Destruction of targeted lesions by localized phototoxic insult has become a clinically relevant modality, called photodynamic therapy (PDT), which is being increasingly used for the treatment of cancer and other indications.1–3 The phototoxic effect is achieved with PDT by using drugs (photosensitizers) that have the capacity of capturing energy delivered by light irradiation and transferring it to molecular oxygen with consequent local generation of singlet oxygen and other reactive species.1, 4 However, therapy outcome following cancer treatment by PDT is a result of a complex interaction of multiple contributory elements besides the direct phototoxic cancer cell kill; this includes the effects on tumor blood vasculature and other stromal constituents as well as the elicited host response.4, 5 Of relevance to the latter, PDT produces in treated solid tumors a strong oxidative stress with rapidly induced massive damage to cellular membranes and/or cytoplasmic structures predominating either in the parenchyma or in vascular endothelium (depending on the type of photosensitizer used). This results in a prompt and abundant release of proinflammatory mediators (alerting the host of the presence of local trauma threatening the integrity and homeostasis at the affected site), which provokes a strong acute inflammatory reaction.6 In addition, newly expressed PDT damage-related self molecules include those representing endogenous danger signals recognized by the immune system as altered self; the provoked response can culminate in the development of T cell-mediated tumor antigen-specific immunity.6, 7
In addition to inflammation, acute phase response8, 9 could be another major effector process mobilized by the host in response to tumor-localized insult inflicted by PDT. Systemic mobilization of host resources executed through the acute phase response could be important for optimal implementation of the host-protecting mechanism mounted following tumor PDT. Major manifestations of acute phase response are the activation of pituitary-adrenal hormonal axis, production and release of acute phase reactants and elevation in the levels of peripheral leukocytes (particularly neutrophils).8, 10 In recent studies, we have obtained evidence supporting the activity of adrenal hormones after tumor PDT and the existence of PDT-induced neutrophilia.11–14 To definitely establish that tumor PDT induces a fully engaged acute phase response, as well as to characterize it better and understand its role, it is necessary to document the participation of major acute phase reactants. Since our studies are based on mouse tumor models, we focused the investigation on serum amyloid P component (SAP) because this pentraxin protein is the prototypic acute phase reactant in the mouse.15, 16 Mannose-binding lectin-A (MBL-A), another important acute phase reactant with functional attributes similar to SAP,17, 18 was also included in this study.
Material and methods
Tumor model and therapy
Subcutaneous FsaR fibrosarcoma,19 the tumor model already used in our related studies,11, 14 was implanted into lower dorsal region of syngeneic C3H/HeN mice (7–9-week-old females) by injecting 1 × 106 cells from suspensions prepared by enzymatic digestion of tumor tissue.20 Cohorts of mice with these FsaR tumors were treated 7–8 days after implantation when their largest diameter reached 8 mm. The animal protocols were approved by the Animal Ethics Committee of the University of British Columbia. For PDT treatment, mice received Photofrin (provided by Axcan Pharma, Mont-Saint-Hilaire, Quebec, Canada) injected intravenously at 10 mg/kg, and 24 hr later their tumors were submitted to superficial illumination while the mice were restrained unanesthetized in holders exposing their backs. The light was produced by a FB-QTH high throughput illuminator (Sciencetech, London, Ontario, Canada) based on a 150 W QTH lamp and equipped with integrated ellipsoidal reflector and 630 ± 10 nm interference filter, and it was delivered through a 8-mm core diameter liquid light guide (model 77638 by Oriel Instruments, Stratford, CT) with a fluence rate 80–90 mW/cm2. The PDT dose used was 150 J/cm2. In some experiments, groups of mice were treated with either rat anti-mouse IL-6 antibody (Pierce Endogen, Rockford, IL) at 60 μg/mouse or mifepristone (Sigma Chemical, Saint Louis, MO) at 50 mg/kg; both were injected intraperitoneally at 30 min before the onset of PDT light treatment.
Serum (from whole blood collected by cardiac puncture) and tumor tissue samples were obtained from mice sacrificed at various time points after PDT. Serum aliquots and tumor tissue homogenates were used for determining the content of SAP, MBL-A or pentraxin 3 (PTX3) based on enzyme-linked immunosorbent assay (ELISA). Total protein content was determined from the aliquots of the tumor tissue samples using Pierce Modified Lowry Protein Assay kit (Fisher Canada, Nepean, Ontario, Canada) in order to express measured tumor levels of investigated proteins in micrograms per milligram of total tumor protein.
