Summary:

In this issue, Weeraratna and colleagues demonstrate that observed differences in melanoma aggressiveness in younger versus older patients can be explained not just by cell-intrinsic alterations over time, but by age-dependent changes in fibroblasts and the extracellular matrix they help create. Their findings identify novel cellular targets for melanoma therapy, as well as candidate prognostic biomarkers to better inform clinical decisions for patients with melanoma.

See related article by Kaur et al., p. 64.

See related article by Ecker et al., p. 82.

Cancer and aging are interminably linked, forging a vexing and uneasy pathologic marriage. Advances in medicine that extend the human lifespan have led to the daunting realization that “if you live long enough, you will get cancer.” The reasons for this relationship have consumed researchers worldwide, but are mainly focused on the cancer cell itself. An aspiring cancer cell will, for example, acquire an increasing number of mutations over time including those that confer an oncogenic advantage, epigenetically adapt toward a chromatin configuration that favors malignancy, “learn” to avoid senescence and the inevitable time-dependent shortening of telomeres, and metabolically adjust to handle the rigors of tumorigenesis. However, much less effort has been spent on microenvironmental aging. The obvious age-related change that is known to influence the rise and aggressiveness of cancer is in the immune system, which substantially diminishes in potency over time. But beyond immune cell potency, how does the aging microenvironment contribute to tumor progression? In this issue of Cancer Discovery, a group led by Ashani Weeraratna describes two studies that tackle this question head-on, focusing on extracellular matrix (ECM)–associated mechanisms occurring in the dermis and in the lymphatic vasculature (1, 2).

As a driving force for their two studies, Weeraratna and colleagues noted an apparent paradox: Aged patients with melanoma have reduced sentinel lymph node metastasis (assessed through lymph node biopsy) compared with younger patients, despite having a worse prognosis and generally harboring more aggressive melanomas with distant metastases (3, 4). They therefore began to search for factors in aged patients that might explain these discordant observations. Like a wrinkle on your grandparent's face, the answer was hiding in plain sight: A fundamental difference in aged ECM was found to promote melanoma metastasis. Skin melanocytes reside at the epidermal–dermal junction but can invade vertically downward into the dermis after transformation and progression (vertical growth phase), significantly enhancing patient risk. The dermis is primarily built from ECM constituents secreted from dermal fibroblasts, including fibrous connective tissue comprised largely of collagen. ECM proteins form a dense basket-weave structure in young skin, but a gap-filled structure in old skin (apparent as wrinkles). Kaur and colleagues asked whether this change in ECM structure was intrinsically linked to dermal fibroblast age and, reasoning that a gap-filled ECM structure would facilitate melanoma cell invasion, whether the aged ECM structure could account for the enhanced metastatic propensity of melanoma in aged patients (1). Analysis of the secretomes of dermal fibroblasts from aged versus young healthy donors indeed showed a striking loss of ECM remodeling proteins in the aged secretome, including agrin, laminin β2, fibula, tenascin, and most dramatically HAPLN1, which was a focus for much of their studies.

To address how the age of the microenvironment can influence collagen conformation, the authors used a comprehensive array of in vitro techniques, in vivo mouse models, and human tissue samples. In vitro cell-derived matrix (CDM) assays proved a powerful and versatile technique allowing the authors to confirm that matrices formed by young fibroblasts (dermal or lymphatic) were denser isotropic collagen structures, whereas matrices formed by aged fibroblasts (dermal or lymphatic) had a higher degree of alignment in anisotropic collagen structures. Moreover, this technique lent itself to manipulation by the addition of signaling molecules (TGFβ), chemical inhibitors (TGFβ inhibitors), ECM remodeling proteins [recombinant HAPLN1 (rHAPLN1)], and genetic manipulation [short hairpin knockdown of HAPLN1 (shHAPLN1)]. Taken together, these studies demonstrated that: the secretion of high levels of HAPLN1 observed in young dermal/lymphatic fibroblasts was responsible for the dense “basket-weave” structure characterizing young collagen matrices; rHAPLN1 was sufficient to rescue the loose, aligned structure of aged matrices; and genetic knockdown of HAPLN1 in young fibroblasts resulted in a decreasingly dense matrix. Moreover, through the addition of exogenous TGFβ, the authors could draw parallels between aged dermal fibroblasts and the established TGFβ-stimulated cancer-associated fibroblast (CAF) phenotype (5).

