ATRA enhances the bystander effect of suicide gene therapy driven by the specific promoter LEP 503 in human lens epithelial cells

Construction of recombinant lentiviral vectors

We amplified the LEP503 promoter with KOD-PLUS kit (Toyobo, Osaka, Japan) according to the manufacturer’s procedure guide. A volume of 1×10⁶ HLECs was collected and lysed, then genomic DNA was isolated with Viogene VioFast Blood & Cell Genomic DNA Extraction Minprep System (Viogene-Biotek Corporation, Taibei, China) according to the manufacturer's instructions. The LEP503 promoter was cloned from HLEC genomic DNA by PCR, using the following set of primers: forward primer 5′-atc gat ctc cag cct ggg caa caa aac aag a-3′ and reversed primer 5′-gga tcc gtg ggc aac agt tcc gag gaa ggg t-3′. The products contained Cla I site at their 5′-ends and BamH I site at their 3′-ends. The amplification conditions were as follows: 94 °C for 2 min, 40 cycles of 94 °C for 15 s, 59 °C for 2 min and 68 °C for 2 min; and finally 72 °C for 10 min. The PCR products, 2,439 base pairs (bp), were incompletely double-digested with Cla I and BamH I restriction enzymes. The digested products (2,436 bp) were ligated into the promoter-less vector double-digested with Cla I and BamH I enzymes from Lenti-CMV-HSV-tk-EGFP, which has been constructed as previously described [10,11]. The recombinant lentiviral vector (Lenti-LEP503-HSV-tk-EGFP) was sequenced and the sequence was used to BLAST search in gene bank to confirm the nucleotide constitution.

Lentiviral vector preparation and virus infection

As previously described [10,11], the transfer vectors and the packaging system pSPAX2 and pMD2.G vectors (kindly provided by Dr. Didier Trono, University of Geneva, Geneva, Switzerland) were produced by transient transfection into 293T cells. Around 5×10⁶ 293T cells were seeded onto 10-cm-diameter dishes 24 h before transfection. A total of 20 μg of plasmid DNA was used to transfect one dish: 5 μg of the envelope plasmid pMD2.G, 5 μg of packaging plasmid pSPAX2, and 10 μg of transfer vector plasmid (LEP503-HSV-tk-EGFP or CMV-HSV-tk-EGFP). Precipitates were formed by adding the plasmids to: 250 µl 0.5 M CaCl₂ and 0.1× TE (pH, 8.0), final volume is 500 μl, mixing well, then adding 500 μl of 2× HEPES-buffered saline dropwise containing 281 mM NaCl, 50 mM HEPES, and 1,500 mM Na₂HPO₄ (pH 7.12), while vortexing and immediately adding the precipitate to the cultures. The medium (10 ml) was replaced 14~16 h after transfection. The conditioned medium was collected after another 24 h and cleared by low-speed centrifugation, followed by filtration through 0.45-μm-pore-size cellulose acetate filters. Virus was collected by ultracentrifugation for 90 min at 80,000× g at 4 °C. The pellet was re-suspended in 1 ml phosphate buffered saline (PBS) and stored at −80 °C. The concentrated supernatants were titered by serial dilutions of vector stocks on 1×10⁵ Hela cells, followed by fluorescence-activated cytometric analysis according to the formula: 1×10⁵ Hela cell × %EGFP positive cell × 1,000/μl virus. Titers of lentiviral vectors were 1×10⁸ ~1×10⁹ TU/ml.

