Journal of Cancer 2017, Vol. 8
http://www.jcancer.org
761
J
J
o
o
u
u
r
r
n
n
a
a
l
l
o
o
f
f
C
C
a
a
n
n
c
c
e
e
r
r
2017; 8(5): 761-773. doi: 10.7150/jca.17648
Review
Role of tumor microenvironment in tumorigenesis
Maonan Wang
1,2
, Jingzhou Zhao
2
, Lishen Zhang
2
, Fang Wei
2
, Yu Lian
2
, Yingfeng Wu
2
, Zhaojian Gong
2
,
Shanshan Zhang
1
, Jianda Zhou
3
, Ke Cao
3
, Xiayu Li
3
, Wei Xiong
1,2,3
, Guiyuan Li
1,2,3
, Zhaoyang Zeng
1,2,3
,
Can Guo
1,2,3
1. Key Laboratory of Carcinogenesis of Ministry of Health, Xiangya Hospital, Central South University, Changsha, Hunan 410078, China;
2. Key Laboratory of Carcinogenesis and Cancer Invasion of Ministry of Education, Cancer Research Institute, Central South University, Changsha, Hunan
410078, China;
3. Hunan Key Laboratory of Nonresolving Inflammation and Cancer, Disease Genome Research Center, The Third Xiangya Hospital, Central South
University, Changsha, Hunan 410013, China.
Corresponding authors: Zhaoyang Zeng, Email: zengzhaoyang@csu.edu.cn; or Can Guo, Email: guocde@csu.edu.cn
© Ivyspring International Publisher. This is an open access article distributed under the terms of the Creative Commons Attribution (CC BY-NC) license
(https://creativecommons.org/licenses/by-nc/4.0/). See http://ivyspring.com/terms for full terms and conditions.
Received: 2016.09.20; Accepted: 2016.12.22; Published: 2017.02.25
Abstract
Tumorigenesis is a complex and dynamic process, consisting of three stages: initiation,
progression, and metastasis. Tumors are encircled by extracellular matrix (ECM) and
stromal cells, and the physiological state of the tumor microenvironment (TME) is closely
connected to every step of tumorigenesis. Evidence suggests that the vital components of the
TME are fibroblasts and myofibroblasts, neuroendocrine cells, adipose cells, immune and
inflammatory cells, the blood and lymphatic vascular networks, and ECM. This manuscript, based
on the current studies of the TME, offers a more comprehensive overview of the primary functions
of each component of the TME in cancer initiation, progression, and invasion. The manuscript a
lso
includes primary therapeutic targeting markers for each player, which may be helpful in treating
tumors.
Key words: cancer-associated fibroblasts (CAFs), neuroendocrine cells, adipose cells, immune-inflammatory
cells, angiogenesis
Introduction
Currently, ten major characteristics of cancer
have been universally recognized, including
unlimited multiplication, evasion from growth
suppressors, promoting invasion and metastasis,
resisting apoptosis, stimulating angiogenesis,
maintaining proliferative signaling, elimination of cell
energy limitation, evading immune destruction,
genome instability and mutation, and tumor
enhanced inflammation (Figure 1) [1]. Although
researchers now have an understanding of most
characteristics of cancer [2-30], the characteristics
regarding cancer formation, which is the focus of the
current study, remains unknown. After the ‘ecological
therapy’ strategy was widely employed [31], much
effort has been devoted to determining how cellular
and noncellular components of the tumoral niche help
tumors to acquire these characters. These cellular and
noncellular components of the tumoral niche
comprise tumor the microenvironment (TME). The
TME consists of extracellular matrix (ECM) as well as
myofibroblasts and cellular players, such as
fibroblasts, neuroendocrine (NE) cells, adipose cells,
immune-inflammatory cells, and the blood and
lymphatic vascular networks [32]. Furthermore, TME
has increasingly been shown to dictate aberrant tissue
function and play a critical role in the subsequent
evolution of more stubborn and advanced
malignancies [33]. Oncologists have also found that
when the microenvironment in a healthy state, it can
help protect against tumorigenesis and invasion. By
contrast, if it is not in a healthy state, it will become an
accomplice.
The intent of this paper was to summarize the
existent knowledge on the potential role of each TME
Ivyspring
International Publisher
Journal of Cancer 2017, Vol. 8
http://www.jcancer.org
762
component in tumorigenesis: initiation, progression,
and metastasis, respectively. We have also
summarized some of the main cellular players, such
as cancer-associated fibroblasts, immune and
inflammatory cells, blood and lymphatic vascular
networks, adipose cells, neuroendocrine cells and
ECM in the TME, as well as their corresponding
targets in TME, in the hope of providing some clues
for future TME research. We have also introduced the
therapeutic target markers for various parts of TME
based on the current research results.
