INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Pertussis toxin (PT), the major virulence determinant of Bordetella pertussis (35), is composed of two distinct functional units, the A protomer, consisting of a single polypeptide (S1) which mediates adenosine diphosphate ribosylation of host G proteins, and the B oligomer, which mediates the binding of the toxin to host cells and the translocation of toxic S1 to its target (8, 28). The B oligomer is a complex pentamer composed of subunits S2, S3, S4, and S5 in a respective molar ratio of 1:1:2:1, with S2 and S3 occuring as heterodimers each with S4, i.e., dimer S2-S4 and dimer S3-S4, held together by S5 (28, 32). The effects of the toxin on cells of the immune system are multiple and include induction of lymphocytosis, inhibition of macrophage migration, adjuvant activity, and T-cell mitogenicity (18). A number of the biological activities of PT, such as lymphocytosis and adjuvant activity, implicate the enzymatic activity of PT in its toxicity and can be abrogated by inactivation of the S1 subunit (1, 5, 13). In contrast, PT-associated T-cell mitogenicity is mediated by the B oligomer (9, 31, 36) and appears to be independent of the enzymatic activity of the toxin, as inactivation of the S1 subunit by mutation has no effect on the mitogenic activity of PT (13, 36), while alterations in the B oligomer can totally abrogate the mitogenic activity of PT (15, 16, 20-22). In addition, the B oligomer devoid of S1 induces T-cell proliferation to the same extent as PT (22, 29, 31, 32, 36). However, although the mitogenic effect of the B oligomer is well known, the roles of its individual components in the mitogenic function have not been extensively studied. Both dimers have been implicated in the binding of PT to cells via interaction of the B oligomer with glycoproteins and glycolipids on many types of eukaryotic cells (10, 26, 37), seemingly via carbohydrate-recognizing domains on subunits S2 and S3 (11, 26, 34, 37). However, there is evidence of a difference between the binding specificities of the two dimers (26, 37), which may account for observations leading to the suggestion that the B oligomer must bind to the cell surface to permit translocation of the A protomer into the cell in a manner different from its binding leading to T-cell mitogenicity, since these two types of binding displayed different susceptibilities to chemical modification of the molecule (22; see Discussion). Experiments with hybrid PTs composed of various combinations of chemically modified dimers indicate a differential role of the two dimers in T-cell stimulation and suggest that the S3-S4 dimer is more relevant to the binding of the B oligomer which results in T-cell stimulation than to the binding which results in translocation of the S1 subunit (20-22). A mitogenic effect for the S3-S4 dimer as an isolated dimeric molecule was not demonstrated.

We have been investigating PT as an immunomodulatory agent for experimental autoimmune encephalomyelitis (EAE), the well-accepted animal model for multiple sclerosis, and found that the protective effect imparted by PT against EAE could be fully attributed to the B oligomer (3). To understand the mechanism by which the B oligomer immunomodulates EAE, the role of its components in the biological activity of the B oligomer should be investigated.

In the B oligomer, S2 and S3 occur naturally as parts of heterodimers, each with S4. These dimers can only be dissociated into monomers under strong denaturing conditions, suggesting that the dimeric conformation is functionally important. Isolation of functional dimers is necessary if we are to investigate their role in the biological activities of the B oligomer. We have now purified the S3-S4 dimer to homogeneity under conditions which favor preservation of the native conformation and investigated its biological activity in vitro. We show that the S3-S4 dimer is as strong a mitogen for T cells as is the B oligomer, in contrast to previous reports suggesting that a combination of S2-S4 and S3-S4 dimers is absolutely necessary to induce T-cell proliferation and that neither dimer is mitogenic by itself (31, 36). The heterodimeric molecule appears to be essential for the mitogenic activity, as recombinant preparations of individual subunits could not significantly induce T-cell proliferation. Most striking was the in vitro modulatory effect of the purified S3-S4 dimer, as well as that of the B oligomer, in reversing the CD4⁺/CD8⁺ T-cell ratio upon stimulation of naive lymph node (LN) cells.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