Cell culture

The HLEC line SRA 01/04 (transformed by large T antigen), human adult retinal pigment epithelium cell line D407 (RPECs), HeLa, and 293T cell strains were purchased from the American Type culture Collection (ATCC, Manassas, VA). Human adult skin fibroblast cells (ASFCs) were a gift from the central laboratory of the Shanghai Eye and ENT Hospital, and these cells were cultured in DMEM (Invitrogen, Grand Island, NY) with 10% FBS (Invitrogen). The expression of EGFP was analyzed by plating cells at 2×10⁵ cells/well in six-well plates. After cells adhesion, the lenti-CMV-HSV-tk-EGFP or lenti-LEP503-HSV-tk-EGFP vectors were added (the multiplicity of infection [MOI] was 10 and 40, respectively). polybrene (Sigma, St. Louis, MO; 8 µg/ml) was added to all lentiviral vectors infections dishes. Transfection efficiency was determined by visualizing EGFP (enhanced green fluorescent protein) expression by fluorescence microphotography or by western blot assays. The mRNA transcription of HSV-tk was analyzed by RT–PCR assay, after infected cells were cultured for 48 or 72 h. The cytotoxicity of GCV (InvivoGen, San Diego, CA) was evaluated by plating cells into six-well dishes at a density of 5×10⁵ cells/well and incubating at 37 °C in 5% CO₂ for 24 h. The cells were then infected with lenti-LEP503-HSV-tk-EGFP and lenti-CMV-HSV-tk-EGFP (MOI=40 and 10, respectively). Twenty-four hours after infection, cells were treated with GCV (20 µM) and incubated at 37 °C in 5% CO₂. Morphological changes were observed by fluorescence microscopy and cell viability was determined using the Cell Counting Kit-8 Cell Proliferation Assay (CCK-8; Dojindo, Kumamoto, Japan) or the MTT assay at 12, 24, 36, 48, 60, and 72 h after treatment.

The in vitro cell killing effect was investigated by seeding HLECs infected by lenti-LEP503-HSV-tk-EGFP at a density of 5×10³ cells/well in 96-well plates and incubating for 24 h. The cells were then exposed to GCV (ganciclovir; 20 µM) and incubated at 37 °C in 5% CO₂ for 12, 24, 36, 48, 60, and 72 h. The bystander effect was assayed with the CCK-8 (cell counting kit-8 cell proliferation) or MTT ( 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assays. For co-culture, 24 h after being infected with lenti-LEP503-HSV-tk-EGFP vector, HLECs were trypsinized and mixed with uninfected cells at ratios of 0, 20, 40, 60, and 80%, and then were re-plated in 24-well culture plates. The cells were simultaneously exposed to 20 μM GCV, 10⁻⁷ M ATRA (all trans-retinotic acid), and (or) 40 μM 18-α-glycyrrhetinic acid (18-α-GA; Sigma-Aldrich, St. Louis, MO) for 72 h.

RNA extraction and RT–PCR analysis

Three weeks after stable expression of HSV-tk was observed in HLECs and Hela cells, total RNA was extracted from HLECs and Hela cells and reverse transcribed according to the protocol described for the RT–PCR kit (Promega, Madison, WI). Relative transcription of HSV-tk and EGFP was determined by performing a semi-quantitative PCR for HSV-tk and β-actin (ACTB; internal control) analysis with PCR amplification for 28 cycles. For HSV-tk, the forward primer was 5′-atc cca tcg ccg ccc tcc tgt gc-3′ and the reverse primer was 5′-cgc ccc cga aag ctg tcc cca atc-3′, for EGFP, the forward primer was 5′-cga gct gga cgg cga cgt aaa c-3′ and the reverse primer was 5′-gcg ctt ctc gtt ggg gtc ttt g-3′, and for ACTB, the forward primer was 5′-aac gag cgg ttc cga tgc cct gag-3′ and the reverse primer was 5′-tgt cgc ctt cac cgt tcc agt t-3′. The amplification conditions were as follows: 94 °C for 5 min, 28 cycles of 94 °C for 1 min, 58 °C for 2 min and 72 °C for 1 min; and final 72 °C for 10 min. The expected lengths of the HSV-tk, EGFP, and ACTB products were 613, 597, and 590 bp, respectively. A total of 10 µl of each PCR product was loaded onto a 1.5% agarose gel containing 0.5 µg/ml of ethidium bromide and separated by electrophoresis. The expression differences between HSV-tk and EGFP were quantified with Image-Pro Plus 5.0 software (Media Cybernetics, Silver Spring, MD) and normalized with respect to ACTB expression.

