JOURNAL OF INTERFERON RESEARCH 11:371-377 (1991) Mary Ann Liebert, Inc., Publishers
The Induction of Interferon-a and Interferon-ß mRNA in Human Natural Interferon-Producing Blood Leukocytes Requires De Novo Protein Synthesis BRITTA
CEDERBLAD,1
ANDERS E.
GOBL,1
and GUNNAR V.
ALM1'2
ABSTRACT The induction of interferon-a (IFN-a) and IFN-ß mRNA in natural IFN producing (NIP) cells in cultures of human peripheral blood mononuclear cells (PBMCs), stimulated by glutaraldehyde-fixed Herpes simplex virus type 1 (HSV)-infected WISH cells, was studied. The protein synthesis inhibitor cycloheximide (CHX) totally prevented the appearance of both IFN-a and IFN-ß mRNA, also in cultures supplemented with a conditioned medium (CM) assumed to contain soluble factors necessary for the IFN induction. However, when PBMCs were preincubated for 4 h in medium supplemented with fetal bovine serum (FBS) with or without addition of CM, the subsequent induction of IFN-a/ß mRNA became partially resistant to CHX. In serum-free medium containing interleukin-3 (IL-3) or granulocyte-macrophage colony-stimulating factor (GM-CSF), the early induction of IFN-a mRNA became resistant to CHX, and, in contrast to FBS and CM supplemented medium, this was observed also without a preincubation of the PBMCs. In contrast, IL-1, IL-2, IL-4, IL-6, tumor necrosis factor-a (TNF-a), IFN-a, or IFN-7 had no such effects. Our results suggests that de novo synthesis of proteins normally is required for the induction of IFN-a/ß mRNA. Such proteins might be cytokines, possibly CSFs, which in turn also may require protein synthesis for their actions. In contrast, the actual triggering signal provided by the HSV-inducer is independent of protein synthesis.
proposed,"'2'4'2'"
INTRODUCTION
and also other cell interactions may be inwell as production of soluble factors.'6' The latter may include cytokines such as interleukin-3 (IL-3), granuloCyte-macrophage colony-stimulating factor (GM-CSF), and iFN_a which all have marked enhancing effects on the HSVinduced IFN-a response.'2" ln the present study we examined if induction of IFN mRNA jn hurnan PBMCs by HSV required de novo protein synthesis, as wou)d be expected if, e.g., production of cytokines was essential for the IFN induction. Since the protein synthesis inhibitor cycloheximide (CHX) prevented the induction of IFN_a and 1FN.p mRNA, we examined whether this block cou,d be cjrCumvented by soluble factors in a conditioned medium (CM) or by defined cytokines.
volved,
Peripheral
blood mononuclear cells (PBMCs) can be induced in vitro to produce interferon-a (IFN-a) by a vari4' Fixed ety of glutaraldehyde-fixed virus-infected cells." WISH virus 1 cells induce type (HSV)-infected Herpes simplex the production of IFN-a by infrequent, but efficient, IFN-aproducing cells (IPCs) among the PBMCs.'5'6' These HSVinduced IPCs contain relatively high levels of IFN-a mRNA, but less mRNA for IFN-ß.17' The HSV-induced IPCs, also referred to as natural IFN-producing (NIP) cells,'8'9' express MHC class II antigens and low levels of CD4, but lack other markers of B-, T-, or NK cells. The NIP cells are not typical monocytes, since they do not express CD14 or CD1 lb/c antigens, and are not phagocytic or adherent to plastic or to nylon wool.'9'2' Probably the same category of IPCs is induced by other virus-infected cells, bacteria, or tumor cells.'813"'9' The mechanism by which the NIP cells are induced to produce IFN is not well understood, but may be complex. A receptor-mediated recognition of virus-infected cells has been
1 Interferon Laboratory, Uppsala University, Agricultural Sciences, Uppsala, Sweden.
and
as
MATERIALS AND METHODS
Preparation of PBMCs: Human PBMCs were prepared by Ficoll-Hypaque (Pharmacia, Uppsala, Sweden) density gradi-
^Department of Veterinary Microbiology, 371
Division of
Immunology,
Swedish
University
of
CEDERBLAD ET AL.
372
centrifugation of the buffy coats from normal blood donors. were washed three times in phosphate-buffered saline (PBS), at room temperature (RT) or at +4°C (PBMCs in Fig. 4), and resuspended in RPMI-1640 medium (Flow Laboratories, Irvine, UK) supplemented with penicillin (60 p,g/ml), streptomycin (100 |i.g/ml), L-glutamine (2 mM), 20 mM HEPES, and either 10% fetal bovine serum (FBS; Cytosystems, ent
The cells
Castle Hill, N.S.W., Australia) or 0.1% (wt/vol) human albumin (HSA; Kabi-Vitrum, Stockholm, Sweden).
serum
IFN-inducer Cells: The IFN-inducer cells were prepared as previously described.'6' Briefly, confluent monolayers of mycoplasma-free human WISH cells (Flow), grown in Dulbecco's modified Eagle medium (DMEM) supplemented as described above and with 10% FBS, were established in 24-well plates or in 175-cm2 tissue culture flasks (Nunclon; Nunc, Roskilde, Denmark). The cell monolayers were fixed with 0.05% glutaraldehyde and incubated overnight at 37°C with HSV, about 1.3 X 108 pfu/ml in 0.2- or 10-ml volumes. Control, mockinfected cells were incubated with medium alone. The monolayers were then fixed again, blocked with 3% glycine and finally rinsed with PBS.
