Skip to main content

Tumor extracellular matrix: lessons from the second-harmonic generation microscopy

Abstract

Extracellular matrix (ECM) represents more than a mere intercellular cement. It is physiologically active in cell communication, adhesion and proliferation. Collagen is the most abundant protein, making up to 90% of ECM, and 30% of total protein weight in humans. Second-harmonic generation (SHG) microscopy represents an important tool to study collagen organization of ECM in freshly unfixed tissues and paraffin-embedded tissue samples. This manuscript aims to review some of the applications of SHG microscopy in Oncologic Pathology, mainly in the study of ECM of epithelial tumors. It is shown how collagen parameters measured by this technique can aid in the differential diagnosis and in prognostic stratification. There is a tendency to associate higher amount, lower organization and higher linearity of collagen fibers with tumor progression and metastasizing. These represent complex processes, in which matrix remodeling plays a central role, together with cancer cell genetic modifications. Integration of studies on cancer cell biology and ECM are highly advantageous to give us a more complete picture of these processes. As microscopic techniques provide topographic information allied with biologic characteristics of tissue components, they represent important tools for a more complete understanding of cancer progression. In this context, SHG has provided significant insights in human tumor specimens, readily available for Pathologists.

Introduction

Non-cellular components of tissue, or extracellular matrix (ECM), provide biochemical and biomechanical support for cellular constituents. More than a mere intercellular cement, ECM is physiologically active in cell communication, adhesion and proliferation (Frantz et al. 2010). It is composed of approximately 300 proteins, the proportion between which gives the precise composition to specific structures (Naba et al. 2012). Fibrous proteins, such as collagen, elastin, fibronectins and laminins, are the major constituents of the ECM, which, together with proteoglycans, that are locally secreted and assembled, form the structural framework of most tissues (Frantz et al. 2010). Among the fibrous proteins, collagen is the most abundant in ECM, making up to 90% of ECM and 30% of total protein weight in humans (van der Rest and Garrone 1991).

Molecular approaches have highlighted genes that encode ECM components, which have been correlated with tumor behavior and clinical outcomes (Finak et al. 2008; Ramaswamy et al. 2003). Increased expression of genes encoding proteins that mediate extracellular remodeling has been associated with increased mortality in patients with breast, lung and gastric cancer. Such studies corroborate histological findings that show association between excessive ECM deposition and worse prognosis in solid tumors (Chang et al. 2005; Chang et al. 2004). Further, increased collagen deposition is the most well recognized ECM alteration that occurs inside the tumor tissue (Hasebe et al. 1997; Colpaert et al. 2003; Huijbers et al. 2010; Kauppila et al. 1998; Zhu et al. 1995; Gould et al. 1990; Boyd et al. 2002).

In the context of cancer biology, collagen regulates the biophysical and biochemical properties of tumor microenvironment (TME), which modulate cancer cell polarity, migration and signaling (Fraley et al. 2012; 2010; Levental et al. 2009a; Paszek et al. 2005; van Kempen et al. 2003). Thus, careful evaluation of ECM collagen could provide important information about the tumor.

The purpose of the present review is to discuss the role of a significant tool to characterize collagen in TME, by second-harmonic generation (SHG) microscopy, presenting some of the experience gained in Oncologic Pathology.

Multiphoton microscopy and second-harmonic generation microscopy

Multiphoton microscopy, a form of laser-scanning microscopy, utilizes nonlinear excitation to generate signal only within a thin raster-scanned plane (Tajik 2018). Its applications are diverse, as this device has become a choice for fluorescence microscopy, mainly in thick tissues and live animals.

One of its most striking features is its intrinsically focal-plane selectivity, without need of any confocal optical tool (e.g., pinhole). It offers a modest resolution improvement over conventional microscopy at the same wavelength since the point of spread function is squared, besides improved penetration and reduced damage conferred by the longer excitation wavelength (Cox 2011).