Protein collection and western blot analysis

Protein expression of Cx43 was analyzed by seeding 1×10⁶ HLECs in 25 mm² cell culture flasks for 24 h and then exposing them to 10⁻⁷ mol ATRA l⁻¹. The cells were collected and lysed with RIPA buffer including protease inhibitor cocktail (Pierce, Rockford, IL) at 0, 12, 24, 36, and 48 h after ATRA treatment, or at 72 h after exposure to different doses of 18-α-GA. EGFP expression was analyzed by collecting and lysing the cells with the same RIPA buffer at 0, 24, 48, and 72 h after infection. A total of 100 μg cell protein were separated by standard sodium dodecyl sulfate–PAGE under reducing conditions and blotted to polyvinylidene difluoride membranes using a semidry blotting apparatus. After blocking (0.5% defatted dry milk in TBS with 0.1% Tween-20 (TBST)), the membranes were incubated with the primary monoclonal mouse anti-Cx43 antibody (1: 2000; Sigma), goat anti-EGFP (1:1,000; Abcam, San Francisco, CA), mouse anti-GAPDH antibody (1:5,000; Santa Cruz Biotechnology, Santa Cruz, CA), and mouse anti-β-actin (1:5,000; Santa Cruz Biotechnology) in blocking buffer for 1 h, and were washed three times with TBST. The secondary peroxidase-conjugated donkey anti-mouse or goat antibodies in TBST were added for 1 h. Finally, the members were washed with TBST for 20 min, and immunoreactive proteins were visualized using the Western Blot Chemiluminescence System (Amersham™ ECL; GE Healthcare, Little Chalfont, Buckinghamshire, UK). Primary and secondary antibodies were diluted in 5% defatted dry milk in TBST. The spot densities for EGFP, Cx43, β-actin, and GAPDH were quantified with Image-Pro Plus 5.0 software and normalized by the respective GAPDH or β-actin.

Flow cytometry analysis

The specificity of the LEP503 promoter was analyzed by trypsinizing and collecting HLECs, RPECs and ASFCs 48 h after infection with lenti-LEP503-HSV-tk-EGFP (MOI=50) or lenti-CMV-HSV-tk-EGFP vector (MOI=5). Cells were washed with cold PBS and analyzed by flow cytometry (Beckman LSRII; Beckman Coulter, EI Cajon, CA). For analysis of cell death, cells were infected with lenti-LEP503-HSV-tk-EGFP or lenti-CMV-HSV-tk-EGFP vector for 24 h and following exposure to GCV (20 μM) and ATRA (10⁻⁷ M) for 72 h. Cells were then trypsinized, collected, and re-suspended in cold PBS (10⁶ cells/ml). After washing twice with cold PBS, cells were re-suspended in 100 μl binding buffer, then 2 μl Annexin-V-APC (20 μg/ml) was added and the cells were incubated for 15 min on ice. Then samples were transferred into detective tubes and 400 μl PBS was added. Cellular DNA was stained with 1 μl propidium iodide (PI, 50 μg/ml) for 2 min. Cell necrosis and apoptosis were determined by flow cytometry (Beckman LSRII). Cells without Annexin-V-APC and PI treatment were used as blank controls.

CCK-8 and MTT assay

Time- and dose-dependent cell viability was monitored using CCK-8 or MTT assay. CCK-8 is a sensitive nonradioactive colorimetric assay for determining the number of viable cells in cell proliferation and cytotoxicity assays. The solution is added directly to the cells and no pre-mixing of components is required. CCK-8 uses Dojindo's tetrazolium salt, WST-8[2-(2-methoxy-4-nitrophenyl)-3-(4-nitrophenyl)-5-(2, 4-disul-fophenyl)-2H-tetrazolium, monosodium salt], which produces a water-soluble formazan upon reduction in the presence of an electron carrier. The amount of yellow formazan produced in cells is directly proportional to the number of living cells. Briefly, at the end of the culture period, 10 μl of the CCK-8 solution was added to each well of the culture plate. After a 4-h incubation, absorbance at 450 nm was measured with a universal microplate spectrophotometer (Biotek Instruments, Gene Company Limited, Winooski, VT). The cytotoxic effect was indicated as a percentage of surviving cells (ratio of surviving cells after treatment and without treatment) using the following formula: Cell viability=(absorption of sample - absorption of background)/ (absorption of control - absorption of background) × 100%.

For the MTT assay, 20 μl MTT stock solution (0.5 mg/ml) was added to cells in each well of a 96-well tissue culture plate, which was then incubated at 37 °C for 4 h. The medium containing MTT was removed and then 200 μl DMSO (DMSO) was added per well to stop the reaction and dissolve the formazan crystals. After 15 min, the absorbance in each well was measured at 570 nm using the universal microplate spectrophotometer (Gene Company). Cultures without cells were used as blank controls. The percentage of surviving cells was calculated using the following formula: Cell viability=(absorption of sample - absorption of background)/ (absorption of control - absorption of background) × 100%.