Conditioned Medium (CM): The CM was collected from cultures of T-cell-deprived nonadherent PBMCs, stimulated for 6 h by fixed HSV-infected WISH cells in RPMI-1640 medium supplemented with 10% FBS.'6' The CM was used at a concentration of 25% during preincubations, except for the Northern blot analysis, where 12.5% was used.
Preparation of Probes: The human cDNA for IFN-a2, IFN-ß, and ß-actin have been described before.'7' DNA probes for Northern blot analysis were purified and labeled with [ct-32P]dCTP (3,000 Ci/mmole; Amersham) by the Multiprime labeling reaction (Amersham). The specific activity was normally 1-2 x 109 cpm/p-g DNA. The cRNA probes used for in situ hybridization were transcribed as described before,(7> using [a-35S]UTP (1000 Ci/mmole; Amersham) and T7 RNA polymerase. Specific activities were 1-2 X 109 cpm/|jig RNA. RNA Isolation and Analysis: RNA for Northern blot analysis prepared from PBMCs,stimulated for 5 h by the IFNinducer cells in 175-cm2 tissue culture flasks. The induced PBMCs were chilled on ice, washed once in cold PBS, and total cellular RNA was prepared by the guanidinium thiocyanate/ cesium chloride method.(22) Northern blot analysis was performed as described before.'23' A total of 40 tig RNA was electrophoresed through 1% agarose gels, containing 0.2 M formaldehyde. The RNA was then transferred to a nylon filter (Hybond N+, Amersham) and hybridized for 12-16 h at 43°C in 50% formamide and 5 x SSC to the human cDNA probes described above. The filter was washed in 0.1 x SSC and 0.1% NaDodS04 at 65°C for 1 h, then exposed for 3 days using one intensifying screen (Du Pont Cronex, Dreieich, Germany) and XAR-5 film (Eastman Kodak). The same filter was used for hybridization with the three probes. was
In Situ Hybridization: PBMCs used for in situ hybridization stimulated for 6 h by the IFN-inducer cells in duplicates in 24-well plates. The induced PBMCs were chilled on ice, harvested, centrifuged, and resuspended in 1% paraformaldehyde in PBS. The fixed PBMCs were then put onto glass slides in wells (8 mm diameter) surrounded by a hydrophobic coating. The cells were allowed to settle and excess liquid was removed with a filter paper. The slides were dried and stored at —80°C until hybridization. In situ hybridizations were carried out as described before.'7'24' A 10-ixl volume of 35S-labeled cRNA probe (0.5-1 ng of RNA) was added per well and hybridized to cells for 3 h at 50°C, followed by treatment with RNase A and Tl. Slides were autoradiographed using NTB-2 nuclear track emulsion (Eastman Kodak) for 5 days, and examined in a light microscope at a magnification of 400x. For each treatment, two wells (50-100 x 103 PBMCs) were scanned. A majority of the IPCs were heavily labeled with more than 50 grains per IPC, and cells with more than 5 grains were scored as positive. Sense probes were used as a control of the specificity of the hybridizations'7' and never resulted in labeled cells. were
Cytokines: All cytokines were from Genzyme Corp. (Boston, MA), except for the IFN-y (Amersham International, Amersham, UK) and the recombinant IFN-a2b (Intron-A; Schering Corp., Kenilworth, NJ). The following recombinant human cytokines were tested at the indicated concentrations in the experiments summarized in Fig. 3 (U/ml): IL-lß (50, 5, 0.5), IL-2 (10\ 100. 10), IL-3 (500, 50, 5, 0.5, 0.05), IL-4 and IL-6 (2 x 10\ 200, 20), tumor necrosis factor-ot (TNF-a; 10 x 10\ 10\ 100), GM-CSF (500, 50, 5, 0.5, 0.05), IFN-a, and IFN-y (10 x 10\ 103, 100). The concentrations given in italics were used in all other experiments, and usually coincided with the concentrations recommended by the manufacturers. Preincubation and Induction of IFN in PBMCs: The PBMCs preincubated for 4 h at 37°C in medium supplemented as indicated, using 50-ml polypropylene tubes (Nunc) with loose caps. Equal volumes of preincubated PBMCs, without prior washing, and medium with or without 100 p.g CHX/ml (Boehringer Mannheim GmbH, Mannheim, Germany) were then added to the IFN-inducer cells. The final concentration of CHX, 50 Lig/ml, inhibits the total protein synthesis by more than 98% (results not shown). For the experiments in Fig. 3, also recombinant IFN-a2b was included at a final concentration of 103 U/ml. PBMCs for Northern blot analysis were stimulated by HSV at a concentration of 8 x 106/ml and in a volume of 50 ml. PBMCs used for in situ hybridizations were stimulated at 2 X 106/ml in a volume of 1 ml. The PBMCs were cultured with the HSV-infected WISH cells for 5 or 6 h (as indicated) at 37°C and 7% CO, in air. were
RESULTS De novo protein synthesis is of IFN-a and IFN-ß mRNA
required for the
induction
The importance oíde novo protein synthesis for the induction of IFN-a and IFN-ß mRNA in PBMCs was studied by performing the IFN-induction in the presence of the protein synthesis inhibitor CHX. Northern blot analysis (Fig. 1) showed that the
CYCLOHEXIMIDE SENSITIVE INDUCTION OF
1
2
3
4
IFN-a/ß
5
mRNA
6
373
7
8
9
10 11
12
§ 00 »
-SS?