PT and recombinant subunits. PT was obtained commercially (catalog no. P 9452; Sigma, St. Louis, Mo.). Mutant PT (PTmu), a genetically derived mutant of PT, PTX-9K/129G, was prepared as described previously (24). This mutant contains two amino acid substitutions, Arg⁹ with Lys and Glu¹²⁹ with Gly, in the S1 subunit which abolish its enzymatic activity (24); the B oligomer is unaffected, and PTmu retains its mitogenic activity (13; this study). The DNAs corresponding to PT subunits S2, S3, and S4 cloned into expression vector pEx31 or pEx34 (19) were subcloned into expression vector pRSET (Invitrogen Corporation, San Diego, Calif.) after amplification by PCR using specific primers in which NheI and BamHI restriction sites or NheI and BglII restriction sites had been included to enable ligation at the 5' and 3' ends of S2 and S4 or S3, respectively. The primers used were as follows: 5' S2 primer, CATGGTATGGCTAGCACGCCAGGCATCGTCATTCCG; 3' S2 primer, GCAGCCGGATCCTCAGCATAAGGATGATCCAGGATT; 5' S4 primer, CATGGTATGGCTAGCGACGTTCCTTATGTGCTG; 3' S4 primer, GAGCTCGGATCCTCAGGGGCACTGCTTGCCGCT; 5' S3 primer, CATGGTATGGCTAGCGTTGCGCCAGGCATCGTCATC; 3' S3 primer, TGAGCTCAGATCTCAGCATATCGACGCTGCCGGGTT. The nucleotide sequence of the PCR product within each construct was confirmed by direct sequencing using primers derived from pRSET (forward primer, 5'-ATGCGGGGTTCTCATCAT-3'; reverse primer, 5'-TAGCAGCCGGATCAAGCT-3') and a 373A DNA sequencer (Applied Biosystems, Foster City, Calif.). The DNA sequences obtained confirmed for each construct an open reading frame for the relevant subunit, preceded by (Met)-Arg-Gly-Ser-(His)₆-Gly-Met-Ala-Ser. Expression of the recombinant subunits was induced in the host Escherichia coli BL21(DE3) (catalog no. C6000-03; Invitrogen Corporation), and the recombinant protein was purified to homogeneity by metal chelate affinity chromatography on Ni²⁺ nitrilotriacetic acid-agarose (catalog no. 30230; Qiagen, Chatsworth, Calif.) according to the manufacturer's protocol.

Preparation of antisera against PT and recombinant subunits. Antisera against PT and the recombinant S2 subunit (rS2) were prepared in SJL/J mice by subcutaneous injection of PT (500 ng of PT in incomplete Freund adjuvant) or purified rS2 (50 µg of rS2 in complete Freund adjuvant), followed by two booster injections of 25 µg of rS2 administered subcutaneously at weekly intervals in the flank. The anti-PT serum which reacts predominantly with the S1 subunit (see Results) is henceforth referred to as anti-S1 serum. Rabbit antisera were prepared against rS3 and rS4 by standard procedures.

Separation of PT components by HPLC. PTmu was dissociated by incubation in 5 M cold urea by the procedure of Tamura et al. (32), which yields products corresponding to the S1 and S5 subunits and the S2-S4 and S3-S4 dimers (32). The dissociated PTmu (100 µg in 500 µl of 5 M urea) was separated by high-pressure liquid chromatography (HPLC) using a Spectra Physics SP8750 HPLC system and a Superdex 75 HR 10/30 gel filtration column equilibrated with 100 mM phosphate buffer, pH 7.0, containing 25 mM NaCl. Elution (flow rate, 0.5 ml/min; 264 lb/in²) was carried out in the same buffer. Protein elution was monitored at 220 nm with a Waters model 441 detector. The protein concentration was estimated according to the area of the peaks.

SDS-PAGE and Western blotting. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was carried out according to the procedure of Laemmli (14) on 12% gels (henceforth referred to as "Laemmli gels") or according to the procedure of Schägger and von Jagow (27) on 10% gels (henceforth referred to as "Tricine gels"). All gels were stained with silver. Western blotting onto nitrocellulose membrane was carried out according to the procedure of Towbin et al. (33). Dilutions of 1:500, 1:1,000, 1:8,000, and 1:10,000 were used for anti-S1, anti-S2, anti-S3, and anti-S4 primary antisera, respectively. Reactive bands were detected by enhanced chemiluminescence (ECL) using an ECL analysis system (catalog no. RPN 2108; Amersham, Amersham, England) according to the manufacturer's protocol.