However, although multiphoton microscopy offers many advantages compared to conventional microscopy, it also has limitations and drawbacks. One obvious limitation is still the high cost of the appropriate ultrafast laser.

A second limitation pertains to the accelerated photobleaching in some nonlinear optical microscopies, like two-photons excited fluorescence (TPEF). Although TPEF produces photobleaching only in the focal plane, within the focal volume, high order photobleaching is observed. Compared to confocal microscopy, where the photobleaching rate increases quasi linearly with the excitation power, in multiphoton microscopy, the higher photon density activates more photobleaching pathways, resulting in accelerated effect (Kalies et al. 2011; Patterson and Piston 2000), however this effect it not observed in other non-linear microscopies, such as SHG.

Finally, specific limitations arise if the multiphoton laser beam interacts linearly with chromophores in the sample. For example, near infrared light is absorbed by the photosynthetic complex; other naturally occurring chromophores, such as the pigment melanin, can limit tissue imaging by causing thermal and mechanical damage in an experiment (Ustione and Piston 2011).

TPEF, SHG (Campagnola et al. 2002; 2001; Dombeck et al. 2003; Freund and Deutsch 1986) and third-harmonic generation (THG) (Müller et al. 1998) are examples of multiphoton microscopy and can also be used for specific imaging. SHG was one of the earliest forms of biological nonlinear microscopy proposed (Sheppard and Kompfner 1978). Figure 1 demonstrates the physical representation of each process.

Fig. 1
figure 1

Excitation and emission processes of the same hypothetical molecule in the various microscopy regimes represents: a conventional confocal microscopy: where a single photon of wavelength equal to λ0 excited molecule and produces a wavelength emission equal to λ1, where λ1 > λ0; b two-photons excited fluorescence: where two photons with wavelength equal to \( \raisebox{1ex}{${\lambda}_0$}\!\left/ \!\raisebox{-1ex}{$2$}\right. \) excited the molecule, producing an emission wavelength equal to λ2, where λ2 < \( \raisebox{1ex}{${\lambda}_0$}\!\left/ \!\raisebox{-1ex}{$2$}\right. \); c second harmonic generation microscopy: where two photons with wavelength equal to \( \raisebox{1ex}{${\lambda}_0$}\!\left/ \!\raisebox{-1ex}{$2$}\right. \) excited the molecule, producing an emission equal to λ3, where λ3 = \( \raisebox{1ex}{${\lambda}_0$}\!\left/ \!\raisebox{-1ex}{$2$}\right. \); and d third-harmonic generation microscopy: where three photons with wavelength equal to \( \raisebox{1ex}{${\lambda}_0$}\!\left/ \!\raisebox{-1ex}{$3$}\right. \) excited the molecule, producing an emission equal to λ4, where λ4 = \( \raisebox{1ex}{${\lambda}_0$}\!\left/ \!\raisebox{-1ex}{$3$}\right. \). The color arrows have length proportional to the energy of the photons and black arrows indicate vibrational energy loss. Solid lines represent real energy levels and dashed lines represent virtual energy levels

The SHG microscope has been successfully integrated into biological research. Therefore, in the past two decades, as femtosecond laser sources have become more robust and commercially available, publications related to this application have grown exponentially (Fig. 2). In biological and medical sciences, the role of SHG has been used in high-resolution optical microscopes. SHG imaging modality can probe molecular organization, symmetry, orientation, alignments and ultrastructures on the micro as well as the nanoscale. Because most biological structures are not highly ordered, they are optically isotropic and do not produce any SHG signal. Only those few biological structures that are ordered or that involve some spatial organizations that break the optical centro-symmetry can produce harmonic signal. Namely collagen, the major protein of the extracellular matrix, is one of the best-known SHG structures in biology. Collagen fibrils often aggregate into larger, cable-like bundles, several micrometers in diameter. This regularly staggered packing order provides the needed structural conditions for efficient SHG (Williams et al. 2005). An example of detailed schematic workstation setup has been previously shown (Natal et al. 2018a).