Immunocytochemistry

Cells (seeded on cover glasses) were fixed with 4% paraformaldehyde for 15 min at room temperature then washed 3 times (3 min/per time) with PBS. Cells were perforated with 0.5% Triton X-100 in PBS for 15 min and then blocked in PBS containing 10% goat serum, 0.1% gelatin and 0.05% Triton X-100 for 30 min. Cells were incubated overnight at 4 °C with mouse antibodies against Cx43 (1:1,000; Sigma). After washing 3 times with PBS, cells were incubated for 30 min at 37 °C with Cy3-conjugated goat anti-mouse secondary antibodies. The nuclei were counterstained with DAPI (4,6-diamino-2-phenyl indole).

DNA rragmentation assay

Cells (1×10⁶) infected with Lenti-LEP503-HSV-tk-EGFP or Lenti-CMV-HSV-tk-EGFP were exposed to 20 µg GCV ml⁻¹. A sufficient number of cells (1×10⁷) were collected at 72 h and washed in PBS before resuspension in 1 ml of fresh lysis buffer (50 mM, pH 8 Tris, 100 mM EDTA, 1% SDS, 100 μg proteinase K ml⁻¹, and 1 mM NaCl). Samples were incubated at 37 °C for 16 h with gentle rocking. The DNA was extracted with an equal volume of phenol/chloroform/isoamyl alcohol solution (25:24:1) and precipitated with 3 M sodium acetate (pH 5.2) and 100% ice-cold ethanol. The pellets were dried by evaporation for 10 min, and resuspended in T-E buffer (10 mM Tris-HCl and 1 mM EDTA, pH 8.0) overnight at room temperature. The DNA was quantified by ultraviolet absorbance at 260 nm. Approximately 10 μg of DNA was added to each well after mixing with 5 μl loading buffer. Horizontal electrophoresis of DNA was performed for 2 h at 75 V on 1.5% agarose gels.

Statistical analysis

All data were expressed as means±SD and analyzed with SPSS version 16.0. A Dunnett t-test and a one-way ANOVA were performed to assess the statistical significance between different groups. Significance for all tests was established at p<0.05.

LEP503 promoter mediated bicistronic lentiviral vector construction

We generated the target plasmid using conventional molecular biologic techniques. First, we cloned the LEP503 promoter from the mRNA of HLECs by RT–PCR, and designed the full length of LEP503 (2,439 bp) containing double enzyme sites of Cla I and BamH I (Figure 1B). The digested product of LEP503 (Figure 1B) was ligated onto the promoter-less vector lenti-(CMV)-HSV-tk-EGFP, which was also digested by Cla I and BamH I (Figure 1A). The recombinant lentiviral-LEP503-HSV-tk-EGFP vector was sequenced and the assumption diagram of the vector was showed in Figure 1B.

We verified whether the LEP503 promoter was subcloned into the lenti-LEP503-HSV-tk-EGFP vector by first amplifying the LEP503 promoter and EGFP with the PCR assay using the lenti-LEP503-HSV-tk-EGFP vector as the template. The full length of LEP503 product, using its previous clonal primers (2,439 bp) and the EGFP product (597 bp) were detected by agarose gel electrophoresis (Figure 1C). Second, the lenti-LEP503-HSV-tk-EGFP vector was further confirmed by double-digestion with Mlu I and Xba I. We found that the fragment of HSV-TK (~900 bp) was digested from the lenti-LEP503-HSV-tk-EGFP vector (Figure 1D). DNA sequencing, PCR results, and analysis of restriction enzyme digests verified the correct orientation of the insert in the constructs; i.e., the human LEP503 promoter was subcloned into the expressional vector of lenti-HSV-tk-EGFP.

