**2.2. High dose LA significantly increased frequencies of intestinal and systemic CD4+CD25-FoxP3+ Treg cells whereas low dose LA decreased TGF-β and IL-10 producing Treg cell responses**

Dose Effects of LAB on Modulation of Rotavirus Vaccine Induced Immune Responses 267

These data clearly demonstrated that low dose LA promoted IFN-γ producing T cell and down-regulated Treg cell responses, whereas high dose LA induced a strong Treg cell response and promoted the regulatory cytokine production by tissue-residing Treg cells postchallenge in gnotobiotic pigs. Studies of other lactobacilli strains have reported similar findings. A mixture of *L. plantarum* CEC 7315 and CEC 7316 at high dose (5×109 CFU/day) resulted in significant increases in the percentages of activated potential T-suppressor and NK cells, while at low dose (5×108 CFU/day) increased activated T-helper cells, B cells and antigen presenting cells (APCs) in institutionalized seniors (Mane et al., 2011). High concentration (≥ 1×106 colony forming unit [CFU]/ml) of a combination containing LA and *Bifidobacterium* or *B. infantis* attenuated mitogen-induced immune responses by inhibiting cell proliferation and arresting the cell cycle at the G0/G1 stage in both mitogen-stimulated spleen cells and peripheral blood MNC. However, low concentration (≤ 1×106 CFU/ml) promoted a shift in the Th1/Th2 balance toward Th1-skewed immunity by enhancing IFN-γ and inhibiting IL-4 response (Li et al., 2011). The differences between the "low dose" and "high dose" LAB in these studies are small, yet the immunomodulatory effects are

Dose effects may explain some of the controversies that result from the same probiotic strain used by different research groups in animal studies showing opposite immunomodulatory functions. For example, administration of *L. casei* suppressed pro-inflammatory cytokine expression by CD4+ T cells and up-regulated IL-10 and TGF-β levels in rats (So et al., 2008a; So et al., 2008b). On the contrary, another study found that *L. casei* was a pure Th1 inducer in mice. In addition to the difference in animal species, the *L. casei* doses used by the different studies differed significantly, with much higher doses used in the studies of rats (So et al., 2008a; So et al., 2008b). In the studies of rats, the amount of *L. casei* was 5 × 109 or 2 x1010 CFU/dose per rat, three times per week for 11-12 weeks. In the study of mice, the amount of *L. casei* was 2 x 108 CFU/dose per mouse, twice per week for 8 weeks (Van Overtvelt et al., 2010). Thus, different dose and frequency of administration of the same LAB strains may

The dose effects of LA on immune responses to the AttHRV vaccine in pigs may also partly explain why the efficacies of oral rotavirus vaccines are significantly reduced in low-income countries compared to developed countries. The two licensed rotavirus vaccines, RotaTeq and Rotarix have a protective efficacy of >85% against moderate to severe rotavirus gastroenteritis in middle and high-income countries (O'Ryan et al., 2009). However, the protective efficacy of RotaTeq vaccine is only 39.3% against severe rotavirus gastroenteritis in sub-Saharan Africa (Armah et al., 2010) and 48.3% in developing countries in Asia (Zaman et al., 2010). Rotarix vaccine showed a similar disparity in efficacy in low-income countries in Africa (O'Ryan & Linhares, 2009). In addition to other factors that contribute to the reduction in rotavirus vaccine efficacy (e.g., higher titers of maternal antibodies, malnutrition), during the initial colonization of human infants, exposure to high doses of commensal bacteria (common in countries with lower hygiene standards) would have a suppressive effect on IFN-γ producing T cell responses and promote Treg cell responses,

thus leading to the lowered protective immunity after rotavirus vaccination.

qualitatively different.

result in totally different *in vivo* effects.

Frequencies and cytokine production of Treg cells in pigs were differentially modulated by low versus high dose LA at both prechallenge and postchallenge. HiLA+AttHRV pigs had significantly higher frequencies of CD4+CD25-FoxP3+ Treg cells (ranging from 6- to 86-fold higher) in all the tissues compared to LoLA+AttHRV and AttHRV only pigs pre- and postchallenge (Table 2).

Because Treg cells exert regulatory functions through mechanisms involving TGF-β and IL-10, we also compared the effects of high and low dose LA on frequencies of the Treg cell subsets that produced TGF-β or IL-10 among the AttHRV-vaccinated pigs.


(Summarized from Wen et al., 2012)

**Table 2.** Effect of low dose vs. high dose LA on frequencies of Treg cells

Low dose LA reduced frequencies of TGF-β producing CD4+CD25+FoxP3+ and CD4+CD25- FoxP3+ Treg cells compared to high dose LA and AttHRV only pigs in all tissues pre- and postchallenge (Table 3; data for CD25+ Treg cells are not shown). Low dose LA also reduced pre- and postchallenge frequencies of IL-10 producing CD4+CD25+FoxP3+ and CD4+CD25- FoxP3+Treg cells compared to high dose LA and AttHRV- only (except for CD4+CD25- FoxP3+ Treg cells in ileum and spleen postchallenge) (Table 3). High dose LA induced 2.6 fold and 20-fold, respectively higher frequencies of IL-10 producing CD4+CD25-FoxP3+ Treg cells in ileum and spleen postchallenge compared to AttHRV only.