Slot blot analysis for the presence of PT dimers. A glycoprotein binding experiment to assay for the presence of PT subunits in the form of dimers in the relevant peak was essentially done according to the procedure of Witvliet et al. (37), with minor modifications. Briefly, proteins resuspended in phosphate-buffered saline (PBS) were applied directly to nitrocellulose using a Bio-Rad Bio-Dot SF microfiltration apparatus (catalog no. 170-6542). After blocking for 2 h in bovine serum albumin (BSA) solution (3% BSA in PBS), the blots were incubated with the various preparations of PT and PT components as indicated in Results, washed with PBS-Tween (37), and reacted with anti-S4 serum; binding by anti-S4 serum was detected by ECL as described above.

In vitro assay of the mitogenic activity of PT and PT components on naive and committed T cells. LN cells from naive SJL/J mice were cultured in microtiter wells (5 × 10⁵ cells/well) as described previously (2), in the presence of PT, PT components, or concanavalin A (ConA), as indicated. The mitogenic effect of PT and PT components on committed T cells was tested on mouse line PL/J T cells specific for pMOG35-55, a myelin oligodendrocyte glycoprotein peptide encompassing amino acids 35 to 55 of the molecule (12). The pMOG35-55-specific T cells (10⁴ cells/well) were cultured as described above, in the presence of irradiated (2,500 rads) syngeneic spleen cells (5 × 10⁵ cells/well) from naive PL/J mice. The cells were incubated for 72 h (naive LN cells) or 48 h (pMOG35-55-specific T cells) at 37°C in humidified air containing 7.5% CO₂. [³H]thymidine (1 µCi/well) was added for the last 16 h of incubation, and the cultures were harvested and counted using a Matrix 96 Direct beta counter (Packard Instruments, Meriden, Conn.). The proliferative response was measured as [³H]thymidine incorporation expressed as mean counts per minute of triplicate cultures.

Cytofluorometric analysis of T-cell subsets in naive LN cells stimulated with PTmu, PT components, and other T-cell mitogens. Naive LN cells (5 × 10⁶ cells/ml) from SJL/J mice were cultured as described previously (2), in the presence of PTmu (250 ng/ml), peak II (see Results) (2 µg/ml), ConA (500 ng/ml), or an anti-CD3 monoclonal antibody (MAb) (30 µl of a 1:2 dilution of supernatant from hybridoma hamster anti-mouse CD3, clone no. 145-2C11, from the American Type Culture Collection). After 2 to 3 days, the cells were split in medium containing interleukin-2. At the time intervals indicated in Results, the cells were analyzed by fluorescence-activated cell sorting (FACS) as described previously (17), using the following MAbs: fluorescein isothiocyanate (FITC)-conjugated anti-mouse TCR (H57-597; catalog no. 01304D; Pharmingen, San Diego, Calif.); FITC-conjugated anti-mouse CD8 (YTS 169.4; catalog no. RM2201; Caltag, San Francisco, Calif.), and phycoerythrin-conjugated anti-mouse CD4 (catalog no. 1447; Becton Dickinson, Mountain View, Calif.).

Cytofluorometric analysis for apoptosis-necrosis with annexin. FACS analysis of apoptotic-necrotic CD4⁺ and CD8⁺ cells was performed by double staining with FITC-conjugated anti-CD4 (catalog no. 09424D; Pharmingen) or anti-CD8 (as described above) MAb and Annexin-V-Alexa 568 (catalog no. 1985 485; Boehringer Mannheim/Roche, Mannheim, Germany) according to the manufacturer's instructions.

Depletion of CD8⁺ and CD4⁺ T cells from naive mice by anti-CD8⁺ and anti-CD4⁺ MAbs. Naive SJL/J mice were injected intraperitoneally on three consecutive days, 7 to 9 days before in vitro analysis of the mitogenic effect of PT, with a rat anti-mouse CD8⁺ (YTS-169) (6) or CD4⁺ (GK 1-5) (7) MAb (1 ml of a 1:10 dilution of ascites fluid per injection). The hybridomas, YTS-169 and GK1-5, secreting anti-CD8 and anti-CD4 Abs respectively, were a kind gift from L. Eisenbach of the Department of Immunology, The Weizmann Institute of Science. Depletion of the relevant T-cell subsets was monitored on the day of in vitro culture with PT or mitogens by FACS analysis of the LN cell population with anti-CD8⁺ and anti-CD4⁺ MAbs as described above; the results obtained showed full depletion of CD4⁺ T cells in mice treated with the anti-CD4 MAb (0.4% CD4⁺ and 50% CD8⁺ TCR⁺ cells in the LN cell population) and of CD8⁺ T cells in mice treated with the anti-CD8 MAb (0.5% CD8⁺ and 65% CD4⁺ TCR⁺ cells in the LN cell population).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Dissociation of PT components with urea and separation by HPLC. Native S2-S4 and S3-S4 dimers can be dissociated from the PT holomer by incubation of the holomer with 5 M ice-cold urea (32). To isolate the dimers, PTmu was incubated for 4 days at 4°C in 0.05 M phosphate buffer (pH 6.0) containing 5 M urea and the dissociation products were separated by gel filtration on HPLC. As can be seen on the typical HPLC elution profile presented in Fig. 1A, three protein peaks were eluted, which were further analyzed for PT components and the presence of dimers.

