer cultures and 3D constructs was assessed by measuring
ATP levels using CellTiter-Glo 3D Viability Assay (Promega, Southampton,
U.K.). At set time points following drug treatment, samples were mixed
with CellTiter Glo Reagent with a (1:1) ratio and then subjected to
vigorous shaking for 5 min using an orbital shaker followed by 25
min of benchtop incubation at room temperature. 100 μL of the
mix was then transferred to a black 96-well plate (Thermo Fisher Scientific,
U.K.) in triplicate, and luminescence was measured by the Tecan Infinite
Lumi plate reader (Männedorf, Switzerland).

2.4
Live-Dead Staining for Fluorescence Imaging
Following
drug treatment, the 3D constructs were stained for viability
imaging using the Live-dead viability kit (Molecular Probes, Thermo
Fisher Scientific, U.K.). Cell culture media was removed, and then
3D constructs were incubated with the Live-dead solution composed
of 4 μM Calcein-AM and 2 μM Ethidium homodimer-1 for 40
min at room temperature before imaging with Zeiss AxioObserver with
ApoTome.2 instrument and software (Zeiss, Oberkochen, Germany) using
green filter (%Ex/Em 495/515 nm) and red filter, respectively (%Ex/Em
495/635 nm).

2.5
Immunofluorescence Imaging
The 3D
constructs were fixed in 4% formalin (Sigma-Aldrich, Darmstadt, Germany)
for 30 min followed by rinsing and storage in phosphate-buffered saline
(PBS) (Fisher Scientific, Loughborough, U.K.). Then, the constructs
were permeabilized and blocked for 1 h at ambient temperature utilizing
a solution composed of 0.3% Triton X-100 and 1% bovine serum albumin
(both sourced from Sigma-Aldrich, Dorset, U.K.), dissolved in PBS.
Primary antibodies: antipancytokeratin (Genetex, Inc., US) at a dilution
of 1:100 and antiplatelet cell adhesion molecule (CD31) mouse JC70/A
(Abcam, Cambridge, U.K.) at a dilution of 1:200, were subsequently
incubated overnight at 4 °C. Following this incubation, primary
antibodies were removed through 3 sequential washes with PBS on a
plate shaker. The subsequent day, constructs were exposed to secondary
antibodies: 1:1000 antimouse Alexa Fluor 488 IgG H&L (ab150113)
and 1:500 antirabbit DyLight 594 (ab96885), both obtained from Abcam
(Cambridge, U.K.) for 2.5 h at room temperature. This was followed
by 3 consecutive washes in PBS and counterstaining with DAPI, using
NucBlue (Invitrogen through Thermo Fisher Scientific, Loughborough,
U.K.) 20 min before imaging.

2.6
Assessment of Oxygen Gradient
Level in 3D
Constructs
Spatio-temporal oxygen gradient levels within
the 3D constructs were examined using the minimally invasive oxygen
mapping technique provided by a VisiSens TD MIC system, which relies
on collecting luminescence signals from an oxygen-sensitive luminescent
foil placed in contact with the construct. The spatially resolved
luminescence signal is captured by quantitative 2D imaging. The 3D
constructs were placed on top of the oxygen sensor foils glued to
the bottom of a 24-well imaging sensor plate ISP24-RPSu4 (PreSens
Precision Sensing GmbH, Regensburg, Germany). 1 mL of cell culture
media was added to each well. Then, the 24-well plate was incubated
at 37 °C, and a 1 h time series of measurements was set for recording
signals over a period of 24 h. For calibration, sodium sulfite 1%
(wt/V) solution was used for the 0% oxygen baseline, and air-equilibrated
distilled water was used for the 100% calibration marker, where 100%
corresponds to air oxygen saturation. VisiSens ScientifiCal Software
Version VS 1.0.1.5 Plugins was used to analyze the data by selecting
the tumor mass as a region of interest, then a graph of the analyte
versus time was plotted. Upon reaching a steady state level, an average
oxygen percentage was calculated and analyzed via the GraphPad Prism
9 software.

2.7
Evaluation of Endothelial
Networks Complexity
Zeiss AxioObserver equipped with the
ApoTome.2 instrument and software
(Zeiss, Oberkochen, Germany) was used for imaging 3D constructs. Four
images were captured at 2.5× magnification, positioned at the
12, 3, 6, and 9 o’clock orientations on the same focal plane,
and subsequently uploaded to Fiji (NIH open-source software). To characterize
the complexity of the endothelial networks present within the stromal
compartment, images were analyzed for the number of loops, junctions,
branches, and total branch lengths with statistical significance determined
using GraphPad Prism 9 software.

