Apoptosis

, Volume 15, Issue 9, pp 1050–1071

Photodynamic therapy: illuminating the road from cell death towards anti-tumour immunity

Authors

  • Abhishek D. Garg
    • Department of Molecular Cell Biology, Faculty of MedicineCatholic University of Leuven
  • Dominika Nowis
    • Department of Immunology, Center of Biostructure ResearchMedical University of Warsaw
    • Department of Immunology, Center of Biostructure ResearchMedical University of Warsaw
    • Institute of Physical ChemistryPolish Academy of Sciences
    • Department of Molecular Cell Biology, Faculty of MedicineCatholic University of Leuven
Clearance of dead cells: mechanisms, immune responses and implication in the development of diseases

DOI: 10.1007/s10495-010-0479-7

Cite this article as:
Garg, A.D., Nowis, D., Golab, J. et al. Apoptosis (2010) 15: 1050. doi:10.1007/s10495-010-0479-7

Abstract

Photodynamic therapy (PDT) utilizes the destructive power of reactive oxygen species generated via visible light irradiation of a photosensitive dye accumulated in the cancerous tissue/cells, to bring about their obliteration. PDT activates multiple signalling pathways in cancer cells, which could give rise to all three cell death modalities (at least in vitro). Simultaneously, PDT is capable of eliciting various effects in the tumour microenvironment thereby affecting the tumour-associated/-infiltrating immune cells and by extension, leading to infiltration of various immune cells (e.g. neutrophils) into the treated site. PDT is also associated to the activation of different immune phenomena, e.g. acute-phase response, complement cascade and production of cytokines/chemokines. It has also come to light that, PDT is capable of activating ‘anti-tumour adaptive immunity’ in both pre-clinical as well as clinical settings. Although the ability of PDT to induce ‘anti-cancer vaccine effect’ is still debatable, yet it has been shown to be capable of inducing exposure/release of certain damage-associated molecular patterns (DAMPs) like HSP70. Therefore, it seems that PDT is unique among other approved therapeutic procedures in generating a microenvironment suitable for development of systemic anti-tumour immunity. Apart from this, recent times have seen the emergence of certain promising modalities based on PDT like-photoimmunotherapy and PDT-based cancer vaccines. This review mainly discusses the effects exerted by PDT on cancer cells, immune cells as well as tumour microenvironment in terms of anti-tumour immunity. The ability of PDT to expose/release DAMPs and the future perspectives of this paradigm have also been discussed.

Keywords

Photodynamic therapyDAMPsImmunologyAntitumour immunityNeutrophilsImmunotherapy

Abbreviations

AIF

Apoptosis-inducing factor

ALA

Amino levulinic acid

APCs

Antigen-presenting cells

APR

Acute-phase response

ATP

Adenosine triphosphate

CD

Cluster of differentiation

DAMP

Damage-associated molecular patterns

DC

Dendritic cell

ER

Endoplasmic reticulum

GRP

Glucose-regulated protein

HIF

Hypoxia-inducible factor

HMGB1

High-mobility group box-1

HSP

Heat shock protein

IFN

Interferon

IL

Interleukin

MAPK

Mitogen-activated protein kinase

MHC

Major histocompatibility complex

MMP

Mitochondrial membrane permeabilization

NK cells

Natural killer cells

NO

Nitric oxide

PDT

Photodynamic therapy

RIP1

Receptor interacting protein 1

ROS

Reactive oxygen species

TAA

Tumour-associated antigen(s)

TGF

Transforming growth factor

TLR

Toll-like receptor(s)

TNF

Tumour necrosis factor

Introduction

Photodynamic therapy (PDT) is a multi-step procedure involving the selective uptake of a photosensitiser (usually porphyrins, porphyrin analogs or other agents with suitable photophysical properties) by the tumour tissue, followed by illumination of the neoplastic lesion with a light of appropriate wavelength able to trigger photochemical reactions that lead to the generation of singlet oxygen (1O2) and other reactive oxygen species (ROS) [1]. Photogenerated ROS cause damage at sites of photosensitiser accumulation and ultimately lead to tumour destruction. Clinically used agents in PDT can associate with plasma membrane or intracellular membranes of the endoplasmic reticulum (ER), mitochondria, lysosomes, or combinations of these sites [1]. Since the nucleus is not a primary site of photosensitisers’ localization this anticancer therapy is thought to be less genotoxic as compared with conventional chemotherapy or radiotherapy. The extent of photodamage and cytotoxicity after PDT in vivo is multifactorial and can depend on the type of photosensitizing agent used, its subcellular localization, the time between administration of the photosensitiser and irradiation as well as on different illumination conditions. Moreover, also the type of tumour and its level of oxygenation are determining factors.

PDT: basic mechanisms of anti-tumour activity

PDT-based anti-tumour effects are multifactorial and include: (1) direct tumour cell kill, (2) damage to the vasculature, (3) cytotoxic effects towards tumour-infiltrating immune cells and (4) rapid recruitment and activation of immune cells that can facilitate development of anti-tumour adaptive immunity (see Fig. 1) [13]. PDT is not particularly selective in terms of the cytotoxic effects exerted against the cells that form the tumour. Therefore, both transformed as well as normal cells that form the tumour stroma are lethally damaged (Fig. 1) [25]. Simultaneously, PDT stimulates recruitment of immune cells into the treated area leading to development of a local inflammatory reaction, which sometimes is accompanied by induction of systemic neutrophilia [25]. It has also come to light recently that PDT can induce local microenvironmental changes that facilitate development of anti-tumour adaptive immune response (Fig. 1).
https://static-content.springer.com/image/art%3A10.1007%2Fs10495-010-0479-7/MediaObjects/10495_2010_479_Fig1_HTML.gif
Fig. 1

Schematic representation of PDT-induced effects on tumour cells, tumour tissue and immune system. Light-based illumination of photosensitiser-loaded tumour/cancer cells leads to production of reactive oxygen species (ROS) within these cells thereby leading to cell death (predominantly apoptotic and necrotic). This cell death is usually accompanied by release/exposure of damage-associated molecular patterns (DAMPs). Tumour cell kill is further augmented by PDT-induced vascular shut-down, which basically pushes the tumour towards starvation. This is simultaneously accompanied by, (1) direct effects of PDT on tumour-associated/-infiltrating cells like macrophages, (2) orchestrating of immune cells recruitment manoeuvres, e.g. high neutrophilia as well as dendritic cell (DCs) chemoattraction and (3) miscellaneous inflammation-related processes. PDT tends to generate a lot of tumour debris which is phagocytosized by various immune cells and probably processed further. Here, DCs being the principle antigen-presenting cells could take up the tumour-associated antigens (TAAs), become mature and present the same to the cells adaptive immune system in order to instigate an anti-tumour immune response. Also, the production of various cytokines and chemokines in the tumour microenvironment plays an important role in modulating a plethora of innate and adaptive immune processes

PDT and tumour/cancer cell kill: mechanisms of cell death

Over the last decades, along with a rising interest in PDT as a promising anticancer treatment, numerous in vitro as well as in vivo studies have been documented on the ability of PDT to induce direct tumour/cancer cell kill (i.e. photokilling). It is now accepted that photokilling (induced in cultured cells) may involve all three main cell death morphologies described, i.e. apoptotic, necrotic and autophagic cell death [1, 6]. Furthermore, new imaging technologies, which have been made available in the last years have confirmed indeed that apoptosis is a major pathway of PDT-mediated cell death in vivo, thus justifying major efforts bestowed in understanding the biochemistry of the apoptotic photokilling process [710]. The mechanisms and signalling pathways activated by PDT are remarkably complex and have been the subjects of recent extensive reviews [1, 11, 12].

The induction of apoptosis by PDT entails the activation of two apoptotic cascades known as the extrinsic or death-receptor (DR) and the intrinsic or mitochondrial pathways [1]. The involvement of DR-mediated signalling appears to be engaged mainly following the release of cytokines by the PDT-stressed or dying cells [1]. On the other hand, apoptotic photokilling has been shown to involve the mitochondrial pathway in virtually all PDT paradigms studied so far [1]. Induction of intrinsic apoptotic photokilling critically depends on either direct mitochondrial damage or is secondary to signalling pathways activated by the photodynamic injury to other subcellular sites. As in other apoptotic paradigms mitochondrial membrane permeabilization (MMP) after PDT is a crucial lethal event, which is tightly controlled by Bcl-2 family members and is thought to be largely p53-independent [1]. In this pathway, multidomain Bax and Bak pro-apoptotic proteins are essential for apoptotic photokilling, since their genetic loss in cancer cells or in fibroblasts inhibits all of the canonical hallmarks of apoptosis [13, 14], whereas rescuing Bax expression is both necessary and sufficient to fully restore caspase activation and apoptotic photokilling. PDT has been shown to directly damage mitochondria-associated anti-apoptotic Bcl-2 proteins [15], thus facilitating Bax/Bak-mediated MMP and the subsequent release of caspase activators, such as cytochrome c and Smac/DIABLO or other pro-apoptotic molecules, including apoptosis-inducing factor (AIF). On the other hand, when primary photodamage involves predominantly other organelles, such as the ER or the lysosomes, different pathways involving the activation of upstream pro-apoptotic BH3-only proteins is required for MMP and apoptotic photokilling [1]. Lysosomal photosensitisers may activate intrinsic apoptosis following the release of cathepsins from photodamaged lysosomes resulting in the cleavage of the BH3-only Bid and consequent MMP [1]. With ER-associated photosensitisers, PDT has been shown to cause the up-regulation of several ER chaperones, including glucose-regulated protein/binding protein (GRP78/Bip), GRP94 and protein disulfide isomerase (PDI), along with activation of the ER stress response [1]. In this pathway CCAAT/enhancer-binding protein (C/EBP) homologous protein (CHOP), a key pro-apoptotic transcription factor activated following severe ER stress may convey MMP through the expression of the BH3-only pro-apoptotic Bim [16].