(Summarized from Wen et al., 2012)

**Table 3.** Effect of low dose vs. high dose LA on CD25-FoxP3+ Treg cell cytokine production

These data clearly demonstrated that low dose LA promoted IFN-γ producing T cell and down-regulated Treg cell responses, whereas high dose LA induced a strong Treg cell response and promoted the regulatory cytokine production by tissue-residing Treg cells postchallenge in gnotobiotic pigs. Studies of other lactobacilli strains have reported similar findings. A mixture of *L. plantarum* CEC 7315 and CEC 7316 at high dose (5×109 CFU/day) resulted in significant increases in the percentages of activated potential T-suppressor and NK cells, while at low dose (5×108 CFU/day) increased activated T-helper cells, B cells and antigen presenting cells (APCs) in institutionalized seniors (Mane et al., 2011). High concentration (≥ 1×106 colony forming unit [CFU]/ml) of a combination containing LA and *Bifidobacterium* or *B. infantis* attenuated mitogen-induced immune responses by inhibiting cell proliferation and arresting the cell cycle at the G0/G1 stage in both mitogen-stimulated spleen cells and peripheral blood MNC. However, low concentration (≤ 1×106 CFU/ml) promoted a shift in the Th1/Th2 balance toward Th1-skewed immunity by enhancing IFN-γ and inhibiting IL-4 response (Li et al., 2011). The differences between the "low dose" and "high dose" LAB in these studies are small, yet the immunomodulatory effects are qualitatively different.

266 Lactic Acid Bacteria – R & D for Food, Health and Livestock Purposes

**producing Treg cell responses** 

postchallenge (Table 2).

(Summarized from Wen et al., 2012)

(Summarized from Wen et al., 2012)

**2.2. High dose LA significantly increased frequencies of intestinal and systemic CD4+CD25-FoxP3+ Treg cells whereas low dose LA decreased TGF-β and IL-10** 

Frequencies and cytokine production of Treg cells in pigs were differentially modulated by low versus high dose LA at both prechallenge and postchallenge. HiLA+AttHRV pigs had significantly higher frequencies of CD4+CD25-FoxP3+ Treg cells (ranging from 6- to 86-fold higher) in all the tissues compared to LoLA+AttHRV and AttHRV only pigs pre- and

Because Treg cells exert regulatory functions through mechanisms involving TGF-β and IL-10, we also compared the effects of high and low dose LA on frequencies of the Treg cell

HiLA+AttHRV 2.96 10.34 9.00 1.56 3.78 3.65 LoLA+AttHRV 0.24 0.54 0.10 0.08 0.22 0.05 AttHRV only 0.23 0.59 0.51 0.27 0.26 0.17

Low dose LA reduced frequencies of TGF-β producing CD4+CD25+FoxP3+ and CD4+CD25- FoxP3+ Treg cells compared to high dose LA and AttHRV only pigs in all tissues pre- and postchallenge (Table 3; data for CD25+ Treg cells are not shown). Low dose LA also reduced pre- and postchallenge frequencies of IL-10 producing CD4+CD25+FoxP3+ and CD4+CD25- FoxP3+Treg cells compared to high dose LA and AttHRV- only (except for CD4+CD25- FoxP3+ Treg cells in ileum and spleen postchallenge) (Table 3). High dose LA induced 2.6 fold and 20-fold, respectively higher frequencies of IL-10 producing CD4+CD25-FoxP3+

PID 28 PCD 7

Frequencies of IL-10+ cells among CD4+CD25-FoxP3+ Treg cells

HiLA+AttHRV 4.51 1.13 0.92 4.87 10.05 1.39 LoLA+AttHRV 0.32 0.15 0.31 0.00 0.17 0.38 AttHRV only 4.19 1.19 1.55 4.34 0.79 1.53

HiLA+AttHRV 4.21 2.97 0.88 10.68 17.70 2.51 LoLA+AttHRV 0.92 0.15 0.00 5.22 6.79 0.00 AttHRV only 4.47 3.01 1.77 4.08 0.90 1.44

**Table 3.** Effect of low dose vs. high dose LA on CD25-FoxP3+ Treg cell cytokine production

Ileum Spleen Blood Ileum Spleen Blood Frequencies of TGF-β+ cells among CD4+CD25-FoxP3+ Treg cells

Frequencies of CD4+CD25-FoxP3+Treg cells among total MNC

Ileum Spleen Blood Ileum Spleen Blood

PID 28 PCD 7

subsets that produced TGF-β or IL-10 among the AttHRV-vaccinated pigs.