View larger version (37K): [in this window] [in a new window]   FIG. 1.   Fractionation of urea-dissociated PT by HPLC. Panels A, chromatographic profile; B, silver-stained 10% Tricine gel of PT and peaks I through III (protein amounts loaded: PT, 4 µg; peak I, 1 µg; peak II, 1.5 µg; peak III, 1.5 µg); C, electrophoresis profiles of PTmu on Tricine and Laemmli gels (lanes a) and corresponding Western blots (lanes b) indicating that the S3 subunit migrates as a doublet on Tricine gels but not on Laemmli gels. Lanes a are silver-stained gels, and lanes b are the corresponding Western blots probed with anti-S3 serum. Rabbit anti-S3 serum was used at a dilution of 1:8,000, and reactive bands were detected by ECL. Two micrograms of PTmu was loaded on the gels. OD, optical density.

Characterization of HPLC gel filtration peaks: isolation of S3-S4 dimer. (i) SDS-PAGE. Because of its greater power of separation, the Tricine SDS-PAGE method of Schägger and von Jagow (27) was used to analyze the HPLC fractions obtained from the dissociated PT (Fig. 1B). Although PT migrates as five bands in the Laemmli (14) SDS-PAGE system (32) (Fig. 1C), an additional band running close to the S3 subunit could be observed on a 10% Tricine gel (Fig. 1C). Interestingly, the S3 subunit was demonstrated by Western blotting with a highly specific anti-S3 Ab (see below) to run as a doublet in this gel system but not on a 12% Laemmli gel (Fig. 1C). The significance of this observation is unclear. The electrophoretic pattern of peak I on the 10% Tricine gel (Fig. 1B) was similar to that of undissociated PT, except that the density of the band corresponding in molecular size to S1 was decreased, suggesting that peak I represents a mixture of the B oligomer and whole PT. A doublet corresponding to S3 and one band corresponding to S4 were seen in peak II (Fig. 1B). Peak III contained very small amounts of protein. Upon concentration, most of the protein aggregated (Fig. 1B).

(ii) Western blotting. The specificity of the polyclonal Abs to be used was determined by Western blotting of native subunits of PTmu and of the recombinant proteins used for Ab production. The Laemmli SDS-PAGE system (14) was used throughout in Western blotting analyses to ascertain Ab specificity and to determine the composition of the HPLC fractions. On SDS-PAGE, rS2, rS3, and rS4 migrated with slightly higher molecular weights than the native subunits, due to the His tag added to enable purification of the recombinant proteins by nickel affinity chromatography. As can be seen in Fig. 2, single bands reacted when native PT was analyzed with anti-S2 (Fig. 2A), anti-S3 (Fig. 2B), and anti-S4 (Fig. 2C) Abs, respectively, indicating that despite the high degree of homology between S2 and S3 (19), the Abs raised recognize the native subunits in a highly specific manner and can be used to analyze the HPLC peaks for the presence or absence of the various subunits. A slight cross-reactivity was observed when the Abs were tested against the recombinant subunits. Such cross-reactivity, observed mainly with the anti-S2 and anti-S4 Abs (see also below), is likely to be due to the presence, within the polyclonal antisera, of Abs which react to the common His tag portion of the recombinant proteins used for immunization. The presence of S1 was tested with a mouse anti-PT serum which had shown a predominant high reactivity to S1 (see below).

View larger version (29K): [in this window] [in a new window]   FIG. 2.   Western blot analysis of the specificity of Abs raised to rS2, rS3, and rS4. rS2, rS3, and rS4 (0.5 µg/lane) and PTmu (1 µg/lane), electrophoresed on a 12% Laemmli gel and Western blotted onto nitrocellulose, were reacted with anti-S2 serum (dilution, 1:500; panel A), anti-S3 serum (dilution, 1:10,000, panel B), or anti-S4 serum (dilution, 1:10,000; panel C), and reactive bands were detected by ECL.