While caspase are critical mediators of apoptosis following PDT, caspase inhibition usually delays photokilling without preventing it, thus indicating that PDT mediates cell death also through caspase-independent or non-apoptotic pathways. As observed in other cellular apoptotic paradigms, caspase inhibition or genetic deficiency in key components of the apoptotic machinery converts the cell death modality into necrosis [1]. Despite the idea that necrosis is provoked passively by enhancing the intensity of the photodynamic injury, certain forms of necrotic photokilling can be modulated [17], and should therefore be considered as propagated through signal transduction pathways. Although the molecular mechanisms underlying programmed necrosis are still elusive, activation of receptor interacting protein 1 kinase (RIP1), severe mitochondrial ROS production, lysosomal damage and intracellular Ca2+ overload are molecular events associated with necrotic cell death [17]. Although the involvement of RIP1 in PDT has not been explored yet, severe mitochondrial damage with loss of mitochondrial inner membrane permeability, increased ROS production accompanied by a drastic drop in ATP production, intracellular Ca2+ overload and ROS-mediated caspase inhibition are all PDT-induced processes that could favour necrotic rather than apoptotic photokilling [1]. The observations that typical features of apoptotic and necrotic cell death coexist in the same photosensitized cell suggest that during photokilling there is a continuum of apoptosis and necrosis. In spite of this complexity, clarification of the molecular events favouring necrotic signalling in photosensitized cells could be particularly important, when considering the development of pharmacological strategies able to modulate the pro-inflammatory effects associated with necrotic photokilling in vivo.

Additionally, the mechanisms of photokilling have been recently shown to be influenced by the stimulation of macroautophagy (hereafter referred to as autophagy) in the photosensitized cells [1, 18]. Autophagy is a lysosomal degradation pathway for the bulk destruction and turnover of long-lived proteins and damaged organelles, which is stimulated by various cellular stresses including oxidative stress. Recent studies on the role of this catabolic pathway in mammalian cells indicate that autophagy can activate both pro-survival mechanisms and act as a major quality control pathway for the clearance of damaged proteins and organelles, as well as lethal programs, especially under conditions of enduring organellar damage. Given this functional dichotomy, activation of autophagy may either impede or facilitate PDT-mediated cell killing.

Based on the recent observations indicating that the knock-down of essential autophagy genes (i.e. atg7, atg5) in cancer cells lowers the threshold for apoptotic photokilling, it can be postulated that autophagy is a mechanism to preserve cell viability following photodynamic injury [1, 18]. As a consequence, blocking autophagy pathways could increase the therapeutic benefit of PDT.

On the other hand, defects in key components of the apoptotic machinery have been shown to burst autophagy in photosensitized cells and promote a form of ‘autophagic cell death’ with a necrotic morphology [18]. In this case, enhancing autophagic degradation may provide a backup mechanism of photokilling in cancer cells with defects in apoptotic signalling [1]. The mechanisms shifting the intrinsic cytoprotective role of autophagy into a pro-death process are currently not known and are being intensively investigated. Nonetheless, these studies suggest that the functional outcome of autophagy after PDT may be crucially influenced by the availability of functional apoptotic machinery. Definitely, a better understanding of the interplay between autophagy, apoptosis and necrosis at the intracellular level, along with increased knowledge of their impact on the development of the immune response, are requisite to devise better therapeutic strategies in PDT.

PDT and its effects on tumour micro-environment: an overview

PDT-inflicted insult usually results in a massive tissue injury that initiates complex host-tumour interactions that from an evolutionary perspective are cytoprotective and aimed at confining the trauma to prevent further spreading or infection. With some oversimplification, PDT triggers a host response that resembles wound healing. The complex reactions of the host dictate the extent of the damage playing either protective or destructive role. For example, PDT-induced damage to the tumour vasculature initiates robust angiogenesis that might facilitate survival of residual tumour cells. Hence, not surprisingly, antiangiogenic or antivascular therapies combined with PDT were shown to markedly improve local tumour control [19, 20]. During effective PDT, over 90% of tumour cells become lethally damaged within several hours after illumination, which in a long run is insufficient for tumour eradication. Therefore, cytotoxicity alone cannot account for frequently observed complete and long-lasting antitumor responses (post PDT) and hence, other indirect antitumor mechanisms must be involved. Comparison of the antitumor effects of PDT in normal and immunodeficient mice revealed that despite comparable short-term outcomes complete antitumor responses were observed only in immunocompetent animals [21] indicating that the engagement of the immune system is what makes PDT so effective.

It has been observed that, changes in the tumour microenvironment induced by PDT seem to profoundly affect the activity of innate immune cells. PDT leads to a rapid oxygen consumption [22, 23], although this effect seems to be dependent on the fluence rate of light [24]. Together with the destruction of blood vessels this treatment favours development of profound hypoxia within illuminated tumour [25]. A stress triggered by low oxygen levels activates hypoxia-inducible factor (HIF), a transcription factor that regulates the expression of more than 100 genes [26]. Several preclinical and clinical studies have shown that PDT can activate HIF-driven responses in tumour microenvironment [20, 27, 28], and recent observations indicate that myeloid cells are evolutionarily adapted to execute their effector functions in hypoxic microenvironment [29]. Markedly lowered levels of oxygen (<1%) are typically found within wounds and at sites of infection as compared with normal conditions and conditional knock-out of HIF1α gene in myeloid cells impairs their phagocytic and bactericidal capacities [30]. Also secretion of numerous inflammatory mediators such as TNF, IL-1, IL-12, nitric oxide (NO) or expression of Toll-like receptors (TLR) are regulated by HIF in myeloid cells [31, 32]. Finally, HIF regulates production of cytokines and co-stimulatory molecules in dendritic cells, thereby enhancing their ability to present antigens and induce lymphocyte proliferation [30, 32]. Therefore, hypoxia seems to be a suitable environment that bridges innate and adaptive immune responses [29]. In the later sections, the impact of PDT on immune cells and tumour cells in terms of immunomodulation will be discussed in details.

PDT and anti-tumour immunity: past and present trends in pre-clinical and clinical research

Pre-clinical research in various animal models has contributed immensely to our understanding of PDT-induced host immune responses as well as direct local cellular and vascular effects [33]. Most of these pre-clinical studies have involved utilization of transplantable tumours grown in syngeneic rats or mouse strains [2]. Preclinical studies have also convincingly established that PDT can also promote ‘anti-cancer vaccine effect’ and due to this the concept of ‘PDT-based vaccines’ has materialized strongly. The most prominently used rodent models in pre-clinical PDT research have been briefly described in Table 1.
Table 1

Most prominently used rodent models in PDT research and their main characteristics

Mice model

Main/principal characteristics

Refs.

Nude mice

Nude mice are hairless mice possessing congenital absence of a thymus (because of homozygous Foxn1nu mutationa) which severely affects the number of mature T cells in their body. However, the B cells and NK cell levels are normal. Levels and activity of myeloid cells are also normal however epithelial cell development is abnormal

[134, 135]

SCID mice

SCID (Severe Combined Immune Deficiency) mice have a homozygous mutation in Prkdcscid genea which encodes for phosphatidylinositol kinases (PIK)-related kinase, a recombinase that plays an important role in rejoining of double-stranded DNA breaks. This defect affects the process of V(D)J recombination thereby affecting the rearrangement of T-cell receptor genes as well as immunoglobulin genes, which in turn hampers normal T and B cell maturation. Thus, SCID mice lack any mature T or B cells. However, occasional low levels of CD4+ cells might be found. Levels and activity of myeloid cells is normal

[135, 136]

BALB/c mice

These albino mice are the most widely used general purpose strain in laboratory for various kinds of in vivo studies. In immunological terms, they tend to show resistance to experimental—allergic encephalomyelitis (EAE), autoimmune thyroiditis and autoimmune prostatitis. These mice tend to be susceptible to ethyl nitrosourea based mutagenesis

[137141]

C3H mice

These are agouti mice which are also amongst the most widely used mouse strains. These mice tend to be susceptible to ethyl nitrosourea based mutagenesis. In case of PDT, the C3H/HeN sub-strain is widely used. This strain tends to show a high incidence of hepatomas. In immunological terms, C3H/HeN mice tend to show lower leukocyte count and susceptibility to experimental autoimmune orchitis

[141144]

Fischer rat

Fischer rat has become a very commonly used rat model for carcinogenesis tests, over a period of last two decades. Most important factors behind its popularity are—longevity, resistance towards infectious diseases and low incidence of spontaneous neoplasms. In case of PDT, Fischer rat has been used as a model for in vivo Hypericin-PDT studies

[145, 146]

aJAX Mice Database, URL: http://jaxmice.jax.org (as on 15th Dec 2009)

PDT-based activation of anti-tumourigenic adaptive immune response: a bird’s eye view

