• Open Access

Antitumor effect of berberine against primary effusion lymphoma via inhibition of NF-κB pathway


To whom correspondence should be addressed.

E-mail: okadas@kumamoto-u.ac.jp


Primary effusion lymphoma (PEL) is an infrequent and distinct entity among the aggressive non-Hodgkin B cell lymphomas that occurs predominantly in patients with advanced AIDS. It shows serous lymphomatous effusion in body cavities, and is resistant to conventional chemotherapy with a poor prognosis. Thus, the optimal treatment for PEL is not well defined and there is a need for novel agents. PEL has been recognized as the tumor caused by Kaposi sarcoma-associated herpes virus/human herpes virus-8 (KSHV/HHV-8), and nuclear factor (NF)-κB activation plays a critical role in the survival and growth of PEL cells. In this study, we assessed the antitumor effect of berberine, a naturally occurring isoquinoline alkaloid, on this pathway. The methylthiotetrazole assay showed that cell proliferation in the PEL cell lines was inhibited by berberine. Berberine also induced caspase-dependent apoptosis and suppressed NF-κB activity by inhibiting IκB kinase (IKK) phosphorylation, IκB phosphorylation and IκB degradation, upstream targets of the NF-κB pathway, in PEL cells. In a xenograft mouse model that showed ascites and diffuse organ invasion of PEL cells, treatment with berberine inhibited the growth and invasion of PEL cells significantly compared with untreated mice. These results show that the suppression of NF-κB is a molecular target for treating PEL, and berberine is a potential antitumor agent for PEL. (Cancer Sci 2012; 103: 775–781)

Primary effusion lymphoma (PEL) is a rare and distinct subtype of non-Hodgkin lymphoma that was originally identified in patients with advanced AIDS.[1, 2] PEL arises exclusively in body cavities (pleura, peritoneum and pericardium) and is caused by Kaposi sarcoma-associated herpes virus/human herpes virus-8 (KSHV/HHV-8).[2] It is generally resistant to chemotherapy, with a median survival of only 3 months;[3] therefore, there is a need to develop new therapies. PEL displays constitutive activity of many signaling pathways in survival and growth, including the NF-κB, JAK/STAT and PI3K/Akt pathways.[4-6] Inhibition of NF-κB induces apoptosis in PEL cells and this pathway represents a molecular target for this disease.[4, 7]

Berberine, an isoquinoline alkaloid from a plant used in traditional Chinese and Ayurvedic medicine, is an active component of Berberis aquifolium (Oregon grape), Berberis aristata (turmeric tree), Berberis vulgaris (barberry), Coptis chinensis (coptis or golden thread) and Hydrastis canadensis (golden seal). Berberine has a wide range of biological effects, including antidiarrheal, antihypertensive, antiarrhythmic, cholesterol-lowering, antimicrobial and anti-inflammatory activities.[8-13] In addition, berberine possesses antitumor activities against various tumor cells.[14-17] The suppression of NF-κB by berberine has been demonstrated in several tumor cell lines;[18-21] however, the specific target of the NF-κB pathway is not fully understood, and the antitumor ability in vivo of berberine is limited.[21-23] Further studies in animal models are required to identify the potential effects and clinical application of berberine.

In this study, we investigated the effect of berberine on proliferation and apoptosis in PEL cells and clarified the target molecules of berberine in the NF-κB pathway against PEL cells in vitro. The suppression of upstream molecules of the NF-κB pathway led to the inhibition of NF-κB activity. We also assessed the in vivo effect of berberine, showing the rationale for a clinical study. These findings provide insights into the molecular target of PEL and the antitumor mechanism of berberine against PEL cells.

Materials and Methods

Cell lines and reagents

Human PEL cell lines, BC-1,[24] BCBL-1,[25] TY-1[26] and human non-PEL cell line, U937,[27] were maintained in RPMI1640 supplemented with 10% heat-inactivated FCS, penicillin (100 U/mL) and streptomycin (100 μg/mL) in a humidified incubator at 37°C and 5% CO2. Berberine chloride was obtained from Sigma-Aldrich (St. Louis, MO, USA).