**Table 2.** Effect of low dose vs. high dose LA on frequencies of Treg cells

Treg cells in ileum and spleen postchallenge compared to AttHRV only.

Dose effects may explain some of the controversies that result from the same probiotic strain used by different research groups in animal studies showing opposite immunomodulatory functions. For example, administration of *L. casei* suppressed pro-inflammatory cytokine expression by CD4+ T cells and up-regulated IL-10 and TGF-β levels in rats (So et al., 2008a; So et al., 2008b). On the contrary, another study found that *L. casei* was a pure Th1 inducer in mice. In addition to the difference in animal species, the *L. casei* doses used by the different studies differed significantly, with much higher doses used in the studies of rats (So et al., 2008a; So et al., 2008b). In the studies of rats, the amount of *L. casei* was 5 × 109 or 2 x1010 CFU/dose per rat, three times per week for 11-12 weeks. In the study of mice, the amount of *L. casei* was 2 x 108 CFU/dose per mouse, twice per week for 8 weeks (Van Overtvelt et al., 2010). Thus, different dose and frequency of administration of the same LAB strains may result in totally different *in vivo* effects.

The dose effects of LA on immune responses to the AttHRV vaccine in pigs may also partly explain why the efficacies of oral rotavirus vaccines are significantly reduced in low-income countries compared to developed countries. The two licensed rotavirus vaccines, RotaTeq and Rotarix have a protective efficacy of >85% against moderate to severe rotavirus gastroenteritis in middle and high-income countries (O'Ryan et al., 2009). However, the protective efficacy of RotaTeq vaccine is only 39.3% against severe rotavirus gastroenteritis in sub-Saharan Africa (Armah et al., 2010) and 48.3% in developing countries in Asia (Zaman et al., 2010). Rotarix vaccine showed a similar disparity in efficacy in low-income countries in Africa (O'Ryan & Linhares, 2009). In addition to other factors that contribute to the reduction in rotavirus vaccine efficacy (e.g., higher titers of maternal antibodies, malnutrition), during the initial colonization of human infants, exposure to high doses of commensal bacteria (common in countries with lower hygiene standards) would have a suppressive effect on IFN-γ producing T cell responses and promote Treg cell responses, thus leading to the lowered protective immunity after rotavirus vaccination.

### **3. Dose effects of LA on antibody and B cell responses**

Probiotics are known to modulate both humoral and cellular immune responses. Probiotics can induce antigen-specific and non-specific IgA antibody responses at mucosal surfaces (Perdigon et al., 2001; Wells & Mercenier, 2008) to prevent invasion by pathogenic microorganisms. Oral administration of *L. acidophilus* L-92 strain led to a significant increase of IgA production in Peyer's patches in mice (Torii et al., 2007). *L. casei* CRL 431 strain increased induction of intestinal IgA secreting cells in mice (Galdeano & Perdigon, 2006). *L. acidophilus* La1 strain and bifidobacteria enhanced specific serum IgA titers to *S. typhi* strain Ty21a and also total serum IgA in humans (Link-Amster et al., 1994). *L. rhamnosus* GG enhanced rotavirus-specific IgA ASC responses in humans and promoted recovery from rotavirus diarrhea (Kaila et al., 1992). In our earlier study of gnotobiotic pigs, a mixture of LA strain and *L. reuteri* strain did not alter virus-specific intestinal and systemic antibody and ASC responses, but they significantly enhanced total intestinal IgA secreting cell responses and total serum IgM and intestinal IgM and IgG titers in rotavirus infected pigs (Zhang et al., 2008a).

Dose Effects of LAB on Modulation of Rotavirus Vaccine Induced Immune Responses 269

and systemic lymphoid tissues of gnotobiotic pigs induced by rotavirus vaccination were examined. The animal treatment groups were the same as listed above in the studies of T cell responses. High dose LA did not significantly alter the HRV-specific antibody responses in serum and ASC responses in any tissue at PID 28 and PCD 7 in the AttHRV-vaccinated pigs (Figs 1 and 2), except to reduce the IgG ASC response in ileum of the mock-vaccinated pigs postchallenge (Fig. 2c). In contrast, low dose LA significantly reduced the HRV-specific IgA antibody titers at PID 7 and 14 (Fig. 1a) and reduced or significantly reduced IgG ASC responses in blood pre- and postchallenge as well as IgA ASC responses in spleen and blood postchallenge (Fig. 2a and 2b). Low dose LA also significantly reduced the IgA and IgG ASC

**Figure 1.** Rotavirus-specific serum IgA and IgG antibody responses in Gn pigs vaccinated with AttHRV with or without high or low dose LA feeding. Rotavirus-specific serum IgA (a) and IgG (b) antibody were measured by an indirect isotype-specific antibody ELISA. Error bars indicate the standard error of the mean. Different capital letters (A, B) indicate significant difference among different pig groups at the same time point (Kruskal Wallis Test, p<0.05, n=3-27), whereas shared letters or no letters on top