2.8
Experiments
in 2D
2.8.1
Drug Uptake in 2D
Cellular uptake
was assessed by using fluorescence microscopy. MCF-7 and MDA-MB-231
breast cancer cell lines were cultured at a density of 4 × 103 per well on a coverslip within a Petri dish (Fluorodish,
World Precision Instruments, Hitchin, U.K.). Following a 48 h incubation
period, a 1 μM solution of TMPyP4 (Sigma-Aldrich, Dorset, U.K.)
was incubated for another 24 h. After that, TMPyP4 was removed, and
the cells were washed with PBS (Gibco, U.K.). The cells were then
subjected to a 1 h incubation with 100 nM of Lysotracker Green (Invitrogen,
U.K.) and 100 nM of BioTracker 405 Blue Mitochondria Dye (EMD Millipore,
Merck, U.K.) prior to imaging. Fluorescence images were subsequently
obtained by using a Leica SP8 confocal microscope with a 25×
objective. The excitation wavelength used for detecting TMPyP4 was
405 nm, with emission between 650 and 700 nm. For imaging the mitochondria,
the excitation wavelength was set to 405 nm, with emission captured
between 420 and 450 nm. Lysosomes were imaged by using an excitation
wavelength of 488 nm and an emission range of 510–540 nm.

2.8.2
Drug Efficacy in 2D and PDT Treatment
For PDT, a water-soluble porphyrin photosensitizer was employed,
TMPyP4 (meso-Tetra­(N-methyl-4-pyridyl) porphine tetra
tosylate), where the tosylate serves solely as the counterion (Sigma-Aldrich,
Dorset, U.K.). To evaluate the cytotoxic effects, MCF-7, MDA-MB-231,
HUVECs, MSCs, and HNLF were cultured in a 96-well plate at a density
of 4 × 103 cells per well for 48 h. This was followed
by a further 48 h incubation with increasing concentrations of TMPyP4,
ranging from 0.1 to 4 μM. The cells were washed twice with PBS,
and fresh phenol-free cell culture media was subsequently added. For
the illumination of cells treated with TMPyP4, a filtered xenon-arc
lamp was utilized, featuring a light guide to filter out ultraviolet
and infrared emissions (Paterson Lamp, Phototherapeutics Ltd., U.K.).
This lamp emitted light within a wavelength range of 420 ± 20
nm at a fluence rate of 6 mW/cm2. The irradiation lasted
for 5 min, resulting in a cumulative light dose of 1.8 J/cm2 (calculated as total light dose (J/cm2) = fluence rate
(W/cm2) × treatment time (s)). Furthermore, experiments were conducted in which samples
were incubated with TMPyP4 in the absence of light (dark conditions)
to compare the cytotoxic effects of both treatments. After 24 h, cell
viability was evaluated using the CellTiter-Glo 3D Viability Assay.

2.9
Experiments in 3D
2.9.1
Drug
Uptake in 3D
In order to assess
the uptake of TMPyP4 in 3D tumoroids of MCF-7 and MDA-MB-231, 5 μM
TMPyP4 was added on the 18th day after seeding, for a period of 48
h. Following this incubation period, fluorescence imaging was conducted
with a confocal Leica SP8 inverted microscope, using a 10× objective
and an excitation wavelength of 405 nm, with an emission range set
between 650 and 700 nm.

2.9.2
Drug Efficacy in 3D
Compartmentalized
3D tumoroids of MCF-7 and MDA-MB-231 were engineered, incorporating
two distinct stromal compartments that simulate primary breast cancer
and metastatic breast cancer in the lungs. On the 18th day postseeding,
the 3D compartmentalized tumoroids were divided into 4 groups as presented
below.a.Control:
No treatment (just cell culture
media).

b.Photosensitizer-treated:
received 5
μM of TMPyP4.

c.Doxorubicin-treated: received 10 μM
of doxorubicin.

d.Photosensitizer
+ Doxorubicin-treated:
5 μM of TMPyP4 + 10 μM of doxorubicin

Following a 48 h incubation period, the drugs were washed
twice with PBS, after which fresh phenol-free cell culture media was
added. The irradiation protocol described earlier for the 2D culture
experiments (light dose of 1.8 J/cm2) was also applied
to the 3D complex tumoroids. Additionally, experiments were carried
out in which samples were incubated with drugs without undergoing
irradiation (dark condition). The samples were then allowed to incubate
for an additional 24 h postirradiation or no irradiation before the
endpoint experiments were conducted. Therapeutic efficacy was assessed
via CellTiter Glo 3D Viability Assay, live-dead fluorescence imaging,
vascular network complexity, and real-time monitoring of oxygen gradient
levels, which are previously described.
To evaluate the potential
synergistic effects of doxorubicin and
TMPyP4, we employed the following equation to compute the α
(α) value

In this equation, FPDT and Fcytotoxin in
the numerator represent the fractional cell viability obtained from
each individual treatment. The denominator represents the fractional
cell viability resulting from combination treatment. An α value
greater than 1 implies a synergistic effect, a value of 1 corresponds
to an additive effect, and a value less than 1 indicates an antagonistic
effect.