Mouse SAP ELISA was performed following the protocol described by Taktak and Stenning.21 Briefly, multi-well plates were coated with sheep anti-mouse SAP (Calbiochem, Merk KGaA, Darmstadt, Germany) at 1:1,000 dilution in PBS, test samples were added after blocking procedure for 1 hr incubation at 37°C, followed by staining with rabbit anti-mouse SAP antibody (Alpha Diagnostic International, San Antonio, TX) at 1:2,000 dilution in PBS, and then with HRP-conjugated donkey anti-rabbit IgG (Oncogene Research Products, Boston, MA, USA, 1:5,000 dilution). Finally, 0.4 mg/ml o-phenylenediamine (OPD, Sigma) was added for 20 min incubation that was stopped with a 150 μl aliquot of 1 M sulphuric acid. The mouse SAP standard was purchased from Calbiochem.
Mouse MBL-A ELISA kit (HyCult Biotechnology, Uden, The Netherlands) was used for determining the levels of this protein following manufacturer's instructions. The kit uses wells coated with rat anti-mouse MBL-A monoclonal antibody, and captured MBL-A from the samples is detected by staining with biotinylated monoclonal rat antibody to mouse MBL-A (directed against a different epitope than the capture antibody) followed by streptavidin peroxidase conjugate. The substrate in the last step is tetramethylbenzidine.
Mouse PTX3 ELISA was performed by coating (0.5 μg/well) with rabbit anti-mouse PTX3 (Alpha Diagnostic) and detection with goat anti-mouse PTX3 (R&D Systems, Minneapolis, MN, 1:1,000 dilution) followed by peroxidase-conjugated donkey anti-goat IgG (Jackson ImmunoResearch Laboratories, West Grove, PA, 1;5,000 dilution). The OPD substrate was employed in the last step as with the SAP ELISA. Recombinant mouse PTX3 (R&D Systems) was used for standard.
Sections (5 μm) cut from paraffin-embedded tumors were first rehydrated and treated with antigen retrieval citrate buffer. They were than stained with sheep anti-mouse SAP polyclonal antibody (Alpha Diagnostic, 1:500) or control sheep IgG (Jackson ImmunoResearch) followed by peroxidase-conjugated rabbit anti-sheep IgG (Calbiochem, 1:5,000). Sections were stained using a DAB chromagen and hematoxylin counterstain.
Real time RT-PCR
Expression levels of mouse genes encoding SAP, MBL-A and C-reactive protein (CRP) were determined in a quantitative real-time PCR by using primers designed based on complete genomic sequences found at the Entrez database search engine, constructed to be 17–22 nucleotides in length and devised to produce 75–200 bp long amplicons (Table I). Their suitability rating was checked using a web-based program NetPrimer (http://www.premierbiosoft.com/netprimer/netprlaunch/netprlaunch.html). The primers (ordered from Invitrogen) had amplification efficiencies ranging 90–100% determined based on standard curves for the assessment of the reaction optimization. Immediately upon collection, liver and tumor tissues were placed in ice-cold Trizol reagent (Invitrogen Canada, Burlington, Ontario, Canada) and homogenized. The homogenates were processed for total RNA isolation and phenol-chloroform purification followed by DNAse digestion and subsequent removal of protein and ion contaminants by another round of phenol-chloroform extraction. The RNA pellet was washed in 75% ethanol, air dried and dissolved in UltraPure DNase/RNase-free distilled water (Invitrogen). Two-step real-time quantitative reverse transcription PCR was performed with the total RNA samples (1 μg) using SuperScript III Platinum Two-Step qRT-PCR kit with SYBR Green (Invitrogen) following the manufacturer's instructions. In brief, first-strand cDNA was synthesized from total RNA using random hexamers and superscript III reverse transcriptase and employing a PCR 9700 thermocycler (Applied Biosystems, Foster City, CA) to provide the set parameters (10 min at 25°C, 50 min at 42°C and 5 min at 85°C). The RNA template was removed from cDNA:RNA hybrid by a 20 min incubation in presence of E. coli RNase H. For subsequent RT-PCR amplification run on the 7,500 Real-Time PCR System (Applied Biosystems), cDNA was mixed in SYBR Green qPCR SuperMix with forward and reversed gene-specific primers. The qPCR cycling parameters were 50°C for 2 min (uracil DNA glycosylase incubation), 95°C for 2 min (Taq DNA polymerase activation), and 50 cycles at 95°C for 15 sec and 60°C for 60 sec. Cycle threshold (cycle number at which the fluorescence signal from the amplified product surpasses background fluorescence) was determined for each reaction and the comparative threshold method22 was used for normalizing all the results to the expression of housekeeping gene glyceraldehide-3-phosphate dehydrogenase (GAPDH).