The negative effects of low dose LA on the HRV-specific serum antibody responses and ASC responses induced by the AttHRV vaccine were undesirable for the vaccine's immunogenicity; however it is consistent with the strong pro-Th1 effect of the low dose LA. The skewed balance toward a Th1 type immune response in the low dose LA group may have resulted in the weakened antibody responses. In the subsequent studies, we evaluated the effects of a low dose and an intermediate dose of *L. rhamnosus* GG on the effector T cell,

indicate no significant difference.

responses in ileum of the mock-vaccinated pigs postchallenge (Fig. 2c).

The first reported adjuvant effect of probiotic LAB in vaccination was from a human clinical trial in which *L. rhamnosus* GG was shown to enhance rotavirus-specific IgM secreting cells and rotavirus IgA seroconversion in infants receiving a live oral rhesus-human rotavirus reassortant vaccine (Isolauri et al., 1995). In recent years, an increasing number of human clinical trials have demonstrated adjuvant effects of probiotics in enhancing vaccine-induced antibody responses. In a double-blind randomized controlled trial, *L. rhamnosus* GG or *L. acidophilus* CRL 431 increased serum poliovirus neutralizing antibody titers and poliovirusspecific IgA and IgG titers 2- to 4-fold in adult human volunteers vaccinated with the live oral polio vaccine (de Vrese et al., 2005). In another human clinical trial, six out of the seven probiotic strains tested enhanced cholera-specific IgG antibody concentration in serum; for the *B. lactis* Bl-04 and *L. acidophilus* La-14 strains the increase was more significant (Paineau et al., 2008). Daily consumption of a fermented dairy drink (*L. casei* DN-114 001 and yoghurt ferments, Actimel) was shown to increase virus specific antibody responses to the intramuscular inactivated influenza vaccine in individuals of over 70 years of age (Boge et al., 2009). In a randomized, double-blind placebo-controlled pilot study, *L. rhamnosus* GG significantly improved the development of serum antibody responses to the H3N2 strain influenza virus (84% receiving *L. rhamnosus* GG versus 55% receiving placebo had a protective titer 28 days after vaccination) in healthy adults receiving the live attenuated influenza vaccine (FluMist, Medimmune Vaccines, Gaithersburg, MD, USA) (Davidson et al., 2011). Thus, specific strains of probiotics can act as adjuvants to enhance humoral immune responses following not only mucosal (oral or intranasal) but also parenteral vaccination. Yet, dose effects of probiotics on antibodies responses have not been well studied.

In our studies, we demonstrated that high dose LA did not significantly alter the HRVspecific antibody responses whereas low dose LA had negative effects on the antibody responses. The effect of high and low dose LA NCFM strain on HRV-specific serum IgA and IgG antibody levels and HRV-specific ASC and memory B cell responses in the intestinal and systemic lymphoid tissues of gnotobiotic pigs induced by rotavirus vaccination were examined. The animal treatment groups were the same as listed above in the studies of T cell responses. High dose LA did not significantly alter the HRV-specific antibody responses in serum and ASC responses in any tissue at PID 28 and PCD 7 in the AttHRV-vaccinated pigs (Figs 1 and 2), except to reduce the IgG ASC response in ileum of the mock-vaccinated pigs postchallenge (Fig. 2c). In contrast, low dose LA significantly reduced the HRV-specific IgA antibody titers at PID 7 and 14 (Fig. 1a) and reduced or significantly reduced IgG ASC responses in blood pre- and postchallenge as well as IgA ASC responses in spleen and blood postchallenge (Fig. 2a and 2b). Low dose LA also significantly reduced the IgA and IgG ASC responses in ileum of the mock-vaccinated pigs postchallenge (Fig. 2c).

268 Lactic Acid Bacteria – R & D for Food, Health and Livestock Purposes

(Zhang et al., 2008a).

**3. Dose effects of LA on antibody and B cell responses** 

effects of probiotics on antibodies responses have not been well studied.