View larger version (48K): [in this window] [in a new window]   FIG. 3.   Western blot analysis of peak II components by sequential probing with specific Abs. PTmu (1 µg) and peaks I (1 µg) and II (1.5 µg) from the HPLC fractionation of PTmu were electrophoresed on a 12% Laemmli gel and Western blotted onto nitrocellulose. The blot was then sequentially probed with anti-S3 serum (dilution, 1:10,000; panel A), anti-S2 serum (dilution, 1:500; panel B), and anti-S4 serum (dilution, 1:10,000; panel C). Reactive bands were detected by ECL after each probing.

View larger version (73K): [in this window] [in a new window]   FIG. 4.   Western blot analysis for the presence of S1 in peaks I and II. S1 (approximately 100 ng), purified by preparative gel elution as described previously (3), and PT (1 µg), peak I (1 µg), and peak II (0.5 µg), electrophoresed on a 12% Laemmli gel and Western blotted, were reacted with anti-S1 serum (dilution, 1:1,000). The reactive bands were detected by ECL.

(iii) Slot blot analysis for dimer binding to glycoprotein. The assay of Witvliet et al. (37) is based on the well-known property of PT, or the S2-S4 and S3-S4 dimers, to bind glycoproteins, in particular, fetuin and haptoglobin, via the S2 or S3 subunit, both of which have a lectin-like property (26, 31). Purified glycoproteins are blotted onto nitrocellulose and reacted with the PT preparations, and the bound dimers are detected by recognition of the S4 subunit with an anti-S4 serum (37). This assay was adapted to determine whether the S3 and S4 subunits comprising peak II were present as a dimer. The slot blots to be reacted contained the relevant purified glycoproteins; fetuin and haptoglobin; ovalbumin (OVA) and BSA as negative controls; and rS2, rS3, and rS4 (Fig. 5). Incubation of the blots with PT, followed by anti-S4 serum, confirmed the ability of PT to bind to haptoglobin and fetuin (Fig. 5); rS4 reacted strongly with the anti-S4 serum (a slight reactivity of the anti-S4 serum with rS2 and rS3 was observed as noted above). No reactivity with OVA or BSA could be detected. When the blots were probed with peak II containing S3 and S4, a pattern of reactivity identical to that of blots probed with PT was observed (Fig. 5), indicating that peak II contains the S3 and S4 subunits as a dimer which binds to fetuin and haptoglobin via the S3 subunit and is detected by antigen-antibody recognition of the S4 subunit. The lower intensity of binding to fetuin by the S3-S4 dimer compared to PT is related to the presence of the S2-S4 dimer in PT and its absence in peak II, as the S2-S4 dimer has a stronger affinity for fetuin than does the S3-S4 dimer (37). That the S4 subunit by itself does not bind to fetuin or haptoglobin and therefore could not be responsible for the pattern observed with peak II is demonstrated in the blot probed with the rS4 subunit and the anti-S4 Ab, which displayed a pattern commensurate only with reactivity of the anti-S4 Ab with rS4 (Fig. 5); no reactivity could be observed with fetuin, haptoglobin, OVA, or BSA.

View larger version (63K): [in this window] [in a new window]   FIG. 5.   The S3 and S4 subunits are present in peak II in the form of a dimer. Fetuin (1 µg), haptoglobin (1 µg), OVA (1 µg), BSA (1 µg), and rS2, rS3, and rS4 (0.5 µg of each) were adsorbed onto nitrocellulose using a Bio-Rad Bio-Dot SF microfiltration apparatus. The slot blots were incubated with PTmu (2 µg/ml), peak II (2 µg/ml), or rS4 (2 µg/ml), as indicated. Binding of dimers within the preparations was detected by reaction of the slot blots with anti-S4 serum (dilution, 1:10,000) developed with ECL. Binding to fetuin by the S3-S4 dimer is weaker than to haptoglobin (37). Accordingly, the band of peak II reactivity to fetuin is faint, albeit highly reproducible.

