ABSTRACT
Human T-cell leukemia virus type 1 (HTLV-1) infection causes T-cell leukemia and inflammatory diseases, most notably including HTLV-1-associated myelopathy/tropical spastic paraparesis (HAM/TSP). The underlying mechanism for the pathogenesis of HAM/TSP remains unclear. According to a recent clinical trial, a humanized antibody that targets CCR4+ cells ameliorates inflammation by reducing the number of infected cells in the central nervous system; this result suggests that the transmigration of HTLV-1-infected cells plays a crucial role in HAM/TSP. Partly due to the blood-brain barrier, current treatments for HAM/TSP are mostly palliative. Pentosan polysulfate (PPS), a semisynthetic glycosaminoglycan, has recently been used to treat HAM/TSP and was found to alleviate the symptoms. In this study, we investigated the effect of PPS on HTLV-1-infected cells and provide evidence for its efficacy in HAM/TSP. PPS was cytotoxic to certain HTLV-1-infected cells and significantly suppressed HTLV-1 virion production. PPS also efficiently inhibited HTLV-1 cell-cell transmission in T cells. In addition, PPS blocked HTLV-1 infection of primary endothelial cells (human umbilical vascular endothelial cells) and suppressed the subsequent induction of proinflammatory cytokine expression. Furthermore, PPS was found to inhibit the adhesion and transmigration of HTLV-1-infected cells. We also confirmed the anti-HTLV-1 effect of PPS in vivo using two mouse models. PPS blocked HTLV-1 infection in a mouse model with peripheral blood mononuclear cell (PBMC)-humanized NOD-scid IL2Rgammanull (huPBMC NSG) mice. PPS was also found to suppress the development of dermatitis and lung damage in HTLV-1 bZIP factor (HBZ)-transgenic (HBZ-Tg) mice, an HTLV-1 transgenic mouse model in which the mice develop systemic inflammation.
IMPORTANCE HTLV-1 is the first human retrovirus to have been identified and is endemic in certain areas worldwide. HTLV-1 infection leads to the development of an inflammatory disease called HAM/TSP, a myelopathy characterized by slowly progressive spastic paraparesis. There have been no effective therapeutics available for HAM/TSP, but recently, a semisynthetic glycosaminoglycan, named pentosan polysulfate (PPS), has been found to alleviate the symptoms of HAM/TSP. Here we conducted a comprehensive study on the effect of PPS both in vitro and in vivo. PPS demonstrated anti-HTLV-1 potential in infected cell lines, as shown by its suppressive effects on HTLV-1 replication and transmission and on the transmigration of infected T cells. Moreover, results obtained from two HTLV-1 mouse models demonstrate that PPS inhibits HTLV-1 infection and inflammation development in vivo. Our work offers insights into the treatment of HAM/TSP by PPS and also suggests its possible use for treating other HTLV-1-induced inflammatory diseases.
INTRODUCTION
Human T-cell leukemia virus type 1 (HTLV-1) is a retrovirus that can infect various cell types in vitro yet preferentially infects and immortalizes CD4 T cells in vivo (1). Approximately 10 million people are infected with HTLV-1 throughout the world (2). HTLV-1 infection is highly endemic in areas such as southwestern Japan, sub-Saharan Africa, South America, and the Caribbean (2). About 5% of infected individuals develop adult T-cell leukemia-lymphoma (ATL) after a long latency. HTLV-1 also causes inflammatory diseases, such as uveitis, polymyositis, dermatitis, and HTLV-1-associated myelopathy/tropical spastic paraparesis (HAM/TSP), at a frequency lower than that at which it causes ATL (3).
HAM/TSP is characterized by slowly progressive spastic paraparesis and is the most notable inflammatory disease caused by HTLV-1. Although viral regulatory genes, including Tax and HTLV-1 bZIP factor (HBZ), have been implicated (4–7), the exact mechanism for the pathogenesis of HAM/TSP remains elusive. Multiple pathogenic models have been proposed so far (8). A direct toxicity theory suggests that HTLV-1 infects central nervous system (CNS)-resident cells, such as glial cells, and causes tissue damage (8). A bystander hypothesis suggests that HTLV-1-infected CD4 T cells transmigrate through the blood-brain barrier (BBB) and release tissue-toxic cytokines (8). Recently, a phase I/IIa clinical trial of the anti-C-C chemokine receptor 4 (CCR4) antibody mogamulizumab was carried out in HAM/TSP patients (9). Mogamulizumab improved the symptoms of HAM/TSP patients by decreasing the number of infected cells in the CNS and suppressing inflammatory cytokine production (9). This result appears to support the bystander model for the pathogenesis of HAM/TSP (8).
Effective therapies for HAM/TSP are still unavailable, and current therapeutics are mostly palliative treatments intended to suppress inflammation. Recently, a semisynthetic glycosaminoglycan (GAG), pentosan polysulfate (PPS), has been found to be effective in a clinical trial for HAM/TSP patients (10). GAGs are long unbranched polysaccharides that are commonly present in mammalian tissues, specifically, on the cell surface or in the extracellular matrix (ECM) (11). They play important roles in a broad range of physiological activities, including inflammation responses (12). In particular, GAGs can assist with the recruitment of leukocytes to inflammatory tissues by modulating their actions, such as rolling, adhering, and transmigrating through the endothelial cell barrier (12). Accordingly, it is presumed that under inflammatory conditions, exogenously introduced free GAGs likely act as competitors to endogenous GAGs and may suppress leukocyte migration (12, 13).
PPS has already been used to treat certain types of inflammatory conditions in humans, such as arthritis and interstitial cystitis (IC) (14–16). Laboratory studies using animal models also demonstrated that PPS was effective against various inflammatory conditions in vivo. For instance, PPS was discovered to be anti-inflammatory in a mouse model of diabetic nephropathy (17), in rat models of mucopolysaccharidoses (18) and allergic encephalomyelitis (19), and in rabbit models of arthritis (20) and atherosclerosis (21). Mechanistically, PPS was found to either suppress the expression or secretion of proinflammatory molecules, such as tumor necrosis factor alpha (TNF-α) or interleukin-6 (IL-6), or hinder the adhesion and migration of leukocytes (17–21). In this study, we tested the effect of PPS on HTLV-1-infected cells and found that PPS could interfere with both HTLV-1 virion production and de novo infection of T cells and endothelial cells. We also confirmed that PPS suppressed HTLV-1 infection in vivo using the peripheral blood mononuclear cell (PBMC)-humanized NOD-scid IL2Rgammanull (huPBMC NSG) mouse model. Moreover, we found that PPS inhibited dermatitis development in HBZ-transgenic (Tg) mice, providing further evidence of its potential in treating HTLV-1-related inflammatory diseases.