2.10
Statistical
Analysis
Analysis of
the data was carried out using GraphPad Prism 9 software (GraphPad,
San Diego, CA). The data were initially tested for normality using
the Shapiro-Wilk test. It was determined that two-way ANOVA was justified
because multiple parameters were compared. The two key independent categorical variables, which were simultaneously
in play, were the different treatment groups and the light/dark conditions.
These combination variables were then used to study their effect on
3 separate outcome variables when subjected to the drug intervention.
These included invasion, cell death, and subsequent hypoxia gradient
formation. A Two-way ANOVA test of significance was applied to determine
if there was an interaction effect between the two factors. A significance threshold of p < 0.05 was used. Data points are presented as means with standard
error of the mean (SEM). All experiments were conducted with a minimum
of 4 repeats.

3
Results
3.1
Uptake
of Photosensitizer in 2D and 3D
MCF-7 cells take up TMPyP4
more readily than MDA-MB-231 cells in
2D demonstrated by increased red fluorescence inside cells (Figure
). Photosensitizers
are known to accumulate in different organelles, significantly influencing
the activation of specific signaling pathways during PDT. To investigate
the subcellular distribution of TMPyP4, fluorescent probes LysoTracker
(green) and Mitotracker (blue) were used to label the lysosomes and
mitochondria. TMPyP4 was present in both lysosomes and mitochondria,
though it appeared to cluster more in the mitochondria (Figure
). Fluorescence imaging of
3D tumoroids incubated with 5 μM of TMPyP4 confirmed that MCF-7
cells in 3D tumoroids demonstrated a greater uptake of TMPyP4 compared
to MDA-MB-231 (Figure
).

3.2
Photodynamic Therapy Efficacy in 2D
Breast
cancer cell lines MCF-7 and MDA-MB-231 exhibited a significant
concentration-dependent cytotoxicity when exposed to light-activated
TMPyP4, compared to the control and nonactivated TMPyP4. The calculated
IC50 for light-activated TMPyP4 was determined to be 0.12 μM
for MCF-7 cells and 0.44 μM for MDA-MB-231 cells. Furthermore,
light-activated TMPyP4 also induced significant cytotoxicity in mesenchymal
stem cells, human lung fibroblasts, and endothelial cells compared
to the control and nonactivated TMPyP4. Human lung fibroblasts demonstrated
the greatest resistance to light-activated TMPyP4 (IC50 = 0.84 μM)
compared to mesenchymal stem cells (IC50 = 0.13 μM) and endothelial
cells (IC50 = 0.14 μM) (Figure
).

3.3
Photodynamic Therapy Efficacy in 3D
3.3.1
Breast Cancer Surrounded by a Primary Stromal
Compartment
Light-activated TMPyP4 exhibits a diminished
capacity to induce cell death in 3D tumoroids of breast cancer compared
to 2D. In MCF-7 3D tumoroids, light-activated TMPyP4 exhibited a statistically
significant mean reduction (64.7%) in viable cancer cells within the
TM, which corresponds to a 1.5-fold reduction, compared to the controls.
The combination of doxorubicin with light-activated TMPyP4 produced
a significant mean decline (36.8%) in viability of cancer cells within
the TM relative to the control (mean = 98.81%) (P-value < 0.0001), light-activated TMPyP4 (mean = 64.7%), and the
combination of doxorubicin with nonactivated TMPyP4 (mean = 73.2%)
(Figure
(2)).
In tumoroids where the stroma is representative of the primary
stromal tissue site (breast tissue), MCF-7 cells within the TM exhibited
a more pronounced reduction in the percentage of living cells compared
to the stromal compartment when treated with light-activated TMPyP4,
either alone or in combination with doxorubicin (Figure
(2 and 4)). Light activation
of TMPyP4 produced a notably greater cytotoxicity (P-value = 0.01) than the nonactivated TMPyP4, corresponding to the
live/dead assay. Light-activated TMPyP4 alone reduced mean cell viability
to 66.4%, and when combined with doxorubicin, mean cell viability
further dropped to 53.6%. Using the previously mentioned equation,
the calculated α value for the combined doxorubicin and light-activated
TMPyP4 was found to be 0.8. Moreover, light-activated TMPyP4 increased
oxygenation levels (41.1%), though not significantly, over the untreated
(9.7%) and nonactivated groups (17.7%). However, the combination of
doxorubicin and light-activated TMPyP4 significantly increased oxygen
saturation to 73.1% (P < 0.0001), indicating alleviation
of tumor hypoxia. Findings of the CellTiter Glo assay corresponded
well with real-time oxygen saturation reading (Figure
(5–7)) and demonstrated an inverse
dependence, i.e., where cell viability was high, oxygen levels were
low, and following drug action that reduced cell viability, oxygen
levels increased, and hypoxia gradients were diminished. MCF-7 tumoroids
support the formation of endothelial/vascular network by HUVECS in
3D (Figure
(8)). Significant
decreases in vascular networks in the stroma were observed following
treatment with light-activated TMPyP4 (3-fold decrease), doxorubicin
(9-fold decrease), and doxorubicin with light-activated TMPyP4 combination
(40-fold decrease) (Figure
(9)).
In MDA-MB-231 3D tumoroids, monotherapy with either
light-activated
TMPyP4 or doxorubicin alone elicited limited efficacy (Figure
(2)). However, the combination
of doxorubicin and light-activated TMPyP4 demonstrated a significant
mean decline in the percentage of live cancer cells in the TM (P-value = 0.02), from 99% to 71.06%. In the stromal compartment
of the MDA-MB-231 3D tumoroids, combining doxorubicin and light-activated
TMPyP4 resulted in a significant mean reduction (59.1 %) of viable
cells compared to (97.3 %) in the untreated group and (101.4 %) in
the light-activated TMPyP4 group, as shown in Figure
(4).
The CellTiter Glo assay
demonstrated that the only significant
mean decline (60.6%) in cell viability among the different treatment
regimens was the combination of doxorubicin and light-activated TMPyP4.
MDA-MB-231 3D tumoroids treated with light-activated TMPyP4 or doxorubicin
exhibited a nonsignificant reduction in cell viability (Figure
(7)). The calculated value
of α (α = 1.002) indicated that the combination of doxorubicin
and light-activated TMPyP4 only produced an additive effect. This
contrasts with the MCF-7 calculated value (α = 0.8), reinforcing
cell line-specific differences in therapeutic interaction. Real-time
monitoring of oxygen levels in MDA-MB-231 3D tumoroids revealed that
light-activated TMPyP4 led to a slight improvement in oxygenation
(12.9%) compared to the control group (3.2%). It was also found that
doxorubicin, in conjunction with light-activated TMPyP4, significantly
increased oxygenation levels from 3.2% to 73.8% (P-value < 0.0001) (Figure
(6)). It should be noted that MDA-MB-231 3D tumoroids did
not support the formation of vascular networks, thus rendering vascular
network analysis redundant.