Tetrazolium dye MTT assay

The antiproliferative activities of berberine against PEL cell lines were measured using the MTT method (Sigma-Aldrich). Briefly, 2 × 104 cells were incubated in triplicate in a 96-well microculture plate in the presence of different concentrations of berberine in a final volume of 0.1 mL for 24 h at 37°C. Subsequently, MTT (0.5 mg/mL final concentration) was added to each well. After 3 h of additional incubation, 100 μL of a solution containing 10% SDS plus 0.01 N HCl was added to dissolve the crystals. Absorption values at 595 nm were determined with an automatic ELISA plate reader (Multiskan; Thermo Electron, Vantaa, Finland). Values are normalized to untreated (control) samples.

Annexin V assay

Apoptosis was quantified using Annexin V-Alexa fluor 647 (AF647) (BioLegend, San Diego, CA, USA). Briefly, after treatment with berberine for 24 h, cells were harvested, washed and then incubated with Annexin V-AF647 for 60 min in the dark, before being analyzed on an LSR II cytometer (BD Bioscience, San Jose, CA, USA). Data were analyzed using FlowJo software (Tree Star, San Jose, CA, USA).

Analysis of DNA fragmentation by agarose gel electrophoresis

To detect apoptosis and DNA damage, DNA ladder assays were performed as previously described.[28] Briefly, BCBL-1 cells were cultured in the presence or absence of berberine at 37°C for 48 h. After incubation, 1 × 106 cells were lysed in 100 μL of 10 mM Tris–HCl buffer (pH 7.4) containing 10 mM EDTA and 0.5% Triton X. After centrifugation for 5 min at 20 000g, supernatant samples were treated with RNase A (Qiagen, Valencia, CA, USA) and Proteinase K (Wako Pure Chemical, Osaka, Japan). Subsequently, 20 μL of 5 M NaCl and 120 μL isopropanol were added to the sample and kept at −20°C for 6 h. Following centrifugation for 15 min at 20 000g, the pellets were dissolved in 20 μL TE buffer (10 mM Tris–HCl and 1 mM EDTA) as loading samples. To assess the DNA fragmentation pattern, samples were loaded onto 1.5% agarose gel and electrophoretically separated.

Western blot analysis

For whole cell extraction, BCBL-1 cells treated with 100 μM berberine for 0, 1, 3 and 6 h were collected and washed in cold PBS before the addition of 100 μL cold lysis buffer (25 mM HEPES, 10 mM Na4P2O7·10H2O, 100 mM NaF, 5 mM EDTA, 2 mM Na3VO4 and 1% Triton X-100). After rotation for 2 h at 4°C, whole cell extracts were obtained by centrifugation at 20 000g for 15 min. For nuclear extraction, BCBL-1 cells treated with 100 μM berberine for 0, 1, 3 and 6 h were collected and washed in cold PBS before the addition of 400 μL cold buffer A (10 mM HEPES-KOH pH 7.9, 1.5 mM MgCl2, 10 mM KCl, 0.1% NP-40, 0.5 mM DTT, 0.5 mM PMSF, 2 μg/mL pepstatin A, 2 μg/mL aprotinin and 2 μg/mL leupeptin). After incubation on ice for 10 min, the samples were vortexed for 10 sec. Nuclei were pelleted by centrifugation at 2000g for 1 min and washed once with buffer A. Then, 50 μL of buffer C (50 mM HEPES-KOH pH 7.9, 10% glycerol, 420 mM KCl, 5 mM MgCl2, 0.1 mM EDTA, 1 mM DTT, 0.5 mM PMSF, 2 μg/mL pepstatin A, 2 μg/mL aprotinin and 2 μg/mL leupeptin) was added to the nuclei, before incubating on ice for 30 min. Nuclear extracts were obtained by centrifugation at 20 000g for 15 min. Whole or nuclear extracts (40 μg protein) were separated by 10% SDS-PAGE and blotted onto a PVDF membrane (GE Healthcare, Tokyo, Japan). Detection was performed using the Enhanced Chemiluminescence Western Blotting Detection System (ECL; GE Healthcare Bio-Science, Buckinghamshire, UK).

Primary antibodies were as follows: anti-cleaved caspase-3 (9661), anti-caspase 9 (9502), anti-phospho (Ser180/181)-IKKα/β (2681), anti-phospho (Ser32/36)-IκBα (9246), anti-p65 (536) (Cell Signaling Technology, Danvers, MA, USA), anti-IKKα/β (H-470), anti-IκBα (C-21) and anti-γ tublin (C-20) (Santa Cruz Biotechnology, Santa Cruz, CA, USA), anti-Hsc70 (SPA-815) (Stressgen Bioreagents, Ann Arbor, MI, USA).