RESULTS
Cytotoxicity of PPS against HTLV-1-infected cells.The HTLV-1 proviral load (PVL) is an important risk factor for HAM/TSP and is also correlated with disease severity (22). Because the HTLV-1 PVL increases mainly via the clonal expansion of infected cells, we first tested whether PPS is cytotoxic against HTLV-1-infected cells. Six of eight of the HTLV-1-infected T-cell lines that we tested showed reduced viability upon PPS treatment for 2 days (Fig. 1A). The cytotoxicity of PPS was dose dependent, and the highest concentration of PPS that we used was 100 μg/ml, which has been widely used for studying the effect of PPS (10, 23, 24). PPS-induced apoptosis might account for its cytotoxicity, as apoptosis was seen in PPS-sensitive cells, such as ED cells, but not PPS-insensitive cells, such as MT-2 cells (Fig. 1B). In addition, PPS also suppressed the viability of two of four HTLV-1-negative T-cell lines (Fig. 1A). However, the 50% cytotoxic concentration (CC50) of PPS in HTLV-1-infected cells was generally lower than that in uninfected cells (Table 1). On the other hand, PPS did not affect the survival of normal CD4 T cells (Fig. 1C). Overall, the above-described cytotoxicity of PPS observed seems to correspond with its general anticancer activities reported previously (25–27).
PPS demonstrates cytotoxicity against HTLV-1-infected cells. (A) An MTT assay was performed on four HTLV-1-uninfected T-cell lines and eight infected T-cell lines, and the absorbance (optical density) at 595 nm was measured from day 0 to day 3. Five concentrations of PPS were used: 1, 3, 10, 30, and 100 μg/ml. The optical density at 595 nm of the PPS-treated cells was normalized to that for the control group without PPS to represent relative viability. The MTT assay was performed in triplicate, and the results are shown as the average with standard deviation (SD). (B) Both ED and MT-2 cells were treated with 10 or 100 μg/ml PPS for 48 h and then subjected to 7-AAD and annexin V staining. The results show that PPS induced profound apoptosis in ED cells but only marginal apoptosis in MT-2 cells. Numbers indicate percentage (%) of each population. (C) Normal CD4 T cells were isolated from two healthy donors and activated by Dynabeads. An MTT assay was performed in triplicate in activated CD4 T cells treated with or without PPS as described in the legend to panel A. Two concentrations of PPS, including 10 and 100 μg/ml, were used. PPS did not show any inhibitory effect on the viability of normal CD4 T cells.
CC50 of PPSa
PPS inhibits HTLV-1 virion production.It has been found that PPS has antiviral activities against various enveloped viruses, such as HIV-1 (28, 29). Therefore, we asked whether PPS could suppress HTLV-1 replication. To this end, we measured the production of HTLV-1 virions in the presence of PPS in Tax-positive (Tax+) HTLV-1-infected T cells and pX1MT-M-infected HEK293FT (293FT) cells. pX1MT-M is a plasmid with a complete HTLV-1 proviral sequence and is infectious through transfection into cells. It is originally derived from the HTLV-1 infectious clone pX1MT and has been modified to obtain better infectivity (30–32). We collected supernatants at 12 h and 24 h after PPS treatment, when viability was not affected by PPS, for the five Tax+ HTLV-1-infected T-cell lines (Fig. 1A). We also confirmed that PPS did not suppress the viability of 293FT (Fig. 2B). A dose-dependent inhibition of HTLV-1 virion production by PPS was observed in all cells that we tested (Fig. 2A). This inhibition occurred without PPS affecting HTLV-1 sense or antisense transcription, as there was no alteration of the Tax or HBZ mRNA level (Fig. 2C). Furthermore, neither the expression of the Tax protein nor the ability of Tax to activate HTLV-1 transcription was affected, as demonstrated by Western blotting (Fig. 2D) and HTLV-1 5′ long terminal repeat (LTR) reporter assays (Fig. 2E). These results suggest that PPS suppressed HTLV-1 production without affecting viral transcription.
PPS suppresses HTLV-1 production. (A) pX1MT-M-infected 293FT cells (24 h posttransfection) and five HTLV-1 producing T-cell lines were incubated with various concentrations of PPS for 12 or 24 h, and then the supernatants were collected for quantification of HTLV-1 virions by HTLV-1 p19 ELISA. A dose-dependent inhibition of HTLV-1 production was seen for each cell line. ELISA was performed in duplicate, as recommended by the manufacturer. Statistical analyses were carried out by Student's t test. *, P < 0.05; **, P < 0.01. (B) The MTT assay was performed in triplicate in 293FT cells treated with or without PPS as described in the legend to Fig. 1A. Three concentrations of PPS, including 1, 10, and 100 μg/ml, were used. The result indicates that PPS does not inhibit 293FT growth. (C and D) pX1MT-M-infected 293FT cells (at 24 h posttransfection) and MT-2 or MT-4 cells were treated with 100 μg/ml PPS for 24 h and then subjected to RNA extraction and protein lysate preparation to quantify Tax and HBZ mRNA (C) and Tax protein (D). qPCR was performed in triplicate. No difference in the expression of Tax or HBZ RNA was observed (P > 0.05 for all six results). 18s indicates 18s rRNA that was used as the internal control for qPCR. The band intensity shown in panel D was quantified by ImageJ software. No difference in the expression of the Tax protein was observed. (E) 293FT cells were transfected by the HTLV-1 5′ LTR-Luc reporter plasmid with or without a Tax expression vector to test the effect of PPS on the transactivating ability of Tax. MT-2 and MT-4 cells were transfected with only the HTLV-1 5′ LTR-Luc reporter. Cells were treated with 100 μg/ml PPS for 24 h before being subjected to a dual-luciferase reporter assay. The reporter assay was performed in triplicate, and the values were derived by normalizing firefly luciferase units to Renilla luciferase units. The results indicate that Tax activation of the HTLV-1 5′ LTR was not affected by PPS (P > 0.05 for all three results).
Interestingly, pX1MT-M-infected 293FT cells formed significantly fewer and smaller syncytia in the presence of PPS (Fig. 3A and B). HTLV-1 is known to induce cell fusion to form a syncytium via its envelope proteins, including gp46 (33–35). We indeed observed syncytium formation in 293FT cells transfected with an HTLV-1 Env expression vector, pCAG-Env (Fig. 3A and B). Similarly, this syncytium formation caused by Env expression was also greatly inhibited by PPS (Fig. 3A and B), suggesting a possible targeting effect of Env by PPS. So, we further stained cell surface HTLV-1 gp46 in pX1MT-M-infected 293FT cells and two HTLV-1-producing cell lines, MT-2 and MT-4. As shown in Fig. 3C, surface gp46 detection was dose-dependently suppressed by PPS. On the other hand, the total amount of gp46 remained unaffected by PPS (Fig. 3D). Collectively, these data suggest that PPS significantly hinders HTLV-1 viral particle production (Fig. 2A), likely by interfering with HTLV-1 envelope protein expression on the cell surface (Fig. 3).
PPS targets HTLV-1 surface gp46 and inhibits syncytium formation. (A) Syncytia were observed in pX1MT-M-transfected (top) or pCAG-Env-transfected (bottom) 293FT cells. PPS treatment for 24 h reduced the number and size of syncytia. Black arrows indicate the syncytia. (B) The number of syncytia in pX1MT-M- or pCAG-Env-transfected 293FT cells upon PPS treatment for 24 h was counted. The results are shown as the average number of syncytia in at least five random areas with the SD. (C) Surface gp46 on pX1MT-M-infected 293FT or MT-2 and MT-4 cells was stained and analyzed by flow cytometry. The top histograms represent the staining results for the control cells and cells treated with 100 μg/ml of PPS for clarity. The bottom bar graphs summarize and represent the percentage of gp46-expressing cells for all concentrations of PPS. Results are shown as the average percentage of duplicates with the SD. (D) Whole-cell lysates were prepared from pX1MT-M-infected 293FT cells or MT-2 and MT-4 cells, and Western blotting was performed to detect total gp46 expression. The band intensity was calculated by ImageJ software, and the normalized results are shown as bar graphs below. All statistical analyses of the data presented in this figure were carried out by Student's t test. *, P < 0.05; **, P < 0.01.