The S3-S4 dimer is mitogenic for T cells. The mitogenic effect of PT is known to reside with the B oligomer (9, 31, 36). In a preliminary attempt to investigate the possibility that such mitogenic activity of the B oligomer involves the S3-S4 dimer, the effect of peak II on naive and committed T cells was assessed, compared to the effects of PTmu, the B oligomer, rS2, rS3, and rS4. As can be seen in Fig. 6, culture of naive LN cells from SJL/J mice in the presence of 250 ng of PT or the B oligomer resulted in extensive proliferation; the same amount of peak I or peak II induced an even stronger response, commensurate with the proliferative response to the T-cell mitogen ConA. Naive LN cells did not proliferate in response to 250 ng of rS2, rS3, or rS4 (Fig. 6) or in response to lower concentrations (50 and 100 ng/well) of the individual recombinant subunits (data not shown).

View larger version (18K): [in this window] [in a new window]   FIG. 6.   The S3-S4 dimer of PT is mitogenic for naive and committed T cells. Cells were cultured as described in Materials and Methods with PT, the B oligomer, peak I, peak II, rS2, rS3, or rS4 (250 ng/well); ConA (100 ng/well); or pMOG35-55 (1 µg/well). The stimulation index (in parentheses) was calculated as the mean stimulation in the presence of the relevant protein preparation divided by the mean stimulation in the absence of the added protein (None). n.t., not tested; n.a., not applicable; SD, standard deviation.

Reversal of the CD4⁺/CD8⁺ T-cell ratio upon stimulation of naive LN cells with the S3-S4 dimer, as well as PTmu. LN cells isolated from naive SJL/J mice were cultured (5 × 10⁶ cells/ml) for 96 h in the presence of PTmu (250 ng/ml), the S3-S4 dimer (peak II) (2 µg/ml), ConA (500 ng/ml), or the anti-CD3 MAb (30 µl of a 1:2 dilution of culture supernatant/ml), and the proportion of TCR⁺ T cells which expressed CD4⁺, CD8⁺, or CD4⁺ CD8⁺ was estimated by FACS analysis. As can be seen in Table 1, the ratio of CD4⁺ to CD8⁺ T cells in the TCR⁺ LN cell population which were stimulated in vitro with ConA or the anti-CD3 MAb (53 and 44% CD4⁺ and 21 and 20% CD8⁺) did not differ significantly from the ratio of CD4⁺ to CD8⁺ T cells estimated in naive TCR⁺ cells (50% CD4⁺ and 28.5% CD8⁺), indicating that ConA and the anti-CD3 MAb have a nondiscriminatory mitogenic effect on T cells. In contrast, the proportions of CD8⁺ T cells and CD4⁺ T cells changed dramatically following stimulation with PTmu (Table 1), resulting in a complete reversal of the CD4⁺/CD8⁺ ratio (19.9% CD4⁺ and 65.9% CD8⁺). A similar reversal of the CD4⁺/CD8⁺ ratio was obtained upon stimulation with native PT or with the B oligomer (data not shown). As shown in Table 1, stimulation of naive LN cells with the S3-S4 dimer (peak II) also resulted in reversal of the CD4⁺/CD8⁺ ratio (24% CD4⁺ and 58.2% CD8⁺), indicating that the mitogenic S3-S4 dimer is sufficient to reproduce the effect of PTmu on the CD4⁺/CD8⁺ ratio. Similar results were obtained whether interleukin-2 was present or not for the last 2 to 3 days of culture (data not shown). To further assess the effect of PT and the S3-S4 dimer on T-cell subsets, the proportions of CD4⁺ and CD8⁺ T cells in naive LN cells were determined by FACS analysis at various intervals during incubation with PTmu, the S3-S4 dimer, the anti-CD3 MAb, and ConA. As can be seen in Fig. 7, an increase in the proportion of CD8⁺ T cells, in parrallel with a decrease in the proportion of CD4⁺ T cells, was observed upon incubation with PTmu or with the S3-S4 dimer; the extent of the reversal of the CD4⁺/CD8⁺ ratio peaked after approximately 72 h of incubation and leveled out thereafter (Fig. 7). The proportion of CD4⁺ or CD8⁺ T cells did not change at any time during incubation with the anti-CD3 MAb or ConA (Fig. 7). The reversal of the CD4⁺/CD8⁺ ratio in naive LN cells incubated with PTmu or the S3-S4 dimer does not appear to be due to preferential stimulation of CD8⁺ T cells by PTmu or the S3-S4 dimer; indeed, the extent of proliferation in response to PTmu, as standardized to the proliferation in response to the anti-CD3 MAb, did not differ between LN cells isolated from mice fully depleted of CD8⁺ T cells by treatment with the anti-CD8 MAb and LN cells isolated from mice fully depleted of CD4⁺ T cells by treatment with the anti-CD4 MAb (Fig. 8). To assess the possibility that the increase in the proportion of CD8⁺ T cells upon stimulation with PT or the S3-S4 dimer was a result of preferential apoptosis-necrosis of CD4⁺ T cells, LN cells from naive mice and from mice depleted of CD8⁺ or CD4⁺ T cells were stimulated with PTmu, the anti-CD3 MAb, or ConA and analyzed by FACS for annexin staining at various intervals during a 5-day culture. FACS analysis of cells double stained with annexin and the anti-CD4 or anti-CD8 MAb did not indicate any preferential apoptosis-necrosis of CD4⁺ T cells over CD8⁺ T cells in naive LN cells incubated with PTmu (Fig. 9, Undepleted), nor was there any difference in the extent of apoptosis-necrosis in the naive LN cells incubated with the anti-CD3 MAb or ConA, compared with the PTmu incubation (Fig. 9, Undepleted). These observations were corroborated by data obtained with LN cells from CD4⁺ T-cell-depleted or CD8⁺ T-cell-depleted mice (Fig. 9); indeed, the extent of apoptosis-necrosis of the T cells was essentially the same in cultures stimulated with PTmu, the anti-CD3 MAb, or ConA, regardless of the origin of the LN cells (Fig. 9).