Studies performed in various murine models have divulged considerable evidence in the direction of immune system activating capabilities of PDT [3]. Research involving various alteration tactics like impairing the function of a variety of immune cells, compromising the activity of chemokines/cytokines and modulating the activity of complement system or coagulation cascade; has shown that intervention with the host immune system components tends to introduce dramatic variations in tumour cure rates [3]. One of the first studies highlighting PDT-induced anti-tumour immunity in mice models was performed by Canti et al. [34]. These authors have shown that re-challenge with (MS2 fibrosarcoma) tumour cells, of immunosuppressed and normal surviving mice that were previously treated with aluminium disulfonated phthalocyanines (AlS2Pc)-PDT, leads to death of the immunosuppressed surviving mice while normal surviving mice tend to resist the re-challenge. Subsequently, Korbelik et al. in a series of elegant experiments unequivocally demonstrated that activation of adaptive immunity is necessary for the most effective tumor control [21, 35]. PDT used at curative doses in normal mice provided only short-term effects in immunodeficient scid mice, and adoptive transfer of bone marrow cells from immunocompetent animals restored complete and long-lasting cures in mice with impaired immune responses [21]. Moreover, adoptive transfer of lymphocytes from normal mice, which showed complete tumor regression following PDT to immunodeficient mice allowed complete restoration of the curative antitumor effects of PDT to the same but not third party tumors [35], thus indicating that PDT is capable of generating immune memory. The activation of tumor-specific lymphocytes by PDT was also observed in rats as adoptive transfer of splenocytes harvested from animals successfully treated with PDT conferred complete and long-lasting antitumor immunity in untreated recipients [36]. Overall, these results strongly suggested that the contribution of host lymphoid populations is vital for prevention of tumour recurrence after PDT treatment and added ‘anti-tumour immunity’ in the résumé of PDT-based modalities. The effect of PDT discussed above heavily depends on the participation of adaptive immune system. Hence, not very surprisingly it has been shown that PDT has the ability to instigate anti-tumour T-cell specific immunity which also leads to generation of immune memory cells that are recoverable from distant sites [3]. Also, it has been proposed that a vital factor behind PDT-evoked adaptive immunity might be the accentuated phagocytosis of dying/dead tumour cells by cells of innate immune system which might eventually lead to a successful presentation of tumour associated antigens (TAAs) to adaptive immune cells [3]. In fact it has been shown that PDT-treated tumour sites have the ability to attract the attention of dendritic cells (DCs) [3]. Although PDT is an approved modality for clinical use and several thousand patients underwent PDT over a period of the last three decades only rarely analysis of the host immune response within routine clinical practice has been carried out [2]. Nonetheless, clinical data seem to support the important role played by the immune response in the therapeutic outcome of PDT. A study in patients with vulval intraepithelial neoplasia (VIN) revealed better response to PDT if the tumors expressed MHC class I molecules. Moreover, treatment efficacy correlated with increased tumor infiltration with CD8+ T cells post PDT [37]. PDT with Talaporfin in patients with inoperable hepatocellular carcinoma showed responses in both treated and untreated lesions [38]. In patients with nonresectable cholangiocarcinoma undergoing PDT with Photofrin tumor reduction was observed not only within illuminated lesions but also at intrahepatic bile ducts, which were not reached via cholangioscopy [39]. ALA-PDT treatment of actinic keratosis and Bowen’s disease in immunosuppressed organ-transplant recipients and normal patients produced similar clinical responses observed at 4 weeks post PDT, but transplant recipients showed much worse long-term outcomes than patients with intact immune system [40]. PDT with topical ALA followed by illumination with a blue (400–450 nm) light was not effective in preventing the recurrence of new cutaneous squamous-cell carcinomas in organ-transplant recipients [41]. Moreover, several clinical observations indicated that PDT induces local or systemic release of pro-inflammatory mediators [42, 43]. Also in a case report it was observed that PDT induced regression of the treated as well as untreated tumor lesions in a patient with multifocal angiosarcoma [44]. However, only recently a definitive proof for the induction of tumor-specific immunity against tumor-associated antigen (HIP1) has been provided in studies performed in patients with basal cell carcinoma of the skin [45].

In vivo PDT-induced acute-phase response and neutrophilia

Acute-phase response (APR), also considered to be an integral part of the PDT-based inflammation, is basically a complex immunological process involving all four classical signs of inflammation, i.e. swelling, redness, heat and pain as well as signs of ‘sickness behaviour’ (a term defining the systemic psychological effects of local inflammation like loss of appetite, lethargy, increased pain sensitivity and apathy) [46]. There are various prominent markers associated with acute-phase response including cytokines like IL-1β, IL-6 and IL-2R [46]. APR is regarded as a stress response on the level of an organism that is initiated in order to tackle tissue injury. PDT has been found to initiate APR which is characterized by increased/decreased levels of certain plasma proteins, leukocytosis and enhanced adrenal/pituitary hormonal production [3]. PDT-associated APR was first documented in a study involving treatment of normal mouse peritoneum wherein high leukocytosis (mainly nautrophilia) was prominently observed [47]. Since then, neutrophilia (see Fig. 1) has been established to be a strong PDT-associated phenomenon based on studies carried out in rodents [3]. Further research has shown that the process of PDT-mediated neutrophilia is orchestrated by various mediators such as IL-1β, TNF, various chemokines, prostaglandins, leukotrienes, IL-6, histamine, IL-10, G-CSF, PTX3, components of the complement and coagulation cascades [3]. Neutrophilia is also influenced by the adrenal–pituitary axis, an observation derived from the studies showing that PDT-induced neutrophilia is less intense in adrenalectomized as compared with normal mice [48]. In fact, neutrophils are considered to be among the first cells of the innate immune system that enter PDT-treated tumours. These cells adhere to the vascular wall within 5 min after the start of illumination [49] and neutrophil-derived myeloperoxidase (MPO) can be detected in tumours just 2 h following PDT, reaching concentrations almost 200-fold greater than in non-treated controls [50]. Here, MPO increase has been calculated to account for about 9 million neutrophils infiltrating every 100 mg of tumour tissue after PDT [50]. Once at the treated site, neutrophils can produce various factors like lysosomal enzymes, MPO and ROS, which tend to destroy residual tumour cells and the vasculature [50]. The vital role of neutrophils is further supported by the observations that neutrophil depletion in host mice (after tumour-PDT) tends to negatively affect the therapy outcome [51]. Moreover, decreased tumour cure rates have also been observed after blocking the mediators of neutrophilia, including complement components, IL-1β, IL-6, histamine, thromboxane, chemokines or xanthine oxidase, an enzyme induced by ischemia-reperfusion injury and G-CSF [50, 5254]. On the other hand, administration of G-CSF [4, 55], granulocyte macrophage colony-stimulating factor (GM-CSF) [56] or anti-platelet serum that increases neutrophil accumulation in the tumour [57] ameliorates tumour response rates of PDT. Similarly, increase in neutrophil accumulation spurred by thrombocyte antiserum administration in PDT-treated tumours tends to improve anti-tumour response [50]. In fact recently it has been demonstrated that generation of tumour-specific primary as well as memory CD8+ T-cell response following PDT depends largely on the higher neutrophil infiltration levels [58]. This conclusion was reached upon after it was observed that CXCR2−/− mice which have defective neutrophil homing to peripheral tissues or mice with depleted neutrophils are not able to put-up considerable CD8+ T-cell based anti-tumour response, following 2-[1-hexyloxyethyl-]-2-devinyl pyropheophorbide-a (HPPH)-PDT [58]. Besides, neutrophils are a rich source of pro-inflammatory cytokines that participate in the recruitment and activation of other types of immune cells.

The importance of PDT-induced neutrophilia in overall anti-tumour immunity is highlighted by the observation that active neutrophils have the ability to exert maturational effects on DCs [59]. It has been proposed that these alarmins help in establishing a link between neutrophils and DCs such that they facilitate local recruitment as well as phenotypic/functional maturation of DCs [59]. Moreover, neutrophils are themselves capable of processing and presenting antigens derived from tumour cells to CD4+ T cells [60]. This process might be of little significance during primary PDT, but could play a major role in orchestrating the local immune response if PDT is performed for the second or third time, when adaptive immune response has been already triggered. Additionally, neutrophils have the ability to secrete various factors called alarmins at the inflamed site. Most prominent alarmins secreted by neutrophils include α-defensins, cathelicidins, lactoferrin and high mobility group box 1 (HMGB1) protein [59]. These alarmins are capable of chemoattracting various immune cells including naive T cells, DCs and monocytes as well as capable of inducing expression of different pro-inflammatory cytokines and chemokines [59].

In vivo PDT-induced activation of the complement cascade

PDT can rapidly activate the complement system [61] thereby inducing a number of microenvironmental changes that include early tumour damage recognition, instigation/accretion of inflammation, clearing up of tumour cells and promotion of anti-tumour adaptive immune response [62]. Macrophages infiltrating PDT-treated sites tend to have up-regulated expression of genes encoding complement components such as C3, C5 and C9 [63]. Moreover, certain membrane-bound complement regulatory proteins (mCRPs), which resist complement deposition in cell membranes tend to be down-regulated by PDT [64]. Specifically, the expression of decay accelerating factor (DAF/CD55), protectin (CD59) and mouse-specific Crry present on murine SCCVII tumour cell surface is diminished by in vivo Photofrin-PDT thereby making these cells susceptible to complement components deposition and efficient removal by phagocytes [3, 64]. Similarly, it has been shown that FUT-175-based inhibition of classical and alternative complement pathways tends to have a derogatory effect on the antitumor efficacy of benzoporphyrin derivative (BPD)-PDT [3]. Recently, Photofrin-PDT was shown to initiate a signalling cascade involving TLR2 and TLR4 receptors along with NF-κB activation which ultimately leads to complement protein production by tumour-associated macrophages (TAMs) [62].