Table I. Description of Oligo-Nucleotide Primer Pairs Used in qPCR Reactions
Primer sequence (5′–3′)
Amp. Size (bp)
Listed are the investigated genes, their NCBI accession numbers, their respective primer pair sequences and alignments, their melting temperatures (Tm) in degrees Celsius, respective ratings based on NetPrimer software shown in percentage; optimal PCR MgCl2 concentrations and their resulting amplicon size.
For each time point, the samples were collected from 4 mice. The results were statistically evaluated based on Mann-Whitney test and significance level threshold of 5% was set for determining if the groups were statistically different.
Blood samples, collected from mice bearing subcutaneous FsaR tumors treated by PDT at different time intervals after the termination of photodynamic light treatment, were used for determining serum SAP levels using the ELISA assay. The results reveal that tumor localized PDT (using a dose producing >50% tumor cures) induced a rise in serum SAP levels of the host mice that was evident at 18 hr and peaked at 24 hr after therapy with around 50-fold increase compared to the level found in untreated mice (0.53 mg/ml) (Fig. 1a). Serum concentration of this key acute phase reactant then gradually decreased but still remained elevated at 3 days post PDT, and declined to the pretreatment range around 1 week after therapy.
The profile of serum concentration changes of another important acute phase reactant, MBL-A, in mice bearing PDT-treated FsaR tumors was also determined by ELISA (Fig. 1b). The rise in MBL-A levels peaked earlier than SAP at around 18 hr after tumor PDT light treatment, but the increase was markedly more modest than SAP (up to 3-fold). However, the elevated MBL-A levels persisted even after 1 week post therapy.
Examined next was the concentration of these 2 acute phase reactants in PDT-treated tumors. For that purpose, mice were sacrificed and their FsaR tumors excised at different times after PDT. The homogenates of these tumor tissues were then used for SAP and MBL-A ELISA assays. The results for SAP reveal that tumor levels of this protein showed signs of increase already at 2 hr post PDT and reached 7–8-fold rise at 18 hr post PDT (Fig. 2a). Since tumors completely regressed after 24 hr post PDT, samples were not collected past that time point. The accumulation of SAP in PDT-treated FsaR tumors was also demonstrated by immunohistochemistry. In contrast to a faint SAP staining in sections from untreated tumors, a strong positive staining was evident with samples prepared from tumors excised at 24 hr after PDT; the localization of widely distributed SAP deposits appears either cytoplasmic (extranuclear) or more likely on cellular surfaces (Fig. 2b). The specificity of SAP immunoreactivity was verified by the absence of positive color in controls stained with nonspecific sheep IgG or with omitted primary antibody (not shown).
Tumor MBL-A content, assessed by ELISA, became significantly elevated already at 3 hr post PDT and reached a maximum (over 3-fold) increase 2 hr later; this was followed by a decline and return to pretreatment levels at 24 hr post PDT (Fig. 2c).
Related to SAP is another protein of pentraxin family PTX3, whose upregulation in acute phase response can be induced by inflammatory mediators.23 Measurement of PTX3 content by ELISA established that no significant serum release of this protein was induced by PDT and no accumulation in PDT-treated FsaR tumors after therapy (not shown).
To better understand tumor PDT-induced upregulation of acute phase reactants SAP and MBL-A, we used the RT-PCR based analysis to examine the expression of genes encoding these proteins in liver and tumor tissues excised from mice bearing PDT-treated FSAR tumors that were sacrificed at various time intervals after the therapy. An increase in the expression of SAP gene in both liver and tumor tissue was evident already at 4 hr after PDT and became strongly pronounced at 8 hr after PDT with the extent of upregulation about twice greater in the liver than in the tumor (Fig. 3a). A further dramatic increase in the expression of this gene was detected at 24 hr after PDT reaching over 170-fold in the tumor and almost 150-fold increase in the liver. Similar upregulation kinetics was found for the MBL-A gene (albeit on a more than 10 times lower scale), with the exception of the expression in the liver at 24 hr after PDT which subsided to the pretreatment levels (Fig. 3b).