Electrophonic mobility shift assay

An EMSA was performed using a second generation DIG Gel Shit Kit (Roche Diagnostics, Mannheim, Germany). Briefly, double-stranded oligonucleotide probes containing the immunoglobulin kappa (Igκ) light chain NF-κB site and the Oct-1 binding site were purchased from Promega (Madison, WI). The oligonucleotide was 3′ end-labeled with a digoxigenin-11-ddUTP. The nuclear extract (10 μg protein) from BCBL-1 cells was incubated with 1 μg poly[d(I-C)], 0.1 μg poly-L lysine and DIG-labeled oligonucleotide in binding buffer (20 mM HEPES pH 7.6, 1 mM EDTA, 10 mM (NH4)2SO4, 0.2% Tween20 and 30 mM KCl) for 15 min at 25°C. After incubation, 5× loading buffer (0.25× TBE and 60% glycerol) was added, and the samples were separated on 6% acrylamide gel in 0.5× TBE buffer. The oligonucleotide was electroblotted onto a positively charged nylon membrane (Roche Diagnostics) and immunodetected using anti-digoxigenin-AP.

Xenograft mouse model

NOD Rag-2-deficient (Rag-2−/−) mice and NOD Jak3-deficient (Jak3−/−) mice were established by crossing Rag-2−/− mice or Jak3−/− mice with the NOD strain for 10 generations, respectively. NOD Rag-2/Jak3 double-deficient (Rag-2−/−Jak3−/−) mice (NRJ mice) were established by crossing NOD Rag-2−/− mice and NOD Jak3−/− mice, and were housed and monitored in our animal research facility according to institutional guidelines. All experimental procedures and protocols were approved by the Institutional Animal Care and Use Committee at Kumamoto University. Twelve week-old NRJ male mice were inoculated i.p. with 7 × 106 BCBL-1 cells suspended in 200 μL PBS. The mice were then treated with i.p. injections of PBS or berberine (10 mg/kg, three times a week). Tumor burden was evaluated by measuring the body weight and volume of ascites. For assessment of overall survival, Kaplan–Meier analysis was performed and P-values were determined by two-tailed analysis using the log-rank test.


To investigate the expression of KSHV/HHV-8 ORF73 (LANA) protein, tissue samples were fixed with 10% neutral-buffered formalin, embedded in paraffin and cut into 4-μm sections. The sections were deparaffinized by sequential immersion in xylene and ethanol, and rehydrated in distilled water. They were then irradiated for 15 min in a microwave oven for antigen retrieval. Endogenous peroxidase activity was blocked by immersing the sections in methanol/0.6% H2O2 for 30 min at room temperature. Affinity-purified PA1-73N antibody,[29] diluted 1:3000 in PBS/5% BSA, was then applied, and the sections were incubated overnight at 4°C. After washing in PBS twice, the second and third reactions and the amplification procedure were performed using kits according to the manufacturer's instructions (Catalyzed Signal Amplification System; DAKO, Copenhagen, Denmark). The signal was visualized using 0.2 mg/mL diaminobenzidine and 0.015% H2O2 in 0.05 M Tris–HCl, pH 7.6.


Total RNA was extracted from the cells using Trizol (Invitrogen, Carlsbad, CA, USA). First-strand cDNA was synthesized from RNA using a PrimeScript RT-PCR kit (Takara Bio, Otsu, Japan) with random primers. The PCR products were analyzed by 1.5% agarose gel electrophoresis and ethidiumbromide staining. Primer sequences were as follows:


Statistical analysis

Data are expressed as the mean ± SD. The statistical significance of the differences observed between experimental groups was determined using Student's t-test, and P < 0.05 was considered significant.