Figure 1. The tumor microenvironment and characteristics of
cancer. It is currently widely recognized that tumor microenvironments are
wildly influenced by the ten main characteristics of cancer: A. unlimited
multiplication; B. escaping from growth suppressors; C. promoting invasion and
metastasis; D. resisting apoptosis; E. stimulating angiogenesis; F. maintaining
proliferative signaling; G. elimination of cell energy limitation; H. evading
immune destruction; I. genome instability and mutation; J. tumor-enhanced
inflammation. Lower cure rate and poor prognosis of cancer patients are closely
related to these ten characteristics of cancer. These ten characteristics make
cancer more mysterious within the complex tumor microenvironments.
Cancer-associated fibroblasts (CAFs)
A sub-population of fibroblasts with a
myofibroblastic phenotype in cancerous wounds is
distinguished as cancer-associated fibroblasts (CAFs).
After activation, fibroblasts are known as CAFs or
myofibroblasts [34-36]. During natural wound repair,
myofibroblasts are transiently present [37]. Unlike the
process of wound healing, CAFs at the site of a tumor
remain perpetually activated, as in tissue fibrosis.
Several studies have demonstrated that only the
activated fibroblasts are required to initiate and
promote tumor growth [38-40]. Fibroblast activation
may be induced through various impetuses when
tissue lesions occur, including growth factors, direct
cell-cell communication, adhesion molecules
contacting with leukocytes, reactive oxygen species
[41], and microRNA [42, 43]. When the fibroblasts
remain activated after the initial insult has regressed,
these activated fibroblasts may work with other
molecular pathways to boost neoplasm initiation.
These CAFs have a significant impact on cancer
progression through remodeling ECM, inducing
angiogenesis, recruiting inflammatory cells, and
directly stimulating cancer cell proliferation via the
secretion of growth factors, immune suppressive
cytokines, and mesenchymal-epithelial cell
interactions [41, 44]. For instance, Galectin-1
overexpression in CAFs advances the development of
abutting cancer cells [45] and is correlated with poor
prognosis in several types of cancer, including breast
and prostate cancer and laryngeal carcinoma [46-49].
Chemokine (C–X–C motif) ligand 12 (CXCL12),
violently uttered in CAFs, may induce
epithelial-mesenchymal transition (EMT) of cancer
cells to promote cancer progress in gastric and
prostate cancers [50, 51]. Moreover, one team
discovered that MMP-2, derived from senescent
CAF-CMs, induced epithelial invasion and
keratinocyte discohesion into collagen. Interleukin-22
(IL-22) is also expressed by CAFs to encourage gastric
cancer cell invasion through STAT3 and ERK
signaling [52]. Using a 3D invasion model, another
study found that HCT116 cells manifested a
substantially invasive phenotype, while media
originated from human dermal fibroblasts (HDF) [53].
Since myofibroblasts can be distinguished by
alpha-smooth muscle actin (α-SMA), laminin-1,
transforming growth factor beta (TGF-β1), vascular
endothelia growth factor A (VEGF-A), etc. [54, 55],
CAFs have been recognized as playing an essential
role in the metastasis and development of cancer [56].
Oncologists have found that through HGF, TGF-β,
platelet-derived growth factor (PDGF) etc., CAFs may
promote tumor growth and invasion (Figure 2).
Through fibroblast growth factor 2 (FGF2), VEGF, etc.,
CAFs may promote tumor development by
promoting angiogenesis [57]. Additionally, CAFs also
interact with immune-inflammatory cells and
neuroendocrine cells through different cell factors and
cytokines to jointly promote the initiation,
progression, and invasion of cancer [58-62]. However,
many of the markers that have been gradually proven
to be unable to identify all of the CAFs, are not unique
to the CAFs [63]. The cardinal functions and the
primary markers of CAFs are illustrated in Table 1.
For instance, α–SMA, one of the previous major
markers of CAFs, was found to be expressed in
normal fibroblasts [64], pericytes, and smooth muscle
Journal of Cancer 2017, Vol. 8
http://www.jcancer.org
763
cells [65]. The cell-surface serine protease fibroblast
activation protein α (FAPα), which is highly
expressed in quiescent mesodermal cells in multiple
tissue types [66], is also not specific to CAFs.
Additional markers fibroblast-specific protein 1
(FSP-1) [67], vimentin, and certain proteins, including
NG2 (Neuroglial Antigen-2), platelet-derived growth
factor receptor-β (PDGFR-β), fibroblast-associated
antigen, and prolyl 4-hydroxylase have been shown to
be expressed in cells other than CAFs [63].