PPS interferes with HTLV-1 transmission in T cells.Since PPS affects cell surface HTLV-1 envelope protein gp46 (Fig. 3), we wondered whether PPS would interfere with HTLV-1 transmission, which also requires the participation of the viral envelope protein. We used a reporter cell line for HTLV-1 infection, JETWT35, in which a red fluorescent protein, tdTomato, is expressed under the control of Tax-responsive elements (36). It is well-known that HTLV-1 transmission relies mostly on cell-cell contact (37). So, we conducted the experiment by coculturing JETWT35 cells with an HTLV-1 donor cell line, ATL-2, as previously described (36). While the ATL-2/JETWT35 cell coculture successfully infected JETWT35 cells, the immediate addition of PPS to the coculture greatly blocked the infection or transmission of HTLV-1 (Fig. 4A). This effect was not a result of the cytotoxicity of PPS because it did not affect the viability of JETWT35 cells (Fig. 4B). We further confirmed that this inhibition of HTLV-1 transmission by PPS was dose dependent (Fig. 4C). Moreover, similar results were obtained by coculturing JETWT35 cells with MT-2 cells (Fig. 4C), which is another donor cell line widely used for HTLV-1 infection (38–43). Interestingly, this inhibition of HTLV-1 transmission by PPS became inefficient when PPS was added to the coculture 12 h later (Fig. 4D). This indicates that the effect of PPS was likely limited to blocking cell-cell contact-mediated HTLV-1 transmission but not intracellular Tax expression or function, which also agrees with our earlier observation (Fig. 2D and E).
PPS blocks HTLV-1 transmission in T cells. (A) Cells of the HTLV-1-infected T-cell line ATL-2 were mixed 1:3 with cells of the HTLV-1 reporter T-cell line JETWT35 in the presence or absence of PPS. Forty-eight hours later, cells were analyzed with a fluorescence microscope and images were obtained by a Keyence imaging system. Red fluorescence indicates infected JETWT35 cells. (B) The MTT assay, performed in triplicate, suggests that the viability of JETWT35 cells was not affected by PPS at 10 or 100 μg/ml. (C) Cells of the HTLV-1-infected T-cell line ATL-2 or MT-2 were cultured with JETWT35 cells in the presence or absence of various concentrations of PPS, and fluorescence was analyzed by use of a FACSVerse flow cytometer 48 h later. (Left) Representative histograms of FACS; (right) average rates of JETWT35 cells infection. (D) The ATL-2/JETWT35 cell or MT-2/JETWT35 cell coculture was either treated with PPS for 36 h, followed by PPS removal, or left without PPS until 12 h later, before FACS analysis for HTLV-1 infection at 48 h. The schemes of PPS administration are illustrated on the left. Representative histograms are shown in the middle, and the percentages of infected cells are shown on the right. All FACS experiments whose results are presented in this figure were performed in duplicate, and the results are shown as the average percentage with the SD. PE, phycoerythrin. All statistical analyses were carried out by Student's t test. *, P < 0.05; **, P < 0.01; n.s., not significant.
PPS suppresses HTLV-1 infection of primary endothelial cells.The blood-brain barrier (BBB) has been implicated in the pathogenesis of HAM/TSP. Now that we have found that PPS blocks HTLV-1 transmission in T cells (Fig. 4), we next continued to investigate if PPS could prevent HTLV-1 transmission to BBB cells. To this end, we used human umbilical vascular endothelial cells (HUVEC) because they are primary cells that resemble BBB cells and can be infected by HTLV-1 (44, 45). We used lethally irradiated MT-2 cells as the donor cells to infect HUVEC, which is a widely used method of HTLV-1 infection (38–42). We first confirmed that there were no viable MT-2 cells 3 days after lethal irradiation by a trypan blue exclusion assay (data not shown). To test its effect on HTLV-1 transmission, PPS was immediately added to the MT-2 cell/HUVEC coculture. Three days later, PPS was removed along with dead MT-2 cells by use of a medium change. Another 4 days later, we lysed HUVEC for RNA extraction and measured Tax mRNA expression as a marker of HTLV-1 infection efficiency. As shown in Fig. 5A, while irradiated MT-2 cells successfully infected HUVEC, addition of the reverse transcriptase inhibitor azidothymidine (AZT) greatly blocked the infection. This suggests that the Tax mRNA that we detected was expressed by infected HUVEC rather than the contaminating MT-2 cells. On the other hand, the simultaneous addition of PPS at 100 μg/ml also efficiently blocked the infection of HUVEC (Fig. 5A). We further proved that this inhibition by PPS was dose dependent (Fig. 5B) and was not because of the reduced viability of HUVEC (Fig. 5C). Tax has been reported to induce cytokine expression in CNS cells and to contribute to the elevated cytokine level in HAM/TSP patients (46, 47). We also observed the induction of proinflammatory cytokines in HTLV-1-infected HUVEC, likely due to Tax expression (Fig. 5D). This induction was greatly suppressed by PPS (Fig. 5D), probably as a result of suppressed Tax expression (Fig. 5B). Moreover, we confirmed that IL-6 secretion by infected HUVEC was dose-dependently inhibited by PPS (Fig. 5E, left), in agreement with the quantitative real-time PCR (qPCR) result (Fig. 5D). In addition, the PPS added 12 h later lost the ability to inhibit IL-6 secretion (Fig. 5E, right), which also agrees with our finding in T cells that PPS was less efficient or not efficient at blocking HTLV-1 transmission when added at a later point (Fig. 4D). Note that lethally irradiated MT-2 cells did not produce any detectable IL-6 (Fig. 5E), indicating that the result was not because of MT-2 cell contamination. The above-described observations may further imply that PPS has the potential to suppress local inflammation by preventing HTLV-1 infection of primary endothelial cells and subsequent proinflammatory cytokine induction.
PPS inhibits HTLV-1 infection of HUVEC and subsequent proinflammatory cytokine expression. (A) HUVEC were infected by lethally irradiated MT-2 cells in the presence of DMSO, the reverse transcriptase inhibitor azidothymidine (AZT; 5 μM), or 100 μg/ml PPS and then subjected to RNA extraction and qPCR quantification of Tax expression. The results for HTLV-1-uninfected HUVEC are shown as a negative control (NC). Suppressed Tax expression in cells treated with AZT suggests successful infection of HUVEC. A decline in Tax expression in HUVEC was seen in the presence of PPS. (B) An experiment similar to that whose results are presented in panel A was performed with various concentrations of PPS. A dose-dependent inhibition of HTLV-1 infection by PPS was observed. (C) Results of the MTT assay performed in triplicate. The results indicate that HUVEC viability was not inhibited by 10 or 100 μg/ml PPS. (D) The same cDNA samples from the experiment whose results are presented in panel B were used to quantify six proinflammatory cytokine genes. Data are presented after normalization to the expression level in the negative-control cells. The expression of all cytokines was dose-dependently reduced in the presence of PPS. (E) The MT-2 cell/HUVEC coculture was treated with PPS for 60 h, followed by PPS removal, or the culture was left without PPS until 12 h later for a total 60 h of PPS treatment. At 72 h, all media were replaced by fresh media and HUVEC were further cultured for 4 days. Finally, the supernatants were collected for IL-6 measurement by ELISA. Supernatants from irradiated MT-2 cells and uninfected HUVEC were also collected for IL-6 ELISA at day 7 (7 d) after a medium change as described for infected HUVEC at day 3. ELISA was performed in duplicate, as recommended by the manufacturer. All qPCR experiments whose results are presented in this figure were performed in triplicate. Results are shown as the average with the SD. All statistical analyses were carried out by Student's t test. *, P < 0.05; **, P < 0.01.