Berberine inhibits proliferation and induces apoptosis in primary effusion lymphoma cells

The chemical structure of berberine is shown in Figure 1 and has a molecular weight of 371.8. We first determined whether treatment with berberine leads to the inhibition of PEL cell proliferation using MTT assay. Three PEL cell lines (BC-1, BCBL-1 and TY-1) were cultured with varying concentrations of berberine (0, 3, 10, 30 and 100 μM) for 24 h, and proliferation was analyzed by MTT assay. Figure 2 shows that as the dose of berberine increased from 3 to 100 μM, cell growth inhibition increased in a dose-dependent fashion in all PEL cell lines. The IC50 (50% inhibitory concentration) for BC-1, BCBL-1 and TY-1 were 13.56, 29.17 and 32.82 μM. In contrast, the IC50 value is >100 μM for non-PEL cell line, U937. In subsequent experiments, we determined whether the observed suppressive effects of berberine in MTT assay were due to the induction of apoptosis. We used Annexin V staining and DNA ladder formation to detect apoptosis. As shown in Figure 3(a), 30 and 100 μM berberine treatment for 24 h caused apoptosis in BCBL-1. As shown in Figure 3(b), berberine treatment for 48 h caused DNA fragmentation, which is a characteristic of apoptosis cell death. Next, we analyzed cleaved caspase 3 and cleaved caspase 9 to further confirm that berberine induced apoptosis in PEL cells. As shown in Figure 3(c), berberine treatment of BCBL-1 induced time-dependent cleavage of caspase 3 and caspase 9, hallmarks of cells undergoing apoptosis, in western blotting.

Figure 1.

Chemical structure of berberine chloride.

Figure 2.

Berberine inhibits the proliferation of primary effusion lymphoma (PEL) cells. PEL cell lines (BC-1, BCBL-1 and TY-1) and non-PEL cell line (U937) were incubated with 0, 3, 10, 30, 10 and 100 μM berberine for 24 h. A cell proliferation assay was carried out using MTT as described in the 'Materials and Methods'Materials and Methods. One representative result from three independent experiments is shown.

Figure 3.

Berberine induces apoptosis in BCBL-1 cells. (a) Berberine-induced apoptosis was detected by Annexin V staining. A primary effusion lymphoma cell line, BCBL-1, was treated with berberine (30, 100 μM) for 24 h and was subsequently stained with Annexin V-Alexa fluor 647 before being analyzed by flow cytometry. (b) Berberine caused DNA fragmentation of nuclei. BCBL-1 cells were treated with berberine (30, 100 μM) for 48 h. (c) BCBL-1 cells were treated with 100 μM berberine for 0, 1, 3 and 6 h and total proteins were extracted for western blotting.

Berberine suppresses NF-κB activity in PEL cells

It was reported previously that NF-κB was required for the survival and proliferation of PEL cells.[4, 32, 33] Because NF-κB is constitutively active in PEL cells,[34] we examined whether berberine inhibited NF-κB activation. BCBL-1 constitutively expressed both total and phosphorylated IKK and IκB, upstream of NF-κB. When BCBL-1 was treated with 100 μM berberine for 0, 1, 3 and 6 h, berberine treatment reduced phosphorylated IKK and phosphorylated IκB, whereas total IκB was increased (Fig. 4a), suggesting that inhibition of IKK phosphorylation leads to the accumulation of IκB by blocking the phosphorylation and degradation of IκB protein. Next, we fractioned nuclear protein and analyzed the expression of p65 by western blotting (Fig. 4b) to confirm p65 NF-κB suppression by berberine. When PEL cell lines were treated with 100 μM berberine for 6 h, the amount of nuclear p65 NF-κB protein was reduced, indicating that berberine suppresses NF-κB activity. Thus, berberine mainly suppresses p65 NF-κB nuclear translocation by inhibiting the upstream of NF-κB. To confirm that berberine could inhibit NF-κB activity in PEL cells, we performed EMSA with DIG-labeled double-stranded NF-κB oligonucleotides. Berberine also suppresses constitutive NF-κB binding activity for 24 and 48 h (Fig. 4c). These results demonstrate that berberine inhibits the constitutive NF-κB activity of PEL cells.

Figure 4.

Inhibitory effects of berberine on the expression of NF-κB pathways. (a) A primary effusion lymphoma cell line, BCBL-1, was treated with 100 μM berberine for 0, 1, 3 and 6 h and total proteins were extracted for western blotting. (b) BCBL-1 cells were treated with 100 μM berberine for 0, 1, 3 and 6 h and nuclear proteins were extracted for western blotting to detect NF-κB p65. (c) BCBL-1 cells were treated with 100 μM berberine for 0, 12, 24 and 48 h and assessed for NF-κB DNA binding activity by EMSA using an NF-κB-specific oligonucleotide probe.