Fig. 2
figure 2

Graphical representation of the number of publications per year on biological applications of second harmonic generation microscopy between 1990 and 2020. A more significant increase is noted from year 2000 on

Second-harmonic generation microscopy in oncology pathology

Gynecological pathology

Breast cancer

In breast cancer was described three forms of collagen deposits in breast cancer, known as tumor associated collagen signature (TACS). In TACS-1, dense collagen was deposited in areas around the tumor, which was indicated by increased signal intensity, as hallmark for locating small tumor regions. In TACS-2, taut (straightened) collagen fibers were stretched around the tumor, indicating growth and increasing in tumor volume. Finally, in TACS-3, radially aligned collagen fibers can be identified, facilitating cell motility, and pointing to the invasive and metastatic growth potential of the tumor (Provenzano et al. 2006a) (Fig. 3). Additionally, tumor progression and the risk of metastases was directly associated to the increased density of collagen (Provenzano et al. 2008).

Fig. 3
figure 3

Second-harmonic generation images of breast cancer. Representation of tumor associated collagen signatures (TACS, see text): a TACS-1; b TACS-2; and c TACS-3

Considering the experimental findings in mice, TACS-3 phenotype facilitated tumor metastasis, a parameter that could predict survival outcomes in human patients diagnosed with breast carcinoma. Patients diagnosed with invasive breast cancer with TACS-3 phenotype was associated with poor disease-specific and disease-free survival. This biomarker was also confirmed to be an independent prognostic indicator regardless of tumor grade, size, and estrogen receptor, progesterone receptor, or human epidermal growth factor receptor-2 (HER-2) immunohistochemical status. Surprisingly, TACS-3 was also prognostically independent of lymph node status (Conklin et al. 2011). It is noteworthy that in the model using mice, breast tumors were all of the luminal molecular subtype (Hollern and Andrechek 2014). In addition, in women with the luminal subtype of invasive ductal carcinoma, higher deposition of collagen was associated with poorer prognosis (Natal et al. 2018b), (Fig. 4), including lymph node involvement (Kakkad et al. 2012).

Fig. 4
figure 4

Two cases of invasive ductal carcinoma of the breast. Images a and c show, respectively, two-photon-excited fluorescence microscopy (TPEF) and second harmonic generation (SHG) of a patient (P1) who presented disease progression (metastasis/recurrence). b and d show, respectively, TPEF and SHG of a patient (P2) whose disease did not progress. In graphs E and F, the black line corresponds to patient P1 and the red line to patient P2. Notice that in graph e collagen fibers are arranged between − 70° to − 20° (variation of 50°) for patient P1 and between − 50° to 60° (variation of 110°) for patient P2. In graph f it is shown that patient P1 had total amount of collagen fibers greater than patient P2. This is in line with previous evidence that intratumoral collagen with less angulation, that is, more organized, and that greater amounts of intratumoral collagen are associated with higher risk of disease progression

Of note, changes in collagen quantity and organization were seen when areas of invasive ductal carcinoma were compared with areas of breast tissue without alterations (Natal et al. 2018b), as well as within the same areas along tumor progression (Brabrand et al. 2015). In addition, polarimetric SHG succeeded to reveal ultrastructural disorganization in the focal volume amongst the different molecular subtypes of breast carcinoma, therefore it could be used for the molecular classification (Golaraei et al. 2016).

In special types of invasive breast cancer, tubular carcinoma and invasive lobular carcinoma presented higher collagens fibers organization, generally these fibers are arranged parallel to the tumor invasion. Although, medullary, mucinous and papillary carcinomas showed an intense collagen fibers disorganization (Adur 2012; Natal et al. 2019).