PPS inhibits adhesion and transmigration of HTLV-1-infected T cells.HTLV-1 transmission is not the only consequence of the contact between HTLV-1-infected T cells and BBB (45). It has been reported that HTLV-1-infected CD4 T cells can also adhere to and migrate through the BBB (48, 49), a process called transmigration. Both viral transmission and the transmigration of infected T cells are likely to happen as a result of the contact between infected T cells and the BBB. Since we have proven that PPS suppresses HTLV-1 transmission to HUVEC (Fig. 5), we next set out to investigate the role of PPS in the other event, the transmigration of infected T cells. However, because PPS was found to suppress lymphocyte migration (19, 50, 51), we first decided to see whether PPS would inhibit the migration of HTLV-1-infected cells. The conventional transwell migration assay did not reveal any inhibition of the migration of MT-2 or MT-4 cells by PPS (Fig. 6A). Similarly, PPS did not inhibit the migration of HTLV-1-negative T-cell lines, such as Jurkat and Molt4 cells (Fig. 6A). Transmigration requires lymphocytes to migrate through endothelial cells of blood vessels and into inflamed tissues in order to contribute to innate or adaptive immunity (52). To examine the effect of PPS on transmigration, we seeded HUVEC in a Boyden chamber and let the cells grow to reach 100% confluence (53). Then, we seeded MT-2 or MT-4 cells onto the HUVEC monolayer and counted the cells that migrated through the monolayer 4 h later. As shown in Fig. 6B, both MT-2 and MT-4 cells demonstrated less transmigration through the HUVEC monolayer in the presence of 100 μg/ml PPS. On the other hand, PPS did not affect the transmigration of Jurkat or Molt4 cells (Fig. 6B). Since transmigration is a stepwise process where adhesion is a key initial step (52), we next performed an adhesion assay to see if PPS might affect T-cell adhesion. Carboxyfluorescein succinimidyl ester (CFSE)-labeled T cells were allowed to adhere to the HUVEC monolayer, and adhesion was evaluated by measuring the green fluorescence of cells adhering to HUVEC (54). As shown in Fig. 6C, PPS inhibited the adhesion of MT-2 and MT-4 cells but not Jurkat or Molt4 cells to HUVEC in a dose-dependent manner. Therefore, PPS suppresses the transmigration of HTLV-1-infected cells, likely by inhibiting their adhesion to endothelial cells.
PPS suppresses adhesion and transmigration of HTLV-1-infected cells. (A) PPS at 100 μg/ml did not affect the general migrating ability of HTLV-1-infected (MT-2 and MT-4) and uninfected (Jurkat and Molt4) T cells. Migration was performed in triplicate at 24 h after seeding as previously described (71). (B) The transmigration of the above-mentioned HTLV-1-infected and -uninfected T cells through a 100% confluent HUVEC monolayer was investigated as described in Materials and Methods and a previous report (53). Transmigration was performed in triplicate, and the results are shown as the average number (with the SD) of cells that transmigrated into the bottom chamber. (C) The adhesion of the above-mentioned HTLV-1-infected and -uninfected T cells to a 100% confluent HUVEC monolayer in the presence or absence of PPS was quantified as previously described with adaptations (54). The experiment was performed in triplicate, and the results are shown as the average fluorescence units (with the SD) of cells that adhered to HUVEC. All statistical analyses were carried out by Student's t test. *, P < 0.05; **, P < 0.01.
PPS blocks HTLV-1 infection in huPBMC NSG mice.Next, we tested the effect of PPS on HTLV-1 infection of PBMC ex vivo and in vivo. We isolated PBMC from two healthy donors, infected the PBMC with HTLV-1 ex vivo, and cultured the cells in the presence or absence of PPS for 10 days. PPS-treated PBMC demonstrated better viability during the infection course (Fig. 7A) and also had much lower PVL (Fig. 7B). To evaluate the effect of PPS on HTLV-1 infection in vivo, we generated PBMC-humanized NOD-scid IL2Rgammanull (huPBMC NSG) mice and infected the mice with lethally irradiated MT-2 cells (Fig. 7C) (55). Ten days later, we sacrificed all the mice and recovered peritoneal cavity cells and splenocytes. PPS-treated mice had more cells (Fig. 7D) with higher percentages of human CD4 T cells (Fig. 7E) but lower HTLV-1 PVL in the peritoneal cavity (Fig. 7F). In addition, we found that peritoneal cavity cells from PPS-treated mice expressed lower levels of HTLV-1 genes, including Tax and HBZ (Fig. 7G). Consistent with the PVL result in the peritoneal cavity, PPS-treated mice also had lower PVL in the spleen (Fig. 7F). Importantly, all mouse samples were free of MT-2 cell contamination, as no MT-2 cell integration-specific product was detected (Fig. 7I). The results presented above collectively demonstrate that PPS is able to block HTLV-1 infection ex vivo and in vivo.
PPS inhibits HTLV-1 infection in huPBMC NSG mice in vivo. PBMC (107) from donor 1 or 2 were mixed with 106 lethally irradiated (150 Gy) MT-2 cells in the presence or absence of PPS (100 μg/ml) and cultured in full RPMI medium supplemented with 50 units/ml IL-2. (A) Viable cells were counted at days 3, 6, and 10 by a trypan blue exclusion assay. (B) Genomic DNA was extracted from the remaining cells at day 10, and the HTLV-1 PVL was quantified by qPCR. qPCR was performed in triplicate, and the results are normalized to those for the control (Ctrl) PBMC of donor 1 and shown as the average with the SD. (C) Scheme of HTLV-1 infection of huPBMC NSG mice. i.p., intraperitoneal. (D) Cell count of recovered peritoneal cavity cells. (E) Anti-human CD3/CD4 staining of recovered mouse peritoneal cavity cells. Percentages for each of the 10 mice are shown on the right. (F) HTLV-1 PVL in peritoneal cavity cells recovered from HTLV-1-infected huPBMC NSG mice. PVL was measured by qPCR in triplicate, and the results are normalized to those for MT-2 cells. Note that PVL results for only four NSG mice in the PPS group are shown because the PVL of the last one was not detected. (G) Tax and HBZ mRNA expression in peritoneal cavity cells. qPCR was performed in triplicate, and the results are normalized to those for a control mouse. (H) HTLV-1 PVL in splenocytes isolated from HTLV-1-infected huPBMC NSG mice. PVL was measured by qPCR in triplicate, and the results are normalized to those for MT-2 cells. (I) Threshold cycle (CT) values of qPCR amplification of RAG1 and MT-2 cell integration-specific (MT2-IS) products from mouse samples and cell lines. C1 to C5, HTLV-1-infected control mice; P1 to P5, PPS-treated infected mice. Mice C1 to C3 and P1 to P3 were injected with PBMCs from donor 1, while the rest of the mice were injected with PBMCs from donor 2. ATL-2 and Jurkat cells were negative controls for MT2-IS product amplification. Amplification of human RAG1 DNA suggests successful PBMC humanization in NSG mice, while no amplification of MT2-IS DNA in all isolated mouse samples indicates they had no MT-2 cell contamination. Statistical analyses of the data presented in panels A and B were carried out by Student's t test, and mouse data from panels D to H were analyzed by the Mann-Whitney test. *, P < 0.05; **, P < 0.01.