In vivo PDT-induced production of cytokines

PDT research has also highlighted the vital role played by cytokines in the overall tumour curative outcome. Usually it is the interplay between pro-inflammatory and anti-inflammatory cytokines that ultimately affects the overall outcome of the anti-tumoural immune response. And this interplay has been observed to affect the outcome of PDT-induced antitumour effects. For instance, neutralisation of TGFβ or IL-10 with monoclonal antibodies significantly potentiated antitumor effects of PDT in a subcutaneous model of FsaR fibrosarcoma in mice [3]. In stark contrast, blocking of the activity of pro-inflammatory IL-1β tends to trim down the tumour curative rate [50]. Similarly, it has been observed that certain factors including vascular endothelial growth factor (VEGF), matrix metalloproteinases (MMP) and cyclooxygenase-2 (COX-2), which are produced in response to PDT tend to negatively affect the therapeutic outcome [20, 65, 66]. This outlines the significant role played by cytokines in shaping the optimal conditions for generating maximum tumour responses following PDT. Clinical studies have also shown that cancer patients who have undergone PDT might have elevated levels of cytokines like IL-6, IL-8, IL-1β and IL-10 in their serum [42]. Here, it is very important to generate more clinical data on the cytokine profile of tumour sites treated with PDT since this could help us in deciding various parameters that might be used to augment anti-tumour immunity both on the levels of innate as well as adaptive immune system.

PDT and immunomodulation: effects exerted by PDT-stressed cancer cells on immune system

Damage to the illuminated cells resulting from the photo-oxidative stress is associated with the formation of a number of lesions and structural rearrangements within cytoplasmic and membranous components of tumour, vascular and stromal cells. Numerous molecules (categorised as DAMPs and discussed in details later) that are normally confined to the intracellular compartments are released or translocated to the surface of damaged cells and play an important role in immunomodulation. Simultaneously, in response to PDT, a large amount of tumour cell debris becomes available for phagocytic cells (including macrophages, neutrophils and dendritic cells) in a relatively short time interval. These cells can then become loaded with released tumour antigens and, in the inflammatory microenvironment accompanied by the appearance of “danger signals” (or DAMPs) they become activated to produce more inflammatory mediators, to process and to present tumour-derived antigens to host lymphoid cells. In this aspect it should be emphasized that not only dendritic cells and macrophages but to some extent also neutrophils and mast cells can become antigen presenting cells [60, 67, 68]. Of these only DCs are capable of getting into local lymph nodes to initiate an immune response. The remaining antigen presenting cells can function locally to sustain an effective immunity. Several independent observations seem to support that the above described scenario takes place during PDT. DCs incubated with PDT-generated (murine EMT6/P815) tumour lysates secrete IL-12 and express more MHC class II and co-stimulatory (CD86) molecules [69]. Moreover, they efficiently phagocytose PDT-treated (C-26) tumour cells [70], and when isolated from tumour draining lymph nodes of PDT-treated mice exhibit enhanced ability to stimulate T cell proliferation and IFN-γ secretion [71]. Local inoculation of immature DCs into PDT-treated tumours impairs the growth of tumours inoculated at a distinct region [70]. Besides this there is also a possibility of PDT-stressed cells themselves modulating the immune system (and the nature of tumour microenvironment) by affecting the expression/production of immunomodulatory genes/proteins; for example, it has been observed increasingly that in response to PDT cancer cells can produce prostaglandin E2 (PGE2), which in turn might modulate the immune response [2]. This point has been discussed in details (in the next section) with the perspective of Hypericin-PDT up-/down-regulating the expression of immunomodulatory genes in treated cancer cells in vitro.

Immunomodulatory potential of Hypericin-PDT: a microarray-based glimpse!

Hypericin is considered to be one of the most powerful photosensitisers available from nature [6] thereby making Hypericin a valuable photosensitiser for PDT. It is derived from Hypericum plants (27 out of 36 evaluated species), more specifically the Hypericum perforatum or St. John’s wort, an erect perennial herb [6]. Hypericin exhibits potent photosensitizing characteristics, minimal dark toxicity and high singlet oxygen quantum yield [6]. Several studies published over the period of last decade have shown that Hypericin-PDT is a very effective anti-cancer modality in terms of cancer cell killing and cell death [6, 13, 14, 72, 73]. One of these studies also evaluated the overall effect of Hypericin-PDT on gene expression in T24 human bladder carcinoma cells via microarray-based analysis [72]. Although microarray based analysis is ‘blind’ to certain aspects such as intracellular compartmentalization and translational control yet it gives a comprehensive evaluation of steady state coding mRNA amounts thereby providing direct representation of transcriptional as well as post-transcriptional regulation [74]. Thus, microarray based data is considered vital in terms of drawing a picture of immunologically relevant genes (or “immunological genome”) that are affected at a certain point of time [74]. The analysis of changes in gene expression after Hypericin-PDT of T24 cells showed up-/down-regulation of various genes coding for pro-inflammatory cytokines, chemokines, anti-inflammatory cytokines, immunologically relevant receptors and other proteins capable of modulating the immune response (see Table 2) [72]. Hypericin-PDT was found to upregulate expression of genes coding for potent pro-inflammatory cytokines/proteins like GM-CSF, IL-6, IL-1α, IL-15, IL-18, IL-24 and 4-1 BB ligand as well as those coding for potent chemokines like CXCL2, CXCL3, CKLF and IL-8/CXCL8. On the other hand this treatment was found to downregulate expression of genes coding for anti-inflammatory mediators like thrombospondin-1 and TGFβ2 (see Table 2). This clearly shows that Hypericin-PDT is probably capable of inducing potent inflammatory response proficient in exerting anti-tumourigenic effect by attracting attention of immune cells (via chemokines) as well as by activating them (via increased pro-inflammatory proteins and decreased tolerogenic/anti-inflammatory proteins). On the flip side, Hypericin-PDT treatment of T24 cells also leads to upregulation of certain genes coding for potentially anti-inflammatory proteins like COX-2, DAF/CD55 and PTX3 (see Table 2). Considering cancer cell’s knack to negate anti-tumour immune response [75], this could hardly be considered as surprising. Yet interestingly, Hypericin-PDT treated T24 cells were also found to up-regulate a gene coding for the enzyme, hydroxyprostagladin dehydrogenase-15, which has been very recently found to be capable of degrading prostaglandin E2, thereby reducing its immunosuppressive effects [76].
Table 2

Various immunomodulatory genes whose expression is affected in Hypericin-PDT treated T24 human bladder carcinoma cells

Upregulation of genes coding for various pro-inflammatory cytokines and/or inflammation/immune system augmenting proteins

 CSF2 (Granulocyte-macrophage colony stimulating factor 2 or GM-CSF): Enhances anti-tumour immunity by facilitating tumour antigen presentation, DC maturation, steady-state differentiation of alveolar macrophages/iNKT cells, priming of macrophages/neutrophils/eosinophils and mobilization of myeloid populations into blood [147, 148]

 IL6 (Interleukin-6), IL1A (Interleukin-), IL15 (Interleukin-15) and IL18 (Interleukin-18), IL24 (Interleukin-24): IL-6 facilitates B-cell proliferation, antibody production and T-cell proliferation/differentiation/cytotoxicity; IL-1α facilitates chemokine release, monocyte cytokine release, iNOS activation and prostaglandin release [149]; IL-18 facilitates the Th1 cell-based cytotoxicity/immunity and angiogenesis inhibition [148]; IL-15 enhances T-cell chemokinesis/activation/memory maintenance, B-cell differentiation/isotype switching and NK-cell activation/cytotoxicity [149]; IL-24 exhibits potent tumour suppressive activity [150] and stimulates IL-6, TNF-α and IFN-γ production from peripheral blood mononuclear cells (PMBC)

 TNFSF9 (TNF ligand superfamily member 9 or 4-1 BB ligand): 4-1BB ligand interacts with the 44-1BB molecule on naive T cell surface which activates the T-cells and facilitates potent anti-tumour response [151]

 HPGD (Hydroxyprostaglandin dehydrogenase-15 or 15-PGDH): 15-PGDH can exert potent tumour suppressive effects [152]. Its overexpression/upregulation can lead to increased degradation of PGE2, a potent immunosuppressor and tumour progression augmenter thereby assisting in increasing anti-tumour immune response [76]

Upregulation of genes coding for chemokines

 CXCL3 (Chemokine C-X-C ligand 3 or Macrophage inflammatory protein 2-β), CXCL2 (Chemokine C-X-C ligand 2 or Macrophage inflammatory protein 2α) and IL-8 (Interleukin-8): CXCL2 (or GRO-β) and CXCL3 (or GRO-γ) can chemotactically attract neutrophils while IL-8 (or CXCL8) can chemotactically attract neutrophils, T cells, NK cells, basophils [153] and dendritic cells (DCs) [154]

 CKLF (Chemokine-like factor): A potent chemokine for various different leukocytes [155]

Downregulation of genes coding for proteins capable of contributing towards tolerogenicity/anti-inflammatory processes

 THBS1 (Thrombospondin 1 or TSP-1): An important mediator of apoptotic cell engulfment via phagocytosis. TSP-1 exerts immunosuppressive effects (by assisting in production of tolerogenic immature DCs) thereby assisting in the immunologically ‘tolerogenic’ nature of apoptosis [156]

 TGFB2 (Transforming Growth Factor β2) and TGFBR2 (Transforming Growth Factor β receptor 2): TGF-β exerts immunosuppressive effects [157, 158] by preventing DC maturation [159] and suppressing NK-cell proliferation/effector function as well as iNOS expression [149]. It facilitates proliferation of immunosuppressive T cell types, e.g. Th17 and Treg cells [149]. Here, TGF-β receptor 2 is an important TGF-β receptor [160]