Table 1. The function of cell players in the tumor
microenvironment.
Cell players
Main markers
Primary functions
Cancer-associated
fibroblasts (CAFs)
PDGF*; FAP*;
FGFR*; VDR*
Regulating inflammation;
Participating in wound healing;
Integrating collagen and protein to
form the ECM fiber network;
Escaping damage;
Immune &
Inflammatory cell
TNF-α; IL-10; IL-12;
TGF-β; Foxp3+*;
HMGB1*; CD163+*;
KIR*; PD-1+*
Treatment of wound healing and
infection; Clearing dead cells and
cellular debris; Having a double
effect on tumor formation
The blood &
lymphatic
vascular
networks
VEGRF3; LYVE-1;
CD31; CD34; VEGF*;
PlGF*; VEGF-B*;
VEGF-C*; VEGF-D*
Require nutrients and oxygen;
Evacuating metabolic wastes and
carbon dioxide; Helping to escape
immune surveillance.
Adipose cell
AIs*; MBD6*
Producing circulating blood
estrogen; A major energy source;
Relating with inflammation;
Recruiting immune cells; Support
vasculogenesis.
Neuroendocrine
cell
NSE; CgA; K18&K8
cytokeratins; PGP9.5;
Ki-67; IL-2; KE108*;
DLL3*; EGF*
Extending lumina and adjacent
epithelial cells; Regulating
secretion and motility; Controlling
lung branching morphogenesis;
Providing a protective niche for a
subset of lung stem cells.
Note: *, the targeting markers.
Similarly, although there is no unique marker,
there are still some targets for significant help in
cancer treatment. CAFs work in two main ways in
cancer treatment. One method is by directly reversing
CAFs into the normal fibroblasts or inhibiting their
growth. This method highlighted that efforts such as
reconstituting miRNA expression had been proven to
deactivate CAFs [68-70] and inhibit PDGF signaling in
the mouse model of cervical carcinogenesis; it can also
reduce tumor proliferation [71]. Additionally, the
fibroblast growth factor receptor (FGFR) signaling
pathway may be one of the therapeutic
objectives in
gastric cancer [72].The other objective is
dedifferentiating CAF into a quiescent state. One data
set showed that Vitamin D receptor (VDR) ligands
promoted the dedifferentiation of satellite cells and
abrogated fibrosis [73]. Using a murine xenograft
model of colon carcinoma, another recent study found
that when targeting fibroblast activation protein
(FAP), the accumulation of CAFs was markedly
reduced [74]. FAP is expected to become another
marker of CAFs targeted therapy.
Immune and inflammatory cell
The main function of the mammalian immune
system is to monitor tissue homeostasis, to protect
against invading or infectious pathogens and to
eradicate damaged cells [75]. The primary theory
advises that immune surveillance has significant
roles in recognizing and eradicating a large part of
nascent tumor cells [1]. However, unlike normal
functions, immune-inflammatory cells would persist
in sites of chronic inflammation, linked to diverse
tissue pathologies, including
fibrosis, aberrant
angiogenesis, and neoplasia
[76]. In light of recent
discoveries in immune system
research, it is difficult to
ignore the crucial issue that
immune-inflammatory cells
may be the early cradle of
cancer [77-83].
Several studies have
revealed the contribution of
adaptive and innate
immunity in cancer
immunoediting, including the
unmanipulated innate
immune system without
adaptive immunity [84].
Dunn et al. divided the
dynamic process of cancer
immunoediting into three
steps: elimination,
Figure 2. The inactive network of cancer cells and the tumor microenvironment.
Journal of Cancer 2017, Vol. 8
http://www.jcancer.org
764
equilibrium, and escape [85]. These three stages can
also be used to express the role of the immune system
in cancer initiation, progression, and invasion. He
pointed out that in the elimination phase, the
immunologic system can defeat nascent tumors. This
is accomplished by different inflammatory cells
[86-88] and signaling molecules [88, 89]. Once cancer
cells have been completely eliminated, these active
factors and immune cells may have an additional role
in producing “immunologic sculpting” or
“immunoediting” [85]. This means that in the
equilibrium stage, tumor cells cannot be completely
cleared, but the tumor growth can controlled. In order
to escape the immune surveillance, cancer cells tend
to evolve a number of phenotypic changes in this
stage, such as EMT [90]. These cells, with survival
advantage, would eventually develop into the
primary solid tumor. In other words, the immune
microenvironment helps cancer cells to select the
dominant cells so that the tumor can progress at the
fastest rate in a limited environment. Several studies
in mice have revealed that the depletion of
macrophages during tumor induction restrained
tumor growth [91, 92]. Regarding the escape stage, the
immune system may help tumor cells to format the
clinical characterization of tumor immune escape
mechanisms, while many experiments have proven
that immune cells can reduce anti-cancer proteins or
cytokines to promote cancer invasion [93, 94]. In
addition, one experiment also found that immune
cells may be related to the regulation of apoptosis [95].