3.3.2
Breast
Cancer Surrounded by a Metastatic
Stromal Compartment
In MCF-7 3D tumoroids, which had a metastatic
stromal compartment, live-dead imaging revealed that treatment with
light-activated TMPyP4 led to a substantial mean reduction (22.4%)
in the proportion of viable cancer cells in the TM corresponding to
∼5-fold reduction compared to the control group and the group
treated with nonactivated TMPyP4 (89.5%), as shown in (Figure
(2)). Although the combination
of doxorubicin and light-activated TMPyP4 also reduced the percentage
of live cancer cells in the TM to 20.7%, no differences were observed
between the individual and combination treatment groups. HLFs within
the stromal compartment of MCF-7 3D tumoroids were impacted by the
application of light-activated TMPyP4 with a mean reduction (9.2%)
in viability compared to (100%) in the control group and (87.9 %)
in the nonactivated TMPyP4 group. Treatment with doxorubicin resulted
in a notable mean decrease (41.4%) in live cells, and the combination
of doxorubicin and light-activated TMPyP4 led to a further mean reduction
(15%) relative to the control group. It is noteworthy that the group
treated solely with light-activated TMPyP4 showed a lower percentage
of live cells compared to those treated with either doxorubicin or
the combined treatment (Figure
(2 and 4)).
The CellTiter Glo assay correlated
with the live-dead imaging results,
with the exception of the control group exposed to light, which had
reduced viability (Figure
(7)) and all treatments resulted in significantly reduced
viabilities, most notably for PDT treatment at 9.2 %. The calculated
α value indicated that there was no synergistic effect observed
between the combinations of doxorubicin and light-activated TMPyP4.
Interestingly, real-time oxygen monitoring of MCF-7 3D tumoroids revealed
that the baseline oxygen levels in the untreated control group remained
relatively stable, unlike those in the MCF-7 3D tumoroids, containing
the primary stromal compartment. This denotes that oxygen consumption
and metabolic activity in MCF-7 3D tumoroids with primary stroma are
higher than those containing the metastatic stroma, which could be
attributed to differences in the stromal cells. However, the groups
treated solely with doxorubicin and light-activated TMPyP4 demonstrated
a statistically significant enhancement in oxygenation levels, with
means of 91.05 and 98.9%, respectively, in comparison to the control
group (Figure
(6)).
In the 3D tumoroids of MDA-MB-231, light-activated TMPyP4 resulted
in a statistically significant mean decrease (59.5%) in live cancer
cells within the TM which corresponds to a 1.7-fold reduction compared
to (99.4%) in the control, as shown in (Figure
(2)). Additionally, doxorubicin alone resulted
in a mean survival rate of 62.5%, while the combination of doxorubicin
with light-activated TMPyP4 reduced live cell proportions to 55.8%.
Nevertheless, no significant differences were found between the groups
treated with the combination of doxorubicin and light-activated TMPyP4
and those treated with doxorubicin alone or light-activated TMPyP4
alone. In the stromal compartment of MDA-MB-231 3D tumoroids, the
group treated with light-activated TMPyP4 demonstrated a significant
mean reduction (57.4%) in viable cells, compared to (100%) in the
control group and (99.1%) in the nonactivated TMPyP4 group. Furthermore,
both the doxorubicin-treated group and the group receiving a combination
of doxorubicin and light-activated TMPyP4 showed significant mean
reductions in live cells, with mean values of 62 and 54.2%, respectively,
relative to 99.4% in the control group (Figure
(2 and 4)).
According to the CellTiter
Glo assay, light-activated TMPyP4 treatment
induced a significant mean decline of 69.5% in cell viability, as
shown in Figure
(7),
which corresponds to a 1.5-fold reduction compared to the control
group (99.9%). The mean viabilities observed for the primary vs metastatic
stroma tumoroids were comparable at 76.1 and 69.5% respectively, corresponding
to a 1.1-fold reduction in viability for the metastatic stroma. In
contrast to the MCF-7 studies, a markedly larger difference was observed
for primary (66.4%) vs metastatic stroma (9.2%), corresponding to
a 7.2-fold reduction in viability. Treatment of the MDA-MB-231 tumoroids
with doxorubicin, as well as the combination of doxorubicin and light-activated
TMPyP4 resulted in markedly lower percentages of viable cells, averaging
43.8 and 44.9%, respectively. The calculated α value showed
that there was no synergistic effect evident on combining doxorubicin
and light-activated TMPyP4 (Figure
(7)).
Real-time oxygen monitoring of MDA-MB-231
tumoroids, revealed a
statistically significant enhancement in oxygenation levels within
the light-activated TMPyP4-treated group, reaching an average of 30.3%
compared to the control group (7.3%). Furthermore, the combination
of doxorubicin and light-activated TMPyP4 demonstrated a remarkable
improvement (P-value < 0.0001) in oxygenation
levels, reaching a mean of 71.1% air oxygen saturation, although this
was lower than the doxorubicin-only group, which reached 82.2% air
oxygen saturation (Figure
(6)). Both MCF-7 and MDA-MB-231 3D tumoroids with metastatic
stroma showed the absence of vascular networks formation.