Upregulation of genes coding for proteins that could contribute towards anti-inflammatory/immune-suppressive processes

 PTGS2 (Prostaglandin-endoperoxide synthase 2 or cyclooxygenase-2 or COX-2): COX-2 enzyme plays a vital role in biosynthesis of prostaglandin E2 (PGE2) [161]. PGE2 is overproduced in various tumours and tends to assist in tumour angiogenesis, metastasis and tumour-induced immune dysfunction [76]. Thus, PGE2 is regarded to be an important immunosuppressive molecule

 DAF or CD55 (Decay accelerating factor): It is a glycosylphosphatidylinositol-anchored membrane inhibitor of the complement pathway and a potent suppressor of T-cell immunity, in vivo [162]

 PTX3 (Pentraxin-related protein 3 or TNF-inducible gene 14 protein): Involved in apoptotic cell clearance [163], modulation of classical complement activation pathway [164] and might act as an “opsonin” [158]. Its expression occurs late in apoptosis and tends to inhibit recognition of apoptotic cells by DCs [165]. PTX3 may enhance nitric oxide and cytokine production from innate immunity cells [166]

Other interesting and potentially ‘immunomodulatory’ genes upregulated by Hyp-PDT

 TLR2 (Toll-like receptor 2); THBD (Thrombomodulin); IL11 (Interleukin-11); SQSTM1 (Sequestosome-1 or Ubiquitin-binding protein p62); SOCS1 (Suppressor of cytokine signalling 1); TNFRSF11B (TNF receptor superfamily member 11b); KLF11 (Kruppel-like factor 11 or TGFβ-inducible early growth response protein 2); IRF1 (Interferon regulatory factor-1); IRF7 (Interferon regulatory factor-7); IFI16 (Interferon gamma-inducible protein 16); IL6R (Interleukin-6 receptor); IL13RA2 (Interleukin 13 receptor α2); TRAF6 (TNF receptor-associated factor 6); ANXA 10 (Annexin A10)

Other interesting and potentially ‘immunomodulatory’ genes downregulated by Hyp-PDT

 GBP1 (Guanylate binding protein 1); TLR4 (Toll-like receptor 4); HLA-DPA1 (Major Histocompatibility Complex, Class II, DP alpha 1); CCL2 (Small inducible cytokine A2); ISG15 or G1P2 (ISG15 ubiquitin-like modifier or interferon-induced 17 kDa protein); TNFAIP1 (Tumour Necrosis Factor, α-induced protein 1); CD40 (TNF receptor superfamily member 5); IL13RA1 (Interleukin 13 receptor, α1); TGFBI (Transforming Growth Factor β-induced 68 kDa)

Data derived from Buytaert et al. [72]

It is also worth mentioning here that even pro-inflammatory proteins/cytokines can sometimes have dual roles in tumour progression, a notion that is especially applicable to IL-6, one of the most pro-inflammatory cytokines. Apart from its direct inflammatory activity, IL-6-based signalling has another interesting facet resulting from binding to IL-6R/gp130 membrane receptor complex, which triggers Janus (JAK) kinases followed by downstream activation of STAT3, PI3K/Akt and SHP-2/Ras [77]. Thus, if IL-6R/gp130 receptor complex is present on the surface of a cancer cell than IL-6 could interact with it (in autocrine or paracrine manner) thereby leading to increased tumour growth/proliferation [77]. The impact of this pathway could be made out from the fact that neutralizing IL-6 in such a scenario could effectively block the self-renewal capabilities of tumour cells [78]. In the present case, Hypericin-PDT is responsible for upregulation of genes coding for both IL-6 and IL-6R in T24 cells (see Table 2) thereby opening the door for potential presence of autocrine IL-6 signalling. However, no upregulation of gp130 was noticed. It is interesting to note that most normal epithelial cells are known to express gp130 but not significant amounts of IL-6R [77] hence in the present case, more studies are required to ascertain whether Hypericin-PDT facilitates the execution of autocrine IL-6 signalling. Another intriguing observation is that gene coding for IL-11 is also upregulated by Hypericin-PDT (see Table 2) and IL-11 along with IL-22 has been shown to contribute towards STAT3 activation and tumour development. These observations if proven could also open doors for Hypericin-PDT based photoimmunotherapies involving anti-cytokine antibodies.

Another captivating observation derivable from these microarray data is the expression profile of genes coding for TLRs in T24 cancer cells undergoing Hypericin-PDT. While TLR expression by cells of the immune system is well known only recently it has been shown that TLRs might be expressed in certain normal as well as cancerous cells (e.g. either in patient-derived cells of cancers like gastric carcinoma, laryngeal carcinoma, colorectal cancer, ovarian cancer, cervical cancer and lung cancer or in established cell lines like HeLa, Jurkat, U937, CaP, ES2, NIH:OVCAR3, SKOV3, CAOV3, A2780, CP70, MDA-MB-231, MCF-7, KG-1, Raji, A549, H1299, 95D) [79, 80]. Normal epithelial cells express TLRs and participate in a first line of defence against pathogens at invasion portals of the digestive system, respiratory or female reproductive tracts [79]. Not unexpectedly it turned out that also cancer cells express at least five (TLR2, TLR4, TLR5, TLR9 and TLR3) of the 11 known human TLR types [79, 80]. While the exact role of TLRs expressed by cancer cells is still being investigated there is circumstantial evidence indicating a pro-tumourigenic function for these receptors [79]. TLRs expressed by cancer cells might promote cancer metastasis, anti-apoptotic activity, tumour angiogenesis and immunoregulation [79, 80]. On the other hand, there is also considerable evidence suggesting that TLRs expressed by cancer cells might govern the status of tumour microenvironment. For instance, it has been shown that cancer cells might secrete a number of immunoregulatory mediators in response to TLR ligands [79, 81]. In case of Hypericin-PDT treated T24 cancer cells, another intriguing observation is that, while TLR2-coding gene is up-regulated after PDT, the TLR4-coding gene is downregulated (see Table 2) and moreover p38MAPK inhibition (via PD169316) further up-regulates TLR2-coding gene [72]. These observations are very unique since p38MAPK has not been implicated yet in governing TLR-expression in cancer cells and moreover, the significance of TLR4-coding gene being downregulated after PDT treatment is as of now, unknown. It is quite lucid from the above discussion that Hypericin-PDT mediated modulation of TLR-coding genes is an unexplored area that warrants further investigation; also because it has been proposed that these TLRs on cancer cells might act as receptors for potential damage-associated molecular patterns or DAMPs [79] secreted/exposed by stressed cancer cells, a unique but unconfirmed conjecture which is important in light of the fact that the existence of DAMP-receptors on the target cancer cells is still more or less a mystery.

Apart from the TLR2-coding gene’s upregulation, inhibition of the p38MAPK protein has also been found to upregulate the Kruppel-like factor 11-coding gene as well as down-regulate genes coding for GM-CSF, sequestosome-1, IL-8, TNF receptor superfamily member 11b, COX-2 and annexin A10 (or annexin 14) [72]. This bestows upon p38MAPK a potential ability to modulate the immune response on the level of PDT-treated cancer cells. However whether p38MAPK inhibition could effectively augment anti-tumour immunity after PDT is an intriguing question that deserves in vitro as well as in vivo investigation. All in all, it is plausible to say that Hypericin-PDT holds great potential in terms of activating the anti-tumour immune response and thus could be one of those therapies that could amalgamate the tumour cell killing potential with anti-tumour immunity. Thus, further (pre-clinical) investigation is required to ascertain whether this PDT modality has the potential to induce ‘anti-cancer vaccine effect’ and revive tumour immunosurvelliance.

PDT and DAMPs: a newfangled paradigm

DAMPs: an introduction

Damage-associated molecular patterns (DAMPs) which are also sometimes called alarmins [82] are essentially molecules retained within the cells under normal conditions but exposed on the surface or released from cells in response to damage or physical/chemical insult/stress [75]. The term DAMPs was coined to serve as an eukaryotic synonym for pathogen-associated molecular patterns (PAMPs). The latter are defined as a set of molecular signatures that represent patterns of a whole classes of microbes capable of evoking an array of immune defence responses when exposed or released from pathogens or pathogen-infected cell [83]. Similarly, DAMPs also usually have immunostimulatory properties once exposed or released, since they tend to interact with various (vesicular or membrane-bound) pattern-recognition receptors (PRRs), e.g. RIG-I-like receptors (RLRs), the NOD-like receptors (NLRs) and Toll-like receptors (TLRs) [84] thereby exerting various pro-inflammatory effects including maturation, activation and antigen processing capabilities of antigen presenting cells (APCs) including the most professional of these, i.e. dendritic cells (DCs) [75]. With some oversimplification it might be stated that the actual diversity of DAMPs may depend upon a number of variables such as a type of cell death pathway, cell type and the extent of tissue injury [75, 84].

The types and diversity of DAMPs released in response to various cell death pathways seems to be intimately related to the biochemistry of that cell death pathway. Necrosis evokes the most diverse DAMP profile followed by apoptosis, while autophagic cell death (i.e. non-apoptotic programmed cell death accompanied by autophagy) has a largely ill-characterized DAMP profile. For extensive discussion of various DAMPs and their association with different cell death pathways please see the review by Garg et al. [75]. One of the cell death pathways, whose immune-stimulating profile has been significantly revamped, is apoptosis, which under particular conditions is associated with the appearance and/or release of a set of DAMPs like HSP90, calreticulin, HMGB1 [75] and a recently added ATP/UTP [85]. These exciting new studies have placed DAMPs at the cutting edge of the anti-cancer therapeutics research. With respect to these trends, PDT is not left far behind as PDT-induced cancer cell death has also been shown to associate with the exposure of DAMPs to the cells of the immune system. This association is discussed extensively in the following sub-sections.