View larger version (21K): [in this window] [in a new window]   FIG. 7.   Time course analysis of reversal of the CD4+/CD8+ T-cell ratio upon in vitro stimulation of naive LN cells with PT and the S3-S4 dimer (peak II). Naive LN cells were cultured as described in Materials and Methods in the presence of the indicated mitogens. At the time intervals indicated, 0.2 × 106 cells were tested for CD4 and CD8 expression by double-staining FACS analysis. The data shown are means ± standard deviations of three separate experiments. An asterisk indicates CD4+ or CD8+ cells among TCR+ cells.

View larger version (30K): [in this window] [in a new window]   FIG. 8.   Reversal of the CD4+/CD8+ ratio in LN cells stimulated with PTmu is not due to preferential stimulation of CD8+ T cells. The proliferative response to the anti-CD3 Ab (5 µl/ml) or PTmu (250 ng/ml) by LN cells isolated from normal mice, mice depleted of CD4+ T cells, or mice depleted of CD8+ T cells was assayed as described in Materials and Methods and is expressed as a mean stimulation index (SI) ± the standard deviation (SD). Standardization of PTmu stimulation to anti-CD3 Ab stimulation equals 0.85, 0.77, and 1.02 for normal, CD4-depleted, and CD8-depleted mice, respectively.

View larger version (13K): [in this window] [in a new window]   FIG. 9.   Time course analysis of apoptosis-necrosis in CD4+ and CD8+-expressing cells upon in vitro stimulation with different mitogens of LN cells from naive (Undepleted), CD4+ T-cell-depleted, and CD8+ T-cell-depleted mice. LN cells from naive (Undepleted), CD4+ T-cell-depleted, and CD8+ T-cell-depleted mice were cultured as described in Materials and Methods in the presence of PTmu, the anti-CD3 MAb and ConA. At the time intervals indicated, 0.2 × 106 cells were tested by double-staining FACS analysis for the proportion of CD4+ or CD8+ apoptotic-necrotic T cells in the cultures. The mean data of three experiments are shown for each time point; the standard deviation (not shown) varied between 5 and 15% of the mean. An asterisk indicates the percentage of CD4+ or CD8+ T cells which were also stained with annexin.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The data presented in this report provide direct evidence that the S3-S4 dimer of PT is mitogenic and suggest, moreover, that the S3-S4 dimer would be sufficient to account for the mitogenic activity of the B oligomer. To the best of our knowledge, this study is also the first to demonstrate that in vitro stimulation with PT or its S3-S4 dimer component, of naive cells derived from a mature lymphoid organ, results in a reversal of the CD4⁺/CD8⁺ T-cell ratio.