• ••• FIG. 1. Northern blot analysis of the effects of CHX on the induction of IFN-a and IFN-ß mRNA in PBMCs by HSV-infected WISH cells. PBMCs were either preincubated for 4 h or directly stimulated by the IFN-inducer cells for 5 h in the presence or absence of CHX, and total RNA was extracted from the PBMCs. PBMCs in lanes 1-4 were not preincubated, and either induced in FBS-supplemented medium without (lane 1) or with CHX (lane 2), or induced in conditioned medium (CM; see Materials and Methods) without (lane 3) or with CHX (lane 4). PBMCs in lanes 5-8 were preincubated and induced either in medium with FBS without (lane 5) or with CHX (lane 6), or in CM without (lane 7) or with CHX (lane 8). Controls were PBMCs neither preincubated nor induced (lane 9), PBMCs induced by mock-infected WISH cells without (lane 10) or with CHX (lane 11 ), or PBMCs induced by mock-infected WISH cells in the presence of CM (lane 12).
normal induction of IFN-a and IFN-ß mRNA by HSV (lane 1 ) was abolished by CHX (lane 2). To examine whether preformed soluble factors could eliminate the requirement for de novo protein synthesis, the IFN-induction cultures were supplemented with a CM (see Materials and Methods). As seen (lane 3), CM increased IFN-a and IFN-ß mRNA levels. However, addition of CHX to such cultures still completely inhibited the IFN-a/ß mRNA induction (lane 4). In contrast, the IFN-a/ß mRNA induction was not inhibited by CHX when the PBMCs had been incubated in the CM for 4 h prior to stimulation by the IFN-inducer cells (cf. lanes 7 and 8). Also the preincubation of PBMCs in FBS-supplemented medium alone induced some resistance to CHX (cf. lanes 5 and 6). No mRNA for IFN-a or IFN-ß was detected in cultures of PBMCs cocultured with mock-infected WISH cells in the absence or presence of CHX or CM (lanes 9-12; see legend to Fig. 1 ).
Development of resistance individual IPCs
to
CHX
at
the level
of
The time course of the development of resistance to CHX studied by preincubating PBMCs for up to 4 h in CM or in medium alone. The PBMCs were then stimulated for 5 h by the HSV-infected WISH cells in the absence or presence of CHX, and IFN-a mRNA containing IPCs was identified by in situ
few IPCs developed in the presence of CHX. In contrast, PBMCs preincubated for 4 h in either medium alone or with CM displayed significant and similar degrees of resistance to CHX, and more than 60% of the IPCs could be induced. However, PBMCs preincubated in CM were more heavily labeled by the in situ hybridization than cells preincubated in medium alone (results not shown). Induction of PBMCs (without CHX) in CM caused an approximately twofold increase in IPC fre-
Ü
2 CD a.
O
was
hybridization.
As shown in Fig. 2, the development of IPCs became gradually more resistant to CHX as the preincubation time increased. Preincubations exceeding 2 h often by itself reduced the numbers of IPCs, especially in PBMCs preincubated in medium
alone. Table 1 summarizes the frequencies of IPCs 5 h after induction. The results demonstrate that without a preincubation very
Time of
preincubation (h)
FIG. 2. Time course of the development of CHX resistance of the IFN-a-mRNA response in NIP-cells. The PBMCs were preincubated for varying periods of time either in medium with FBS (o, •) or with conditioned medium (CM, see Materials and Methods; A A ). The PBMCs were then stimulated by HSVinfected WISH cells for 5 h either in the absence (o, A ) or presence (•, A ) of CHX. Frequencies of IFN-a mRNA containing IPCs were then determined by in situ hybridization. Results from one of two experiments are shown. ,
374
CEDERBLAD ET AL. Table 1. Effects of Preincubation of PBMCs on the Dependence of De Novo Protein Synthesis for Induction of IFN-a mRNA by HSV-Infected WISH Cells3
of CHX No
Yes
Preincubation
No preincubation
Presence
Medium
CM
Medium
19.7 ± 13.7b 0.1 ±0.2
39.0 ± 18.2 0.2 ± 0.3
15.8 ± 8.4 9.8 + 8.6
CM 41.5 27.2
14.8 12.9
"The PBMCs were either preincubated for 4 h in FBS-supplemented medium only or in conditioned medium (CM; see Materials and Methods), or directly stimulated for 5 h by HSV-infected WISH cells in the presence or absence of CHX. b Number of IFN-a mRNA containing cells (IPCs) per 1 X 104 PBMCs, as determined by in situ hybridization. Figures are means ± SD, n—1.
quencies compared to in medium alone, regardless of whether the PBMCs were preincubated or not. No IPCs were detected in PBMCs either preincubated or not in medium with or without CM and then stimulated by mock-infected WISH cells (results not
shown).