PDT and DAMPs: what we know…

Since PDT can induce all three cell death modalities depending upon the parameters of treatment [1], it could be interesting to see if there exists a concept of ‘PDT-induced cell death associated DAMPs’. In case of necrosis, agent-to-agent variations have not been yet observed to cause changes in spectra of necrotic DAMPs. Hence when higher PDT doses kill cancer cells via necrosis the DAMPs that would be produced would be expected to be more or less the same as formed in most cases of necrosis (see Fig. 2). However what might be changed in case of DAMPs associated with PDT is the overall chemical state modifications. Photodamage of cellular macromolecules including DAMPs might lead to the formation of unique oxidative modifications of the latter. However, the concept of modified DAMPs being released after PDT-induced necrosis has not been investigated yet and hence could be an interesting paradigm to explore. On the other hand, nothing is known about the DAMPs associated with PDT-induced autophagic cell death or exposed/released through autophagy that is associated to cell death.
https://static-content.springer.com/image/art%3A10.1007%2Fs10495-010-0479-7/MediaObjects/10495_2010_479_Fig2_HTML.gif
Fig. 2

An overview of changing trends in DAMP profile of three main cell death pathways. Many new DAMPs have been added to the overall DAMP repertoire of necrosis (left), making its profile one of the most diverse amongst the three. It has been observed that necrotic cell is capable of releasing various DAMPs such as uric acid/MSU, HSP70, HSP90, HMGB1, ATP, DNA, RNA, HDGF, SAP130 and S100 protein family members. This is accompanied by secretion of IL-1α and IL-6 in a passive and active manner, respectively. These events, coupled with certain other factors may lead to neutral or pro-inflammatory immune response. On the other hand, the DAMP profile of apoptosis (middle) has only recently been consolidated. The “classical” apoptotic DAMP profile consisted of different “eat me” and “find me” signals which were usually associate with induction of immune tolerance. DAMPs like heat shock proteins (HSPs) were observed only during secondary necrosis. However, the “new” apoptotic DAMP profile (induced by specific chemotherapeutic/apoptotic agents) consists of various DAMPs like calreticulin, ATP/UTP and HSPs which lead to an immunogenic form of apoptosis capable of activating the immune system. In fact, a new DAMP, i.e. HMGB1 has been reported to be associated with secondary necrosis. In contrast to the above two pathways, autophagic cell death (right) has the least characterized DAMP profile with no DAMP implicated in the “classical” model. However, recently HMGB1’s active release has been associated with this cell death pathway paving way for further research

One group of DAMPs that predominantly associate with PDT-induced apoptosis are heat shock proteins (HSPs). Because of their widespread association with apoptotic cell death HSPs are called seasoned apoptotic DAMPs [75]. HSPs are a family of highly conserved chaperones playing an important role in structural folding of both newly synthesized as well as stress-modified proteins [86]. While intracellularly these HSPs might exhibit cancer augmenting or pro-survival activity (e.g. HSP70 being considered as one of the most powerful anti-apoptotic proteins [87]) yet their presence outside the cells (either on the membrane or in extracellular space) might be immunostimulatory. Membrane (outer leaflet) association of HSPs has attracted special attention due to their ability to interact with innate immunity cells like DCs and NK cells. Moreover, membrane HSPs can stimulate these cells to produce pro-inflammatory cytokines and to process/react to tumour antigens [75]. For more on DAMP-like properties of surface HSPs please see the review by Garg et al. [75]. It has been observed that HSP70 might be translocated onto the outer leaflet of plasma membrane of cancer cells (murine SCCVII cells) in response to Photofrin-PDT [88]. Intriguingly, ecto-HSP70 translocation seems to be more robust in apoptotic than in non-apoptotic cells [88]. Also, the supernatant of cells undergoing Photofrin-PDT was found to contain soluble HSP70 [88]. However, whether this release is an ordered secretion or just a consequence of membrane permeabilization/secondary necrosis/primary necrosis is unclear. Such possibility seems to be plausible since the authors reported that higher fluencies of PDT lead to earlier extracellular detection of HSP70. At lower fluencies extracellular HSP70 was only detectable at very late stage post-PDT [88]. Apart from HSP70 the authors also reported surface exposure of HSP60 and GRP94 (GRP—glucose-regulated protein) in response to Photofrin-PDT. Interestingly, HSP60 and HSP70 were detected on the surface only by antibodies that bind to their COOH-rather than NH2-terminus [88]. In vivo PDT studies reported by these authors showed that while ecto-HSP70 was expressed prominently on the surface of SCCVII tumour cells as well as leukocytes (retrieved from the treated site) yet ecto-HSP60, ecto-GRP78 and ecto-GRP94 were expressed more prominently on the surface of myeloid rather than SCCVII tumour cells [88]. This study thus reveals that DAMPs expressed on the surface in response to PDT stress might be cell-type dependant. Also there might exist a difference between DAMPs associated with a particular cancer cell and tumour microenvironment as a whole. Moreover, these authors proposed that ecto-HSP70 expression on cancer cells might assist in opsonisation of cancer cells by supporting deposition of complement proteins on PDT-treated cells [89]. Similarly, another investigation [75] revealed that Photofrin-PDT is capable of translocating HSP70 onto the plasma membrane in a PDT-dose dependant manner and this translocation is closely related to changes in mitochondrial transmembrane potential [90].

It is quite important to note that extracellular release of HSPs could be considered to be at par with their membrane-exposure in terms of immunostimulatory activity or by extension, anti-cancer immunosurvelliance. In fact stress proteins of HSP and GRP families are considered to be capable of exerting powerful effects on immune system [91]. This notion becomes even more interesting if we consider that apart from passive extra-cellular release (via necrosis or apoptotic membrane permeabilization), there might also be active (exocytotic) extracellular release of HSPs taking place in certain conditions, via a non-classical ‘secretory’ pathway not inhibited by brefeldin A and orchestrated by exosomes [87, 9193]. It has been found that HSPs such as HSP70 and HSP90 could be complexed to tumour-associated peptides or antigens when in extracellular space and these HSP-peptide complexes could then be taken up by APCs, processed and complexed to MHC molecules and presented to CD8+ cytotoxic T lymphocytes to initiate adaptive immune response [87, 91, 94, 95]; thus extracellularly released HSPs might act as carriers of vital tumour antigens. Here, while extracellular role of HSP70 has been extensively analyzed, evidence supporting strong immunogenic activity of extracellular HSP90 has also started emerging wherein it has also been shown that extracellular HSP90 could interact with various molecules like MMP-2, CD91, TLR4 and annexin II [96]. Besides this, extracellular HSPs (free from any ‘carried’ peptides) could interact with various receptors like CD14 and TLR2/4 on APCs thereby triggering secretion of pro-inflammatory cytokines including GM-CSF, TNF and IL-1β/-12/-6 [9799]. Also, HSP70 has been found to be capable of stimulating DC migration and maturation (upregulation of MHC class II, CD86, CD83 and CD40 molecules) [95, 100] as well as NK cell activation [87]. In another recent interesting study it was shown that HSP70 could be released in association with membranous structures from dying cells and that these HSP70-membrane complexes could be 260-fold more effective than free recombinant protein in inducing TNF production from macrophages [101]. On a different note though, it has been reported that certain effects of HSPs might vary depending upon their source, for instance mycobacterial HSP70 can also stimulate production of CC chemokines and nitric oxide from innate immunity cells, a phenomena not yet reported for mammalian HSPs [102]. However, there are some doubts raised on HSPs’ cytokine-like activity due to possible (contamination-based) association between HSPs and LPS [87], hence such observations have to be treated with care. Nevertheless, such is the impact of tumour-derived extracellular HSPs that patient tumour-derived HSP70/gp96–peptide complexes are usable as vaccines to treat and prevent cancer yet HSPs derived from normal tissues are not capable of inducing anti-cancer immune response [87]. On a more cautionary note it is important to consider that even in extracellular space, HSPs might be capable of playing a dual role, i.e. while on one side they could help in anti-tumour immune response (as discussed above) yet on another they could probably assist in cancer metastasis, as has been suggested for HSP90 [103]. Although this dual activity still requires further confirmation yet the possibility cannot be ruled out.

PDT-associated DAMPs: exploring the uncharted territories!

Apart from the exploration of PDT-associated spectra of DAMPs, another interesting avenue that warrants further study is whether PDT could actually help us explore certain glitches and conundrums associated with DAMPs as well as help us answer certain questions that have not been asked emphatically yet, e.g. is release/exposure of particular DAMPs dependent on stress exerted over a particular sub-cellular organelle? Since PDT is a unique sort of therapy compared with most chemotherapeutic regimes in terms of its overall execution, it is an impending possibility that PDT could help us explore certain ‘uncharted’ territories.

