Construction of recombinant myxoma viruses expressing foreign genes from different intergenic sites without associated attenuation

Two myxoma virus transient dominant selection vectors were constructed and used to generate recombinant viruses expressing single and double foreign gene insertions from intergenic sites. The intergenic insertion sites were located between the myxoma virus genes MJ2 (thymidine kinase) and MJ2a, and MA24 (beta-subunit RNA polymerase) and MA27 (fusion protein) located approximately 60 and 113 kb from the left-end of the viral genome, respectively. Recombinant myxoma viruses expressing the lacZ gene from either intergenic insertion site retained wild-type virulence. However, expression of the gus gene reduced the virulence of the recombinant viruses in vivo. Northern blot analysis indicated that the major late mRNAs encoding the viral RNA polymerase subunit and fusion protein are both of discrete size. Insertion of a foreign gene under the control of a synthetic late promoter between the MA24 and MA27 genes results in a specific-sized major late transcript for the inserted foreign gene. The MA27 gene transcripts directed by these recombinant viruses are heterogeneous in size, implying the typical pattern of poxvirus late transcription by random 3'-termination prior to polyadenylation. The transcription studies suggest signals located downstream of the insertion site direct 3'-processing of late transcripts irrespective of the gene immediately upstream.

Luminometry: a novel bioluminescent immunoassay enhances the quantitation of mucosal and systemic antibody responses

We have directly compared enzyme-linked immunoassays (ELISAs) with bioluminescent immunoassays employing derivatives of the bioluminescent molecule aequorin, and have shown that detection of mucosal and serum antibodies is considerably more sensitive when detected by luminometry. Luminometry is based upon counting photons of light via phototubes and is generally similar to scintillation spectrometry. Current commercial luminometric technology employs a phototube which is most efficient for light emission in the 400-420 nm wavelength range. For this reason, we have chosen the bioluminescent molecule, aequorin, which upon the addition of Ca2+ undergoes a conformational change resulting in the emission of blue light at 469 nm. The high quantum yield is reflected by the fact that addition of Ca2+ to 1 ng of recombinant streptaequorin, a covalent conjugate of streptavidin and aequorin, resulted in the production of 7 x 10(8) relative light units. In this study, we show the superior sensitivity of biotin-streptaequorin when directly compared with biotin-streptavidin linked horseradish peroxidase commonly used for ELISA. For example, mice orally immunized once with cholera toxin (CT) did not exhibit detectable fecal IgA antibodies as determined by ELISA, whereas use of streptaequorin and the bioluminescent immunoassay revealed fecal IgA anti-CT-B subunit antibody titers of 1:24 500. In addition, no detectable anti-CT-B antibodies were noted in saliva samples by ELISA 7 days following oral immunization with CT, while IgA endpoint titers could be extrapolated to 1:393 000. The 21 day fecal IgA anti-CT-B titers were 1:512 by ELISA, whereas titers determined by luminometry reached 1:10(7) when Neutralite avidin and biotinylated aequorin were employed. In general, the bioluminescent immunoassay was > 10(4)-fold more sensitive when compared with ELISA for detection of mucosal and serum antigen- and isotype-specific antibody responses. Thus, the bioluminescent immunoassay is a more sensitive assay for detection of antibodies in dilute external secretions.

Epitope maps of the Escherichia coli heat-labile toxin B subunit for development of a synthetic oral vaccine

Linear B- and T-cell epitopes spanning all 103 amino acids of the Escherichia coli heat-labile toxin B subunit (LT-B) were assessed in mice orally immunized with native LT or with recombinant Salmonella enteritidis expressing LT-B. Oral administration of native LT induced mucosal immunoglobulin A (IgA) antibodies reactive with an epitope at residues 85 to 91, while IgA induced by recombinant Salmonella LT-B reacted with an epitope at residues 36 to 44. Serum IgG anti-LT-B antibodies from mice orally immunized with either LT or with recombinant Salmonella LT-B were directed to both epitopes. A single T-cell epitope spanning residues 34 to 42 was identified by T-cell proliferative and cytokine responses. When a 20-mer peptide (residues 26 to 45) with B- and T-cell epitopes was given orally to BALB/c (H-2(d)) and B10 congenic (I-A(d), I-A(b), and I-A(k)) mice, significant fecal IgA and serum IgG anti-LT-B antibodies were induced. The peptide also induced LT-B-specific T-cell proliferative responses in these mice. Orally administered LT-B peptide (residues 26 to 45) induced a cytokine profile indicative of both T helper 1- and 2-type cells. The remarkable immunogenicity of this 20-mer peptide makes it a candidate for a vaccine to protect against enterotoxigenic E. coli.