LEP503 promoter specifically drives the downstream targeting genes in HLECs

We assessed the relative specificity of the LEP503 promoter by infection of the lenti-LEP503-HSV-tk-EGFP and lenti-CMV-HSV-tk-EGFP vectors into human RPECs, ASFCs, Hela cells, and HLECs. We found that significantly more EGFP-positive cells were observed in HLECs than in RPECs and ASFCs as determined by fluorescence microscopy at 72 h after infection by lenti-LEP503-HSV-tk-EGFP (Figure 2A), whereas more EGFP-positive cells were observed in HLECs, RPECs, and ASFCs after infection by lenti-CMV-HSV-tk-EGFP (Figure 2B). We further confirmed the relative specificity of the LEP503 promoter in HLECs by detection of expression of EGFP protein with flow cytometry and western blotting assays, and the mRNA transcription of HSV-tk with RT–PCR assay (product: ~613 bp). As shown in Figure 2C, the percentages of LEP503-driven EGFP-positive cells in total cells were 62.7% in HLECs, 15.2% in RPECs, and 17.7% in ASFCs at 48 h after infection, whereas those of CMV-driven EGFP-positive cells were 90.2% in HLECS, 85.4% in RPECs, and 86.1% in ASFCs at 48 h after infection (Figure 2C). Western blotting results showed that the time-course of EGFP expression driven by LEP503 promoter was substantially more rapid in HLECs than in RPECs and ASFCs (Figure 2D,E). In addition, we found that the mRNA transcription of HSV-tk driven by LEP503 promoter was ~10 fold higher in HLECs than in Hela cells (p<0.05; Figure 3A,B). These findings indicated that the LEP503 promoter could specifically initiate the transcription of downstream genes including EGFP and HSV-tk in HLECs. However, we found that, like most other tissue specific promoters used in gene therapy, the LEP503 promoter mediated lower expression than some widely expressed promoters such as the CMV promoter (Figure 2), consistent with our previous report [11]. These results suggested that the LEP503 promoter can specifically express the downstream operating genes in HLECs, but it has a lower promoter capability than the CMV promoter.

Lenti-LEP503-HSV-tk-EGFP/GCV system inhibits HLECs proliferation

Next, we evaluated the effect of lenti-LEP503-HSV-tk-EGFP/GCV system on HLECsby infecting HLECs with the vector at a MOI=10 and lenti-CMV-HSV-tk-EGFP at MOI=5. After these cells were infected for 24 h and exposed to GCV (20 μM) for 48 h, we observed fewer EGFP-positive cells and most EGFP-positive cells were shrunken and floating (Figure 3D), whereas the vehicle control contained more EGFP-positive cells (Figure 3C). We assessed the cytotoxicity of GCV with the MTT assay after these cells were infected for 24 h and exposed to GCV (from 0~50 μM) for 72 h. We found an obvious cytotoxicity at doses of GCV 25 μM and higher (Figure 3E), but 20 μM GCV had no cytotoxicity for HLECs . We further measured the proliferation of cells after infection for 24 h, HLECs and Hela cells were exposed to GCV (20 μM) and viability was tested with the CCK-8 assay. The time course of the inhibitory effects of the system on HLECs and Hela cells is shown in Figure 3F. The kill effects were ~10, ~25, and ~32% at 36, 48, and 72 h after GCV treatment, respectively, whereas the system had no inhibitory effects on Hela cells (Figure 3F). These results indicated that the lenti-LEP503-HSV-tk/GCV system specially induced an inhibitory effect in HLECs and that the inhibitory effect was not directly produced by cytotoxicity of GCV itself.

ATRA enhances the proliferative inhibition effect of lenti-LEP503-HSV-tk/GCV system against HLECs

We next determined whether ARTA affected the observed inhibition of HLEC proliferation. We assessed the bystander effect of the system by infecting HLECs with Lenti-LEP503-HSV-tk-EGFP and exposing them to GCV or (and) ATRA. The CCK-8 assays showed ~8, ~32, ~55, ~65, ~68, and ~71% inhibition at 12, 24, 36, 48, and 72 h after GCV and ATRA treatment, respectively, and ~0, ~0, ~10, ~25, and ~32% at 12, 24, 36, 48, and 72 h after GCV and vehicle control treatment, respectively (Figure 4A). We further confirmed the inhibitory effects using the MTT assay, the classical approach for measuring cell proliferation. We found that ATRA could dose-dependently enhance the killing effect of lenti-LEP503-HSV-tk/GCV system (Figure 4B) and that the inhibitory effect at 20 μM GCV and ATRT was similar to that observed with the CCK-8 assay (Figure 4A). These data indicate that the treatment with ATRA exerted a strong bystander effect when compared to the vehicle control treatment (Figure 4A,B).