However, the greatest divergence between these
transitional immune cells and the other stroma cells is
that these immune cells can be redesigned toward the
tumor destruction in therapies. How to activate the
normal function of immune cells will be the focus of a
future study.
As depicted in Figure 2, CAFs [96], NE cells [97],
adipose cells [98-100], and inflammatory cells
[101-103] in the TME can affect the role of immune
cells through the secretion of different cytokines, cell
factors, or interacting proteins, which adds several
difficulties to the search for markers and targets for
cancer therapy [104, 105]. In different cell and tumor
types, the complexity and heterogeneity of immune
factors also provides a further complication to finding
more specific markers of the immune cells. The
immune system is divided into adaptive immunity
and innate immunity. Adaptive immune cells include
thymus-dependent lymphocytes (T cells), and
bursa-dependent lymphocytes (B cells). Innate
immune cells consist of dendritic cells (DC), killer
lymphocytes, natural killer (NK) cells, hyaline
leukocyte/macrophage, granulocytes, and mast cells
[106]. According to the different clusters of
differentiation, T cells are divided into CD4
+
T (helper
T cells, Th) and CD8
+
T (cytotoxic T cells, Tc) cells.
These secrete IFN-γ, TNF-α, and IL17, which have
antitumor effects. B cells are mainly marked by
different antigens in different physiological periods,
such as mainly expressing CD19 and CD20 in pre-B
cells, immature B cells, and plasma cells, mainly
expressing IgM, IgD, and CR1 in mature B cells, and
mainly expressing IgM, IgD, IgA, IgG in memory B
cells. A key feature of human NK cells, which could
efficiently recognize infected and malignant target
cells, is the expression of HLA class Ⅰ-specific
receptors of the KIR and NKG2 gene families [107].
DCs express co-stimulatory molecules and innate
inflammatory cytokines, such as IL-12, IL-23, and IL-1,
that promote IFN-γ-secreting CD4
+
T cells and
cytotoxic T lymphocyte responses [108]. DCs
represent key targets for 1,25-dihydroxyvitamin D
3
(1,25(OH)
2
D
3
), which can directly induce T cells [109].
After the addition of immune-inflammatory cells
into an emerging hallmark of cancer [1], studies have
shown that immune cells were associated with
immunosuppression. Some types of immune cells
have an innate function of immune suppression, and
some cytokines can also activate them through
different signaling pathways. The main
immunosuppressive cells are regulatory T (Treg) cells
marked by Foxp3
+
[110], myeloid-derived suppressor
cells (MDSC) marked by HMGB1 [111], and M2
macrophages marked by CD163
+
[108, 112], which
have emerged as a leading method in the
development of new immunotherapeutics. Studies
have also found that some granulocytes can promote
cancer development through the expression of
cytokines, such as the hematopoietic growth factor
(HGF), granulocyte colony-stimulating factor
(G-CSF), or inducing changes in stromal cells
[113-115]. PD-1
+
and cytotoxic T-lymphocyte
antigen-4 (CTLA-4) expressed by “exhausted” CD8
+
T
cells are also targeting markers in treating patients
with breast and non-small cell lung (NSCL) cancer
[116-118]. In addition to the mutual activation [119],
the antitumor effects also can be suppressed by some
co-inhibitory molecules expressed by antitumor
immune cells, such as PD-1/PD-L1 [120, 121]. Some
antitumor cytokines can also promote
immunosuppression, such as IL-10 and TGF-β
secreted by DCs, which may activate Treg cells that
are recruited to the tumor under the influence of the
chemokines, including CCL22 and CXCL12 [108]. As
for these immunosuppressive cells, cell depletion
strategies [122-125], tumor vaccines [126],
intratumoral injection with an agonistic antibody
[127], targeting the transcription factor, and
suppression of activated receptors [128, 129] have
Journal of Cancer 2017, Vol. 8
http://www.jcancer.org
765
been widely used in improving immune surveillance
and promoting antitumor immune responses. Agonist
antibodies of CD40, 4-1BB, GITR, and OX-40 can also
be used to enhance antigen-specific T cell responses
[130], CD25 antibody can be used to reduce the
number of Treg cells or inhibit Treg function [131],
and promote the maturation of DC and rational use of
cytokines and antibodies to break the immune
tolerance [132]. The significant roles and the master
markers of the immune and inflammatory cells are
indicated in Table 1.