The same liver and tumor tissue homogenates were used also for determining tumor PDT-induced changes in the expression of gene encoding pentraxin CRP, which is a dominant acute phase reactant in humans in contrast to its reputedly modest role in mice.24 Tumor PDT induced only a minor (less than 2-fold) increase in the expression of liver CRP gene. However tumor CRP expression, after a weak upregulation at 4 hr post PDT, continued to increase at later time points reaching almost 7-fold difference at 24 hr post PDT (Fig. 3c).
Cytokine IL-6 is known as the main mediator of the induction of various acute phase reactants, particularly those produced in the liver.25 The administration of rat antibody neutralizing mouse IL-6 to mice at 30 min before tumor PDT strongly attenuated the induction of liver SAP gene upregulation assessed at 24 hr post therapy (Fig. 4). The same treatment with nonspecific rat immunoglobulin produced no significant effect on liver SAP gene expression (not shown). The treatment with IL-6 neutralizing antibody had no significant effect on PDT-induced tumor SAP gene upregulation at the same time point (Fig. 4). As shown in the same figure, treatment of mice with a standard glucocorticoid receptor antagonist mifepristone given 30 min before tumor PDT strongly inhibited the induction of SAP gene upregulation in the liver and almost completely blocked it in the tumor. The used dose of mifepristone (50 mg/kg) effectively blocks the effects of glucocorticoids in mouse models.26 Glucocorticoid hormones thus emerge as major mediators (together with IL-6) of tumor PDT induced SAP gene upregulation in the liver, and as even more important promoters of the upregulation of this gene in PDT-treated tumors. It remains to be verified whether other signals, including those derived from cancerous membrane lipids, transmitted to the liver from the inflamed PDT-treated tumor have a significant role in the upregulation of various acute phase response genes.
Although PDT used for treatment of solid cancerous lesions is not systemic but localized therapeutic modality delivering light-focused insult at the targeted site, it is becoming increasingly clear that this treatment modality has a pronounced systemic impact. The systemic effects of tumor PDT are invariably associated with the implementation of the elicited host response. The acute inflammatory response, although focused at the PDT-treated tumor, involves sequestration of cellular and acellular inflammatory effectors from the circulation and is associated with the release of cytokines and other mediators from the affected site.6 The engendered immune response recognizing PDT-treated tumor as its target has the capacity to act against remote deposits of the same cancer and is associated with the generation of immune memory cells that can be found at distant lymphoid sites.7, 27–29 The present study definitely establishes the induction by tumor PDT of another important systemic reaction, the acute phase response.
While the existence and consequences of local oxidative stress inflicted at the cellular level at the PDT-treated site have already been recognized for some time, it is now becoming clear that a stress response at the systemic level (which is one of the definitions of acute phase response) is also triggered by this modality. Acute phase response denotes a dynamic homeostatic process involving multiple organ systems with systemic changes distant from the local insult site whose purpose is to create an overall protective systemic environment required for coping with tissue injury.8–10
A major feature of acute phase response is the release in the circulation of acute phase reactants, which are defined as proteins whose plasma concentration increases or decreases by at least 25% but the magnitude of increase with some of them could be well over 100-fold. This mainly reflects radically changed biosynthetic profile of the liver, but the production of acute phase reactants occurs also extrahepatically.9 While the hallmark acute phase reactant in humans is CRP, in the mouse this role is taken over by SAP.16, 30 The results of the present study demonstrate that tumor PDT treatment induced almost 150-fold increase in the expression of SAP gene in the liver of host mice, while serum levels of this protein increased around 50-fold at the peak interval around 24 hr post PDT. Conceivably, the released SAP does not remain in the circulation but is attracted to the PDT-treated tumor (as elaborated below). Notably, serum SAP elevation was also reported following PDT treatment of normal peritoneal tissue of mice, but the increase was around 10 times lower than we documented after tumor PDT treatment.31
While SAP is probably the most dramatically upregulated acute phase reactant following tumor PDT in mouse models, other acute phase proteins are also induced after this therapy. Evidence is now available of the induction after the same therapy treatment of at least several additional well known acute phase reactants. Mice bearing PDT-treated tumors were reported to exhibit elevated serum levels of complement C3 protein32 and haptoglobin,33 and the present study demonstrates the increase post tumor PDT in serum levels of another complement protein, MBL-A.