Mechanisms for mucosal immunogenicity and adjuvancy of Escherichia coli labile enterotoxin

Escherichia coli labile toxin (LT) was assessed as mucosal immunogen and as adjuvant for tetanus toxoid (TT) in mice. After oral administration of LT, C57BL/6 (H-2b) and BALB/c(H-2d) mice were high mucosal and serum antibody responders, while C3H/HeN (H-2k) mice were low responders. High responders exhibited mainly serum IgG (including IgG1, IgG2a, and IgG2b), as well as IgM and IgA, while mucosal responses were IgA. Analysis of LT-B-specific CD4+ T helper (Th) cells from Peyer's patches (PP) or from spleen revealed a mixed Th1 (interferon-gamma) and Th2 (interleukin-4 and -5) cell pattern. Oral LT given with TT induced TT-specific response patterns identical to LT-B. Analysis of mRNA from TT-specific PP CD4+ Th cells also revealed a mixed Th1- and Th2- type response. Thus, antibody response profiles induced by LT are regulated by both CD4+ Th1 and Th2 cell types.

The role of Th1 and Th2 cells for mucosal IgA responses

We have used cytokine-knockout mice to help determine the precise requirements for CD4+ Th cell regulation of IgA responses. In these studies, we have used two different oral delivery systems to induce mucosal and systemic antibody responses to the vaccine TT. In normal mice, oral administration of TT with CT as adjuvant induces Th2 cells and cytokines, which give rise to mucosal IgA and serum IgG1, IgA, and IgE responses. On the other hand, oral immunization with rSalmonella expressing Tox C results in Th1-type responses as well as Th2 cell-derived IL-10 and macrophage-derived IL-6, which correlate with mucosal IgA and serum IgG2a antibody responses. Two major conclusions can be drawn from our studies with these two regimens in normal, IFN-gamma-/-, and IL-4-/- mice. First, oral administration of rSalmonella, which elicits classical Th1-type responses also induces significant mucosal IgA responses when given to mice with defective Th1- (IFN-gamma-/-) or Th2- (IL-4-/-) cytokine pathways. Interestingly, we detect Th2-type cells producing IL-10 and macrophage-secreting IL-6 in both normal and cytokine-deficient mice, and we postulate that these two cytokines are of most importance for murine IgA responses. Second, oral administration of TT plus CT as adjuvant induces classical Th2-type responses in both normal and IFN-gamma-/- mice. Further, lack of IL-4 results in failure to induce mucosal IgA responses. Thus, the IL-4 pathway is necessary for the CT adjuvant effect for mucosal IgA responses after oral immunization with a protein vaccine.

Mucosal immunity: regulation by helper T cells and a novel method for detection

The mechanisms which regulate mucosal IgA responses to orally administered protein vaccines are not yet fully elucidated. We have used two delivery systems, soluble tetanus toxoid (TT) with the mucosal adjuvant cholera toxin (CT) and recombinant Salmonella expressing Tox C, a fragment of TT, to assess the nature of CD4+ T helper (Th) cells and derived cytokines which support mucosal IgA responses in both normal and cytokine knockout (interferon gamma knockout; IFN-gamma-/- and IL-4-/-) mice. Our results provide important new information regarding Th cell and cytokine regulation of mucosal IgA responses. Whereas TT coadministered with CT induces predominant TT-specific Th2-type responses, rSalmonella delivery of Tox C induced dominant Th1-type responses along with synthesis of the Th2-cytokine IL-10. Both vaccine regimen elicited high levels of mucosal S-IgA and IL-6 production by macrophages. Further oral immunization of IFN-gamma-/- and IL-4-/- mice with rSalmonella Tox C also induced macrophage-derived IL-6 and Th2-derived IL-10 as well as S-IgA responses, suggesting that IFN-gamma from Th1-type cells as well as traditional Th2 cells producing IL-4 and IL-5 are not essential for mucosal IgA responses. Rather, induction of second level Th2 cells producing IL-10 together with high levels of IL-6 from other cell sources may be sufficient for mucosal IgA responses in the absence of traditional Th2 cells. These studies were facilitated by the development of a sensitive new luminometry assay which allowed detection of cytokines and cell surface molecules which are below the levels of detection by current solid phase assays.