PPS suppresses dermatitis development in HBZ-Tg mice.Results from many animal models have suggested that PPS suppresses inflammation in vivo (17–21). Next, we tried to see if PPS is anti-inflammatory using an animal model of HTLV-1-associated inflammation. Previously, we generated both Tax- and HBZ-Tg mice that restrict transgene expression in CD4 T cells in order to best recapitulate their roles in human CD4 T cells (56). While our Tax-Tg mice did not develop any inflammation (56), our HBZ-Tg mice developed inflammation in the skin and lung (57), which resembles the symptoms observed in HTLV-1-infected individuals (58). So, we chose HBZ-Tg mice as our mouse model for HTLV-1-related inflammation and monitored them for dermatitis development. At the end of the experiment, all untreated control HBZ-Tg mice (n = 5) developed dermatitis, whereas PPS-treated HBZ-Tg mice (n = 5) did not (Fig. 8A and B). Pathological examination further revealed the absence of inflammation in the lungs of the PPS-treated HBZ-Tg mice (Fig. 8C). To see whether PPS affected the function of HBZ-expressing CD4 T cells, we first performed intracellular cytokine staining. Although there was a difference in IL-2 expression, expression of the other three cytokines, including gamma interferon (IFN-γ) and TNF-α, which are known to be elevated in HAM/TSP patients (59, 60), did not change (Fig. 8D). No inhibition of the major cytokines associated with HAM/TSP suggests that PPS might not suppress inflammation by altering the activities of CD4 T cells. In accordance with this observation, microarray analysis of CD4 T cells isolated from both groups did not identify any clear change in the expression of proinflammatory molecules (data not shown). However, gene ontology analysis of differentially expressed genes demonstrated that PPS instead affects genes encoding proteins associated with the cell membrane or extracellular matrix (Fig. 8E and F), indicating that the modulation of membrane functions in vivo by PPS may contribute to its anti-inflammatory effect.
PPS suppresses dermatitis development in HBZ-Tg mice. (A) Dermatitis was observed in control HBZ-Tg mice (n = 5) but not in PPS-treated HBZ-Tg mice (n = 5). Representative photos for each group are shown. (B) The incidence of dermatitis development in control and PPS-treated HBZ-Tg mice is shown. **, P < 0.01. (C) Histological analysis of lungs from representative control mice (left) or PPS-treated HBZ-Tg mice (right). Boxes indicate regions displayed at a higher magnification in the bottom row (magnification, ×20). (D) Intracellular staining of inflammatory cytokines in CD4 T cells isolated from control or PPS-treated HBZ-Tg mice was performed as previously described (69). There was a statistically significant reduction in IL-2 production. IL-6 production had a declining trend, but it did not reach statistical significance. *, P < 0.05. MFI, mean fluorescence intensity. (E and F) The most enriched functional clusters of the PPS-downregulated (E) and -upregulated (F) genes in HBZ-Tg mice. CD4 T cells were isolated from control and PPS-treated HBZ-Tg mice and subjected to microarray analysis. Genes regulated differentially by PPS were identified by GeneSpring GX software and analyzed by the DAVID tools (74, 75).
DISCUSSION
PPS is already commercially available under the trade name Elmiron to treat the bladder pain and discomfort associated with interstitial cystitis (IC). However, the exact mechanism of its action is unclear. It is believed that PPS can repair the damaged GAG layer on the bladder epithelium and prevent further damage in vivo (16). PPS has also been found to suppress inflammation by reducing histamine secretion from mast cells in vitro (16). It seems that PPS mainly exerts its effects on the extracellular environment by either forming a protective GAG layer or inhibiting inflammatory molecule secretion (16). Our observation of PPS preventing tissue damage in HBZ-Tg mice (Fig. 8) may reflect its GAG layer role. On the other hand, the ability of PPS to suppress HTLV-1 virion production (Fig. 2) seems to resemble its action in the inhibition of histamine secretion (16). A conserved mechanism(s) for PPS to antagonize inflammation may be present under different inflammatory conditions.
It has been reported that PPS has broad antiviral activities against enveloped viruses (28, 29). As with HTLV-1, whose envelope protein was targeted by PPS (Fig. 3), the replication of other viruses, such as HIV-1, influenza A virus, and respiratory syncytial virus, was also hindered by PPS at either the viral attachment or virus-cell fusion steps (29). One intriguing observation is that while PPS decreases surface gp46 to a similar extent in pX1MT-M-infected 293FT cells and HTLV-1-infected T cells, including MT-2 and MT-4 cells (Fig. 3C), its inhibition of HTLV-1 virion production seemed to be more efficient in the former (Fig. 2A). Also, the half-maximal inhibitory concentration (IC50) of PPS in pX1MT-M-infected 293FT cells was the lowest (Table 2). It might be due to unidentified variations in PPS action on distinct cell types, or an additional mechanism other than targeting gp46 by PPS may exist for pX1MT-M-infected 293FT cells. Nevertheless, the fact that PPS mainly disrupts the replication stages where a virus-cell membrane interaction is involved highlights its nature as a GAG. Moreover, PPS is also able to efficiently block HTLV-1 transmission in T cells (Fig. 4). However, the observation that PPS is no longer effective or is less effective when it is used 12 h later (Fig. 4D) indicates that early events of transmission, such as viral entry, may be the key step targeted by PPS for blocking HTLV-1 transmission.
IC50 of PPSa
HTLV-1 infection of a BBB cell line has been reported, and infection of BBB cells has been implicated in HAM/TSP pathogenesis (42, 61). In this study, we demonstrated that HTLV-1 infection of primary endothelial cells (HUVEC) elevates their expression of proinflammatory cytokines (Fig. 5), likely via the expression of Tax (46, 47). On the other hand, PPS was able to block HTLV-1 infection of HUVEC, as demonstrated by suppressed Tax expression (Fig. 5B). As a result, Tax-induced cytokine expression was also efficiently suppressed by PPS (Fig. 5D). It is worth noting that the repressed cytokine expression in the presence of PPS more likely resulted from blocked HTLV-1 transmission than from any direct effect on Tax, since PPS likely has no effect on Tax expression or function (Fig. 2). Nevertheless, the inhibition of HTLV-1 infection of HUVEC and the subsequent inflammatory cytokine production suggest that PPS has the potential to protect the BBB from HTLV-1 infection.
