Enhancement of cisplatin-based TACE by a hemoglobin-based oxygen carrier in an orthotopic rat HCC model

Abstract

Objective: Hypoxic tumor cells are more resistant to standard chemotherapies. A number of studies indicated that improving oxygenation inside the tumor could serve as a potential strategy to target hypoxia-induced chemoresistance. In this study, we examined whether a hemoglobin-based oxygen carrier (OC89) could increase tumor oxygenation and thus enhance the efficiency of transarterial chemoembolization (TACE) in an orthotopic rat HCC model. Methods: Efficiency of the hemoglobin-based oxygen carrier (OC89) in improving tumor oxygenation was examined by OxyLab pO₂. Sensitization of chemotherapy (cisplatin) in TACE by OC89 was evaluated in four different therapeutic regimens including cisplatin (1 mg/kg) + OC89 (0.2 g/kg), cisplatin (1 mg/kg) + OC89 (0.4 g/kg), cisplatin (3 mg/kg) + OC89 (0.2 g/kg), cisplatin (3 mg/kg) + OC89 (0.4 g/kg). For all the therapeutic regimens, a single delivery of OC89 via the tail vein was performed 1 h before TACE. Results: Compared with Ringer's buffer, systemic delivery of OC89 (0.4 g/kg) attenuated tumor hypoxia (p < 0.05). Additionally, partial pressure of oxygen (pO₂) fraction of low readings (0–10 mmHg) inside the tumor decreased from 74.1% to 24.6% after OC89 delivery, while pO₂ fraction of high readings (15–25 mmHg) increased from 22.2% to 41.5%. When cisplatin was combined with OC89, regimen cisplatin (3 mg/kg) + OC89 (0.4 g/kg) resulted in a significant inhibition of tumor growth at Day 21 after therapy (p < 0.05). Further investigation indicated that OC89 delivery influenced anti-apoptotic and pro-apoptotic balance of the UPR pathway in the tumor. Conclusions: Our data suggest that targeting tumor hypoxia with the hemoglobin-based O₂ carrier serves as a promising approach to enhance the efficacy of cisplatin-based chemotherapy in HCC.

Keywords::

• chemoresistance
• GRP78
• hypoxia
• oxygenation
• oxyLab pO₂
• UPR pathway

Introduction

Hepatocellular carcinoma (HCC) is one of the leading causes of cancer death worldwide. Although chemotherapeutic regimens such as transarterial chemoembolization (TACE) have been developed over the past several decades, unfortunately, HCC was observed clinically to be resistance to a wide variety of cytotoxic agents such as cisplatin (Lo et al. 2002, Sangro et al. 2002, Llovet and Bruix 2003, Siddik 2003, Bruix et al. 2004). Therefore, the strategy to improve the efficacy of current chemotherapeutic regimens for HCC is urgently needed.

Hypoxia, as a major component of microenvironment in solid tumors, has been shown to exert important roles in diverse aspects of tumor biology including response to chemotherapeutic agents (Kennedy et al. 1980, Teicher 1994, Knisely and Rockwell 2002, Siddik 2003, Janssen et al. 2005, Song et al. 2006). Emerging lines of evidence demonstrated attenuation of hypoxia inside the tumor tissue with hemoglobin-based oxygen carriers may serve as alternative strategy to improve the chemosensitivity of cancer cells (Teicher et al. 1995, Robinson et al. 1995, Linberg et al. 1998, Yu et al. 2007, Gundersen and Palmer 2008). However, no study has been reported to examine the efficacy of such an approach in the therapy of HCC.

In this study, we evaluated whether the hemoglobin-based oxygen carrier could improve cisplatin-based TACE in an orthotopic rat HCC model. The tested solution OC89 is a novel product of the New A Innovation Limited (Hong Kong). Two preliminary studies performed in a Good Laboratory Practice (GLP)-compliant laboratory in China demonstrated OC89 could safely improve oxygenation in rat and dog models yet showing minimal adverse effects such as vasoconstriction or kidney toxicity.