One of the most important conundrums associated with DAMPs is the possibility of shift in DAMP’s activity from pro-inflammatory/immunity-encouraging to neutral or even anti-inflammatory/immunity-discouraging based upon certain modifications either by the stress agent or by proteases [75]. Specific concerns raised in this regard pertain to the neutralized activity of HMGB1 following oxidative modification [104]. To this end, one cannot help but wonder whether ROS-based PDT could actually modify the DAMPs released from dying tumour cells, in a way that renders them inactive. Although this conjecture is still a matter of debate yet the discussions above have indeed shown that PDT-derived HSPs are usually immunostimulatory. However, on the flip side some evidence that is currently available points towards possible benefits of PDT-based photo-oxidation. For instance, it has been proposed that PDT-mediated photo-oxidation of certain cellular molecules might generate unique tumour antigens [3]. Moreover, there is a possibility that PDT-based membrane changes or protein unfolding events might actually reveal certain neo or cryptic antigens [3]. Thus, there is a likelihood that PDT-based photo-oxidation might actually be immunologically beneficial rather than the other way round. However, this notion requires further investigation and might not be considered to be universally acceptable for every PDT photosensitiser or ROS-based modality. Another emerging question related to DAMPs is that of overlap between DAMPs exposed by different cell death pathways especially between apoptosis and ‘autophagic’ cell death. Though ‘autophagic’ cell death has only been recently shown to be capable of releasing HMGB1 yet it remains to be clarified whether ‘transiently’ activated autophagy could actually assist in or hinder the exposure of certain DAMPs. Here, since PDT is capable of activating both these cell death pathways depending upon the dose or post-PDT recovery time, it could be assumed that PDT-based modalities could probably help us solve the question of overlap between DAMPs exposed/released by apoptosis and/or autophagy.

And last but not the least is the relatively unexplored area of correlation between a particular spectrum of DAMPs and stress on a particular sub-cellular organelle. For a long time, a lot of different DAMPs exposed or released by cells under stressed or dying conditions were being analyzed but in cases of active exposure/release, the mechanisms governing the translocation or secretion were elusive. It was only in last few years that more in-depth knowledge of DAMP translocation and secretion pathways was unearthed thereby giving some idea about correlation between stress targets within the cells and particular DAMPs. For instance, in case of ecto-CRT (anthracyclines, oxaliplatin and ionizing radiations) and ecto-HSP90 (bortezomib), it was shown that agents inducing ER-stress (directly or indirectly) could instigate or assist in their exposure [75, 105], thereby making ER an important sub-cellular organelle for DAMPs. However the same is not applicable to DAMPs like HMGB1 and HSPs, e.g. while in most cases the exact pool from which surface HSPs originate is not known, yet Photofrin-PDT which primarily exerts stress on Golgi apparatus and plasma membrane has been shown to cause ecto-HSP70 exposure [88, 90]. Thus, even though ER has been shown to be potential pool for certain DAMPs yet the conundrum of correlation between particular DAMPs and stress on a particular sub-cellular organelle still exists. We envisage that PDT is one of those modalities that could help us efficiently solve this outstanding question. There is a well characterized panel of photosensitisers available, which tend to localize in various different or a particular organelle; thus, usage of PDT-based stress on these organelle and subsequent analysis of exposed/secreted DAMPs could probably help us in revealing the sub-cellular pool from which a particular DAMP originates.

PDT and its effects on immune system

Direct effects of PDT on immune cells: capable of resetting immunosuppressive tumour microenvironment?

PDT is frequently considered to be a selective antitumor therapeutic procedure. While indeed most of the photosensitisers seem to better accumulate in tumour as compared with normal cells the damage is usually conferred to all types of cells within the illuminated area. Therefore, selectivity of PDT mainly results from the fact that illumination itself limits the area that is being damaged and includes tumour cells, tumour stroma (including vasculature and fibroblasts) and tumour infiltrating cells of the immune system. Numerous studies have documented that photosensitisers efficiently accumulate in normal human lymphocytes [105, 106]. However, resting lymphocytes only poorly accumulate photosensitisers, and the uptake is improved in activated cells [106108]. PDT of resting T-cells leads to MHC class I molecule’s downregulation. Significantly more changes are described in activated T cells that under PDT conditions show decreased CD25 expression and undergo cell cycle arrest [107]. In contrast to lymphocytes, DCs and macrophages accumulate photosensitisers without previous activation [106]. Nonetheless, IFN-γ-activated cells seem to be more vulnerable to PDT-mediated killing than resting cells [109]. PDT-associated oxidative stress also leads to modulation of cell surface molecule’s expression in these cells. For example, it has been proposed that superoxide radicals are responsible for induction of structural changes within FcγRI, a high affinity receptor for IgG that participates in immunophagocytosis and ADCC [110].

PDT of tumours might result not only in killing of cancer cells but also tumour-infiltrating/-associated immune cells. Tumours are usually infiltrated by various populations of leukocytes, prominent among these are myeloid-derived suppressor cells (MDSCs) and regulatory T cells (Tregs). Recent evidence has shown that MDSCs contribute to the negative regulation of immune responses during cancer development [111, 112]. MDSCs are a heterogeneous population of cells that includes myeloid progenitor cells and immature macrophages, granulocytes and dendritic cells. These cells are potent suppressors of various effector T-cell functions. Moreover, they secrete numerous growth factors that facilitate tumour growth and angiogenesis [111, 112]. Similarly, Tregs can migrate into tumours and suppress effective anti-tumour responses within the tumour microenvironment [113]. The mechanisms of action of tumour-infiltrating immunoregulatory cells are briefly summarized in Table 3.
Table 3

Selected regulatory immune cell subsets suppressing anti-tumour immune response that might be effectively eliminated by PDT

Cell type

Mechanisms of action

Outcome

Regulatory CD4+CD25+Foxp3+ T-cells

Release of IDO, arginase-1 and NO

Inhibition of effector T-cells function (e.g. IDO is an enzyme involved in tryptophan catabolism. Lack of tryptophan and accumulation of its metabolites is toxic to T-cells) [167]

Myeloid-derived suppressor cells (MDSCs)

Release of arginase-1 and NO

Inhibition of tumour-specific T-cells [167]

Tumour-associated macrophages (TAMs)

Release of arginase-1, IL-10, TGFβ and PDE2

Inhibition of T-cell function, promotion of Th2-cell response [167]

Alternatively activated DCs

Lack of co-stimulatory signals for T-cell activation

T-cell anergy (lack of function) [168]

Plasmocytoid DCs

Release of IDO, not yet identified mechanisms

Inhibition of effector T-cell function, promotion of Th2-cell response and generation of CD8+ regulatory cells [169]

CD4+ natural killer T-cells (NKTs)

Release of IL-13

Suppression of effector T-cell function, activation of MDSCs function [167]

DCs dendritic cells, Foxp3 forkhead box P3, IDO indoleamine 2,3-dioxygenase, IL interleukin, NKT natural killer cells, NO nitric oxide, PDE2 prostaglandin E2, TGFβ transforming growth factor beta

MDSCs and regulatory T cells are potent suppressors of tumour immunity and therefore a significant impediment to cancer immunotherapy. Although the influence of PDT on these cells has not been studied so far it is possible that their PDT-mediated elimination might contribute to resetting immunosuppressive environment allowing tumour infiltration with new waves of inflammatory cells that in the presence of endogenous DAMPs would find favourable conditions for priming immunity. PDT’s ability to induce recruitment of immune cells is discussed in some details in the following sub-section.

Mechanisms of PDT-induced recruitment of inflammatory cells

Apart from the direct (mostly lethal) effects of PDT on tumour-infiltrating/-associated immune cells, PDT is also capable of instigating active recruitment of inflammatory cells (into the treated site), a process that might be vital for ‘resetting’ the immunosuppressive tumour microenvironment. Multiple interdependent mechanisms seem to be responsible for the rapid recruitment of various populations of host immune cells (Table 4). Here, PDT tends to induce a blood flow stasis associated with vessel constriction, platelet aggregation, thrombus formation and activation of the complement cascade already during light illumination [114, 115]. Activation of the latter leads to formation of the membrane attack complexes (MAC), which directly damage tumour and endothelial cells (as discussed previously) [116]. Additionally, complement degradation products including C5a and C3a (so called anaphylatoxins) are the most potent endogenous chemotactic factors for neutrophils [64]. Another source of signals participating in the attraction of neutrophils are arachidonic acid metabolites released from damaged or dying tumour and endothelial cells [117]. Moreover, PDT-mediated oxidative stress in tumour cells triggers a vast array of signal transduction pathways that activate transcription factors such as NF-κB and AP-1 that induce the expression of genes encoding cytokines, adhesion molecules, co-stimulatory molecules and other inflammatory mediators (Table 4) [118120]. This along with release of large amounts of tumour cell debris, accompanied by the production of eicosanoids, chemotactic factors, induction of tumour hypoxia all contribute to a rapid recruitment of immune cells from the circulation and strong activation of neutrophils, macrophages and mast cells. Here, acute inflammation is not only an efficient effector mechanism responsible for the destruction of residual tumour cells surviving the treatment, but it can also promote development of adaptive immunity (as discussed previously) [121].
Table 4

Signals responsible for the recruitment and activation of innate immunity cells in response to PDT

Cause

Mediators

Effects

Damage to the vasculature

Extravasated plasma proteins, released fragments of extracellular matrix and/or basement membrane, release of DAMPs

Recruitment of neutrophils

Activation of the complement cascade

Generation of anaphylatoxins (C3a, C5a)

Chemotaxis of neutrophils

Induction of hypoxia and hypoxia/reperfusion injury

Induction of xanthine oxidase

 

Tumour cell necrosis

DAMPs

Activation of PRR resulting in the release of pro-inflammatory cytokines

Activation of signalling pathways

Activation of PLA2, induction of COX-2

Generation of prostaglandins and leukotrienes

Activation of transcription factors

NF-κB, AP-1, HIF

Secretion of proinflammatory mediators (cytokines, chemokines), expression of adhesion molecules in endothelial cells