Effects of cytokines on mRNA induction
the CHX
sensitivity of IFN-a
To determine whether the resistance of the IFN-a mRNA induction to CHX could be mediated by cytokines, PBMCs were preincubated for 4 h in IL-1, IL-2, IL-3, IL-4, IL-6, TNF-a, GM-CSF, IFN-a, or IFN-y. Serum-free medium was used, supplemented with 0.1% HSA only, to avoid the CHX resistance caused by FBS-supplemented medium. The PBMCs were then stimulated for 6 h by HSV-infected WISH cells in the presence or not of CHX, and frequencies of IPCs were determined by in situ hybridization for IFN-a mRNA. Figure 3 shows the IPC frequencies in the presence of CHX, expressed as percent of the IPC frequencies obtained in control cultures, i.e., PBMCs preincubated in medium alone and induced in the absence of CHX. Addition of CHX to such control cultures reduced the IPC frequencies by 90%. A similar CHXmediated decrease in the frequencies of IPCs was also observed when the preincubations were carried out in the presence of IL-1, IL-2, IL-4, IL-6, TNF-a, IFN-a, and IFN-y. In contrast, preincubation in the two colony-stimulating factors (CSFs) IL-3 or GM-CSF mediated a considerable resistance to CHX. In these cultures the IPC frequencies were approximately 70 and 80% of the control, respectively. No IPCs were detected in PBMCs preincubated in IL-3 or GM-CSF and then stimulated by mock-infected WISH cells (not shown).
the kinetics of IFN-a mRNA induction in PBMCs exposed to IL-3 or GM-CSF
Effects of CHX on
We examined whether the partial CHX resistance of the IFN-a mRNA response caused by IL-3 and GM-CSF required de novo protein synthesis. The PBMCs were either preincubated for 4 h in serum-free medium with or without IL-3 or GM-CSF, or directly stimulated by HSV-infected WISH cells in the absence or presence of CHX. The PBMCs were harvested after varying periods of time, in case CHX altered the kinetics of the IFN-a mRNA response. The frequencies of IPCs were determined by in situ hybridization.
As seen in Fig. 4 (A, B), IL-3 or GM-CSF caused a generally marked increase of the IPC frequencies, and an earlier appearance of IPCs, compared to induction in medium alone. The two CSFs generated some CHX resistance when present during the IFN-a mRNA induction, i.e., without a preincubation step (Fig. 4A). Preincubation of the PBMCs with the CSFs caused a further moderate increase of their resistance to CHX (Fig. 4B). At least with CSF-preincubated PBMCs, the frequencies of IPCs increased normally during the first 3 h of induction in the presence of CHX, but the further rise of the IPC numbers was blocked.
DISCUSSION The results clearly demonstrate that induction of IFN-a/ß mRNA by HSV in NIP cells was inhibited by CHX and thus
125% >.
o
c
CD
3
a a>
O
a. a> S
' «-• « cu
ce
Cytokine FIG. 3. Effect of a 4-h preincubation of PBMCs in different cytokines on the subsequent development of IFN-a mRNA containing IPCs after stimulation by HSV-infected WISH cells
in the presence of CHX. Different concentrations of the cytokines were used with similar results, and the data shown are from experiments using the cytokine concentrations shown in italics in Materials and Methods. "Medium" denotes preincubation in serum-free medium with HSA-supplement. The bars indicate the CHX resistance, i.e., the IPC frequency in the presence of CHX expressed as percent of that of PBMCs preincubated in serum-free medium and then induced in the absence of CHX. IFN-a mRNA positive IPCs were identified by in situ hybridization. Means ±SD are based on three experiments.
375
CYCLOHEXIMIDE SENSITIVE INDUCTION OF IFN-a/ß mRNA
suggests that the latter do not contain significant quantities of
50 -,
IL-3/GM-CSF or other cytokines with similar direct action on the IFN-a/ß response, but may well initiate their synthesis in the PBMC cultures. In the presence of CHX, the number of IFN-a mRNA containing IPCs increased in a relatively normal fashion during the first 3 h after HSV-stimulation in CSF-supplemented medium. Consequently, the actual triggering of the IFN-a mRNA response by HSV in NIP cells can proceed in the absence of de novo protein synthesis, at least when an accessory signal is provided by the CSFs. Even in CSF-preincubated PBMCs, the later phase of the IFN-a mRNA response (after 3 h) was clearly decreased by CHX. This could be due to a general toxic effect of the CHX, or to a more specific requirement of new proteins to sustain the late IFN-a response. None of the other tested cytokines (IL-1, -2, -4, -6, TNF-a, IFN-a, -7) caused any significant resistance to CHX, not even the IFNs that have a significant stimulatory (priming) effect on the IFN-a/ß response of both NIP cells'21'26'27' and other cells.128 32) Such priming has been reported to require new synthesis,'33' probably of positive transcription facprotein 4
The Induction of Interferon-a and Interferon-ß mRNA in Human Natural Interferon-Producing Blood Leukocytes Requires De Novo Protein Synthesis BRITTA
CEDERBLAD,1
ANDERS E.