Materials and methods

Test solution

OC89 is a cross-linked tetrameric hemoglobin purified from the red blood cells of bovine whole blood (Patent no.: US 7 989 593 B1). In this solution, the tetrameric form of hemoglobin is stabilized by DBSF and the dimeric form is removed by tangential flow filtration. Reactive sulfhydryl groups are blocked by iodoacetamide. The properties of OC89 are: concentration 6.5 ± 0.5 g/dL, methemoglobin: < 5%, endotoxin: < 0.05 EU/mL, osmolality: 300 ± 50 mOsm/kg, pH: 7.4 ± 0.2, average molecular weight: 65 ± 5 kDa; P₅₀: 20–35 mmHg. The solution was kept refrigerated at 4°C and protected from sunlight.

Cell line

Buffalo rat hepatoma cell line McA-RH7777 was purchased from American Type Culture Collection (ATCC, Manassas, VA, USA) and stably-labeled with luciferase gene. Briefly, the cells were transfected with pGL3 vector (Promega) and positive clones were selected according to luciferase activity in Xenogen IVIS 100 (Xenogen). The labeled cells were then maintained in DMEM medium (Gibco^®, Long Island, NY, USA) supplemented with 10% fetal bovine serum (Gibco^®, Long Island, NY, USA) at 37°C in a humidified, 5% CO₂ atmosphere.

Human HCC sample collection

Twenty-five patients with recurrent HCC within 4 years after resection of primary HCC between June 2001 and June 2008 were recruited from Sun Yat-Sen University Cancer Center (Guangzhou, China) (Chen et al. 2011). All these patients underwent prechemotherapy biopsies and then received about three cycles of TACE for the recurrent HCC. TACE treatment was performed with a regimen comprising cisplatin (CDDP), 5-fluorourail (5-FU), and an emulsion of doxorubicin (DOX) and lipiodol. The median dose of lipiodol was 10 mg (5–20 mg). The doses of the chemotherapy agents were: DOX 60 mg/m², CDDP 120 mg/m², and 5-FU 600 mg/m² every month for up to three cycles. Based on the response to TACE, the patients were divided into either a TACE-sensitive (12 patients) group or a TACE-resistant (13 patients) group. All the recruited patients gave their informed consent and the study was approved by the Committees for Ethical Review of Research involving Human Subjects at both the University of Hong Kong and the Cancer Center of Sun Yat-Sen University, China.

Cisplatin-resistance assay under hypoxia condition

Hypoxic condition was achieved with the hypoxic chamber (BioSpherix, OxyCycler-C42). Briefly, McA-RH7777 cells were seeded in 96-well plates. After overnight incubation, cisplatin (final concentration of 2.5 ug/ml for 24 h and 0.625 ug/ml for 48 h, respectively) was added in the culture medium. Then the cells were placed under either normoxic (Nx, 21% O₂) or hypoxic (Hx, 1% O₂) condition for 24 or 48 h.

Cell proliferation assay was performed using Cell Titer 96^® AQ_ueous One Solution Cell Proliferation Assay (Promega, Mandison. WI, USA) as instructed by the manufacturer. Each assay was repeated for thrice in triplicate.

Establishment of an orthotopic HCC model in buffalo rats

Male buffalo rats (4–6 weeks old, 250–300 g) were used. Briefly, McA-RH7777 cells (˜1 × 10⁶) in 0.2 mL culture medium were injected into the subcapsular space of the left liver lobe of buffalo rats (Man et al. 2007). The tumor inoculum was collected 2 weeks later and then minced into small pieces approximately 3 mm cube in size. One small piece of tumor tissue was subsequently implanted into the parenchyma of left liver lobe in the recipient rat. Tumor formation was confirmed with higher-resolution 7T magnetic resonance imaging (MRI) about 2 weeks after implantation. Rats were housed in a standard animal laboratory with free activity and access to water and chow. They were kept under constant environment conditions with a 12 h light–dark cycle. Pentobarbitone sodium (40 mg/kg) was given intraperitoneally before any surgical procedure. Carprofen (0.1 mg in 100 ml drinking water) were used to relieve the pain for the first 3 days after operation. All animal studies were conducted according to the animal ordinance set by the government of Hong Kong.