In our studies, we demonstrated that high dose LA did not significantly alter the HRVspecific antibody responses whereas low dose LA had negative effects on the antibody responses. The effect of high and low dose LA NCFM strain on HRV-specific serum IgA and IgG antibody levels and HRV-specific ASC and memory B cell responses in the intestinal

Probiotics are known to modulate both humoral and cellular immune responses. Probiotics can induce antigen-specific and non-specific IgA antibody responses at mucosal surfaces (Perdigon et al., 2001; Wells & Mercenier, 2008) to prevent invasion by pathogenic microorganisms. Oral administration of *L. acidophilus* L-92 strain led to a significant increase of IgA production in Peyer's patches in mice (Torii et al., 2007). *L. casei* CRL 431 strain increased induction of intestinal IgA secreting cells in mice (Galdeano & Perdigon, 2006). *L. acidophilus* La1 strain and bifidobacteria enhanced specific serum IgA titers to *S. typhi* strain Ty21a and also total serum IgA in humans (Link-Amster et al., 1994). *L. rhamnosus* GG enhanced rotavirus-specific IgA ASC responses in humans and promoted recovery from rotavirus diarrhea (Kaila et al., 1992). In our earlier study of gnotobiotic pigs, a mixture of LA strain and *L. reuteri* strain did not alter virus-specific intestinal and systemic antibody and ASC responses, but they significantly enhanced total intestinal IgA secreting cell responses and total serum IgM and intestinal IgM and IgG titers in rotavirus infected pigs

The first reported adjuvant effect of probiotic LAB in vaccination was from a human clinical trial in which *L. rhamnosus* GG was shown to enhance rotavirus-specific IgM secreting cells and rotavirus IgA seroconversion in infants receiving a live oral rhesus-human rotavirus reassortant vaccine (Isolauri et al., 1995). In recent years, an increasing number of human clinical trials have demonstrated adjuvant effects of probiotics in enhancing vaccine-induced antibody responses. In a double-blind randomized controlled trial, *L. rhamnosus* GG or *L. acidophilus* CRL 431 increased serum poliovirus neutralizing antibody titers and poliovirusspecific IgA and IgG titers 2- to 4-fold in adult human volunteers vaccinated with the live oral polio vaccine (de Vrese et al., 2005). In another human clinical trial, six out of the seven probiotic strains tested enhanced cholera-specific IgG antibody concentration in serum; for the *B. lactis* Bl-04 and *L. acidophilus* La-14 strains the increase was more significant (Paineau et al., 2008). Daily consumption of a fermented dairy drink (*L. casei* DN-114 001 and yoghurt ferments, Actimel) was shown to increase virus specific antibody responses to the intramuscular inactivated influenza vaccine in individuals of over 70 years of age (Boge et al., 2009). In a randomized, double-blind placebo-controlled pilot study, *L. rhamnosus* GG significantly improved the development of serum antibody responses to the H3N2 strain influenza virus (84% receiving *L. rhamnosus* GG versus 55% receiving placebo had a protective titer 28 days after vaccination) in healthy adults receiving the live attenuated influenza vaccine (FluMist, Medimmune Vaccines, Gaithersburg, MD, USA) (Davidson et al., 2011). Thus, specific strains of probiotics can act as adjuvants to enhance humoral immune responses following not only mucosal (oral or intranasal) but also parenteral vaccination. Yet, dose

**Figure 1.** Rotavirus-specific serum IgA and IgG antibody responses in Gn pigs vaccinated with AttHRV with or without high or low dose LA feeding. Rotavirus-specific serum IgA (a) and IgG (b) antibody were measured by an indirect isotype-specific antibody ELISA. Error bars indicate the standard error of the mean. Different capital letters (A, B) indicate significant difference among different pig groups at the same time point (Kruskal Wallis Test, p<0.05, n=3-27), whereas shared letters or no letters on top indicate no significant difference.

The negative effects of low dose LA on the HRV-specific serum antibody responses and ASC responses induced by the AttHRV vaccine were undesirable for the vaccine's immunogenicity; however it is consistent with the strong pro-Th1 effect of the low dose LA. The skewed balance toward a Th1 type immune response in the low dose LA group may have resulted in the weakened antibody responses. In the subsequent studies, we evaluated the effects of a low dose and an intermediate dose of *L. rhamnosus* GG on the effector T cell, antibody and ASC responses induced by the AttHRV vaccine and we found that *L. rhamnosus* GG enhanced the production of a balanced Th1 and Th2 immune responses to the AttHRV vaccine and significantly increased the virus-specific IFN-γ producing T cell responses, the antibody responses and the protection rate of the AttHRV vaccine (manuscripts under preparation).

Dose Effects of LAB on Modulation of Rotavirus Vaccine Induced Immune Responses 271