GOBL,1
and GUNNAR V.
ALM1'2
ABSTRACT The induction of interferon-a (IFN-a) and IFN-ß mRNA in natural IFN producing (NIP) cells in cultures of human peripheral blood mononuclear cells (PBMCs), stimulated by glutaraldehyde-fixed Herpes simplex virus type 1 (HSV)-infected WISH cells, was studied. The protein synthesis inhibitor cycloheximide (CHX) totally prevented the appearance of both IFN-a and IFN-ß mRNA, also in cultures supplemented with a conditioned medium (CM) assumed to contain soluble factors necessary for the IFN induction. However, when PBMCs were preincubated for 4 h in medium supplemented with fetal bovine serum (FBS) with or without addition of CM, the subsequent induction of IFN-a/ß mRNA became partially resistant to CHX. In serum-free medium containing interleukin-3 (IL-3) or granulocyte-macrophage colony-stimulating factor (GM-CSF), the early induction of IFN-a mRNA became resistant to CHX, and, in contrast to FBS and CM supplemented medium, this was observed also without a preincubation of the PBMCs. In contrast, IL-1, IL-2, IL-4, IL-6, tumor necrosis factor-a (TNF-a), IFN-a, or IFN-7 had no such effects. Our results suggests that de novo synthesis of proteins normally is required for the induction of IFN-a/ß mRNA. Such proteins might be cytokines, possibly CSFs, which in turn also may require protein synthesis for their actions. In contrast, the actual triggering signal provided by the HSV-inducer is independent of protein synthesis.
proposed,"'2'4'2'"
INTRODUCTION
and also other cell interactions may be inwell as production of soluble factors.'6' The latter may include cytokines such as interleukin-3 (IL-3), granuloCyte-macrophage colony-stimulating factor (GM-CSF), and iFN_a which all have marked enhancing effects on the HSVinduced IFN-a response.'2" ln the present study we examined if induction of IFN mRNA jn hurnan PBMCs by HSV required de novo protein synthesis, as wou)d be expected if, e.g., production of cytokines was essential for the IFN induction. Since the protein synthesis inhibitor cycloheximide (CHX) prevented the induction of IFN_a and 1FN.p mRNA, we examined whether this block cou,d be cjrCumvented by soluble factors in a conditioned medium (CM) or by defined cytokines.
volved,
Peripheral
blood mononuclear cells (PBMCs) can be induced in vitro to produce interferon-a (IFN-a) by a vari4' Fixed ety of glutaraldehyde-fixed virus-infected cells." WISH virus 1 cells induce type (HSV)-infected Herpes simplex the production of IFN-a by infrequent, but efficient, IFN-aproducing cells (IPCs) among the PBMCs.'5'6' These HSVinduced IPCs contain relatively high levels of IFN-a mRNA, but less mRNA for IFN-ß.17' The HSV-induced IPCs, also referred to as natural IFN-producing (NIP) cells,'8'9' express MHC class II antigens and low levels of CD4, but lack other markers of B-, T-, or NK cells. The NIP cells are not typical monocytes, since they do not express CD14 or CD1 lb/c antigens, and are not phagocytic or adherent to plastic or to nylon wool.'9'2' Probably the same category of IPCs is induced by other virus-infected cells, bacteria, or tumor cells.'813"'9' The mechanism by which the NIP cells are induced to produce IFN is not well understood, but may be complex. A receptor-mediated recognition of virus-infected cells has been
1 Interferon Laboratory, Uppsala University, Agricultural Sciences, Uppsala, Sweden.
and
as
MATERIALS AND METHODS
Preparation of PBMCs: Human PBMCs were prepared by Ficoll-Hypaque (Pharmacia, Uppsala, Sweden) density gradi-
^Department of Veterinary Microbiology, 371
Division of
Immunology,
Swedish
University
of
CEDERBLAD ET AL.
372
centrifugation of the buffy coats from normal blood donors. were washed three times in phosphate-buffered saline (PBS), at room temperature (RT) or at +4°C (PBMCs in Fig. 4), and resuspended in RPMI-1640 medium (Flow Laboratories, Irvine, UK) supplemented with penicillin (60 p,g/ml), streptomycin (100 |i.g/ml), L-glutamine (2 mM), 20 mM HEPES, and either 10% fetal bovine serum (FBS; Cytosystems, ent
The cells
Castle Hill, N.S.W., Australia) or 0.1% (wt/vol) human albumin (HSA; Kabi-Vitrum, Stockholm, Sweden).
serum
IFN-inducer Cells: The IFN-inducer cells were prepared as previously described.'6' Briefly, confluent monolayers of mycoplasma-free human WISH cells (Flow), grown in Dulbecco's modified Eagle medium (DMEM) supplemented as described above and with 10% FBS, were established in 24-well plates or in 175-cm2 tissue culture flasks (Nunclon; Nunc, Roskilde, Denmark). The cell monolayers were fixed with 0.05% glutaraldehyde and incubated overnight at 37°C with HSV, about 1.3 X 108 pfu/ml in 0.2- or 10-ml volumes. Control, mockinfected cells were incubated with medium alone. The monolayers were then fixed again, blocked with 3% glycine and finally rinsed with PBS.