Myxoma virus as a vaccine vector for rabbits: antibody levels to influenza virus haemagglutinin presented by a recombinant myxoma virus

To determine whether myxoma virus (MV) could be used as a vector for antigen presentation to European rabbits (Oryctolagus cuniculus), despite its immunosuppressive properties, a recombinant MV expressing the influenza virus haemagglutinin (HA) was constructed. Domestic rabbits, inoculated with the recombinant virus, developed high plasma antibody titres to the HA and IgG to HA was also detected in vaginal secretions. Antibody titres to HA obtained with recombinant MV were similar to titres obtained in rabbits inoculated with vaccinia virus expressing the HA. These results indicate that MV is an efficient vector for antigen presentation in rabbits.

Defective point mutants of the encephalomyocarditis virus internal ribosome entry site can be complemented in trans

Point mutations were introduced at random into cDNA corresponding to nucleotides 260-833 of the encephalomyocarditis virus (EMCV) 5' noncoding region. This region contains the internal ribosome entry site (IRES). The mutations were identified by sequence analysis and the effect on the activity of the IRES was determined using in vitro translation reactions in rabbit reticulocyte lysate. Significantly defective mutants each contained multiple point mutations. These mutants were constructed into a dicistronic mRNA expression plasmid and the activities of the mutant IRES elements were determined using the vaccinia virus/T7 RNA polymerase transient expression system in vivo. The most severely defective of these mutants displayed about 5% of wild-type activity. The activities, relative to wild type, of these mutant IRES elements determined using in vitro and in vivo assays were similar. Two deletion mutants, lacking sequences from the 5' terminus to nt 411 and 484, were also constructed. Each of these deletions inactivated the IRES in vivo (to less than 1% of wild-type activity). Coexpression within cells of the wild-type EMCV IRES, either alone or linked to another coding sequence, enhanced the activity of each of the defective IRES elements except that deleted to nt 484. The results are consistent with a model in which different regions of the IRES participate in a discontinuous transfer of an initiation complex to the 3' end of the IRES element for initiation of protein synthesis to occur.

Mucosal adjuvant effect of cholera toxin in mice results from induction of T helper 2 (Th2) cells and IL-4

Despite pathophysiologic effects including diarrhea, cholera toxin (CT) is a potent mucosal immunogen and adjuvant. We investigated the influence of CT on T helper (Th)-type 1 (Th1) and Th2 cell-regulated Ag-specific B cell isotype and IgG subclass Ab responses elicited when the toxin was co-administered orally with different protein Ags. When mice were orally immunized with tetanus toxoid (TT) and CT as adjuvant, this regimen induced TT-specific secretory IgA responses in the gastrointestinal tract as well as serum IgG, including IgG1 and IgG2b subclasses, and IgA responses. This oral regimen also induced TT- and CT-B-specific IgE responses. In addition, CT also elicited adjuvant effects for Ag-specific IgG1, IgE, and IgA responses when two other protein Ags, OVA and hen egg white lysozyme, were given by the oral route. Quantitative reverse transcriptase-PCR was performed to assess levels of mRNA for Th1 (IFN-gamma) and Th2 (IL-4) cytokine expression in TT-stimulated CD4+ T cell cultures. Both Peyer's patches and splenic CD4+ T cells expressed markedly increased levels of IL-4-specific message, but did not result in changes in IFN-gamma mRNA expression. To determine whether the route of immunization influenced IgE responses, mice were immunized s.c. with TT and CT as adjuvant. Significant increases in total and TT-specific IgE Abs were induced when CT was co-administered. Taken together, these results show that CT acts as a mucosal adjuvant to enhance Th2-type responses and in particular, the IL-4 produced results in a characteristic Ab isotype pattern associated with this cytokine.