The best-described property of the B oligomer is its mitogenic activity; the catalytic subunit S1 does not play a role in the mitogenicity of PT, as PTmu was as strongly mitogenic as the B oligomer (13; this study). Some of the biological effects and biochemical properties of the B oligomer have been shown to be more strongly associated with the S2-S4 or S3-S4 heterodimer of PT, rather than with the S2 or S3 subunit alone, suggesting that conformation of the subunit as a dimer with S4 is important for activity. Along these lines, we observed a very strong mitogenic effect of the isolated S3-S4 dimer (peak II), on both naive and committed antigen-specific T cells, whereas the recombinant subunits had no such effect. These results are in contrast to previous reports indicating that the S2-S4 and S3-S4 dimers are not mitogenic by themselves (31, 36). The only obvious difference between our study and those previously reported (31, 36) is the absence of urea in our S3-S4 dimer preparation. The possibility that the mitogenic activity of peak II is, in fact, due to contamination with small amounts of the S2-S4 dimer, and therefore to divalent binding on T cells by the S2-S4 and S3-S4 dimers, is highly unlikely, as S2 could not be detected by very sensitive Western blotting of relatively large amounts of peak II, where S3 and S4 appear as very strongly reacting, broad bands. It is more likely that an appropriate conformation of the S3-S4 dimer is necessary for mitogenic activity. Hence, the previous failure to demonstrate the mitogenic activity of the dimer itself may be due simply to the fact that all previous attempts were performed with dimers that were isolated and stored in urea and therefore did not present an appropriate conformation for the biological activity. Isolation of the S3-S4 dimer in the absence of urea in our study would have allowed adequate refolding of the dimer. It is highly relevant to point out the circumstantial evidence suggesting that the S3-S4 dimer is most likely implicated in the mitogenicity of the B oligomer, whereas the S2-S4 dimer may only play a more minor role in this function. Studies have used methylated PT to analyze the involvement of the various PT components in the biological activity of PT (20, 21). Chemical modification of PT by methylation of lysine residues does not significantly interfere with the A protomer-transporting activity of the B oligomer but results in loss of the mitogenic activity of PT (21). Subsequent studies with hybrid, differentially methylated PT holomers indicated that the two dimers play differential roles in the mitogenic activity of the B oligomer; thus, hybrid PT formed with a methylated S3-S4 dimer lost its ability to stimulate T lymphocytes, whereas hybrid PT formed with a methylated S2-S4 dimer was as potent a mitogen as unmodified PT (22). Unfortunately, the direct role of the S2-S4 dimer could not be evaluated in our study, as we were not able to purify the S2-S4 dimer under the chromatography conditions used, possibly as a result of its strong binding to Sephadex.

How stimulation by PT or the S3-S4 dimer results in reversal of the CD4⁺/CD8⁺ T-cell ratio is unclear. From our data, it does not appear to be a result of preferential CD4⁺ T-cell death or preferential stimulation of CD8⁺ T cells. Recent studies suggest the possibility that PTmu or the S3-S4 dimer can selectively induce the generation of CD8⁺ T cells from CD4⁺ CD8⁺ T-cell precursors. Thus, using ZAP-70 mutant thymus organ cultures in which T-cell development is arrested at the CD4⁺ CD8⁺ thymocyte stage, Takahama et al. (30) showed that PT or, more specifically, the B oligomer selectively induced the generation of CD4 CD8⁺ TCR^high cells, possibly by up-regulating Notch expression (30). Notch has been implicated as a participant in the CD4 versus CD8 lineage decision, whereby expression of an activated form of Notch in developing murine T cells was demonstrated to lead to both an increase in CD8 lineage T cells and a decrease in CD4 lineage T cells (25). It is interesting that an increase in the proportion of CD8⁺ T cells was also observed in ex vivo LN cells obtained from mice injected with PT (our unpublished results) and similar observations of sustained increases in CD8⁺ T-cell levels were noted in rhesus macaques as long as 53 days after PT injection (23). The possibility that in vitro PT or the S3-S4 dimer may act on CD4⁺ CD8⁺ T cells to preferentially induce differentiation into CD8⁺ T cells is being further investigated.

Hence, we have demonstrated that the mitogenic activity of PT does not necessarily require the whole B oligomer or divalent binding by the S2-S4 and S3-S4 heterodimers, as shown previously, but that the S3-S4 heterodimer alone can be fully functional in stimulating T cells. In addition, in contrast to other T-cell mitogens, PT or the S3-S4 dimer may exert an immunomodulating effect on the T-cell population by stimulating an increase in the overall proportion of CD8⁺ T cells. We are considering this aspect further in the context of our investigation of the protective effect of PT on autoimmune diseases such as EAE (3, 4).