In vivo effect of berberine in severe immunodeficient mice inoculated i.p. with BCBL-1

As the in vitro results suggest that berberine could be an effective treatment against PEL, we assessed the in vivo effects of berberine in the immunodeficient mouse model. Severe immunodeficient, NRJ mice were inoculated i.p. with 7 × 106 BCBL-1 cells. BCBL-1 produced profuse ascites within 4 weeks of inoculation (Fig. 5a). As patients with PEL show lymphomatous effusion in body cavities without a definable tumor mass,[2, 35] these mice could be clinically equivalent to the PEL model. A dose of 10 mg/kg berberine or PBS alone was administrated via i.p. injection on day 3 after cell inoculation, and then three times a week. The 50% lethal dose of berberine from i.p. injections has already been reported and is 57.6103 mg/kg.[36] Hence, the dosage of berberine in vivo in our experiment was expected to be safe. Berberine-treated mice appeared to stay healthy, and the body weight did not change, whereas the body weight of untreated mice significantly increased compared to that of berberine-treated mice on day 28 (36.6 ± 2.7 g vs 30.7 ± 1.7 g, n = 6, P < 0.01; Fig. 5b). Moreover, the volume of ascites was significantly lower than in untreated mice on day 28 (3.8 ± 0.6 mL vs 0.5 ± 0.8 mL, n = 6, P < 0.01; Fig. 5c). As shown in Figure 5d, treatment with berberine significantly prolonged survival of the mice (log-rank test, P < 0.01). These results indicate that treatment with berberine delays or inhibits the growth of PEL cells and produces a survival benefit.

Figure 5.

Treatment of NOD/Rag-2/Jak-3-deficient mice with berberine suppresses the development of PEL in vivo. (a) Photograph of berberine-treated and untreated ascites-bearing mice 4 weeks after inoculation with BCBL-1 i.p. (b) The body weight of mice 4 weeks after inoculation with BCBL-1 cells in berberine-treated or untreated mice is shown as the mean ± SD of 6 mice. (c) The volume of ascites 4 weeks after inoculation with BCBL-1 cells in mice is shown as the mean ± SD of 6 mice. (d) Overall survival curve. Treatment with berberine prolongs survival in vivo.

Organ invasion by PEL cells on day 28 was evaluated by hematoxylin–eosin staining and LANA immunostaining. We found that mice inoculated i.p. with BCBL-1 exhibited invasion into the liver, lung and spleen without macroscopic lymphoma formation (Fig. 6a). The number of LANA-positive cells in berberine-treated mice was significantly reduced (0–1 cells per field magnification, ×40) compared to untreated mice (10–20 cells per field magnification, ×40). The mRNA expression levels of vFLIP and LANA were downregulated in the spleen of berberine-treated mouse (Fig. 6b). These data demonstrated that berberine significantly inhibits the growth and infiltration of PEL cells in vivo and could be a potentially therapeutic agent in patients with PEL.

Figure 6.

Invasion of primary effusion lymphoma cells into the organs of BCBL-1-inoculated mice on day 28. (a) Hematoxylin–eosin (HE) staining and immunohistochemical staining using anti-LANA (PA1-73N antibody) was performed to detect BCBL-1 in the liver, lung and spleen. (b) Viral gene expression after treatment with berberine. Viral gene expression in spleen of berberine treated and untreated mouse was examined by RT-PCR. A representative result from two experiments is shown.


The clinical course of PEL is very aggressive and generally refractory to conventional chemotherapy; hence, novel therapeutic strategies such us molecular targeting therapy are needed. In the present study, we investigated the antitumor effects of a naturally occurring isoquinoline alkaloid, berberine, on PEL cells both in vitro and in vivo, and showed that berberine inhibited the NF-κB pathway with the suppression of IKK phosphorylation, IκB phosphorylation and IκB degradation. In KSHV/HHV-8-infected cells, vFLIP, a homologue of cellular FLIP protein, has the ability to activate the NF-κB pathway by binding to the IKK complex.[33, 37, 38] Moreover, inhibition of NF-κB activity leads to the apoptosis of KSHV-infected PEL cells.[4, 32] These results suggest that inhibition of NF-κB is an effective target for the treatment of PEL. Activation of NF-κB is involved in various kinds of cancer development and progression,[39, 40] indicating that NF-κB is a good molecular target for cancer treatment.