Mucosal adjuvant effect of cholera toxin in mice results from induction of T helper 2 (Th2) cells and IL-4

Despite pathophysiologic effects including diarrhea, cholera toxin (CT) is a potent mucosal immunogen and adjuvant. We investigated the influence of CT on T helper (Th)-type 1 (Th1) and Th2 cell-regulated Ag-specific B cell isotype and IgG subclass Ab responses elicited when the toxin was co-administered orally with different protein Ags. When mice were orally immunized with tetanus toxoid (TT) and CT as adjuvant, this regimen induced TT-specific secretory IgA responses in the gastrointestinal tract as well as serum IgG, including IgG1 and IgG2b subclasses, and IgA responses. This oral regimen also induced TT- and CT-B-specific IgE responses. In addition, CT also elicited adjuvant effects for Ag-specific IgG1, IgE, and IgA responses when two other protein Ags, OVA and hen egg white lysozyme, were given by the oral route. Quantitative reverse transcriptase-PCR was performed to assess levels of mRNA for Th1 (IFN-gamma) and Th2 (IL-4) cytokine expression in TT-stimulated CD4+ T cell cultures. Both Peyer's patches and splenic CD4+ T cells expressed markedly increased levels of IL-4-specific message, but did not result in changes in IFN-gamma mRNA expression. To determine whether the route of immunization influenced IgE responses, mice were immunized s.c. with TT and CT as adjuvant. Significant increases in total and TT-specific IgE Abs were induced when CT was co-administered. Taken together, these results show that CT acts as a mucosal adjuvant to enhance Th2-type responses and in particular, the IL-4 produced results in a characteristic Ab isotype pattern associated with this cytokine.

Direct evidence that polypyrimidine tract binding protein (PTB) is essential for internal initiation of translation of encephalomyocarditis virus RNA

The requirement of PTB, polypyrimidine tract binding protein, for internal initiation of translation has been tested using an RNA affinity column to deplete rabbit reticulocyte lysates of PTB. The affinity column was prepared by coupling CNBr-activated Sepharose with the segment of the 5'-untranslated region of encephalomyocarditis virus (EMCV) RNA previously shown to bind PTB. Lysates passed through this column were devoid of PTB, and were incapable of internal initiation of translation dependent on the EMCV 5'-untranslated region, while retaining the capacity for translation dependent on ribosome scanning. Full activity for internal initiation was restored by the addition of recombinant PTB at the physiologically relevant concentration of about 5 micrograms/mL. When various PTB deletion mutants were tested, it was found that this activity required virtually the full-length protein. Thus, PTB is an essential protein for internal initiation promoted by the EMCV 5'-untranslated region. However, the PTB-depleted lysate retained the capacity for internal initiation promoted by the 5'-untranslated regions of another cardiovirus, Theiler's murine encephalomyelitis virus, and of the unrelated hepatitis C virus, and in neither case did addition of recombinant PTB stimulate internal initiation. Therefore, PTB is not a universal internal initiation factor that is indispensable in every case of internal ribosome entry.

The protective role of enteral IgA supplementation in neonatal gut origin sepsis

Preterm infants and infants unable to breast feed are particularly susceptible to gut origin sepsis. Many studies have shown the benefits of breast milk in decreasing the incidence of bacterial infections in neonates. Little in vivo work has focused on prevention of neonatal gut origin sepsis with breast milk components. The aim of this study was to determine whether supplementation of a standard neonatal formula with exogenous, luminally administered, human secretory IgA protects against gut origin sepsis in a newborn rabbit model. Sixty New Zealand white rabbit pups were delivered by cesarean section 1 day preterm and divided into two groups--the IgA group (n = 26) and the non-IgA group (n = 34). Animals were gavage-fed a standard artificial formula (KMR) twice daily. The IgA group was supplemented on days 3 and 4 with 6.25 mg/kg of human secretory IgA. The non-IgA group received an equal volume of saline. On the evening of day 3, the animals were orally challenged with Escherichia coli K100. The quantity of bacteria that colonized the cecum was similar in the two groups. The quantity of bacteria that translocated to the mesenteric lymph node, liver, and spleen was significantly lower in the IgA group (P < .05). The incidence of translocation to the organs was also significantly lower in the IgA group (P < .05). The exogenous secretory IgA showed specificity to E coli K100 by ELISA. These data show that neonatal formula supplemented with human secretory IgA decreases the incidence and quantity of bacterial translocation of E coli K100 in a neonatal rabbit model.(ABSTRACT TRUNCATED AT 250 WORDS)

Potential carcinogenicity of chloral hydrate--a review

Chloral hydrate is commonly used to sedate children for diagnostic or therapeutic procedures. The drug has been extensively used for many years, but there are remarkably few data on its long-term health effects. Concern in this regard is raised by recent studies showing chloral hydrate to be genotoxic, causing chromosome changes and other effects in vivo and in vitro. In addition, chloral hydrate is a reactive metabolite of trichloroethylene, a known carcinogen, and is structurally similar to other carcinogenic intermediates. Two carcinogenicity studies performed using the oral route of administration in mice indicate that the drug is potentially carcinogenic--in one case after a single dose lower than the typical dose used for sedation. Practitioners should be aware of chloral hydrate's genotoxicity and potential carcinogenicity. Discretion in its use seems appropriate until further studies clarify its long term health consequences.