Although PPS did not suppress the migration of HTLV-1-infected T cells (Fig. 6A), it significantly impaired their transmigration through HUVEC (Fig. 6B). Whereas migration implies the action of lymphocytes moving toward nutrients or chemoattractants, transmigration describes a more complex process that requires lymphocytes to first adhere to and eventually migrate through endothelial cells (52). When tissues are inflamed, they activate endothelial cells, which then recruit lymphocytes to the inflamed sites by transmigration (52). The interaction of lymphocytes and the endothelium is critical to the successful capture of lymphocytes and is generally initiated by adhesion molecules, such as selectins, on the surface of endothelial cells (52). Our finding that PPS specifically suppresses transmigration but not the migration of HTLV-1-infected T cells again highlights the feature of PPS affecting processes involving cell-cell contact. Indeed, adhesion of HTLV-1-infected T cells to HUVEC is blocked by PPS (Fig. 6C). On the other hand, PPS did not show inhibition of the transmigration of HTLV-1-uninfected T cells, including Jurkat and Molt4 cells, which is a reasonable result, since their adhesion to HUVEC was unaffected by PPS (Fig. 6C). It is not yet clear why PPS did not inhibit the adhesion of HTLV-1-uninfected T cells. One possibility might lie in the distinct adhesion molecules expressed by HTLV-1-infected and -uninfected T cells (62, 63) because these molecules also play important roles in adhesion to endothelial cells (52). For example, cell adhesion molecule 1 (CADM1) is known to be expressed specifically on HTLV-1-infected T cells (64). Nevertheless, impaired transmigration of HTLV-1-infected cells implies that PPS may be able to prevent HTLV-1 dissemination in the CNS (since glial cells are also susceptible to HTLV-1 infection) (47). Furthermore, recent evidence suggests that the cytotoxic T-cell response against infiltrating HTLV-1-infected CD4 T cells could lead to bystander neural damage in the CNS (65). Based on our findings, PPS would also be expected to play a role in countering this bystander effect.
PVL is an important risk factor for disease progression not only to HAM/TSP but also to ATL. However, although we found that PPS inhibited the survival of certain HTLV-1-infected cell lines with relatively low CC50 (Fig. 1 and Table 1), it had a minimal effect on the viability of ATL-2 and MT-2 cells, which may render its use in ATL less promising. Moreover, not all HAM/TSP patients responded to PPS treatment with a PVL decrease (10), which indicates a lesser significance of the cytotoxicity of PPS against HTLV-1-infected cells.
In this study, we used two mouse models to evaluate the effects of PPS on HTLV-1 in vivo. NSG mice have been found to support the engraftment of HTLV-1-infected cells due to their severely deficient immunity (66, 67). Using NSG mice, we have also previously achieved establishment of a PBMC-humanized mouse model for HTLV-1 infection (55). Consistent with our in vitro results that reveal the role of PPS in suppressing HTLV-1 replication and transmission, PPS significantly inhibited HTLV-1 infection in huPBMC NSG mice (Fig. 7). Thus, the interference of PPS with HTLV-1 infection has been confirmed both in vitro and in vivo. For the second mouse model, we used HBZ-Tg mice to study the anti-inflammatory role of PPS in vivo (Fig. 8). Although both Tax and HBZ are likely involved in the pathogenesis of HAM/TSP, we used HBZ-Tg mice instead of Tax-Tg mice for two reasons. First, the in vitro anti-HTLV-1 ability of PPS did not result from a direct alteration in Tax expression or function (Fig. 2). Instead, PPS more likely targets the HTLV-1 envelope gp46 protein to inhibit HTLV-1 production (Fig. 3) and suppresses Tax-induced cytokine expression indirectly through blocking HTLV-1 transmission (Fig. 5). This means that Tax is not a direct target of PPS, even though Tax is indispensable for HTLV-1 replication. Second, our Tax-Tg mice with Tax expression in CD4 T cells did not develop inflammation (56). Although the oncogenicity of Tax has been proven by various Tax-Tg mouse models, Tax expression is not restricted to CD4 T cells in these mice due to the use of promoters like the HTLV-1 LTR, CD3, or the granzyme B promoters for Tax expression (57). When we used a CD4-specific promoter to limit Tax expression in the CD4 T cells of transgenic mice, no leukemia/lymphoma or inflammation was found (56). In contrast, the HBZ-Tg mice that were generated in the same way developed CD4 T-cell lymphoma and inflammation, such as dermatitis. Therefore, we eventually chose to use HBZ-Tg mice as the inflammation model of HTLV-1 and successfully proved the anti-inflammatory effect of PPS in vivo.
So far, PPS has been shown to be effective in various animal models of inflammation (17–21). A recent study demonstrated that PPS is able to reduce alphavirus-induced cartilage damage and inflammation (68). In this study, we found that PPS suppresses dermatitis development in HBZ-Tg mice. However, although HBZ-expressing CD4 T cells are expected to induce inflammation in an intrinsic manner (69), they did not seem to be the direct target of PPS for the suppression of inflammation. Nevertheless, the membranes and extracellular matrix of HBZ-expressing CD4 T cells were affected by PPS (Fig. 8E and F). As a GAG, it is likely that PPS may form an extra layer on tissues of HBZ-Tg mice and provide protection for the resident cells of the skin or lung, resembling its actions in IC (16). Moreover, the presence of PPS may further impair the recruitment and infiltration of lymphocytes to these tissues, as PPS was able to suppress infected T-cell adhesion and transmigration (Fig. 6).
In conclusion, we systemically analyzed the functions of PPS both in vitro and in vivo, present evidence for its administration in HAM/TSP, and further argue for its potential use in other HTLV-1-associated inflammatory diseases. In addition, our in vivo results obtained from HBZ-Tg mice verified the anti-inflammatory role of PPS and are in accordance with the result of the clinical trial on HAM/TSP patients (10). Although PPS is not an HTLV-1-specific antagonist, its safety for long-term administration and its anti-HTLV-1 properties demonstrated in this study may allow it to be a viable option for the treatment of HTLV-1-induced inflammatory diseases.
MATERIALS AND METHODS
Animals.C57BL/6J mice were purchased from CLEA Japan. Transgenic mice expressing the HBZ gene in CD4 T cells (HBZ-Tg mice) were described previously (58). HBZ-Tg mice were maintained in a specific-pathogen-free (SPF) facility. Four-week-old female NSG mice were purchased from Charles River Laboratories Japan and were maintained in microisolator cages under SPF conditions. Animal experiments were performed in strict accordance with the Japanese animal welfare bodies (law no. 105, dated 19 October 1973, modified on 2 June 2006) and the Regulation on Animal Experimentation at Kyoto University. The protocol was approved by the Institutional Animal Research Committee of Kyoto University (permit number A17-1-3).
Reagent.PPS was dissolved in reverse osmosis water and filtered through a 0.22-μm-pore-size filter before use. A final concentration of 0.264 mg/ml of PPS was given to mice based on a previous report (70). For HBZ-Tg mice, PPS treatment started when the mice were 6 weeks old. For huPBMC NSG mice, PPS was given throughout the experiment.