The pDC most effectively sense virus infections and are characterized by their capacity to produce large quantities of IFN-α and the pro-inflammatory cytokines IL-6 and TNF-α. These cytokines promote cDC maturation (Summerfield & McCullough, 2009). MHC II expression in professional APCs is tightly regulated. The MHC II of immature DCs are expressed at low levels at the plasma membrane, but abundantly in endocytic compartments. In the presence of inflammatory cytokines such as IFN-γ, DCs are activated; they stop capturing antigens and markedly increase MHC II expression on their plasma membrane. These MHC II are loaded with peptides derived from antigens captured at the site of inflammation. The mature DCs migrate to lymphoid tissues and up-regulate the costimulatory molecules (CD80/86) necessary to activate naïve T cells (Villadangos et al., 2001). It is known that probiotics can modify the distribution, the phenotype and the function of DC subsets (Grangette, 2012). Both species-specific and strain-specific immunomodulatory effects of different LAB on DCs have been described in a large number of studies and was reviewed previously (Meijerink & Wells, 2010). Among the differential effects, several lactobacilli strains, including *L. acidophilus, L. gasseri, L. fermentum, L. casei, L. plantarum, L. johnsonii,* and *L. rhamnosus* have been reported to stimulate human or murine DCs to produce increased levels of proinflammatory cytokines (IL-2, IL-12, TNF-α) that favored Th1 and cytotoxic T cell polarization, and decreased levels of the regulatory cytokine TGF-β (Chiba et al., 2010; Christensen et al., 2002; Mohamadzadeh et al., 2005; Van Overtvelt et al., 2010; Vitini et al., 2000; Weiss et al., 2010; Yazdi et al., 2010). Such immune stimulating effects are characteristics of adjuvants. However, studies of dose effects of lactobacilli on DC responses are scarce, with most consisting of *in vitro* experiments, and there is a dearth of comparative studies linking *in vitro* and *in vivo* results. *L. rhamnosus* Lcr35 was shown to induce a dose-dependent immunomodulation of human DCs. Lcr35 at 107 CFU/ml (10 multiplicity of infection), but not 104 CFU/ml induced the semi-maturation of the DCs and a strong pro-inflammatory response (Evrard et al., 2011). LA NCFM induced a concentration dependent production of IL-10, and low IL-12p70 in monocyte derived DCs (Konstantinov et al., 2008). Immature DCs incubated with the LA NCFM at a bacterium to cell ratio of 1000:1 ("high dose") produced significantly higher IL-10 compared with the ratio of 10:1. In

contrast, IL-12p70 was up-regulated at a lower concentration of the bacterium (10:1).

or no LA were determined using multi-color flow cytometry.

In our studies, dose effects of LA on pDC and cDC responses after rotavirus vaccination were examined in gnotobiotic pigs. The animal treatment groups were the same as listed earlier in the studies of T cell and B cell immune responses. Porcine pDC (CD172a+CD4+) and cDC (CD172a+CD11R1+) were defined as previously described (Jamin et al., 2006). The frequencies and tissue distribution, MHC II and costimulatory (CD80/86) molecular, TLR (2, 3, 9) and cytokine (IL-6, IL-10, IFN-α, TNF-α) expression by pDC and cDC in ileum, spleen and blood of gnotobiotic pigs vaccinated with the AttHRV and fed with high dose, low dose

The low dose LA group had significantly higher frequencies of pDC in ileum and spleen and cDC in spleen and blood compared to the high dose LA and AttHRV only groups (Fig. 3a). The low dose LA group had overall lower MHC II expression on pDC and cDC in all tissues and lower CD80/86 expression in blood, but significantly higher CD80/86 expression

**Figure 2.** Rotavirus-specific IgA and IgG ASC responses in Gn pigs. Rotavirus-specific IgA and IgG ASC in the MNC isolated from ileum, spleen and blood of AttHRV-vaccinated pigs on PID 28 (PCD 0) (a) and PID 35 (PCD 7) (b) and of mock-vaccinated pigs on PID 35 (PCD 7) (c) were enumerated by using an ELISPOT assay and are presented as the mean numbers of virus-specific IgA and IgG ASC per 5×105 MNC. Error bars indicate the standard error of the mean. Different capital letters (A, B, C) on top of the bars indicate significant difference among the treatment groups for the same isotype in the same tissue (Kruskal Wallis Test, p<0.05, n=3-14), whereas shared letters or no letters on top indicate no significant difference. Note the y axis scale difference (HiLA, high dose LA; LoLA, low dose LA).

### **4. Dose effects of LA on DC responses**

The nature and consequences of a CD4+ T cell response (Th1, Th2, Th17, or Treg type) largely depend on the immune functions of DCs, which are the professional antigen presenting cells that can prime and differentiate naive T cells. Both cDC and pDC are responsible for presenting microbial and dietary antigens to the adaptive immune systems, thereby influencing polarization of the adaptive immune response (Konieczna et al., 2012). The pDC most effectively sense virus infections and are characterized by their capacity to produce large quantities of IFN-α and the pro-inflammatory cytokines IL-6 and TNF-α. These cytokines promote cDC maturation (Summerfield & McCullough, 2009). MHC II expression in professional APCs is tightly regulated. The MHC II of immature DCs are expressed at low levels at the plasma membrane, but abundantly in endocytic compartments. In the presence of inflammatory cytokines such as IFN-γ, DCs are activated; they stop capturing antigens and markedly increase MHC II expression on their plasma membrane. These MHC II are loaded with peptides derived from antigens captured at the site of inflammation. The mature DCs migrate to lymphoid tissues and up-regulate the costimulatory molecules (CD80/86) necessary to activate naïve T cells (Villadangos et al., 2001).