Ductal carcinoma in situ and therapeutics agents also demonstrated changes in collagen parameters (Conklin et al. 2018) (Walsh et al. 2015; Wu et al. 2018). In ductal carcinoma in situ, radially deposition of collagen fibers was associated with a higher recurrence rate, while in chemotherapy, there was a disorganization of collagen fibers. Such changes may shed light in the mechanisms of tumor progression as well as of response to therapy (Bredfeldt et al. 2014).

In other breast lesions, fibroadenoma presented greater amounts of collagen than healthy breast tissue (Nie et al. 2015; Zheng et al. 2011) or phyllodes tumors (Tan et al. 2015). The difference in collagen deposition allowed distinction between these lesions with up to 71.4 and 84.4% sensitivity and specificity, respectively.

Ovary cancer

One of the earliest studies addressing ovarian cancer has demonstrated that neoplastic stroma diverged from healthy tissue. Women at high risk of developing ovarian cancer (i.e., BRCA1 or BRCA2 mutated) presented abnormal changes in the structure of ovarian collagen, similar to those found in neoplasia itself (Kirkpatrick et al. 2007).

Mathematical models applied to images satisfactorily distinguished healthy from neoplastic ovarian stroma (Tilbury et al. 2017; Wen et al. 2014). In this context, ovaries with malignancy could be characterized by lower cell density, denser collagen, as well as higher regularity at both fibril and fiber levels (Nadiarnykh et al. 2010). Further, this method has been applied not only to differentiate adenomas from adenocarcinomas, but also for histological subtyping, such as to distinct serous from mucinous tumors (Adur et al. 2014; Adur et al. 2012; Campbell and Campagnola 2017; Zeitoune et al. 2017).

A study in mice has demonstrated that tumor evolution is marked by progressive alteration of collagen. Normal ovary showed very dense collagen at the surface, especially at what appeared to be the epithelial-stromal boundary. While collagen fibers at the surface were thin, linear and interweaved forming a tight-knit; stromal collagen, beyond these features, showed gaps (up to 400 μm for larger follicles) in areas with functional ovarian structures (e.g. follicles and corpora lutea). Atrophic ovary is quite similar, however the gaps were smaller, corresponding to small cysts. Tubular hyperplasia and granulosa cell tumor showed less collagen, former presented spaced far apart tangled appearance with clumps and latter presented revealed less tightly packed. Further, both cystic tumor and fibrosarcoma presented more linear to wavy collagen fibers, differentiating by slightly thicker collagen fibers and more spread out or tangled compared to normal ovary, respectively (Watson et al. 2014).

In addition to 2D images, 3D images can provide more details of collagen structure. This improvement allowed distinction between conditions, as normal from high risk ovarian stroma, benign from malignant tumors, low grade from high grade neoplasms, serous tumors from endometrioid tumors (Wen et al. 2016).

Gastrointestinal pathology

Upper gastrointestinal tract

Esophagus healthy submucosa was characterized by larger amounts of aligned collagen organized in large and spiral bundles. However, esophageal adenocarcinoma and esophageal squamous cell carcinoma submucosa was described by fine and distorted collagen fibers, with disorganized little collagen fibers (Chen et al. 2014; Xu et al. 2017).

Of note, in Barrett’s esophagus significant changes in collagen structure of the basal layer were found, with increased amount of collagen in the basement membrane, resulting in increased thickness (Mehravar et al. 2016). The progression of dysplasia severity was also associated to increase of basement membrane thickness (Mehravar et al. 2016).

In gastric cancer, collagen structure was evaluated according to the extent of invasion, and it was shown that cases with serosa infiltration by tumor were associated with reduction in amount and increase in disorganization of collagen fibers (Yan et al. 2016). Besides, higher values of collagen signature, based on parameters like quantity and alignment of collagen fibers, predicted early lymph node involvement (Chen et al. 2019).