Cell culture.ATL-43T, ED, Hut102, SLB-1, ATL-2, MT-2, MT-4, and TL-Om1 cells are HTLV-1-infected T-cell lines, while CEM, Jurkat, and Molt4 cells are HTLV-1-uninfected T-cell lines. The cells were cultured in full RPMI 1640 supplemented with 10% fetal bovine serum (FBS) and antibiotics (penicillin and streptomycin). Kit225 is an IL-2-dependent HTLV-1-uninfected T-cell line and was cultured in full RPMI 1640 supplemented with 100 unit of recombinant human IL-2 (Peprotech). JETWT35 is a stable line of Jurkat cells and was cultured in 250 units/ml G418 containing full RPMI 1640 as previously described (36). 293FT cells were cultured in Dulbecco modified Eagle medium (DMEM) supplemented with 10% FBS, 500 μg/ml of G418, and antibiotics. HUVEC were purchased from Lonza and maintained in endothelial cell growth medium 2 (EGM-2; Lonza). HUVEC were used at between passages 3 and 6 in this study. All cells were maintained in a humidified incubator at 37°C with 5% CO2.
MTT assay.All cells were resuspended at 3 × 104 to 5 × 104/ml, and PPS was added at the concentrations indicated above. Then, 100 μl of the cell suspension with or without PPS was seeded into 96-well plates and a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay was performed as previously described (71). For 293FT cells and HUVEC, PPS was added the next day after seeding (considered day 0) to allow the cells to adhere and grow.
CD4 T-cell activation.CD4 T cells from two healthy donors were isolated using a human CD4 T lymphocyte enrichment set (BD Biosciences) and then stimulated using Dynabeads Human T-Activator (Thermo Fisher Scientific) following the manufacturer’s instructions. The medium was supplemented with recombinant human IL-2 (Peprotech) at 50 units/ml. Activated cells (1 × 105/ml) were used in the MTT assay. The experiments using primary samples in this study were conducted according to the principles expressed in the Declaration of Helsinki and approved by the Institutional Review Board of Kyoto University (permit numbers G310 and E2005).
Annexin V and 7-AAD staining.ED or MT-2 cells were cultured in the presence or absence of PPS for 48 h. Then, the cells were stained with 7-aminoactinomycin D (7-AAD; BioLegend) and allophycocyanin-annexin V (BioLegend) following the manufacturer’s instructions. Flow cytometry was performed on FACSVerse flow cytometer (BD Biosciences), and the data were analyzed using FlowJo software (BD Biosciences).
HTLV-1 gp46 staining.pX1MT-M (30)-transfected 293FT or MT-2 and MT-4 cells were cultured in the presence or absence of PPS for 24 h. After removing the medium, the cells were directly resuspended in phosphate-buffered saline (PBS) containing 1% FBS and stained with an antibody to HTLV-1 gp46 (1/100; clone 67/5.5.13.1; Abcam) at room temperature for 15 min. Then, the cells were washed twice with PBS and stained with a fluorescein isothiocyanate (FITC)-conjugated anti-mouse IgG antibody (1/1,000; BioLegend) at room temperature for 15 min. After another PBS wash, the cells were resuspended and loaded onto a FACSVerse flow cytometer for examination (BD Biosciences). The data were analyzed using FlowJo software (BD Biosciences).
ELISA.HTLV-1-infected T cells (2 × 105) were seeded in 12-well plates and cocultured with PPS for 12 or 24 h. Then, the cells were harvested and centrifuged at 3,000 rpm for 5 min. The supernatants were collected and passed through a 0.45-μm-pore-size filter. For 293FT cells, PPS was added 24 h after transfection of pX1MT-M. The supernatants were collected 12 and 24 h after PPS addition. The number of HTLV-1 virions in the supernatant was measured using a Retro-Tek HTLV p19 antigen enzyme-linked immunosorbent assay (ELISA) kit (ZeptoMetrix).
For the IL-6 ELISA, at day 7 after infection, supernatants of HUVEC were collected, centrifuged, and passed through a 0.45-μm-pore-size filter. Then, the IL-6 was quantified using a human IL-6 Quantikine ELISA kit (R&D Systems).
huPBMC NSG mice and infection with HTLV-1.Human PBMCs were isolated from two healthy donors with Ficoll-Paque Plus (GE) and resuspended in PBS. Human PBMC (107) were mixed with 106 lethally irradiated (150 Gy) MT-2 cells in the presence (n = 5) or absence (n = 5) of 100 μg/ml PPS and immediately injected intraperitoneally into the mice. Mice in the PPS group (n = 5) were also given PPS-containing water (0.264 mg/ml) throughout the experiment. Ten days later, all mice were sacrificed and peritoneal cavity cells were recovered as reported previously (72). Peritoneal cavity cells were next subjected to fluorescence-activated cell sorting (FACS) staining and DNA and RNA extraction. Splenocytes were also isolated from each mouse for DNA extraction. The antibodies used for FACS were FITC-CD3 (clone UCHT-1; BD) and peridinin chlorophyll protein-Cy5.5-CD4 (clone OKT4; BioLegend).
qPCR.RNA was extracted using the TRIzol reagent (Thermo Fisher Scientific) as previously described (71). cDNA was synthesized from 1 μg of RNA by using a SuperScript III first-strand synthesis system (Thermo Fisher Scientific) with random hexamers as the primer. cDNA was quantified using SYBR green master mix (Roche Life Science) on a StepOnePlus real-time PCR system (Thermo Fisher Scientific). Endogenous 18S rRNA was quantified to normalize the total cDNA load. The primer sequences (from 5′ to 3′) were as follows: ATGGCCCACTTCCCAGGGTT and CCGAACATAGTCCCCCAGAG for Tax, ATGGCGGCCTCAGG and GCTTTCTCCCCTGGAGGGCC for HBZ, AACCCGTTGAACCCCATT and CCATCCAATCGGTAGTAGCG for 18S rRNA, CCAGACATGTTTGAAGACCT and AGATTGATCCATGCAGCCTT for IL1A, CGCCAGTGAAATGATGGCTT and TCGGAGATTCGTAGCTGGAT for IL1B, CTGCTGCACTTTGGAGTGAT and TACAACATGGGCTACAGGCT for TNFA, TGCTGCTGGTTCTGCTGCCT and GGTCTCCAATGAGGTGAGCA for TNFB, TCTCCTGTTGTGCTTCTCCA and CAATTGCCACAGGAGCTTCT for IFNB, and ATCTCAGCCCTGAGAAAGGA and CAGGCAAGTCTCCTCATTGA for IL6.
ACTB was used as the internal control for quantitative real-time PCR (qPCR) quantification of Tax and HBZ expression in the peritoneal cavity cells of huPBMC NSG mice, and the primer sequences (from 5′ to 3′) were AGGCCAACCGCGAGAAGATG and CTATCCCTGTACGCCTCTGG.
HTLV-1 PVL quantification.Genomic DNA was isolated from PBMC and mouse samples by use of a DNeasy blood and tissue kit (Qiagen). The HTLV-1 PVL was quantified using SYBR green master mix (Roche Life Science) on a StepOnePlus real-time PCR system (Thermo Fisher Scientific). The primer sequences (from 5′ to 3′) were CTCCTCAAGCGAGCTGCATG and CAGCTGGGGCTGTAATCACC for PVL, CCCACCTTGGGACTCAGTTCT and CACCCGGAACAGCTTAAATTTC for RAG1, CCTGTGGTGCCTCCTGAACT and CAGTGTGGTCCACAAACTCTTTG for the MT-2 cell integration-specific product.