4
Discussion
This study explored the therapeutic
efficacy of PDT both as an
independent treatment modality and in combination with the chemotherapeutic
agent doxorubicin utilizing an advanced multicompartment 3D tumoroid
model of breast cancer. The tumoroid model comprises a central TM
which is surrounded by a stromal compartment incorporating a dense
matrix prepared via plastic compression. A comparative study was carried
out using two human breast carcinoma cell lines with either a primary
breast stromal compartment or a metastatic lung stroma compartment.
A previous study also described the development of multicompartment
breast tumor organoids featuring two main ECM compartments: the basement
membrane composed of Matrigel and the stromal matrix composed of collagen
type I. Additionally, researchers incorporated
primary monocytes isolated from PBMCs or immortalized macrophage U87
cell lines into the stroma. However, our tumoroids developed herein
incorporated various stromal cells including adipose tissue-derived
MSCs and HNLF in addition to endothelial cells, which are critical
components of the TME. Also, a variety of extracellular matrix proteins
such as laminin and fibrin in addition to collagen type I were utilized
in our model. Moreover, the plastic compression method employed herein
increases the stiffness of the ECM, which improves the biomimicry
of our model to physiological in vivo tissue stiffness. The photosensitizer
employed in this study is TMPyP4, which is a cationic porphyrin characterized
by a porphyrin ring that incorporates four N-methyl-4-pyridyl
substituents. TMPyP4 displays multiple absorption peaks, with its
most significant peak occurring at 424 nm in the blue region of the
visible spectrum, which corresponds to the Soret band. Its fluorescence emission is characterized by
two prominent peaks within the red spectral region, at 649–655
nm and 715–720 nm.
,
TMPyP4 was selected
in this study due to its known efficacy as a photosensitizer in PDT.
Its advantageous properties include water solubility, effective permeability
across cellular membranes, and a notable quantum yield of singlet
oxygen (1O2) (Φ = 0.77 ± 0.04).
,
Furthermore, TMPyP4 has shown the ability to intercalate into and
stabilize G-quadruplex structures within DNA, leading to the inhibition of telomerase activity.

PDT exerts its primary anticancer effects through
direct destruction
of tumor cells via apoptosis, necrosis, or autophagy, provocation of an inflammatory response recruitment of
leukocytes to the affected area, and activation of antitumor T-lymphocytes. Furthermore, PDT treatment may compromise the
tumor microvasculature, resulting in diminished oxygen and nutrient
supply that is vital for tumor viability. ROS inflict damage on vascular ECs, triggering clotting processes,
promoting platelet aggregation, and leading to vascular obstruction.
This vascular occlusion results in sustained hypoxia within tumor
tissue, ultimately culminating in cell death. However, it is important to note that this effect may also intensify
tumor hypoxia, potentially leading to the development of resistance
to PDT.
Studies were carried out initially in standard 2D culture
and simple
tumoroid models without a stromal component. MCF-7 cancer cells exhibit
a greater uptake of TMPyP4 in both 2D and 3D cultures when compared
to MDA-MB-231 cells (Figures
and ). Electrostatic attraction between
positively charged TMPyP4 and negative cell membrane likely drives
its cellular uptake. However, our results contradict the expectation
that TMPyP4 uptake would be greater in MDA-MB-231 cells, which are
reported to be more negatively charged than MCF-7 cells.
,
Evidence suggests that clathrin-mediated endocytosis is particularly
efficient for positively charged (cationic) molecules due to their
strong interaction with negatively charged cell membranes.
−