Pancreas and liver

Chronic pancreatitis presented a variable collagen structure, with some ducts resembling normal ones, and others, with more abundant collagen, were similar to malignant ducts. In opposition, pancreatic ductal adenocarcinoma showed increased alignment, length, and width collagen around malignant ducts. Collagen fibers varied also with respect to tumor cellularity; fibers were sparsely deposited in highly cellular regions, whereas highly aligned collagen was seen in poor cellular areas. In addition, high alignment of collagen was associated with poor prognosis (Drifka et al. 2015). It was also observed a positive association between high collagen alignment and the increased expression of alpha-smooth muscle actin, which was related to poor prognosis (Drifka et al. 2016a, b).

In liver, the amount of collagen present in hepatocellular carcinomas (HCC) samples could distinguish well-differentiated tumors (histological grade I) from moderately to poorly differentiated tumors (histological grades II and III) (Lin et al. 2018). Histological grade I displayed low amounts of collagen fibers, which were scattered, in a slender strip, similar to the normal liver. In histological grade II, there was an increase in the amount of collagen and the bundles became longer and thicker. In histological grade III there was an even greater increase in the amount of collagen and its deposition was heterogeneous throughout the tumor extension (Lin et al. 2018). Further, COL1A1 was significantly upregulated in HCC tumor tissues in comparison to normal tissues and conferred survival advantage and enhanced oncogenicity, promoting epithelial-to-mesenchymal transition (EMT) (Ma et al. 2019).

Colon and rectum

In colon, SHG imaging revealed different patterns in the outlines of basement membrane throughout normal-to-cancer progression. Normal mucosa presented a honeycomb arrangement of round-shaped regular basement membrane, with a concentrated collagen fiber in a submucosal layer, while precarcinomatous lesions revealed larger size and lower density of the tubular-shaped basement membrane. In cancer, basement membrane was missing (Maier et al. 2021; Xia et al. 2017; Zhuo et al. 2012) and collagen fibers of basement membrane were more disorganized and diminished, assuming varied directions as tumor invaded (Liu et al. 2013; Qiu et al. 2015).

In rectal cancer, cases with non-compromised surgical margins were characterized by collagen fibers around the glands, while in cases with positive surgical margins there was significant decrease or loss of such fibers (Yan et al. 2014). Besides, quantification of collagen structure has shown also potential of monitoring response to neoadjuvant chemotherapy in colorectal cancer, since the adequate response to therapy was accompanied by a greater deposition of collagen (Li et al. 2017).

Urogenital pathology

Deposition of higher amount and more linear collagen fibers throughout the tumor was associated with worse prognosis in invasive bladder cancer (Brooks et al. 2016) and higher histological grade in renal cell carcinoma (Best et al. 2019).

Prostate cancer, in opposition to prostate benign lesions, showed change in collagen pattern from papillary to reticular. Further, with the increase in Gleason score, collagen fibers presented higher orientated and stiffened (Ling et al. 2017; Yuting et al. 2018). Other study demonstrated that shorter collagen fibers and loss of fiber alignment was correlated with unfavorable prognosis (Sridharan et al. 2015).

Lung pathology

Collagen fiber distribution and SHG intensity of fibers vary in different areas of the normal lung. Thus, large areas of the organ should be scanned to allow distinction between tumor and non-tumor regions using SHG microscopy.

Lung squamous cell carcinoma showed fewer elastin and collagen fibers compared to normal tissue. In desmoplastic areas, however, significantly larger total amounts and longer fibers of collagen were detected (Xu et al. 2013). Besides, amount of collagen fibers and the spread of fibers towards the dominant direction represented an indicator of disorder in tumor matrix (Golaraei et al. 2014), promoting higher stiffness and EMT (Pankova et al. 2019).