Induction of endoplasmic reticulum stress

Induction of HSPs expression

The appearance of “danger signals”

AP-1 activating protein 1, COX-2 cyclooxygenase 2, DAMPs damage associated molecular patterns, HIF hypoxia inducible factor, HSP heat shock protein, NF-κB nuclear factor κB, PLA2 phospholipase A2, PRR pattern recognition receptor

Emerging PDT-based therapies: photoimmunotherapy and PDT-produced cancer vaccines

Photoimmunotherapy and PDT-mediated immunomodulation

In spite of its ability to induce inflammatory responses and activation of adaptive immunity, PDT can also be used as an immunomodulatory strategy. It is mainly an anatomic site as well as total area of illumination that highly influence the type as well as application of the immunoregulatory effects of PDT [122, 123]. For example, PDT has been shown to suppress contact hypersensitivity (CHS) reactions in mouse models of type IV hypersensitivity [124, 125]. Some observations indicate that the immunosuppressive environment that develops after PDT results from activation of keratinocytes that secrete IL-10 [126], a cytokine with pleiotropic effects exerted in the cells of the immune system [127]. In a CHS model in mice verteporfin (BPD-MA) administration followed by the whole body irradiation led to induction of IL-10 expression in skin [125]. Moreover, treatment of wild type animals with neutralizing anti-IL-10 antibodies or administration of recombinant IL-12, a cytokine that promotes development of cellular immune response, reversed immunosuppressive effects of PDT [128]. However, not all studies confirm the importance of IL-10 in PDT-mediated suppression of the immune responses as no difference in CHS development in IL-10 knockout versus wild type mice has been observed [129]. The immunomodulatory effects of PDT result from induction of immunoregulatory CD4+ and CD8+ T cells as well as macrophages and can be adoptively transferred to naive recipients [130, 131]. All these observations inspired the use of PDT in the treatment of many dermatological as well as autoimmune diseases.

PDT is now approved for use or undergoing clinical trials in relation to managing a vast array of diseases, where elimination or regulation of the immune cell’s function is anticipated (Table 5). Depending on the disease pathogenesis, PDT serves as a local method of destruction of inflamed or involved tissue (e.g. PDT-mediated synovectomy in rheumatoid arthritis or treatment of psoriatic plaques) as well as immunosuppressive or immunomodulatory strategy. Extracorporeal photochemotherapy (ECP) based on UVA irradiation of 8-methoxypsoralen (8-MOP)-sensitized peripheral blood mononuclear cells (PBMC), serves as a promising and effective strategy in the treatment of graft versus host disease (GVHD, a common complication of bone marrow transplantation), prevention of allograft rejection or as an adjuvant therapy of several autoimmune diseases. However, further prospective, double-blind randomized studies are urgently needed to evaluate the exact role of PDT in the treatment of non-oncological diseases.
Table 5

PDT as a treatment modality in organ transplants and autoimmune diseases—summary of chosen clinical and preclinical trials

Disease/clinical problem

Treatment set-up

Results/comments

Refs.

Graft versus host disease

TH9402 (rhodamine derivative)—PDP

Prevention of GVHD and preservation of GVL effect in mice

[170]

Extracorporeal PBMC exposure (ECP) to 8-MOP and UVA

Clinical improvement in patients with severe steroid-refractory GVHD

[171, 172]

Clinical improvement in both acute and chronic GVHD patients with skin and visceral involvement

[172175]

Psoriasis

ALA—local irradiation

No or very poor clinical efficacy in plaque psoriasis patients. Frequent occurrence of severe pain at the site of illumination

[176, 177]

ALA—local irradiation

Clinical improvement in patients with plaque psoriasis

[178181]

ALA—local irradiation

Mild to moderate improvement in patients with palmoplantar pustular psoriasis

[182]

Methylene blue—local irradiation

Complete response in patients with resistant plaque psoriatic lesions

[183]

Methyl-ALA—local irradiation

No superior results of PDT in the treatment of nail psoriasis comparing with pulse dye laser light exposure in patients

[184]

Psoriatic arthritis

BPD-MA—half or total body irradiation with UVA

Safe and well-tolerated treatment, alleviation of the disease and significant pain reduction in 5 out of 17 patients

[185]

Rheumatoid arthritis

m-THPC—local irradiation

A potent method of synovectomy in a rat model

[186]

Polymeric photosensitiser prodrug targeting thrombin—local irradiation

Selective and minimally invasive synovectomy in mice, promising results in primary human synoviocytes in vitro

[187]

m-THPC—local irradiation

Significant reduction of arthritic score in mice

[188]

BPD-MA—PDT—local irradiation

Selective destruction of synovium in an antigen-induced arthritis in rabbits

[189]

BPD-MA—PDT—total body irradiation

Alleviation of signs of Freund’s complete adjuvant-induced arthritis in mice

[190]

Scleroderma

Topical ALA—local irradiation

Good therapeutic response in PUVA-resistant patients, no carcinogenic potential compared with PUVA

[191193]

ATX-S10(Na)—PDT

Enhanced modulation of dermal matrix components compared with ALA-PDT in vitro.

[194]

Alopecia areata

Topical ALA—local irradiation

An ineffective approach in AA. No hair growth was observed while a significant increase in erythema and pigmentation confirmed phototoxic PDT effects

[195]

Heart, lung and kidney transplantation

Extracorporeal PBMC exposure (ECP) to 8-MOP and UVA

Prevention and improved control of recurrent graft rejection

[172, 196199]

Type I diabetes mellitus

Extracorporeal PBMC exposure (ECP) to 8-MOP and UVA

Weak response in children with diabetes

[172, 200]

Multiple sclerosis

Extracorporeal PBMC exposure (ECP) to 8-MOP and UVA

Well tolerated, reduced relapse rate in 5 patients with relapsing-remitting MS

[201]

Moderate efficacy in chronic progressive MS patients with rebound after treatment discontinuation

[202]

8-MOP 8-methoxypsoralen, AA alopecia areata, ALA 5-aminolaevulinic acid, BPD-MA benzoporphyrin derivative monoacid, ECP extracorporeal photochemotherapy, GVHD graft versus host disease, GVL graft versus leukemia, MS multiple sclerosis, m-THPC meta-tetrahydroxyphenylchlorin, PBMC peripheral blood mononuclear cells, PDP photodynamic purging, PDT photodynamic therapy

PDT-based cancer vaccines

The ability of PDT to induce systemic anti-tumour immunity seems to result from its capacity to induce acute inflammatory reaction accompanied by ‘possible’ induction of ‘immunogenic’ cell death. The latter phenomenon has been exploited in the design of antitumor vaccines based on administration of tumour cell lysates generated by ex vivo PDT. Several studies have demonstrated that immunization with PDT-killed tumour cells (like EMT6, P815, SCCVII and H22) induces strong antitumor immunity which is superior to administration of tumour cell lysates generated by UV and ionizing irradiation or freeze-thawing [69, 89, 132]. Moreover, PDT-treated (SCCVII) tumour cells could also act as therapeutic anti-cancer vaccines in the absence of co-administration of adjuvants, to be effective [89]. The efficacy of PDT-generated anti-cancer vaccines in pre-clinical models has led to initiation of phase I clinical trial in cancer patients [133]. As discussed earlier, DAMPs released by PDT-stressed cancer cells might have a vital role to play in defining PDT’s cancer vaccine production capabilities and hence this angle needs to be studied further.

Conclusions

PDT is a promising anti-tumour therapeutic modality which has recently received considerable attention, as evidenced by the rising number of publications pertaining to it. It is only one of those few therapies which can induce all three types of cell death pathways (at least in vitro) thereby making it very versatile in terms of immune response that would be evoked upon induction of a particular cell death pathway. However, there is much work that needs to be done to address the influence of PDT-induced cell death pathways on the formation of potentially immunogenic environment suitable for the generation of systemic anti-tumour immunity. On the other hand, relatively much is known about the ability of PDT to modulate local inflammatory reaction or to directly influence the activity of immune cells. Mounting pre-clinical and clinical evidence has shown that if all pertinent factors and conditions are optimal then PDT can induce effective anti-tumour immunity. However, whether PDT can induce anti-cancer vaccine effect is still a matter of debate which needs immediate attention. In our opinion, the ability of PDT to induce cell death associated with DAMPs release or exposure in the plasma membrane, accompanied by induction of local inflammatory reaction, elimination of tumour-associated macrophages and/or regulatory T cells could be one of those pre-requisites, which make this treatment unique among other therapeutic procedures. It will be crucial in the nearest future to address all potential variables associated with this complex therapeutic strategy to fully exploit its potential effects on the development of anti-tumour immunity.

Acknowledgements

P.A. and A.D.G. are supported by the grants from the K.U.Leuven (OT49/06) and F.W.O. Flanderen (G.0661.09). This paper presents research results of the IAP6/18, funded by the Interuniversity Attraction Poles Programme, initiated by the Belgian State, Science Policy Office. J.G. and D.N. are supported by grants N40112331/2736 and R0504303 from Ministry of Science and Higher Education in Poland; the European Union within European Regional Development Fund through Innovative Economy grant POIG.01.01.02-00-008/08 and by the Foundation for Polish Science Team Programme co-financed by the EU European Regional Development Fund. J.G. is a recipient of the Mistrz Award from the Foundation for Polish Science. Lastly, all figures were produced using Servier Medical Art (www.servier.com) for which the authors would like to acknowledge Servier.

Copyright information

© Springer Science+Business Media, LLC 2010