Acute phase reactants perform a wide range of functions enabling optimized enactment of elicited host response and regulation of its course. Some of them are proteinase inhibitors that prevent the excess activity of inflammatory enzyme cascades, other act as opsonins of dead cells and debris to facilitate their removal, or influence the initiation and type of adaptive immune response, while some promote wound healing.10 Pentraxins SAP and CRP, as well as complement proteins C3 and MBL-A, belong to acute phase reactants that opsonize dead cells to accentuate their allurement as targets for professional phagocytes securing efficient removal of these corpses.34, 35 With PDT, the host is suddenly faced with a burden of large numbers of killed and dying cells emerging in the targeted tumor within hours after treatment. Their efficient and swift nonphlogistic disposal is a key element of PDT-induced host response and is of critical importance for the therapy outcome.6 The accumulation of SAP and MBL-A proteins in PDT-treated tumors demonstrated in the present study and the same previously shown for C3 protein32 is in the accordance with their role in dead cell removal.
However, our finding that SAP and MBL-A genes are upregulated post PDT not only in the liver but also in the treated tumors suggests that not all of the tumor-concentrated protein is of hepatic origin. At 24 hr post PDT, the extent of tumor SAP gene upregulation exceeds that of liver SAP. Even more dramatically, at this time point the expression of liver MBL-A genes declines to the pretreatment range while tumor MBL-A gene upregulation reaches its peak largely surpassing the maximal extent of liver MBL-A gene upregulation seen at 8 hr post PDT (Fig. 3b). Hence, the production of these acute phase reactants in PDT-treated tumors appears to make important contribution to their local availability. Moreover, MBL-A produced in PDT-treated tumor site may represent an important source of serum concentrations of this protein that remain elevated even at 1 week after therapy, i.e., long after the liver MBL-A gene expression declined to preupregulation levels. Importantly, at these late time points tumor tissue is completely or largely eradicated but the likely source of MBL-A production are activated immune cells accumulated at the PDT treatment site. A related recent work has shown that tumor PDT induces the upregulation of C3 gene in the treated tumors without significant increase in the expression of this gene in the liver, and indicated that the major source are tumor-associated macrophages.36
A significant finding in this study is that the gene encoding pentraxin CRP, which is known as minor acute phase reactant in mice,24 is upregulated almost 7-fold in the tissue of PDT-treated tumors. Such response alerts to possible participation of locally- (treatment site) produced CRP in tumor PDT-induced host response in mice, while in humans a dramatic upregulation of this important immune mediator and dominant involvement following clinical PDT can be predicted. Interestingly PTX3, pentraxin which unlike SAP and CRP directly inhibits the cross-presentation of tumor antigens from captured apoptotic cells to T lymphocytes,37 was found not to be engaged as tumor PDT-induced acute phase reactant in the examined FsaR tumor model.
The identification of IL-6 and glucocorticoid hormones as major mediators responsible for tumor PDT-induced upregulation of liver SAP gene is not surprising. The inflammatory cytokine IL-6, known to be released following tumor PDT,38 is recognized as a principal inducer of acute phase response.25 The lack of the effect of IL-6 at the tumor site is in accordance with the reported PDT-induced loss of membrane receptors for this cytokine in directly treated cells.39 Adrenal corticoid hormones, released as a consequence of hypophyseal-pituitary-adrenal axis activation typical for acute phase response,8 were already shown to be involved in tumor PDT response.11 They mediate well-defined transcriptional effects (including transactivation or transrepression of various inflammatory/immune genes) by binding to a cytoplasm-localized glucocorticoid receptor.40 More striking is the revelation that glucocorticoid hormones emerge as major inducers of SAP gene upregulation in PDT-treated tumors (Fig. 4). This suggests that glucocorticoids induced following tumor PDT in the context of acute phase response impact the transcription of host response-relevant genes in tumor tissue, which (because of the established potency of these hormones) is obviously an aspect of considerable relevance that has not been recognized until now.
In conclusion, it is unambiguously established that tumor PDT induces in the hosts a strong acute phase response. This has become definitely clear from the evidence presented in this study, but is supported by the demonstration of tumor PDT-induced systemic SAP upregulation in a different standard mouse strain and unrelated tumor model Lewis lung carcinoma.41 Also reinforcing this conclusion are recently published findings of the induced activity of glucocorticoid hormones and peripheral blood neutrophilia11, 12 with this tumor treatment modality. This revelation unveils a pivotal element of tumor PDT response whose potential impact on therapy outcome with this cancer treatment modality needs to be taken into consideration. Acute phase response participation in response to some other forms of cancer therapy with inflammatory/immune reaction-inducing attributes should be also considered.
The authors thank Dr. Peter Payne for help in immunohistochemistry experiments.