Brain

Glioblastoma xenograft model has shown that in normal brain parenchyma only low amounts of collagen exist, mainly around part of the blood vessels. Differences in focal and invasive tumors were observed: (1) in focal tumors, collagen appeared to encapsulate neoplasia at the transition with normal tissue, and (2) in invasive tumors, collagen appeared intertwined within the tumor cell groups. Focal tumors showed more organized collagen, based on the presence of more aligned, longer, wider and straighter collagen fibers, than animals with invasive tumors (Pointer et al. 2017). In humans, tumors with less organized collagen were associated with unfavorable outcomes (Pointer et al. 2017).

Thyroid pathology

In normal thyroid tissue collagen was deposited in a fine and linear pattern, while in neoplasia fibers were wavy and thick (Huang et al. 2010). In addition, thyroid carcinoma presented greater disorganization of the fibers, both at structural and ultrastructural levels (Tokarz et al. 2015). Further, collagen deposition pattern in the thyroid capsule could distinguish between benign and malignant nodules, i.e., follicular adenoma and papillary thyroid carcinoma (Hristu et al. 2018). Collagen fiber organization was similar between two histological variants of papillary thyroid carcinoma, the classical and the follicular, both with similar prognoses (Passler et al. 2003; Yu et al. 2013).

Dermatology pathology

In melanoma, collagen evaluation might represent a rapid and reliable method for defining the borders of the lesions and the extent of dermal invasion (Breslow thickness), as collagen density in the transition from melanoma to unaltered tissue was shown to gradually increase. Additionally, the evaluation of collagen fiber structure presented high degree of concordance with the evaluation of H&E and Melan-A stained tissue specimens by the pathologists (Thrasivoulou et al. 2011). Clusters of malignant melanocytes were associated with absence of collagen, probably due to its destruction by metalloproteinases. In areas related to melanoma, collagen morphology was greatly affected, namely, with very short and thin fibers. SHG signals gradually intensified from the transition of neoplasia to normal tissues (Thrasivoulou et al. 2011) (Fig. 5). In a melanoma xenograft model, it was also demonstrated that the amount and orientation of collagen fibers diminished as melanoma progressed (Wu et al. 2015). Similar collagen patterns were described in squamous and basal cell carcinomas (Heuke et al. 2013; Lin et al. 2006).

Fig. 5
figure 5

Melanoma specimen. Yellow dashed line represents cluster of neoplastic cells. Transition with neoplastic cell aggregates shows reduced amount of collagen. Stroma more distant to tumor presents increasing amounts of collagen

Discussion and conclusion

The present manuscript comprises some of the applications of SHG microscopy in Oncologic Pathology, mainly in the study of ECM of epithelial tumors. Herein, it was shown how collagen parameters measured by this technique could aid in the differential diagnosis and in prognostic stratification of different cancer types. There is evidence for the association between higher amount, lower organization and higher linearity of collagen fibers, and tumor progression and metastasizing.

Although increased deposition of collagen fibers may be a predictor of progression (Drifka et al. 2015; Kakkad et al. 2012; Pal et al. 2015; Vargas et al. 2009), it should be noted that post therapeutic tissue changes and desmoplastic reaction may mimic such an increase in collagen fibers (Li et al. 2017). Therefore, other parameters of collagen, as alignment and organization, are essential for a more specific evaluation of collagen structure deposited. Such collagen parameters are better evaluated by SHG microscopy. The explanation on the mechanisms by which structure is associated with biological behavior could be explored in functional studies, briefly discussed in the following paragraphs.