Berberine has long been used as a stomachic, an anti-diarrheal agent, an antibiotic and an anti-inflammatory in Asian countries and has been shown to have few side effects.[8, 9, 13] Berberine has been reported to have various pharmacological effects, including an arresting effect on cell cycle progress, inhibition of tumor cell proliferation and the induction of apoptosis, and the mechanism of antitumor activity differs among cell lines.[14-17, 41, 42] Several reports have demonstrated that berberine inhibits cancer cell migration by suppressing COX-2, MMP-2, MMP-9 and urokinase-plasminogen activator,[19-21] downstream molecules of NF-κB. We showed here that berberine induced the apoptosis of PEL by inhibiting IKK phosphorylation, the upstream target of the NF-κB pathway. Consequently, berberine abrogates the phosphorylation of IκB (Fig. 4a), NF-κB nuclear translocation (Fig. 4b) and DNA binding activity (Fig. 4c). Previously, we reported biscoclaurine alkaloid cepharanthine-induced apoptosis of PEL cells mainly via inhibiting p65 activation. In this study, berberine inhibits IKK activation, the upstream of the NF-κB pathway, and causes efficient apoptosis of PEL cells. Inhibiting IKK activation is also considered to be a rational pharmacologic target because vFLIP activates IKK in PEL cells.

We also confirmed the therapeutic effect of berberine against PEL in a xenograft mouse model. We used NRJ mice, which displayed rapid and efficient engraftment of PEL cells, as a small animal system. NRJ mice display not only complete deficiency in mature T/B lymphocytes and complement protein but also complete deficiency of NK cells, such as in NOD/Scid/common γ-deficient (NOG) mice[43, 44] and NOD/Scid/Jak3-deficient (NOJ) mice,[45] and provide the ideal microenvironment for the propagation and increase of PEL cells. Although both scid and Rag mutations prevent the recombination of genes required for functional B and T cell receptors, the Prkdc gene disrupted by the scid mutation is expressed broadly and is involved in DNA repair, while expression of rag genes is limited to hematopoietic cells and is involved only in the DNA recombination of T and B cell receptor genes. Thus, scid mice are more sensitive to radiation-induced or drug-induced DNA damage than their Rag mutation counterparts. In addition, the scid mutation is known to show a leaky phenomenon in which functional T and B cells are produced with aging and ionized irradiation.[46] Taken together, NRJ mice are expected to be more convenient recipients of human cell xeno-transplantation.

The formation of malignant ascites without solid lymphoma formation displayed in PEL xenograft NRJ mice reflects the clinical nature of human PEL and they could be a quite useful in vivo model for studying PEL and HHV-8 pathogenesis. Berberine has been reported to suppress tumor invasion[21] and phorbol-ester-induced tumor promotion,[22] chemical-induced carcinogenesis[23]in vivo; however, the direct antitumor effect and doses of berberine used in animal studies are unclear. In this study, we observed that administration of 10 mg/kg berberine three times a week showed significant reduction of ascites and tumor invasion with no apparent adverse effects on NRJ mice (Figs 5, 6). Tumor invasion is related to some target genes of NF-κB, such as MMP and vascular endothelial growth factor.[39] We confirmed that suppressing NF-κB was also effective for invasion of PEL cells in vivo. Further studies in animals suggest a new direction in the treatment of refractory malignancies such as PEL.

The effects of berberine on PEL cells other than the NF-κB pathway are expected because berberine also affects NF-κB-independent tumors and exerts diverse pharmacological effects.[9, 13, 15] Elucidating the pharmacological diversity of berberine could lead to the development of novel effective therapies for a variety of malignancies as well as PEL. Berberine has been reported to have antiretroviral activity against HIV[47] and to reduce endoplasmic reticulum stress by preventing an HIV protease inhibitor-induced inflammatory response.[48] In AIDS patients who develop PEL, concomitant treatment with berberine could contribute to not only antitumor and tumor-preventing activities, but also antiretroviral therapy.

In conclusion, our data have shown the ability of berberine to induce cell death by blocking the NF-κB pathway in PEL cells.


We thank Ms I. Suzu for technical assistance and Ms K. Tokunaga for secretarial assistance. This work was supported in part by a Health and Labour Sciences Research Grant from the Ministry of Health, Labour, and Welfare of Japan (H22-AIDS-I-002), and by the Global Center of Excellence program “Global Education and Research Center Aiming at the Control of AIDS,” and Grants-in-Aid for Science Research (Nos. 21107522 and 21591209) from the Ministry of Education, Science, Sports, and Culture of Japan.

Disclosure Statement

The authors have no conflicts of interest to declare.