Partial pressure of oxygen (pO₂) measurement inside rat HCC tissues

Real-time pO₂ measurement was performed with OxyLab pO₂ on Day 14 after tumor implantation (Urano et al. 2002). Briefly, the rat was anesthetized with pentobarbitone sodium (40 mg/kg, i.p.) and placed on a heating pad to maintain a core body temperature of 37°C. Then laparotomy with median incision was made and the tumor in the liver was exposed. To facilitate the penetration of the probe, a 27 G ½ needle was used to create a track of approximately 1 cm in length inside the tumor. The OxyLite probe was then inserted into the track and retracted 1 mm when it reached the opposite end to minimize the compression of tissue on the probe tip. Measurement started when pO₂ values stabilized for 5 min at the first measurement point after insertion. Then the probe was retracted by a NanoDirect positioner at 1 mm intervals and pO₂ values at each point were recorded accordingly. The probe stayed at each point for 10–20 s to allow for stabilization of the measurement before recording. This procedure was repeated till the probe went through the whole track. Usually several tracks (3–5) at different planes were carefully produced in each tumor, and for each track readings at 5–8 points were documented. Then the incision was closed temporarily. Either Ringer's buffer or OC89 was given via the tail vein of the rat. One hour later, pO₂ inside the same tumor was examined with the aforementioned procedures again. The duration of measurement for either pre- or post-Ringer's buffer/OC89 delivery was limited to 30–40 min. The rats were sacrificed with overdose of pentobarbitone sodium (100 mg/kg, i.p.) after measurement.

Cisplatin-based TACE

Cisplatin-based TACE was made as reported previously (Maataoui et al. 2005). Briefly, after exposure of the liver tumor, both the left branch of portal vein and the left branch of hepatic artery were carefully identified and prepared for further manipulation. Then cisplatin was injected via the left branch of portal vein (with the right branch of portal vein being temporarily clamped) and ligation of the left branch of hepatic artery was performed subsequently. Prior to TACE, tumor volume (before treatment) was estimated by the formula, ab²/2, where a and b were the largest and smallest superficial visible diameters on the ventral side of the liver lobe (Carlsson et al. 1983).

Chemotherapeutic regimens

Buffalo rats with implanted tumors were randomized into two groups, with six rats in each group. Then either Ringer's buffer or OC89 was given via the tail vein, respectively. One hour later, cisplatin-based TACE was performed in both groups. Two dosages were selected for either cisplatin (1 and 3 mg/kg) or OC89 (0.2 and 0.4 g/kg), and altogether four different chemotherapeutic regimens were examined: A: cisplatin (1 mg/kg) plus OC89 (0.2 g/kg) or buffer; B: cisplatin (1 mg/kg) plus OC89 (0.4 g/kg) or buffer; C: cisplatin (3 mg/kg) plus OC89 (0.2 g/kg) or buffer; and D: cisplatin (3 mg/kg) plus OC89 (0.4 g/kg) or buffer.

Imaging analysis and sample collection

Tumor growth was monitored by either higher-resolution 7T magnetic resonance imaging (MRI) scanner, PharmaScan 70/16 (Bruker BioSpin), or Xenogen IVIS.

Buffalo rats were sacrificed at Day 1 or Day 21 (endpoint) after TACE. Both tumor and normal liver tissues were collected and stored at − 80°C or fixed in neutral formalin for further analysis. For the endpoint evaluation, the tumors were macroscopically dissected from rat liver after sacrifice and tumor volume was calculated by the formula, πcde/6, where c, d, and e were the three mutually perpendicular diameters.