Given that MCF-7 cancer cells predominantly rely on the clathrin-mediated
endocytosis mechanism,
−

this may explain the observed higher uptake of TMPyP4
in MCF-7 cells in this study.
TMPyP4 at 1 μM incubated
for 24 h predominantly localizes
within the mitochondria with a smaller presence in lysosomes in both
MCF-7 and MDA-MB-231 cells in 2D (Figure
), which may be attributed to the elevated
transmembrane potential of the inner mitochondrial membrane, that
facilitates the uptake of cationic molecules into the mitochondrial
matrix. One study has shown nuclear localization
of 10 μM TMPyP4 in MCF-7 cells after 3 h of incubation, while another study reported cytoplasmic localization
of 1 μM TMPyP4 in MCF-7 cells following 12 h of incubation which concurs with our findings. Additionally,
one study has shown that prior to illumination, TMPyP4 was primarily
confined to the lysosomes of colon carcinoma cells; however, following
irradiation, the photosensitizer was relocated to the nucleus and
nucleoli.

2D cultures of MCF-7
and MDA-MB-231 cancer cells treated with TMPyP4
PDT exhibited significant cytotoxicity in a concentration-dependent
manner (Figure
).
Notably, MCF-7 cells demonstrated greater sensitivity to the cytotoxic
effects of TMPyP4 PDT than MDA-MB-231. Several studies have documented
the anticancer effects of TMPyP4 through telomerase inhibition in
breast cancer cells, particularly at concentrations of 5 μM
or higher.
,,
Additionally, other research investigated its effectiveness as a
photosensitizer for PDT in breast cancer.
,
Interestingly, our findings also indicate that approximately six
times higher TMPyP4 concentration is required to kill 50% of the HNLF
lung fibroblasts compared to MSCs. In contrast, ECs were found to
be the most sensitive to the cytotoxic effects of light-activated
TMPyP4. Aligned with other studies,
,
data presented
here suggest that normal lung fibroblasts may be more resistant to
the effects of PDT.
PDT efficacy was investigated using 3D compartmentalized
tumoroids
incorporating both TM and a stromal compartment mimicking in vivo tumors. PDT studies using a variety of 3D models
have provided valuable insight into a range of therapeutic variables,
including the development of immune responses.
,
In order to replicate primary breast cancer effectively, a stromal
compartment containing adipose tissue-derived MSCs and ECs was engineered
alongside TM in dense collagen to mimic physiological collagen densities
and restrict oxygen diffusion. PDT and
doxorubicin are both oxygen-dependent treatments, and thus, the hypoxia
gradient mimicry within tumoroids is a necessary feature for physiological
relevance. 5 μM TMPyP4 concentration was selected based on preliminary
experiments performed with 3D simple tumoroids of just the TM (Supporting 2). These preliminary experiments
revealed a lower drug susceptibility for cells in 3D models compared
to 2D models, and therefore, a higher TMPyP4 concentration was needed,
as reported previously by Hadi et al. using a disulfonated porphyrin
photosensitizer. The concentration of
doxorubicin used herein was based on a study by Pradhan et al.