Conditioned Medium (CM): The CM was collected from cultures of T-cell-deprived nonadherent PBMCs, stimulated for 6 h by fixed HSV-infected WISH cells in RPMI-1640 medium supplemented with 10% FBS.'6' The CM was used at a concentration of 25% during preincubations, except for the Northern blot analysis, where 12.5% was used.
Preparation of Probes: The human cDNA for IFN-a2, IFN-ß, and ß-actin have been described before.'7' DNA probes for Northern blot analysis were purified and labeled with [ct-32P]dCTP (3,000 Ci/mmole; Amersham) by the Multiprime labeling reaction (Amersham). The specific activity was normally 1-2 x 109 cpm/p-g DNA. The cRNA probes used for in situ hybridization were transcribed as described before,(7> using [a-35S]UTP (1000 Ci/mmole; Amersham) and T7 RNA polymerase. Specific activities were 1-2 X 109 cpm/|jig RNA. RNA Isolation and Analysis: RNA for Northern blot analysis prepared from PBMCs,stimulated for 5 h by the IFNinducer cells in 175-cm2 tissue culture flasks. The induced PBMCs were chilled on ice, washed once in cold PBS, and total cellular RNA was prepared by the guanidinium thiocyanate/ cesium chloride method.(22) Northern blot analysis was performed as described before.'23' A total of 40 tig RNA was electrophoresed through 1% agarose gels, containing 0.2 M formaldehyde. The RNA was then transferred to a nylon filter (Hybond N+, Amersham) and hybridized for 12-16 h at 43°C in 50% formamide and 5 x SSC to the human cDNA probes described above. The filter was washed in 0.1 x SSC and 0.1% NaDodS04 at 65°C for 1 h, then exposed for 3 days using one intensifying screen (Du Pont Cronex, Dreieich, Germany) and XAR-5 film (Eastman Kodak). The same filter was used for hybridization with the three probes. was
In Situ Hybridization: PBMCs used for in situ hybridization stimulated for 6 h by the IFN-inducer cells in duplicates in 24-well plates. The induced PBMCs were chilled on ice, harvested, centrifuged, and resuspended in 1% paraformaldehyde in PBS. The fixed PBMCs were then put onto glass slides in wells (8 mm diameter) surrounded by a hydrophobic coating. The cells were allowed to settle and excess liquid was removed with a filter paper. The slides were dried and stored at —80°C until hybridization. In situ hybridizations were carried out as described before.'7'24' A 10-ixl volume of 35S-labeled cRNA probe (0.5-1 ng of RNA) was added per well and hybridized to cells for 3 h at 50°C, followed by treatment with RNase A and Tl. Slides were autoradiographed using NTB-2 nuclear track emulsion (Eastman Kodak) for 5 days, and examined in a light microscope at a magnification of 400x. For each treatment, two wells (50-100 x 103 PBMCs) were scanned. A majority of the IPCs were heavily labeled with more than 50 grains per IPC, and cells with more than 5 grains were scored as positive. Sense probes were used as a control of the specificity of the hybridizations'7' and never resulted in labeled cells. were
Cytokines: All cytokines were from Genzyme Corp. (Boston, MA), except for the IFN-y (Amersham International, Amersham, UK) and the recombinant IFN-a2b (Intron-A; Schering Corp., Kenilworth, NJ). The following recombinant human cytokines were tested at the indicated concentrations in the experiments summarized in Fig. 3 (U/ml): IL-lß (50, 5, 0.5), IL-2 (10\ 100. 10), IL-3 (500, 50, 5, 0.5, 0.05), IL-4 and IL-6 (2 x 10\ 200, 20), tumor necrosis factor-ot (TNF-a; 10 x 10\ 10\ 100), GM-CSF (500, 50, 5, 0.5, 0.05), IFN-a, and IFN-y (10 x 10\ 103, 100). The concentrations given in italics were used in all other experiments, and usually coincided with the concentrations recommended by the manufacturers. Preincubation and Induction of IFN in PBMCs: The PBMCs preincubated for 4 h at 37°C in medium supplemented as indicated, using 50-ml polypropylene tubes (Nunc) with loose caps. Equal volumes of preincubated PBMCs, without prior washing, and medium with or without 100 p.g CHX/ml (Boehringer Mannheim GmbH, Mannheim, Germany) were then added to the IFN-inducer cells. The final concentration of CHX, 50 Lig/ml, inhibits the total protein synthesis by more than 98% (results not shown). For the experiments in Fig. 3, also recombinant IFN-a2b was included at a final concentration of 103 U/ml. PBMCs for Northern blot analysis were stimulated by HSV at a concentration of 8 x 106/ml and in a volume of 50 ml. PBMCs used for in situ hybridizations were stimulated at 2 X 106/ml in a volume of 1 ml. The PBMCs were cultured with the HSV-infected WISH cells for 5 or 6 h (as indicated) at 37°C and 7% CO, in air. were
RESULTS De novo protein synthesis is of IFN-a and IFN-ß mRNA
required for the
induction
The importance oíde novo protein synthesis for the induction of IFN-a and IFN-ß mRNA in PBMCs was studied by performing the IFN-induction in the presence of the protein synthesis inhibitor CHX. Northern blot analysis (Fig. 1) showed that the
CYCLOHEXIMIDE SENSITIVE INDUCTION OF
1
2
3
4
IFN-a/ß
5
mRNA
6
373
7
8
9
10 11
12
§ 00 »
-SS?