Quantitative real-time PCR

Total RNA of either McA-RH7777 cells or tissues was extracted using Trizole reagent (Invitrogen). One microgram of extracted RNA from each sample was used to perform reverse transcription reaction using the TaqMan Reverse Transcription kit (Applied Biosystems). Real-time quantitative PCR was performed using SYBR Green Reagent (Applied Biosystems) in ABI PRISM7700 Sequence Detection System as described previously (Cheng et al. 2010). Primer sets for rat GRP78 were TCACGTGTCTTGGGGTCAGGG (forward) and CCAGATGTGCATGAC CCAACCGT (reverse); primer sets for rat CHOP were ATGTTAAAGATGAGCGGGTGG CAGC (forward) and TTGAACACTCTCTCCTCAGGTTCCA (reverse). 18S was used as the internal control. All samples were detected in triplicate. Gene expression levels were calculated after normalization to internal control.

Immunohistochemistry

Immunostaining was performed using Envision + System-HRP (DAB) (Dakocytomation, CA, USA). Briefly, formalin fixed, paraffin wax embedded, and 4 μm-thick sections were deparaffinized with xylene and then rehydrated with a graded series of ethanol. Subsequently antigen retrieval was performed with a microwave in the citrate buffer (pH 6.0) at 95°C for 15 min. After cooling down to room temperature, the slides were removed and washed with PBS. Endogenous peroxidase activity was quenched by incubating the slides with 3% H₂O₂ for 10 min. The slides were washed well, then incubated with primary antibodies (Hypoxyprobe-1 against pimonidazole HCl (HPI), 1:100; Ki-67 (BD pharmingen), 1:300; GRP78 (Abcam) 1:100; CHOP (Santa cruz) 1:100), and HRP-labeled polymer using two sequential 40 min incubations at room temperature. The staining was developed by incubation with DAB substrate-chromogen for 5–10 min and a hematoxylin counterstain was applied afterwards. The negative control was set up by replacing primary antibody with PBS. The immunoreactivity with regard to the intensity was assessed on a three-level scale of 0, 1, and 2 (no staining, weak staining, and strong staining, respectively). The proportion of positive cells was assessed semi-quantitatively on a four-level scale of 0, 1, 2, and 3 (0, 1–10%, 11–50%, and 51–100%, respectively). The cumulative scores were calculated by multiplying scales of intensity and percentage of positivity.

TUNEL

Apoptosis in tumor tissues was evaluated using In Situ Cell Death Detection Kit (Roche). Briefly, formalin fixed, paraffin wax embedded, and 4 μm-thick sections were deparaffinized with xylene and then rehydrated with a graded series of ethanol. Subsequently antigen retrieval was performed with protease K (20 ug/ml) digestion at 37°C for 30 min. Then the reaction mixture was applied and the sections were incubated at 37°C for 30 min. After washing with PBS, the activity of endogenous peroxidase was blocked with 0.3% hydrogen peroxide for 30 min at room temperature. The signal was developed by incubation with POD for 30 min followed by addition of freshly prepared DAB solution. The nuclei stained as dark brown were considered positive. Positive tumor cells were counted in 10 different fields at high magnification (400×) and apoptotic index was defined as the number of apoptotic cells per 1000 tumor cells.

Statistical analysis

Data analysis was performed using SPSS^® for Windows version 16.0 (SPSS, Inc., Chicago, IL, USA). Student's t-test was used and p values less than 0.05 were defined as significant.

Results

Hypoxia increases cisplatin-resistance in cancer cell McA-RH7777

To examine whether hypoxia may affect cellular sensitivity to cisplatin, McA-RH7777 cells were placed in either normoxic or hypoxic condition in the presence of cisplatin for 24 or 48 h. Then cell survival was determined by MTT assay. As shown in Figure 1, compared with the normoxic counterpart, more cancer cells survived the cytotoxicity of cisplatin under hypoxic condition (24 h: 94% vs. 57%, p < 0.05; 48 h: 96% vs. 54%, p < 0.05).

Figure 1. Decreased cisplatin-sensitivity of McA-RH7777 cells under hypoxia. Rat hepatoma cell line McA-RH7777 was cultured in either normoxic (Nx) or hypoxic (Hx) condition (1%) in the presence of cisplatin for 24 h or 48 h. Cell survival was then assessed by MTT assay. *p < 0.05.