Monotherapy with light-activated TMPyP4
in MCF-7 3D compartmentalized
tumoroids surrounded by primary stromal compartment produced significant
cytotoxicity, especially within the TM (Figure
). This correlated to a moderate enhancement
of oxygen levels and a significant disruption of the vascular networks
(Figure
). The increase
in oxygenation levels following PDT treatment is interpreted as resulting
from the increased level of cytotoxicity, which leads to reduced oxygen
consumption within the construct owing to cellular respiration. PDT
can also induce oxygen consumption during treatment via photochemical
reactions, such as the reaction of singlet oxygen with cellular substrates.
In these experiments, however, the oxygen measurements were carried
out 48 h post-treatment; therefore, the measurements should primarily
reflect the changes in cell viability induced by the treatment.
Unlike the MCF-7 studies, treatment of MDA-MB-231 3D compartmentalized
tumoroids with light-activated TMPyP4 monotherapy proved less effective.
Combination treatment of MDA-MB-231 tumoroids with doxorubicin and
light-activated TMPyP4 did elicit stronger cytotoxicity, albeit merely
an additive effect, as manifested by a substantial decrease in cell
viability within both the TM and the stromal compartment (Figure
). In summary, for
models with a primary stroma compartment, the findings show that only
the MCF-7 3D tumoroids exhibited a significant response to the cytotoxic
effects of PDT monotherapy, eliciting a pronounced impact on the TM
(P-value < 0.0001) compared to the adjacent stromal
compartment (P-value = 0.01). In contrast, using
combination therapy, treatment of MDA-MB-231 3D tumoroids with PDT
and doxorubicin elicited a significant cell killing in the TM compartment
(P-value = 0.02). The reduced efficacy of light-activated
TMPyP4 treatment in MDA-MB-231 3D compartmentalized tumoroids could
be explained by the lower uptake of TMPyP4 noted in 3D tumoroids (Figure
). Hypoxic tumors
present a huge obstacle that compromises the effectiveness of PDT,
and MDA-MB-231 TMs exhibit significantly greater hypoxia relative
to MCF-7 TM (Figures
and ), which may render them more resistant
to PDT.
Several studies have shown that combining chemotherapeutic
agents
with PDT can enhance the breast cancer treatment efficacy. In MDA-MB-231
cells, treatment with 5-aminolevulinic acid PDT in conjunction with
cisplatin caused a marked reduction in cell viability compared to
either therapy alone. Co-administration
of TMPyP/phthalocyanine photosensitizers alongside doxorubicin facilitated
death in breast cancer cells in vivo. Similarly, light-activated chlorin-vitamin conjugates
and their respective indium complexes, combined with Taxol, exhibited
a pronounced synergistic cytotoxicity against TNBC cell lines. A different combination therapy with pheophorbide
A PDT and doxorubicin showed mainly additive effects in HeLa cells,
although in one case, a synergistic effect was observed. The aforementioned
study was carried out in 2D systems, but in our 3D system where doxorubicin
was administered before PDT was carried out, the effect of combination
treatment was additive rather than synergistic. While doxorubicin
can generate singlet oxygen upon light activation, this is largely
suppressed when doxorubicin is bound to DNA in the nucleus. Herein, we did not observe any significant differences
between the light and dark cytotoxicity levels with doxorubicin monotherapy.
The effect of combined PDT and doxorubicin treatment in MDA-MB-231
3D compartmentalized tumoroids observed herein may arise from complementary
mechanisms, as doxorubicin primarily targets nuclear DNA, whereas
TMPyP4 can photodamage lysosomes or mitochondria (Figure
), facilitating the activation
of separate pro-apoptotic signaling pathways.
−

Thus, PDT can
effectively bypass several key regulatory points associated with resistance
to chemotherapy. Additionally, light activation of TMPyP4 may induce
oxidative damage to the efflux pumps responsible for expelling doxorubicin
from the cells, increasing its intracellular concentration.
,
PDT effectively restores the sensitivity of adriamycin-resistant
breast cancer cells to adriamycin via inhibition of MDR1 gene expression. In the present study, however, PDT was applied
after administration of doxorubicin; therefore, PDT impairment of
efflux pumps may exert a relatively minor effect on overall efficacy.
The treatment efficacy of both PDT and doxorubicin in a 3D compartmentalized
tumoroid model of metastatic breast cancer, containing HNLFs and ECs,
was significantly enhanced (Figures
and ), in comparison to tumoroids
with a primary stromal component. However, no synergistic effects
were detected from the combination treatment. The lung fibroblasts
used in the metastatic model were relatively resistant to PDT when
cultured in 2D (Figure
). The superior efficacy of doxorubicin or light-activated TMPyP4
treatment observed in this modified model may result from ECM differences
between 3D compartmentalized tumoroids with metastatic and primary
breast stroma. Existing evidence indicates that ECM composition, density,
stiffness, and mechanical properties can vary significantly between
primary breast tumors and metastatic lesions in the lungs. Within
the TME, the ECM serves not only as structural support but also as
an active regulator of tumor progression. ECM components incorporated in our model included collagen type
I, laminin, and fibrin. Increased collagen cross-linking (via LOX
enzymes) and fiber alignment enhance mechanical tension, activating
signaling pathways that drive tumor cell proliferation, invasion,
and drug resistance.
,
Laminin and fibrin support angiogenesis
and provide a scaffold for migrating tumor cells.
−

Nevertheless,
it is important to consider that ECM remodeling (ECM deposition and
degradation) is a continuous dynamic process that occurs throughout
tumorigenesis. Another point to consider in our 3D compartmentalized
models is that ECM remodeling differs markedly between stroma containing
MSCs and HNLFs. A recent study investigated ECM remodeling in vivo within the premetastatic lungs of mice harboring
primary breast tumor xenografts. Lung fibroblasts demonstrated an
increase in the levels of collagen (Col) Col4A5, MMP2, MMP3, and MMP14,
associated with reduced levels of LOX, LOXL2, and prolyl 4-hydroxylase
α-1. Additionally evidence of elastin
degradation alongside an upregulation of MMP-2, MMP-4, and MMP-9 expression
within the pulmonary tissue of a murine model for breast cancer was
observed during the third and fourth weeks. Also, stromal cell function and activity can be crucially reprogrammed
by tumor cells such as conversion of fibroblasts into CAFs which show
altered expression of MMPs and LOX enzymes, reflecting ECM reorganization
−