• ••• FIG. 1. Northern blot analysis of the effects of CHX on the induction of IFN-a and IFN-ß mRNA in PBMCs by HSV-infected WISH cells. PBMCs were either preincubated for 4 h or directly stimulated by the IFN-inducer cells for 5 h in the presence or absence of CHX, and total RNA was extracted from the PBMCs. PBMCs in lanes 1-4 were not preincubated, and either induced in FBS-supplemented medium without (lane 1) or with CHX (lane 2), or induced in conditioned medium (CM; see Materials and Methods) without (lane 3) or with CHX (lane 4). PBMCs in lanes 5-8 were preincubated and induced either in medium with FBS without (lane 5) or with CHX (lane 6), or in CM without (lane 7) or with CHX (lane 8). Controls were PBMCs neither preincubated nor induced (lane 9), PBMCs induced by mock-infected WISH cells without (lane 10) or with CHX (lane 11 ), or PBMCs induced by mock-infected WISH cells in the presence of CM (lane 12).
normal induction of IFN-a and IFN-ß mRNA by HSV (lane 1 ) was abolished by CHX (lane 2). To examine whether preformed soluble factors could eliminate the requirement for de novo protein synthesis, the IFN-induction cultures were supplemented with a CM (see Materials and Methods). As seen (lane 3), CM increased IFN-a and IFN-ß mRNA levels. However, addition of CHX to such cultures still completely inhibited the IFN-a/ß mRNA induction (lane 4). In contrast, the IFN-a/ß mRNA induction was not inhibited by CHX when the PBMCs had been incubated in the CM for 4 h prior to stimulation by the IFN-inducer cells (cf. lanes 7 and 8). Also the preincubation of PBMCs in FBS-supplemented medium alone induced some resistance to CHX (cf. lanes 5 and 6). No mRNA for IFN-a or IFN-ß was detected in cultures of PBMCs cocultured with mock-infected WISH cells in the absence or presence of CHX or CM (lanes 9-12; see legend to Fig. 1 ).
Development of resistance individual IPCs
to
CHX
at
the level
of
The time course of the development of resistance to CHX studied by preincubating PBMCs for up to 4 h in CM or in medium alone. The PBMCs were then stimulated for 5 h by the HSV-infected WISH cells in the absence or presence of CHX, and IFN-a mRNA containing IPCs was identified by in situ
few IPCs developed in the presence of CHX. In contrast, PBMCs preincubated for 4 h in either medium alone or with CM displayed significant and similar degrees of resistance to CHX, and more than 60% of the IPCs could be induced. However, PBMCs preincubated in CM were more heavily labeled by the in situ hybridization than cells preincubated in medium alone (results not shown). Induction of PBMCs (without CHX) in CM caused an approximately twofold increase in IPC fre-
Ü
2 CD a.
O
was
hybridization.
As shown in Fig. 2, the development of IPCs became gradually more resistant to CHX as the preincubation time increased. Preincubations exceeding 2 h often by itself reduced the numbers of IPCs, especially in PBMCs preincubated in medium
alone. Table 1 summarizes the frequencies of IPCs 5 h after induction. The results demonstrate that without a preincubation very
Time of
preincubation (h)
FIG. 2. Time course of the development of CHX resistance of the IFN-a-mRNA response in NIP-cells. The PBMCs were preincubated for varying periods of time either in medium with FBS (o, •) or with conditioned medium (CM, see Materials and Methods; A A ). The PBMCs were then stimulated by HSVinfected WISH cells for 5 h either in the absence (o, A ) or presence (•, A ) of CHX. Frequencies of IFN-a mRNA containing IPCs were then determined by in situ hybridization. Results from one of two experiments are shown. ,
374
CEDERBLAD ET AL. Table 1. Effects of Preincubation of PBMCs on the Dependence of De Novo Protein Synthesis for Induction of IFN-a mRNA by HSV-Infected WISH Cells3
of CHX No
Yes
Preincubation
No preincubation
Presence
Medium
CM
Medium
19.7 ± 13.7b 0.1 ±0.2
39.0 ± 18.2 0.2 ± 0.3
15.8 ± 8.4 9.8 + 8.6
CM 41.5 27.2
14.8 12.9
"The PBMCs were either preincubated for 4 h in FBS-supplemented medium only or in conditioned medium (CM; see Materials and Methods), or directly stimulated for 5 h by HSV-infected WISH cells in the presence or absence of CHX. b Number of IFN-a mRNA containing cells (IPCs) per 1 X 104 PBMCs, as determined by in situ hybridization. Figures are means ± SD, n—1.
quencies compared to in medium alone, regardless of whether the PBMCs were preincubated or not. No IPCs were detected in PBMCs either preincubated or not in medium with or without CM and then stimulated by mock-infected WISH cells (results not
shown).