OC89 improves tumor oxygenation and alters pO₂ fraction in tumor

Before evaluating the efficiency of OC89, we examined pO₂ distribution inside tumor tissue in five rats with implanted HCC. Totally measurements at 252 points were recorded. The pO₂ values ranged from 0.1 to 50.9 mmHg with the mean and median values of 7.67 mmHg and 2.62 mmHg, respectively. The majority of measurements were between 0–10 mmHg (79.7%), indicating there was an extensive oxygen deficiency inside the tumor tissue.

To examine whether OC89 could carry O2 to tumor tissues, we monitored pO₂ values inside the same tumor both before and 1 h after OC89 delivery. Ringer's buffer was used as the control for OC89. Attempts were made to ensure all measurement-related parameters such as the numbers of tracks, points to be examined and duration of each measurement were comparable. The recorded readings were summarized in Table I. Compared with Ringer's buffer, administration of OC89 (0.4 g/kg) via the tail vein resulted in a significant improvement of oxygenation in the tumor tissue (Figure 2a). Additionally, pO₂ fraction of low readings (0–10 mmHg) decreased from 74.1% to 24.6% after OC89 delivery, while pO₂ fraction of high readings (15–25 mmHg) increased from 22.2% to 41.5% (Figure 2b). The hypoxic status inside tumor tissues before and after OC89 delivery was also evaluated by detecting the protein adducts of hypoxia marker pimonidazole HCl. As shown in Figure 2c, in contrast to the control rats, less binding of pimonidazole was observed in rats receiving OC89.

Figure 2. Attenuation of hypoxia by OC89 in tumor tissues. The rats were divided into two groups, being given either Ringer's buffer or oxygen carrier via tail vein injection. pO₂ readings were measured both before and 1 h after interference. (a) Improved oxygenation by OC89. The data were shown as mean ± SD. * p < 0.05. (b) Shift of pO₂ fraction from low (0–10 mmHg) to high readings (15–25 mmHg) after OC89 delivery. (c) Expression of hypoxia marker pimonidazole after interference. Immunostaining showed less binding to pimonidazole (lower panel) in tumor tissue after delivery of OC89 (Bar = 100 um).

Table I. pO₂ readings (mmHg) inside tumor tissue before and after OC89 delivery.

	Pre-intervention			Post-intervention
	Ringer's buffer group	OC89 group	P value	Ringer's buffer group	OC89 group	P value
Range	− 0.7–30.9	− 0.6–29.8		− 0.7–26.8	1.7–51.3
Mean^¶	8.1 (± 7.7)	9.2 (± 7.0)	0.368	7.4 (± 6.5)	23.3 (± 12.6)	0.000
Median^§	6.8	9.4	0.231	6.7	22.7	0.000

Rats were divided into two groups (three in each group) and either Ringer's buffer or OC89 was given. pO₂ readings before any interference in the two groups was comparable while a significant upregulation of pO₂ readings in tumor tissues was observed 1 h after intervention with OC89.

^¶T-Test; ^§Mann–Whitney U test.

OC89 enhances efficacy of cisplatin-based TACE in rat HCC model

Based on the above results, we further explore whether OC89 could enhance the efficacy of cisplatin–based TACE in the rat HCC model. Totally four different therapeutic regimens were tested (Table II) and the changes of tumor volume at Day 21 (endpoint) after treatment were evaluated. Of all the regimens, combination of 3 mg/kg of cisplatin with 0.4 g/kg of OC89 demonstrated the most significant effect in delaying tumor growth (p = 0.03) (we performed subsequent functional study by using samples collected from this group unless otherwise indicated). In addition, there was a trend that tumors in rats that received either high dosage of cisplatin (3 mg/kg) or its combination with low dosage of OC89 (0.2 g/kg) also progressed more slowly compared with other groups, even though the volume changes were not statistically significant.

Table II. Tumor volume changes in different therapeutic regimens.