The GJIC is well known as the main mechanism mediating the bystander effect in HSV-tk gene therapy. ATRA has been shown to upregulate the expression of Cx43 and GJIC [21-23]. Therefore, we evaluated the expression of the Cx43 by western blotting assay with a specific monoclonal antibody. As shown in Figure 4C, the combined lenti-LEP503-HSV-tk-EGFP/GCV system and ATRA treatment increased the time-dependent expression of Cx43 in HLECs. Semi-quantitative analysis showed that the Cx43 expressions were ~7, ~8, ~15, and ~16 fold higher than control values at 12, 24, 36, and 48 h, respectively, after the combined treatment with lenti-LEP503-HSV-tk-EGFP/GCV and ATRA (data not shown). We used immunostaining assay to confirm Cx43 expression in HLECs and found significantly higher expression at 72 h after the combined treatment with the lenti-LEP503-HSV-tk-EGFP/GCV system and ATRA when compared to the control (Figure 4D). These findings indicated that ATRA enhanced the expression level of Cx43.

The bystander effect was further evaluated by using the CJIC inhibitor 18-α-GA, which can effectively down regulate the expression of Cx43. We mixed HELCs infected with the lenti-LEP503-HSV-tk-EGFP vector with normal cultured HELCs at ratios of 0, 20, 40, 60, and 80%, and then exposed them to GCV, ATRT and 18-α-GA. As shown in Figure 4F, the killing effect of the combined treatment of lenti-LEP503-HSV-tk-EGFP/GCV system and ATRA was ~38, ~66, ~81, and ~94% for a ratio of 20, 40, 60, and 80% infected cells, respectively, whereas the killing effect of treatment with the lenti-LEP503-HSV-tk-EGFP/GCV system alone was ~29, ~48, ~70, and ~89%, respectively (Figure 4F). In addition, we found that a higher ratio of infected cells in the co-culture produced a greater bystander effect, and the killing efficiency percentage (~38, ~66, ~81, and ~94%) was significantly higher than the for increasing ratios of mixture infected cells to coculture cells (20, 40, 60, and 80%, respectively; Figure 4F), in support of a bystander effect. Notably, when these cells were incubated with the agent, 18-α-GA, a selective inhibitor of gap junctions, the expression of Cx43 was dose-dependently inhibited by 18-α-GA (Figure 4E). We also found that the bystander effect mediated by Cx43 was potently prevented in the presence of 18-α-GA in culture medium (Figure 4F). These data suggest that the bystander effect of the combined treatment of lenti-LEP503-EGFP-HSV-tk/GCV system and ATRA on HLECs is significantly enhanced compared to treatment with lenti-LEP503-EGFP-HSV-tk/GCV alone, and that the bystander effect is primarily mediated through GJIC.

The bystander effect results predominantly in cell necrosis

We investigated whether the cell killing was due to apoptosis or necrosis by determining the cell death type with a flow cytometry assay. The percentage of HLECs distributed in the Q1 zone (necrosis) was ~16.6% at 24 h and ~35.1% at 72 h after the combined treatment with the lenti-LEP503-HSV-tk-EGFP/GCV system with ATRA (Figure 5B), whereas the percentage in Q1 zone was ~3.1% at 24 h and ~13.2% at 72 h when cells were exposed to the lenti-LEP503-HSV-tk-EGFP/GCV system alone (Figure 5A). However, more cells (~10.5%) were located in Q2 zone (apoptosis) at 72 h after combined treatment with the lenti-LEP503-HSV-tk-EGFP/GCV system with ATRA (Figure 5B) than after treatment with the lenti-LEP503-HSV-tk-EGFP/GCV system alone (0.1%; Figure 5A). The HLECs treated with the combination of the lenti-LEP503-HSV-tk-EGFP/GCV system with ATRA also showed more typical DNA ladders than did cells exposed to the lenti-LEP503-HSV-tk-EGFP/GCV system treatment alone (Figure 5C). These data suggest that the bystander effect induced by combination of the lenti-LEP503-HSV-tk-EGFP/GCV system with ATRA inhibited HLEC proliferation predominantly by inducing cell necrosis. However, apoptosis also contributed to the bystander effect, although apoptotic cells comprised a minor proportion of the cells, consistent with our previous report [11].