270 Lactic Acid Bacteria – R & D for Food, Health and Livestock Purposes

(manuscripts under preparation).

antibody and ASC responses induced by the AttHRV vaccine and we found that *L. rhamnosus* GG enhanced the production of a balanced Th1 and Th2 immune responses to the AttHRV vaccine and significantly increased the virus-specific IFN-γ producing T cell responses, the antibody responses and the protection rate of the AttHRV vaccine

**Figure 2.** Rotavirus-specific IgA and IgG ASC responses in Gn pigs. Rotavirus-specific IgA and IgG ASC in the MNC isolated from ileum, spleen and blood of AttHRV-vaccinated pigs on PID 28 (PCD 0) (a) and PID 35 (PCD 7) (b) and of mock-vaccinated pigs on PID 35 (PCD 7) (c) were enumerated by using an ELISPOT assay and are presented as the mean numbers of virus-specific IgA and IgG ASC per 5×105 MNC. Error bars indicate the standard error of the mean. Different capital letters (A, B, C) on top of the bars indicate significant difference among the treatment groups for the same isotype in the same tissue (Kruskal Wallis Test, p<0.05, n=3-14), whereas shared letters or no letters on top indicate no significant difference. Note the y axis scale difference (HiLA, high dose LA; LoLA, low dose LA).

The nature and consequences of a CD4+ T cell response (Th1, Th2, Th17, or Treg type) largely depend on the immune functions of DCs, which are the professional antigen presenting cells that can prime and differentiate naive T cells. Both cDC and pDC are responsible for presenting microbial and dietary antigens to the adaptive immune systems, thereby influencing polarization of the adaptive immune response (Konieczna et al., 2012).

**4. Dose effects of LA on DC responses** 

It is known that probiotics can modify the distribution, the phenotype and the function of DC subsets (Grangette, 2012). Both species-specific and strain-specific immunomodulatory effects of different LAB on DCs have been described in a large number of studies and was reviewed previously (Meijerink & Wells, 2010). Among the differential effects, several lactobacilli strains, including *L. acidophilus, L. gasseri, L. fermentum, L. casei, L. plantarum, L. johnsonii,* and *L. rhamnosus* have been reported to stimulate human or murine DCs to produce increased levels of proinflammatory cytokines (IL-2, IL-12, TNF-α) that favored Th1 and cytotoxic T cell polarization, and decreased levels of the regulatory cytokine TGF-β (Chiba et al., 2010; Christensen et al., 2002; Mohamadzadeh et al., 2005; Van Overtvelt et al., 2010; Vitini et al., 2000; Weiss et al., 2010; Yazdi et al., 2010). Such immune stimulating effects are characteristics of adjuvants. However, studies of dose effects of lactobacilli on DC responses are scarce, with most consisting of *in vitro* experiments, and there is a dearth of comparative studies linking *in vitro* and *in vivo* results. *L. rhamnosus* Lcr35 was shown to induce a dose-dependent immunomodulation of human DCs. Lcr35 at 107 CFU/ml (10 multiplicity of infection), but not 104 CFU/ml induced the semi-maturation of the DCs and a strong pro-inflammatory response (Evrard et al., 2011). LA NCFM induced a concentration dependent production of IL-10, and low IL-12p70 in monocyte derived DCs (Konstantinov et al., 2008). Immature DCs incubated with the LA NCFM at a bacterium to cell ratio of 1000:1 ("high dose") produced significantly higher IL-10 compared with the ratio of 10:1. In contrast, IL-12p70 was up-regulated at a lower concentration of the bacterium (10:1).

In our studies, dose effects of LA on pDC and cDC responses after rotavirus vaccination were examined in gnotobiotic pigs. The animal treatment groups were the same as listed earlier in the studies of T cell and B cell immune responses. Porcine pDC (CD172a+CD4+) and cDC (CD172a+CD11R1+) were defined as previously described (Jamin et al., 2006). The frequencies and tissue distribution, MHC II and costimulatory (CD80/86) molecular, TLR (2, 3, 9) and cytokine (IL-6, IL-10, IFN-α, TNF-α) expression by pDC and cDC in ileum, spleen and blood of gnotobiotic pigs vaccinated with the AttHRV and fed with high dose, low dose or no LA were determined using multi-color flow cytometry.

The low dose LA group had significantly higher frequencies of pDC in ileum and spleen and cDC in spleen and blood compared to the high dose LA and AttHRV only groups (Fig. 3a). The low dose LA group had overall lower MHC II expression on pDC and cDC in all tissues and lower CD80/86 expression in blood, but significantly higher CD80/86 expression on cDC in ileum, compared to the high dose LA and AttHRV only groups (Fig. 3b). High dose LA did not have a significant effect on DC frequencies or activation marker MHC II and CD80/86 expression, except for the significantly increased CD80/86 expression on pDC in ileum compared to the AttHRV only group (Fig. 3b).