		Tumor volume (mm^3)
Groups	Regimens	Pre-treatment	Post-treatment	P value
A	Cisplatin(L) + Ringer(L)*	790 ± 417	4910 ± 1201	0.65
	Cisplatin(L) + OC89(L)	810 ± 403	5520 ± 1738
B	Cisplatin(L) + Ringer(H)^†	790 ± 410	6700 ± 989	0.10
	Cisplatin(L)+ OC89(H)	680 ± 257	3000 ± 1470
C	Cisplatin(H) + Ringer(L)	750 ± 280	1540 ± 998	0.24
	Cisplatin(H) + OC89(L)	500 ± 359	680 ± 570
D	Cisplatin(H)+ Ringer(H)	620 ± 251	1980 ± 1020	0.03
	Cisplatin(H) + OC89(H)	570 ± 294	410 ± 187

Four regimens were tested and the changes of tumor volumes before (Day 0) and after therapy (Day 21) were compared. Cisplatin was given at 1 mg/kg (low dose) or 3 mg/kg (high dose); OC89 was given at 0.2 g/kg (low dose) or 0.4 g/kg (high dose). Ringer's buffer of the same volume as OC89 was delivered accordingly. Pretreatment tumor volume was estimated by ab²/2; posttreatment tumor volume was calculated by πcde/6. The data were shown as mean ± SD and analyzed with T test.

*L: low dose; ^†H: high dose.

Tumor growth was also evaluated by gross and histological examination. The results of two representative cases in Group D (receiving combination of 3 mg/kg of cisplatin with 0.4 g/kg of OC89) are shown in Figure 3. In the rat from control group, the signals of tumor cells were still very prominent 21 days after treatment (Figure 3a). Microscopically, the bulk of tumor was still present and necrosis occurred but was incomplete (Figure 3b, left panel). In contrast, tumor became undetected by imaging systems in rats from OC89 group. Under light microscope, no tumor cells were found inside the liver tissue and only granuloma and proliferative fibrous tissues were identified in the site where the tumor inoculum was originally implanted (Figure 3b, right panel).

Figure 3. Enhanced therapeutic effect in TACE by OC89. (a) MRI or xenogen images in rats before or after TACE with or without OC89 delivery. The images were taken at Day 0 (left panel) or 21 days after TACE (middle and right panel). (b) Macro- and microscopic evaluation of tumors 21 days after TACE. In OC89-treated rat, the implanted tumor was replaced by granuloma (lower right, arrow) and proliferative fibrous tissue; while in rats receiving buffer, living tumor cells (lower left, solid arrow) accompanied with partial necrosis (lower left, dashed arrow) were observed. Bar = 20 um. (c) and (d) Altered expression of GRP78 and CHOP in rats at Day 1 after TACE. Bar = 100 um. (e) and (f) Increase of apoptosis (TUNEL assay, Day 1) and inhibition of proliferation (Immunostaining of Ki-67, Day 21) by OC89 in tumor cells after TACE. Bar = 100 um. *p < 0.05.

OC89 delivery knocks down the balance of UPR pathway

The Unfolded Protein Response (UPR) pathway is an important adaptive program to maintain cellular homeostasis under cellular stresses such as hypoxia. As the real-time measurements indicated that systemic delivery of OC89 could attenuate hypoxia and improve oxygenation of the tumor tissue, we investigated whether OC89 exerted its role via the UPR pathway. Here we evaluated the expression levels of prosurvival factor GRP78 and proapoptotic executioner CHOP in the UPR pathway at both mRNA and protein levels. Real-time PCR demonstrated that there was a decrease in GRP78 mRNA (p < 0.05) and an increase in CHOP mRNA (p < 0.05) in tumor tissues at Day 1 in OC89-treated rats (Figure 3c). Corresponding downregulation of GRP78 and augmentation of CHOP at protein levels were also identified by immunostaining in the tumor tissue (Figure 3d).

OC89 delivery increases tumor apoptosis and inhibits tumor cell proliferation

GRP78 has been suggested as an important gene to regulate cellular apoptosis and proliferation. In this study, we explored whether the observed dysregulation of GRP78 was associated with tumor apoptosis (at Day 1) or proliferation (at Day 21). TUNEL assay revealed that cisplatin-based TACE induced tumor cell apoptosis one day after treatment in rats receiving either Ringer's buffer or OC89. However, more tumor cells became apoptotic when OC89 was predelivered (Figure 3e, f; p < 0.05). Proliferative status of tumor cells at 21 days after TACE was assessed by evaluating the expression of Ki-67. Compared with the control, fewer tumor cells were labeled with Ki-67 in rats receiving OC89 (Figure 3e, f; p < 0.05).