or transformation of adipose tissue-derived stem cells into CAFs. A change in the ECM stiffness in 3D compartmentalized
tumoroids featuring metastatic stroma could potentially influence
drug penetration, tumor oxygenation, and the matrix-driven signaling
pathways in cancer cells. ECM stiffness critically modulates breast
cancer biology, driving cell proliferation, stemness, metastasis,
and invasion through key signaling pathways
such as YAP/TAZ, Wnt/β-catenin,
FAK, TWIST1, PI3K/AKT/mTORC1, and ROCK. Consequently, alterations in matrix stiffness
can profoundly influence these signaling cascades, thereby affecting
the behavior of breast cancer cells, the functions of stromal cells,
and responses to therapeutic interventions. Models similar to this
have shown that the tumor cell transcriptome is altered depending
on the interaction with different stromal cells, highlighting the
key role the TME has on tumor progression. Furthermore, stromal cells within the TME could distinctly affect
the response to therapeutic agents in cases of primary breast cancer
compared to metastatic breast cancer. Karimnia et al. demonstrated
that photodestruction of stromal fibroblasts in a 3D coculture with
PDAC cell lines enhanced tumor response to PDT. Additionally, under hypoxic conditions, typical of dense
3D matrices as demonstrated herein, hypoxia-inducible factors (HIF-1α
and HIF-2α) in cancer cells orchestrate broad genetic programs
that reprogram stromal cells via the expression of TGF-β, IL-1,
PDGF, and bFGF, promoting the differentiation of normal fibroblasts
into CAFs. Hypoxia modulates tumor-infiltrating immune cells, enhances the recruitment of immune-tolerant
cells and directs the transcriptional
activation of immunosuppressive factors which has been shown to reduce
the activity of effector cells such as cytotoxic T lymphocytes (CTLs),
NK cells, and DC cells. At the same time,
hypoxia upregulates immunosuppressive regulatory T cells, MDSCs, and
TAMs, stimulates the secretion of immune-suppressing cytokines and
chemokines.
,

While this study provides
insights into the influence of ECM composition
and stromal context on therapeutic response in 3D breast cancer tumoroids,
certain limitations should be acknowledged to guide future work. The
current model lacks immune and functional vascular components, which
restricts the ability to fully assess tumor–immune interactions
and drug delivery dynamics observed in vivo. The
ECM proteins incorporated in the 3D model do not encompass the full
biochemical complexity of the native tumor tissue.
4.1
Conclusion
and Future Directions
This study utilized a 3D breast tumor-stroma
model to investigate
the efficacy of PDT and chemotherapy. A differential response of distinct
breast cancer subtypes to PDT treatment was demonstrated by enhanced
efficacy in MCF-7 tumoroids in contrast to MDA-MB-231 tumoroids. Moreover,
since the model incorporates a central tumor mass (TM) surrounded
by a stromal component, we were able to identify differential responses
to treatment between the TM and the stromal components.
The
study also explored the impact of different 3D stromal tissues on
the treatment efficacy, which, to the best of our knowledge, has not
been previously explored in PDT. A comparative analysis was conducted
between breast cancer models featuring either a primary breast stromal
compartment or a metastatic lung stroma compartment. For 3D tumoroids
with a metastatic stroma, treatment with either a photosensitizer
or doxorubicin alone demonstrated significant cytotoxic effects when
compared to models incorporating a primary breast stroma.
We
examined the development of hypoxia in the 3D constructs and
the response to treatment using real-time oxygen monitoring, since
the presence of hypoxia is an important factor in determining treatment
efficacy for a range of reasons. These studies generated an intriguing
and potentially useful finding in that we were able to demonstrate
that oxygenation levels measured post-treatment correlated with the
treatment-induced cytotoxicity, which indicates that hypoxia gradients
can act as a surrogate measurement for treatment efficacy. This technique
has not been employed previously for studying PDT or combination therapy,
and our results suggest that it could prove of value to future mechanistic
studies.
In summary, we found that alterations within the TME
significantly
influence the efficacy of PDT, as shown by the higher effectiveness
of PDT-TMPyP4 as a standalone therapeutic agent within the 3D breast
cancer model featuring a metastatic stromal compartment, as opposed
to the model with a primary stromal compartment.
The outcomes
of this study emphasize the importance of a personalized
approach to breast cancer therapy, advocating for the adoption of
therapeutic strategies based on the specific subtypes of the disease,
as well as the location, whether it is a primary or metastatic site.
Future research can involve evaluating the effectiveness of PDT within
such 3D models that incorporate patient-derived tissues, including
both cancer cells and stromal cells, such as cancer-associated fibroblasts,
tumor-associated macrophages, and tumor-associated endothelial cells
from both primary and metastatic sites. 3D models that utilize a dense
collagen matrix should also be useful for nanoparticle studies involving
PDT since diffusion through the matrix is a limiting factor for nanoparticle
delivery. Moreover, combining nanodelivery systems with strategies
for the mitigation of hypoxia in tumor tissues, modification of the
TME through normalization of tumor vasculature and ECM, and inhibition
of hypoxia-related proteins could prove beneficial.

Supplementary Material