Cell-associated collagenolytic activity by group B streptococci

Group B streptococci (GBS) are important pathogens in neonatal sepsis, pneumonia, and meningitis. The ability of GBS to invade the collagen-rich amniotic membrane of the placenta has been shown in vitro. In the presence of GBS, the collagen fibrils of the amnion appear disordered, suggesting a role for GBS in premature rupture of membranes. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis, Sephadex G-200 column chromatography, and gelatin zymograms were used in this study to characterize cell-associated collagenolytic activities of GBS. The synthetic peptide 2-furanacryloyl-Leu-Gly-Pro-Ala (FALGPA), which mimics the primary structure of collagen, was degraded by GBS USF704, a clinical isolate from the placenta of a septic newborn. Cells of GBS USF704 (9 x 10(7) CFU/ml) hydrolyzed 902 nmol of FALGPA over a 24-h period. As reported for zinc metalloenzymes such as collagenase, the hydrolysis of FALGPA by GBS was inhibited by addition of EDTA or 1,10-phenanthroline. Boiling of the cells resulted in loss of activity, while higher activity was observed with crude GBS cell lysates (hydrolysis of 970 nmol of FALGPA in 1.5 h). Antiserum raised against collagenase from Clostridium histolyticum was found to cross-react with cell-associated proteins produced by GBS and to inhibit GBS FALGPA hydrolysis. Twenty-five additional GBS clinical isolates were screened and found to have various levels of FALGPA hydrolytic activity. These observations suggest a cell-associated collagenolytic activity by GBS which may be involved in premature rupture of membranes and neonatal disease.

Hantavirus pulmonary syndrome in California: report of two cases and investigation

We report two cases of hantavirus pulmonary syndrome that were probably acquired in California. Genetic analysis of tissue specimens from one of the patients revealed that the virus isolated is a variant of the strain (Sin Nombre virus) identified in the outbreak of hantavirus pulmonary syndrome that occurred in the Four Corners region of the southwestern United States in 1993. In addition to presenting the clinical features of the two cases, we discuss the possible risk factors for infection.

The role of the mucus gel layer in intestinal bacterial translocation

The mucus gel layer is thought to be a vital component of the intestinal mucosal barrier. The purpose of this study is to determine if decreasing mucus production following ischemia and reoxygenation of the intestinal mucosa would alter bacterial translocation in an in vitro rat mucosal model. Clonidine was used to decrease mucus production associated with an ischemia/reoxygenation insult. Bacterial translocation was studied in a modified Ussing chamber using Escherichia coli K100. The quantity of mucus produced, as well as the incidence and quantity of bacteria translocating was measured. In the clonidine-treated animals, there was a significant decrease in the amount of mucus produced compared to the control animals. The clonidine animals also had a higher quantity of bacteria translocating during the reoxygenation period compared to the control animals. The mucus gel layer is protective against in vitro translocation of bacteria following an ischemia/reoxygenation insult.