Hypothesizing that ECM structural differences surrounding the lymphatic vasculature could also influence vasculature permeability, thereby affecting the extravasation of melanoma cells from lymphatic vessels into the surrounding tissues (distant metastasis), Ecker and colleagues employed CDM assays to generate an in vitro model of endothelial cell permeability using HUVEC cells derived from umbilical cord veins (2). These studies showed that aged fibroblast matrices encouraged permeability of an endothelial cell monolayer via reduced VE-cadherin and integrin expression, and that these were HAPLN1 dependent. From these data, the authors predicted that lymphatic endothelial cell permeability would be similarly influenced by ECM structure in an aged/young microenvironment and went on to corroborate these results with studies in mouse lymph nodes, demonstrating the role of HAPLN1 in vivo.

To complement these data, the authors used organotypic reconstructs to ask how invasive melanoma cells respond to these distinct microenvironmental structures. Three-dimensional skin reconstructs allowed the authors to grow a layer of skin ex vivo, whereas in vitro culture reconstructs combined fibroblasts and collagen matrices with T cells or endothelial cells. Using these methods, the authors could seed either young or aged fibroblasts with either shHAPLN1 genetic knockdown or rHAPLN1 treatment, respectively. As predicted, young fibroblasts or aged fibroblasts treated with rHAPLN1 induced denser isotropic collagen fibril structures, whereas aged fibroblasts or young fibroblasts treated with shHAPLN1 induced looser, aligned anisotropic collagen fibril structures. Three-dimensional skin reconstructs revealed that these ECM structures influenced melanoma metastasis: Melanoma cells could successfully invade into looser anisotropic collagen structures, whereas invasion was inhibited in dense isotropic collagen structures (Fig. 1). Strikingly, the opposite was observed for T-cell migration, for which dense isotropic matrices were optimal. Finally, endothelial cells cultured with aged fibroblasts, or young fibroblasts with shHAPLN1 genetic knockdown, were more permeable to melanoma cell invasion. However, endothelial cells cultured with young fibroblasts or aged fibroblasts treated with rHAPLN1 were less permeable to melanoma cell invasion (Fig. 1).

The choice to study the impact of the aged tumor microenvironment on melanoma metastasis is a fresh approach and provides multiple advantages when considering broader clinical implications. The microenvironment is likely to be less heterogenous than the tumor itself, and thus any translational implications of this work could be more ubiquitously applied. Moreover, the ECM presents an easier target as therapies do not have to pass through the cell membrane, a point that was well demonstrated by the authors' successful use of rHAPLN1 treatment in mouse models. rHAPLN1 could be injected either intradermally to modify the dermal ECM and thus melanoma metastasis, or into the draining lymph nodes resulting in increased lymph node tumor burden but reduced distant pulmonary metastases, suggesting lymphatic permeability was altered by this treatment.

In these two articles, the Weeraratna laboratory has taken observations from the clinic—namely, that aged patients tend to have more aggressive melanoma, but despite this, have reduced sentinel lymph node metastases—and used these to generate hypotheses about aged versus young microenvironments. This bedside-to-bench-to-bedside approach provided translatable insights that could prove useful in the clinic. Treatment with rHAPLN1 was sufficient to remodel the ECM and reduce distant metastasis, a finding that could be exploited in the clinic as something of a fountain of youth for ECM through treatment of patients with rHAPLN1 or HAPLN1 agonists. However, additional studies are needed to determine whether HAPLN1-induced ECM remodeling reverses over time, and if localized injections of HAPLN1 result in any untoward effects to normal local or distant stroma. Kaur and colleagues also demonstrated that ECM structure is a key factor influencing the immune tumor microenvironment (1). Optimal T-cell infiltration occurs in young, HAPLN1-rich, dense isotropic ECM. Moreover, in vivo intradermal injection of rHAPLN1 increased the T cell:Treg ratio, thereby potentiating the immune response. Thus, younger ECM is more likely to encourage immune clearance of melanoma cells than aged ECM. This result could have far-reaching implications beyond melanoma. If modulation of the ECM could be exploited to enhance immune cell infiltration, this approach could be used to potentiate response to immune-based therapeutics.