Effects of cytokines on mRNA induction
the CHX
sensitivity of IFN-a
To determine whether the resistance of the IFN-a mRNA induction to CHX could be mediated by cytokines, PBMCs were preincubated for 4 h in IL-1, IL-2, IL-3, IL-4, IL-6, TNF-a, GM-CSF, IFN-a, or IFN-y. Serum-free medium was used, supplemented with 0.1% HSA only, to avoid the CHX resistance caused by FBS-supplemented medium. The PBMCs were then stimulated for 6 h by HSV-infected WISH cells in the presence or not of CHX, and frequencies of IPCs were determined by in situ hybridization for IFN-a mRNA. Figure 3 shows the IPC frequencies in the presence of CHX, expressed as percent of the IPC frequencies obtained in control cultures, i.e., PBMCs preincubated in medium alone and induced in the absence of CHX. Addition of CHX to such control cultures reduced the IPC frequencies by 90%. A similar CHXmediated decrease in the frequencies of IPCs was also observed when the preincubations were carried out in the presence of IL-1, IL-2, IL-4, IL-6, TNF-a, IFN-a, and IFN-y. In contrast, preincubation in the two colony-stimulating factors (CSFs) IL-3 or GM-CSF mediated a considerable resistance to CHX. In these cultures the IPC frequencies were approximately 70 and 80% of the control, respectively. No IPCs were detected in PBMCs preincubated in IL-3 or GM-CSF and then stimulated by mock-infected WISH cells (not shown).
the kinetics of IFN-a mRNA induction in PBMCs exposed to IL-3 or GM-CSF
Effects of CHX on
We examined whether the partial CHX resistance of the IFN-a mRNA response caused by IL-3 and GM-CSF required de novo protein synthesis. The PBMCs were either preincubated for 4 h in serum-free medium with or without IL-3 or GM-CSF, or directly stimulated by HSV-infected WISH cells in the absence or presence of CHX. The PBMCs were harvested after varying periods of time, in case CHX altered the kinetics of the IFN-a mRNA response. The frequencies of IPCs were determined by in situ hybridization.
As seen in Fig. 4 (A, B), IL-3 or GM-CSF caused a generally marked increase of the IPC frequencies, and an earlier appearance of IPCs, compared to induction in medium alone. The two CSFs generated some CHX resistance when present during the IFN-a mRNA induction, i.e., without a preincubation step (Fig. 4A). Preincubation of the PBMCs with the CSFs caused a further moderate increase of their resistance to CHX (Fig. 4B). At least with CSF-preincubated PBMCs, the frequencies of IPCs increased normally during the first 3 h of induction in the presence of CHX, but the further rise of the IPC numbers was blocked.
DISCUSSION The results clearly demonstrate that induction of IFN-a/ß mRNA by HSV in NIP cells was inhibited by CHX and thus
125% >.
o
c
CD
3
a a>
O
a. a> S
' «-• « cu
ce
Cytokine FIG. 3. Effect of a 4-h preincubation of PBMCs in different cytokines on the subsequent development of IFN-a mRNA containing IPCs after stimulation by HSV-infected WISH cells
in the presence of CHX. Different concentrations of the cytokines were used with similar results, and the data shown are from experiments using the cytokine concentrations shown in italics in Materials and Methods. "Medium" denotes preincubation in serum-free medium with HSA-supplement. The bars indicate the CHX resistance, i.e., the IPC frequency in the presence of CHX expressed as percent of that of PBMCs preincubated in serum-free medium and then induced in the absence of CHX. IFN-a mRNA positive IPCs were identified by in situ hybridization. Means ±SD are based on three experiments.
375
CYCLOHEXIMIDE SENSITIVE INDUCTION OF IFN-a/ß mRNA
suggests that the latter do not contain significant quantities of
50 -,
IL-3/GM-CSF or other cytokines with similar direct action on the IFN-a/ß response, but may well initiate their synthesis in the PBMC cultures. In the presence of CHX, the number of IFN-a mRNA containing IPCs increased in a relatively normal fashion during the first 3 h after HSV-stimulation in CSF-supplemented medium. Consequently, the actual triggering of the IFN-a mRNA response by HSV in NIP cells can proceed in the absence of de novo protein synthesis, at least when an accessory signal is provided by the CSFs. Even in CSF-preincubated PBMCs, the later phase of the IFN-a mRNA response (after 3 h) was clearly decreased by CHX. This could be due to a general toxic effect of the CHX, or to a more specific requirement of new proteins to sustain the late IFN-a response. None of the other tested cytokines (IL-1, -2, -4, -6, TNF-a, IFN-a, -7) caused any significant resistance to CHX, not even the IFNs that have a significant stimulatory (priming) effect on the IFN-a/ß response of both NIP cells'21'26'27' and other cells.128 32) Such priming has been reported to require new synthesis,'33' probably of positive transcription facprotein 4