Oral vaccine models: multiple delivery systems employing tetanus toxoid

We have not yet directly examined the Th cell responses induced by using Salmonella/BRD 847 as a vector nor have we performed these experiments following immunization with microspheres. However, production of high serum levels of antigen-specific IgG1 may be indicative of a Th2-type response, whereas high serum levels of IgG2a may reflect a Th1-type response. An important issue in using various oral delivery systems is whether the system(s) employed affects the Th cell response to the same antigen. We therefore analyzed the serum antigen-specific IgG subclasses induced in each of the model systems we studied. Table 2 presents these results. Clearly, oral administration of soluble TT with CT as an adjuvant induced an IgG1 subclass, and encapsulation of TT within microspheres had no effect on this Th2-type response. On the other hand, oral immunization with live Salmonella expressing fragment C of tetanus toxin induced a strong IgG2a subclass response indicative of a Th1-type response. It should be noted that protection against a lethal TT challenge was afforded by elevated levels of both anti-TT IgG1 and IgG2a subclasses. We intend to examine the cytokine profiles in spleen CD4+ T cells from mice immunized with microspheres or Salmonella following in vitro antigen stimulation to confirm the T helper type responses suggested by the IgG subclass data. It will also be important to examine the cytokine patterns induced in Peyer's patch CD4+ T cells following immunization of C57BL/6 with Salmonella/BRD 847. Whereas analysis of the serum IgG subclass profile indicated a strong IgG2a response and thus a systemic Th1-type pattern, this vector also induced a good mucosal IgA response. If our current hypothesis concerning S-IgA production is correct, we would expect a predominant Th2-type profile in CD4+ T cells from Peyer's patch in these mice. Such a result would emphasize the bifurcation of T-cell responses in the systemic versus the mucosal immune environments. The data obtained to data suggest that adjuvants and various vehicle delivery systems may influence the induction of distinct T helper cell subsets to a specific antigen. The unique cytokine arrays produced by these T-cell subsets influence the immune responses in terms of systemic Ig subclasses produced, cell-mediated immune responses, and the production of mucosal S-IgA antibodies. Although additional studies are necessary, the manipulation of T-cell subsets employing adjuvants, antigen packaging, or perhaps even the addition of individual cytokines to various formulations holds significant promise for optimizing immune responses to orally administered vaccines.

Mucosal immunity to infection with implications for vaccine development

The induction of effective mucosal immunity that also provides systemic immunity is a considerable challenge. Over the past two years, efforts to develop novel mucosal vaccine delivery systems to induce mucosal immunity against bacterial and viral diseases, including HIV, have dramatically increased. Here we cite novel vaccines and delivery systems being used to establish effective mucosal immunity.

The immediate downstream codon strongly influences the efficiency of utilization of eukaryotic translation initiation codons

Nucleotide substitutions were introduced into the initiation site of an influenza virus NS cDNA derivative at the +4, +5 and +6 positions (where the A of the AUG codon is defined as +1), in the background of either AUG or CUG as the initiation codon. Capped transcripts of these constructs were translated in rabbit reticulocyte lysate under conditions where the selection of initiation sites conformed to the scanning ribosome model. With CUG as the initiation codon, the efficiency of initiation was as strongly influenced by the nature of the residue in the +5 position as at +4, whilst the influence of the +6 position was smaller. The residues favourable to initiation were as follows: at +4, only G was stimulatory; at +5, A was strongly stimulatory and C fairly beneficial; and at +6, only U exerted any positive influence. The positive influence of the favourable residues (or the negative influence of unfavourable residues) at each position appeared to be additive. With AUG as the initiation codon, the pattern of response to mutations in the +4 and +5 positions was qualitatively similar, but the quantitative effects were smaller. Thus the optimum downstream context for initiation is A/CUGGAU.

Helper Th1 and Th2 cell responses following mucosal or systemic immunization with cholera toxin

We have used the potent mucosal immunogen cholera toxin (CT) to assess antigen-specific CD4+ T-cell responses, including Th1- and Th2-type cells in mucosa-associated tissues, e.g. Peyer's patches (PP), and systemic tissue, e.g. spleen (SP), for their regulatory role in the induction of CT-specific B-cell antibody responses in the gastrointestinal (GI) tract as well as in systemic sites. The CT was given by either oral or intravenous (i.v.) routes and the mice orally immunized with CT exhibited brisk IgA anti-CT antibody responses in faecal extracts and elevated IgG anti-CT antibody responses in serum. Further, significant IgA anti-CT spot-forming cells (SFCs) were seen in lamina propria lymphocytes (LPLs) from mice orally immunized with CT. In contrast, i.v. immunization with CT induced IgM and IgG anti-CT SFC responses in SP, and serum anti-CT antibodies of these two isotypes; no anti-CT responses were induced in the GI tract after immunization by this route. The CD4+ T cells isolated from PP and SP of mice orally immunized with CT were stimulated in vitro with CT-B-coated latex microspheres for 1-6 days, and the induction of IL-2 and interferon gamma (IFN-gamma) (Th1-type) or IL-4 and IL-5 (Th2-type) producing SFCs were analysed by a cytokine-specific ELISPOT and cytokine-specific mRNA was detected by reverse transcriptase (RT)-PCR assays.(ABSTRACT TRUNCATED AT 250 WORDS)