The blood and lymphatic vascular
networks
Similar to normal tissues, the blood and
lymphatic vascular networks supply oxygen and
sustenance as well as removing carbon dioxide and
metabolic wastes for sustaining the survival of
neoplasm growth. These networks have two main
features. First, the new vessels surrounding tumors
are usually inefficient, tortuous, or leaky [133].
Second, the angiogenic switch is almost always
activated and remains active during the process of the
tumorigenesis, resulting in continued growth of new
natural blood vessels [134]. In the tumor angiogenesis
process, new blood vessels form from pre-existing
vessels, making the blood and lymphatic vascular
network more complex [135].
The blood and lymphatic vascular networks
have different roles during the stages of
tumorigenesis [136]. Tumor normal cells are
confronted with the challenge of hypoxic
surroundings [137]. To survive in hypoxic
circumstances, primary tumor cells may adjust to the
low oxygen setting, or migrate to and recruit blood
vessels [138]. A selection shape can be chosen, which
is more enterprising and metastatic, and is provided
by a chronically hypoxic environment [139, 140]. In
the process of tumor progression, one of the primary
functions of the blood and lymphatic vascular
networks is to help tumor cells escape immune
surveillance. Escape measures are mainly divided into
two categories. Directly, the lymphatic
microenvironment will weaken or eliminate the
normal function of immune cells. For instance, the
myeloid-derived suppressor cells (MDSCs) and the
immature DCs in the sentinel lymph nodes (SLNs)
could restrict the normal operation of T cells [141-143].
When the metastatic tumor enters a novel
environment, CD4
+
and CD8
+
T cells may help them to
evade the host immune system [144, 145]. The
remodeling of unusual endothelial venules (HEVs)
can indirectly influence immune cells to traffic into
lymph nodes [146]. Though some immune cells traffic
into the lymph nodes through the draining afferent
lymphatic vessels, lymphocyte recruitment into the
SLN via HEVs is fundamental [144].
Moreover, lymphatic vessels around the tumor
also provide a physical link between the SLNs and
primary tumor. When lymphatic vessels are
obstructed, collateral lymphatic vessels can make up
for the diminution in lymphatic capacity [144, 147].
This physical connection is like a highway through
which tumor cells can reach other locations. Some
phenomena showed that local tumor invasion
correlated with excellent lymphatic vessel density in
the tumor margin [148, 149]. Figure 2 shows that
through different interacting factors, adipose cells,
inflammatory cells, and CAFs can also closely connect
with lymphatic vascular networks [150-153]. The
physical and chemical connection makes the role of
angiogenesis in tumor formation more of a mystery.
The important functions and the primary
markers of the blood and lymphatic vascular
networks are listed in Table 1. Banerji et al. found that
the lymphatic vessel endothelial HA receptor
(LYVE-1) was expressed predominantly in lymphatic
vessels [154]. Evidence suggests that VEGF is among
the most important factors for anti-angiogenic
treatment [155, 156]. A number of studies have
discovered that VEGF would inhibit the development
of new vessels, block the VEGF or its signaling
pathways, prune pre-existing vessels, and induce
vessel normalization [133, 157, 158]. Furthermore,
clinical trials that targeted VEGF and other markers
showed prolonged survival [155, 159]. Members of the
VEGF family include placental growth factor (PlGF),
VEGF-B, VEGF-C, and VEGF-D, which were also
good candidates for anti-angiogenic treatment [133,
160-164]. Another marker, PDGFβ, secreted by
sprouting ECs, has two sides of the effects on tumors.
It would recruit prostate cancer (PC) signaling
through the presenting PDGF receptor-β (PDGFRβ)
[165]. Studies have revealed that blockage of PDGFβ
makes tumor vessels more sensitive to VEGF
inhibitors [166, 167]. Consequently, the
decrement of
PDGFβ
also enhances the risk of increased
metastasis [158]. Generally, anti-PDGFβ drugs play
an auxiliary role in anti-VEGF treatment. Some other
markers, such as CRISP-3 [94], CCR7 [168], GATA2
[169], Prox1 [170], and Foxc2 [171] have also been
found and are well used in the treatment. In
conclusion, abnormal tumor vasculature exhibited
remarkable spatiotemporal heterogeneousness, and
not only damaged perfusion and drug delivery, but
also made chemoradiotherapy less expeditious.
Adipose cells
Adipose tissue comprises two cell types, white
adipose tissue (WAT) and adipocytes [172]. The