Of course, with new opportunities come new challenges. Notably, the premise of this paper—that the aged microenvironment promotes melanoma metastasis—is not true for all cancers. Breast cancer, for example, is more indolent in elderly patients (6). The Weeraratna laboratory has provided insights that may help us understand these cancer-specific differences. For example, whereas melanoma cells metastasize better through aligned, gap-filled ECM structures, breast cancer more efficiently metastasizes through a stiffer matrix (7–9). Therefore, optimal matrix stiffness for metastasis is relative, and may be dependent upon both cancer type and the tissue into which metastasis occurs. Moreover, other age-related factors contribute to poor patient prognosis, such as tolerance to intensive treatment regimens, comorbidity, and other biological factors of the tumor and microenvironment. Finally, each patient's tumor/microenvironment should be assessed on a case-by-case basis, as not all tumors are likely to respond in the same way to the microenvironment, and not all patients age in the same way. Therefore, screening of patient dermal/lymphatic ECM and lymphatic permeability, in conjunction with sentinel lymph node biopsies, when taken together may provide better prognostic markers to inform clinical decisions. In any case, the studies by the Weeraratna laboratory have helped usher in a new age of enlightenment with respect to the tumor microenvironment.

G. Merlino is a scientific advisory board member of the Melanoma Research Foundation. No potential conflicts of interest were disclosed by the other author.

This work was funded by the Intramural Research Program of the National Institutes of Health.

1.
Kaur
A
,
Ecker
BL
,
Douglass
SM
,
Kugel
CH
,
Webster
MR
,
Almeida
FV
, et al
Remodeling of the collagen matrix in aging skin promotes melanoma metastasis and affects immune cell motility
.
Cancer Discov
2019
;
9
:
64
81
.
2.
Ecker
BL
,
Kaur
A
,
Douglass
SM
,
Webster
MR
,
Almeida
FV
,
Marino
GE
, et al
Age-related changes in HAPLN1 increase lymphatic permeability and affect routes of melanoma metastasis
.
Cancer Discov
2019
;
9
:
82
95
.
3.
Balch
CM
,
Soong
SJ
,
Gershenwald
JE
,
Thompson
JF
,
Coit
DG
,
Atkins
MB
, et al
Age as a prognostic factor in patients with localized melanoma and regional metastases
.
Ann Surg Oncol
2013
;
20
:
3961
8
.
4.
Page
AJ
,
Li
A
,
Hestley
A
,
Murray
D
,
Carlson
GW
,
Delman
KA
. 
Increasing age is associated with worse prognostic factors and increased distant recurrences despite fewer sentinel lymph node positives in melanoma
.
Int J Surg Oncol
2012
;
2012
:
456987
.
5.
Papageorgis
P
,
Stylianopoulos
T
. 
Role of TGFβ in regulation of the tumor microenvironment and drug delivery (review)
.
Int J Oncol
2015
;
46
:
933
43
.
6.
Diab
SG
,
Elledge
RM
,
Clark
GM
. 
Tumor characteristics and clinical outcome of elderly women with breast cancer
.
J Natl Cancer Inst
2000
;
92
:
550
6
.
7.
Nazari
SS
,
Mukherjee
P
. 
An overview of mammographic density and its association with breast cancer
.
Breast Cancer
2018
;
25
:
259
67
.
8.
Kauppila
S
,
Stenback
F
,
Risteli
J
,
Jukkola
A
,
Risteli
L
. 
Aberrant type I and type III collagen gene expression in human breast cancer in vivo
.
J Pathol
1998
;
186
:
262
8
.
9.
Acerbi
I
,
Cassereau
L
,
Dean
I
,
Shi
Q
,
Au
A
,
Park
C
, et al
Human breast cancer invasion and aggression correlates with ECM stiffening and immune cell infiltration
.
Integr Biol
2015
;
7
:
1120
34
.