GRP78 is differentially expressed in human TACE-sensitive or resistance HCC

As the above finding indicated GRP78 may play a role in modulating drug sensitivity under hypoxia, we further determined the expression level of GRP78 in human TACE-sensitive or resistant HCC biospy specimens. Immunostaining results revealed that GRP78 positivity was present in either cytoplasm or (and) nuclear of the tumor cells (Figure 4a). Of the 12 TACE-sensitive patients, 10 showed weak and focal staining (10/12) of GRP78 and two showed weak and diffuse staining. Strong positivity was not observed. Of the 13 TACE-resistant patients, strong staining was seen in seven patients (five focal or two diffuse) while the other patients showed weak (five focal or one diffuse) staining. Semi-quantitative analysis based on immunostaining scoring showed there was a statistical difference between TACE-sensitive and resistant patients on GRP78 expression (Figure 4b, p = 0.012).

Figure 4. Differential expression of GRP78 in human TACE-sensitive or resistant HCC. (a) Weak staining of GRP78 in tumor cells of TACE-sensitive HCC and strong positivity in tumor cells of TACE-resistant HCC (Upper panel, Bar = 20 um; lower panel, bar = 100 um). (b) Scoring of GRP78 staining. The scoring was calculated based on both staining intensity and area of positivity.

Discussion

TACE is the mainstay treatment for HCC patients of advanced stages. The rationale of this technique is that combined targeted chemotherapy and selected ischemic necrosis may produce synergistic effects. The advantage of TACE over supportive care has been reported in two randomized controlled trials (Lo et al. 2002, Llovet et al. 2002). However, the response rate to TACE remained far from expectation, partly due to resistance of cancer cells to current chemotherapeutic agents such as cisplatin. Additionally, it has been suggested that TACE may aggravate hypoxia in the residual cancer cells (if there is any) (Pleguezuelo et al. 2008) and further increase the chemoresistance of tumor cells. Thus, improving the sensitivity of cancer cells to the conventional drugs may be an efficient approach to enhance the outcome of TACE.

In recent years, progress in the field of hemoglobin-based oxygen carriers (HBOCs) has instigated great interests in using such carriers as a new strategy to overcome the problem of hypoxia in cancer treatment. Several studies in rodent tumor models reported encouraging results of HBOCs in enhancing the efficacy of either radio- or chemotherapy, though problems such as vasoconstriction or kidney toxicity incurred by free hemoglobin (a powerful scavenger of nitric oxide) were raised. In this study, we observed that after systemic delivery of OC89 (which is an investigational tetrameric hemoglobin solution and does not bind with nitric oxide) to an orthotopic rat HCC model, there was augmented oxygenation in the implanted tumor, indicating that OC89 could effectively attenuate the hypoxic condition inside the tumor tissue. In addition, we found OC89 delivery in combination with cisplatin-based TACE resulted in improved therapeutic efficacy compared with TACE alone, which suggests that OC89, as an oxygen carrier, could serve as an enhancer in TACE. To the best of our knowledge, this is the first report to examine whether the hemoglobin oxygen carrier could improve the treatment of conventional chemotherapeutic agents such as cisplatin in an orthotopic rat HCC model. Additionally, it is worthy to mention that intrahepatic pO₂ in this study was evaluated in an orthotopic HCC model. Compared with studies on an ectopic xenograft HCC model, evaluation in this study is much more susceptible to influences coming from animal breath, body temperature, assessment duration, and sensitivity of the probe. Accordingly, several strategies were employed to maintain the constancy of the experimental conditions in the study, including induction of stable anesthesia with pentobarbitone, maintenance of the rat body temperature with a heat pad, as well as utilization of a precise pO₂ monitor such as OxyLab pO₂ (Griffiths and Robinson 1999, Braun et al. 2001, Urano et al. 2002, Elas et al. 2006, Vikram et al. 2007, Wen et al. 2008). We believed the data collected in this model supplemented precious information on tumor hypoxia that an ectopic xenograft HCC model may not be able to afford. In this study, we did not observe any adverse reaction to OC89 in the rats throughout the research period. Whether OC89 alone or in combination with cisplatin may cause any side effects in the long run needs additional examination in the future.