Western blotting.293FT cells (4 × 105) were seeded in a 12-well plate and transfected with 1 μg of pX1MT-M on the next day. Twenty-four hours later, the old medium was replaced with fresh DMEM supplemented with or without PPS. For MT-2 and MT-4 cells, 4 × 105 cells were seeded in 12-well plates with or without PPS. Twenty-four hours later, the cells were washed once with PBS and lysed in radioimmunoprecipitation assay buffer (Nacalai Japan) containing a protease inhibitor cocktail (Nacalai Japan). The remaining procedures were performed as previously described (71). Images were captured using an ImageQuant LAS 4000 imaging system (GE Life Sciences). Band intensities were quantified using ImageJ software. Both the Tax and the α-tubulin antibodies have been described previously (73).
Luciferase reporter assay.293FT cells (3 × 105) were seeded in a 12-well plate and transfected with 0.1 μg of 5′ LTR-luciferase (Luc) and 10 ng of pPolIII-Renilla Luc (73) with or without 0.5 μg of pCG-Tax (73) using the Trans-IT LT1 reagent (Mirus Bio). Six hours later, the medium was replaced with fresh DMEM supplemented with or without 100 μg/ml PPS. Twenty-four hours later, the reporter assay was performed as previously described (71). For MT-2 and MT-4 cells, 2 × 105 cells were seeded in 12-well plates and transfected with 0.5 μg of 5′ LTR-Luc and 50 ng of pPolIII-Renilla Luc. The rest of the procedure was performed as described above for 293FT cells.
Infection of HUVEC by HTLV-1.HUVEC (1 × 105) were seeded in a 6-well plate and allowed to grow overnight. On the next day, 1 × 105 lethally irradiated (150 Gy) MT-2 cells were added to the HUVEC. Dimethyl sulfoxide (DMSO), AZT (5 μM), or PPS was simultaneously added to the coculture. Three days later, all old medium was replaced by fresh EGM-2 and the cells were further cultured without PPS for 4 days, before being harvested for RNA extraction.
A method of infection similar to that described above was performed for IL-6 ELISA measurement, except that PPS was added immediately or 12 h later. At day 3, all old medium was replaced by fresh EGM-2. At day 7, supernatants were collected for the IL-6 ELISA.
Infection of JETWT35 cells by HTLV-1.The infection of JETWT35 cells by HTLV-1 was performed as previously described (36). Generally, ATL-2 cells were mixed with JETWT35 cells 1:3 and MT-2 cells were mixed with JETWT35 cells 1:1 for optimal infection efficiency. PPS was added at the time points indicated in Fig. 4, and 48 h later the fluorescence was analyzed on a FACSVerse flow cytometer (BD Biosciences).
Adhesion assay.The adhesion assay was performed as described previously with adaptations (54). First, 1 × 104 HUVEC were seeded in a 96-well plate and allowed to grow overnight to reach 100% confluence. CFSE-stained MT-2 cells (5 × 104), CFSE-stained MT-4 cells (3 × 104), or CFSE-stained Jurkat or Molt4 cells (1 × 105) were then added to HUVEC in the presence of various concentrations of PPS. Then, the plate was incubated at 37°C for 30 min. Next, the plate was inverted and kept in that position for 30 min at room temperature to dislodge the nonadherent cells. Finally, all media were removed by aspiration and the plate was loaded on an Arvo microplate reader (PerkinElmer) for fluorescence measurement.
Migration and transmigration assays.The migration assay was performed as previously described (71). The transmigration assay was performed as previously reported with adaptations (53). A BD BioCoat Matrigel chamber was used, and the inserts were rehydrated according to the manufacturer’s instructions. HUVEC (5 × 104) were seeded in the inserts at day 1. On the next day, TNF-α was added to the inserts at 100 μg/ml to activate the HUVEC. At day 3, the media were removed and 5 × 104 MT-2 or MT-4 cells or 1 × 105 Jurkat or Molt4 cells resuspended in FBS-free RPMI 1640 were added to the inserts with or without PPS. The inserts were then loaded into wells containing RPMI 1640 supplemented with 10% FBS and 100 μg/ml of stromal-cell-derived factor 1-alpha (SDF1α) to allow transmigration to occur. Four hours later, the inserts were removed and the transmigrating cells in the lower chamber were counted with a hemocytometer.
Histological analysis.Tissue samples were fixed in 10% neutral buffered formalin and embedded in paraffin. Hematoxylin and eosin (H&E) staining was performed following standard procedures. Images were captured using a Provis AX80 microscope (Olympus) equipped with an Olympus DP70 digital camera and displayed using the DP manager system (Olympus).
Microarray analysis.Six-week-old HBZ-Tg mice were given either plain water or PPS-containing water for 6 weeks. Then, the mice were sacrificed and CD4 T cells were isolated from the spleen using a mouse CD4 T lymphocyte enrichment set (BD Biosciences). RNA was extracted using an RNeasy minikit (Qiagen) and subjected to microarray analysis with a SurePrint G3 Mouse GE 8x60K microarray following the manufacturer’s instructions (Agilent Technologies). Data were analyzed using GeneSpring GX software (Agilent Technologies) and the DAVID tools (74, 75).
Statistical analysis.The Mann-Whitney test was performed on the mouse data using GraphPad Prism software. The rest of the data were analyzed by Student's t test.
ACKNOWLEDGMENTS
We thank D. Derse for providing pX1MT-M; P. Miyazato, Y. Mitagami, H. Kinosada, C. Ohnishi, and Y. Higuchi for their help in the mouse experiments; K. Shimura for the reagent; and L. Kingsbury for proofreading.
G.M. and J.-i.Y. designed the study, performed experiments, analyzed the data, and wrote the manuscript. K.O. performed experiments and analyzed the data. T.M. conceived the original idea, helped design the study, and provided a key reagent. M.M. designed the study, analyzed the data, and wrote the manuscript.
This study was funded by Japan Society for the Promotion of Science (JSPS) KAKENHI grants JP16H05336 (to M.M.) and JP17K07166 (to J.-i.Y.); Project for Cancer Research and Therapeutic Evolution (P-CREATE) grants from the Japan Agency for Medical Research and Development (AMED; 18cm0106611h0002 and 18cm0106306h0003 to J.-i.Y. and M.M.); a Research Program on Emerging and Re-Emerging Infectious Diseases grant from AMED (18fk0108027h0003 to J.-i.Y. and M.M.); the Adaptable and Seamless Technology Transfer Program through Target-Driven R&D (A-STEP) of the Japan Science and Technology Agency (to T.M.); and a grant from the Princess Takamatsu Cancer Research Fund (to J.-i.Y.). G.M. was supported by a JSPS Postdoctoral Fellowship for Overseas Researchers and a Tokyo Biochemical Research Foundation (TBRF) Postdoctoral Fellowship for Asian Researchers.
The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
FOOTNOTES
- Received 8 March 2019.
- Accepted 27 May 2019.
- Accepted manuscript posted online 5 June 2019.
- Copyright © 2019 American Society for Microbiology.