It was demonstrated that high collagen fiber linearity (stretched collagen fibers) enhanced motility of neoplastic cells, by increasing directional persistence and restricting protrusions along aligned fibers, thus promoting greater distance traveled by cancer cell (Riching et al. 2014). Experimental modulation of collagen fibers organization through an anti-lysyl oxidase-like-2 antibody evidenced that ECM composition remained unaltered, while the resulting alteration in collagen fibers structure interfered with neoplastic cell adhesion and invasion properties, reducing primary tumor growth (Grossman et al. 2016). Also, it was indicated that expression of collagen prolyl-hydroxylases promoted cancer cell alignment along collagen fibers, resulting in enhanced invasion and metastasis to lymph nodes and lungs. Finally, it was established that the expression of collagen prolyl-hydroxylase mRNA in biopsies of human breast cancer was associated with prognosis, and that ethyl-3,4-dihydroxybenzoate, a prolyl-hydroxylase inhibitor, was associated with decreased tumor fibrosis and metastasizing in a mouse model of breast cancer (Gilkes et al. 2013). Moreover, cancer-associated adipocytes (CAAs) were shown to remodel collagen alignment in crosstalk with breast cancer cell, further promoting breast cancer metastasis. Tumor derived plasminogen activator inhibitor-1 (PAI-1) was required to activate the expression of the intracellular enzyme procollagen-lysine, 2-oxoglutarate 5-dioxygenase 2 (PLOD2) in CAAs (Wei et al. 2019). Further, tumor-associated macrophages (TAMs) regulate tumor sustained growth by secreting type I collagen, which can activate prosurvival signals produced by the integrin α2β1/PI3K/AKT signaling pathway (Qiu et al. 2019).

Overall literature data on breast cancer point to the association between higher deposition of collagen and worse prognosis, regardless of the model used (i.e., mice (Provenzano et al. 2008), dog (Case et al. 2017) and human (Natal et al. 2018b)). One possible explanation is that increase in collagen density might reduce activation of STAT5 in neoplastic cells, leading to an increase in phosphorylated ERK 1/2 and Akt proteins that, consequently, would promote tumor growth (Barcus et al. 2017). In this line, the association between increased breast cancer risk and obesity could be at least partially explained by the increased collagen deposition and stiffness in the breast of obese patients (Le et al. 2007). In contrast, reduction of linearity and rigidity of collagen promoted in the nursing breast, could be related to its reduced risk for cancer (Maller et al. 2013). Further, COL1A1 was up-regulated in microinvasive breast carcinoma and its knockdown significantly inhibited the proliferation, migration, and invasion by inhibiting EMT process and the TGF-β signaling pathway (Zhu et al. 2019).

Additionally, it was observed that collagen fibers alignment conferred increased stiffness, thus promoting EMT by an essential mechanomediator, TWIST1, a basic helix-loop-helix transcription factor (Riching et al. 2014; Wei et al. 2015). In this case, high matrix stiffness promoted nuclear translocation of TWIST1 by releasing it from its cytoplasmic binding partner G3BP2. Loss of G3BP2 leads to constitutive TWIST1 nuclear localization and synergizes with increasing matrix stiffness to induce EMT and promote tumor invasion and metastasis. In human breast tumors, collagen fiber alignment, a marker of increasing matrix stiffness, together with reduced expression of G3BP2 predicted poor survival. It was revealed that a TWIST1–G3BP2 mechanotransduction pathway, responsive biomechanical signals from the TME, could drive EMT, invasion, and metastasis 105. Further, it was shown that the oriented fibers greatly enhance and facilitate the metastatic cell intravasation process during metastasis (Han et al. 2016).

In a human non-small cell lung carcinoma cell line model in collagen gel, inhibition of LKB1 or MARK1 increased collagen fiber alignment, that is, increased tumor invasiveness (Lee et al. 2017). Mutations of LKB1 are detected in various tumors. This protein interacts with MARK1 to regulate collagen remodeling and cancer cell motility, thus explaining their influence on tumor behavior. This is probably related to another mechanism of cell migration, which is independent of soluble factors, the Rho-associated kinase (ROCK)-mediated contractility, required for collagen alignment and cell migration (Drifka et al. 2016a, b). In other experiment using human lung adenocarcinoma it was described that promoter methylation of RASSF1A, a key regulator of the Hippo pathway, promoted nuclear accumulation of YAP1, and expression of prolyl 4-hydroxylase alpha-2 (P4HA2) (Pankova et al. 2019). P4HA2 is known to have a major effect on physical properties of tumor-associated ECM, which in turn leads to increased organization and stiffness of collagen structure during cancer progression (Levental et al. 2009b; Provenzano et al. 2006b). Further, it was identified that elevated collagen creates a stiff ECM which in turn triggers cancer stem-like programming and metastatic dissemination. Although, re-expression of RASSF1A or inhibition of P4HA2 activity reverses these effects and increases markers of lung differentiation (TTF-1 and Mucin 5B) (Pankova et al. 2019).