GRP78, as a major chaperone of endoplasmic reticulum (ER), is activated under stressful conditions such as hypoxia to facilitate proper protein folding. Recent studies have established that GRP78 is upregulated in a variety of human cancers and indicated its multiple roles in both antiapoptosis and proliferation (Wang et al. 2007, Pyrko et al. 2007, Huang et al. 2012). Of interest, some of these studies suggested that elevated expression of GRP78 in cancer correlated with chemoresistance and reduction of GPP78 expression was effective in enhancing cancer cell death by sensitizing them to chemotherapy. CHOP is a transcription factor that operates at the convergence of the three arms of UPR pathway (Oyadomari and Mori 2004). First identified as a highly stress-inducible gene to growth arrest or DNA damage, the role of CHOP in ER-stress mediated apoptosis was signified by increasing evidence (Gately et al. 1994, Wang et al. 1996, Zinszner et al. 1998, McCullough et al. 2001, Pino et al. 2009). Several research groups reported that over-expression of CHOP promoted ER stress-induced apoptosis while knock-down of CHOP prevented cellular death caused by ER disruption (Pino et al. 2009, McCullough et al. 2001). It was also revealed that CHOP expression could be suppressed by GRP78 over-expression indicating there was a cross-talk between the two molecules (Fu et al. 2008). Additionally, several recent studies suggested that cisplatin can contribute to cell death through ER-stress pathway in which CHOP serves as a response gene (Oyadomari and Mori 2004, Li and Holbrook 2004, Rabik et al. 2008). In this study, we observed that in contrast to treatment with cisplatin-based TACE alone, pre-delivery of OC89 could alleviate the level of GRP78 in tumor tissue and concurrently increase the expression of CHOP. Furthermore, compared with TACE-sensitive patients, higher expression of GRP78 was validated in patients showing resistance to TACE in a cohort of human recurrent HCC. Taken together, these finding suggest that upregulation of GRP78 is involved in the regulation of tumor resistance to conventional chemotherapeutic agents such as cisplatin. Delivery of oxygen carrier may sensitize tumor cells to cisplatin-based TACE by knocking down GRP78 expression. Whether GRP78 could serve as a novel therapeutic target against chemoresistance in HCC is worthy of further investigation.

Selected ischemic necrosis is one advantage of TACE. In this study, the hemoglobin-based oxygen carrier was delivered before TACE and the subsequent hypoxia attenuation may partly counteract the effect of hepatic artery ligation. Based on the observation that OC89 enhanced the efficacy of TACE, we propose that the increased oxygenation incurred by OC89 was just mild or moderate in degree. This is sufficient to sensitize tumor cells but could not mask off the major effect of blood supply occlusion.

Conclusions

In summary, in this study we observed that a hemoglobin oxygen carrier, OC89, could effectively attenuate the hypoxia of HCC tissue and enhance the efficacy of cisplatin-based TACE in an orthotopic rat HCC model. Mechanistic investigation revealed that delivery of OC89 influenced antiapoptotic and proapoptotic balance of the UPR pathway in tumor cells. These findings indicated that hemoglobin-based oxygen carriers could serve as a promising approach to improve our current modalities for HCC. Meanwhile, whether enhancing chemosensitivity by targeting hypoxia-activated UPR signaling could be an alternative therapeutic strategy is also worthy of further investigation.

Acknowledgements

The authors thank the New A Innovation Limited for its provision of OC89.

Declaration of interest

The authors report no declarations of interest. The authors alone are responsible for the content and writing of the paper.

This study was supported by Innovation Technology Fund (UIM/184, 10/08-02/10) from Innovation Technology Commission, the government of Hong Kong SAR.