The hepatitis B virus X protein (HBx) is essential for HBV replication and is thought to play a major role in HCC (Kremsdorf et al. 2006). HBx expression potentiates liver fibrosis with increased expression of proteins involved in matrix remodeling, such as collagen, and inflammatory cytokines, such as TNF-α. This correlated with a higher expression of tumor progenitor cell markers (AFP, Ly6D and EpCam), indicating a higher risk of progression from fibrosis to HCC (Ahodantin et al. 2019).

However, it was shown that ionizing radiation could reduce the stiffness collagen matrix. When non-irradiated cancer cells were seeded in an irradiated matrix, adhesion, spreading, and migration were reduced (Miller et al. 2018). In contrast, there are cases where the greater amount of collagen fibers deposited or their greater alignment is related to a better prognosis, and their loss may be associated with tumor progression. This is especially true for glioblastoma (Pointer et al. 2017), a tumor with important differences in comparison with epithelial tumors.

In conclusion, cancer progression and metastasizing represent complex processes, in which matrix remodeling plays a central role, together with cancer cell genetic modifications. Integration of studies on cancer cell biology and ECM are highly advantageous to give us a more complete picture of these processes. As microscopic techniques provide topographic information allied with biologic characteristics of tissue components, they represent important tools for a more complete understanding of cancer progression. In this context, SHG has provided significant insights in human tumor specimens, readily available for Pathologists.

Availability of data and materials

The present study represents an overview of specialized medical literature, with articles retrieved by Medline search.

Abbreviations

CAA:

Cancer-associated adipocytes

ECM:

Extracellular matrix

EMT:

Epithelial-to-mesenchymal transition

HCC:

Hepatocellular carcinoma

HER-2:

Human epidermal growth factor receptor 2

P4HA2:

Prolyl 4-hydroxylase alpha-2

PAI-1:

Plasminogen activator inhibitor-1

PLOD-1:

Procollagen-lysine, 2-oxoglutarate 5-dioxygenase 2

ROCK:

Rho-associated kinase

SHG:

Second-harmonic generation

TACS:

Tumor associated collagen signature

TAM:

Tumor-associated macrophages

THG:

Third-harmonic generation

TMA:

Tissue microarray

TME:

Tumor microenvironment

TPEF:

Tow-photons excited fluorescence

References

Download references

Funding

Not applicable.

Author information

Authors and Affiliations

Authors

Contributions

All authors have participated in the selection of the articles cited herein, and in the manuscript writing; all have approved its final version.

Corresponding author

Correspondence to Rodrigo de Andrade Natal.

Ethics declarations

Ethics approval and consent to participate

The study is fully compliant with the Declaration of Helsinki (approved by the “Comitê de Ética em Pesquisa da Unicamp,” approval number 087/2008).

Consent for publication

All authors have participated in the selection of the articles cited herein, and in the manuscript writing; all have approved its final version.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

de Andrade Natal, R., Adur, J., Cesar, C.L. et al. Tumor extracellular matrix: lessons from the second-harmonic generation microscopy. Surg Exp Pathol 4, 7 (2021). https://doi.org/10.1186/s42047-021-00089-0

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s42047-021-00089-0

Keywords

  • Second-harmonic generation microscopy
  • Collagen
  • Tumor extracellular matrix
  • Tumor microenvironment
  • Oncologic pathology