Dose Effects of LAB on Modulation of Rotavirus Vaccine Induced Immune Responses 273

increased TNF-α, IL-10 and IFN-α production in blood compared to the high dose LA or the AttHRV only group. High dose LA did not have a significant effect on IL-6, TNF-α, and IL-10 but lowered or significantly lowered IFN-α production in both pDC and cDC in all

**Figure 4.** TLR expression patterns of pDC and cDC in intestinal and systemic lymphoid tissues of Gn pigs vaccinated with AttHRV vaccine with high dose, low dose or no LA at PID 28. MNC were stained freshly without *in vitro* stimulation before flow cytometry analyses. Data are presented as mean frequency ± standard error of the mean (n = 3-5). Different letters on top of bars indicate signicant differences in frequencies among groups for the same TLR in the same tissue (Kruskal–Wallis test, p <

Therefore, the effects of high versus low dose LA on the frequencies, maturation status and functions of DCs were strikingly different. Low dose LA significantly increased frequencies of both DC subsets, but these DCs were immature because they expressed lower frequencies of activation makers CD80/86 and MHC II and had reduced TNF-α, IL-10 and IFN-α production compared to the high dose LA and AttHRV only groups. Low dose LA promoted a strong IL-6 response in all tissues and increased all the other cytokine TNF-α, IL-10 and IFN-α production in blood for both pDC and cDC. High dose LA did not have such a significant modulating effect on the DC responses compared to the low dose (with a few exceptions). These findings are consistent with the differential effects of low dose versus high dose LA on the adaptive immune responses. The differential modulating effects of high versus low dose LA are intriguing. The biological and immunological implications of these effects and the underlying mechanisms require further investigation. From these data, it is clear that the same probiotic strain at different doses can exert qualitatively different modulating effects

tissues compared to the AttHRV only group.

0.05), while shared letters indicate no significant difference.

**Figure 3.** Frequencies of pDC and cDC (a) and the activation marker (CD80/86 and MHC II) expression (b) in intestinal and systemic lymphoid tissues of Gn pigs vaccinated with AttHRV vaccine with high dose, low dose or no LA at PID 28. MNC were stained freshly without *in vitro* stimulation before being subjected to flow cytometry analyses. Data are presented as mean frequency ± standard error of the mean (n = 3-13). Different letters on top of bars indicate signicant differences in frequencies among groups for the same cell type and tissue (Kruskal–Wallis test, p < 0.05), while shared letters indicate no significant difference.

The low dose LA group had lower or significantly lower frequencies of TLR3 expression in both pDC and cDC in all tissues and significantly lower TLR2 expression on cDC in spleen compared to the high dose LA and AttHRV only groups (Fig. 4). High dose LA did not have a significant effect on TLR expression in ileum and spleen. In blood, high dose LA group had significantly lower TLR3 expression in cDC (and lower in pDC) compared to the AttHRV only group.

The most striking dose effect of LA on the cytokine production profile in DCs is the significantly increased IL-6 in the low dose LA group (Fig. 5). The low dose LA group had significantly higher frequencies of IL-6 producing pDC and cDC in all tissues compared to the high dose LA and AttHRV only groups. Interestingly, the low dose LA reduced or significantly reduced the other cytokine TNF-α, IL-10 and IFN-α production in pDC in ileum and spleen. In contrast to ileum and spleen, the low dose LA increased or significantly increased TNF-α, IL-10 and IFN-α production in blood compared to the high dose LA or the AttHRV only group. High dose LA did not have a significant effect on IL-6, TNF-α, and IL-10 but lowered or significantly lowered IFN-α production in both pDC and cDC in all tissues compared to the AttHRV only group.

272 Lactic Acid Bacteria – R & D for Food, Health and Livestock Purposes

in ileum compared to the AttHRV only group (Fig. 3b).

significant difference.

AttHRV only group.

on cDC in ileum, compared to the high dose LA and AttHRV only groups (Fig. 3b). High dose LA did not have a significant effect on DC frequencies or activation marker MHC II and CD80/86 expression, except for the significantly increased CD80/86 expression on pDC

**Figure 3.** Frequencies of pDC and cDC (a) and the activation marker (CD80/86 and MHC II) expression (b) in intestinal and systemic lymphoid tissues of Gn pigs vaccinated with AttHRV vaccine with high dose, low dose or no LA at PID 28. MNC were stained freshly without *in vitro* stimulation before being subjected to flow cytometry analyses. Data are presented as mean frequency ± standard error of the mean (n = 3-13). Different letters on top of bars indicate signicant differences in frequencies among groups for the same cell type and tissue (Kruskal–Wallis test, p < 0.05), while shared letters indicate no

The low dose LA group had lower or significantly lower frequencies of TLR3 expression in both pDC and cDC in all tissues and significantly lower TLR2 expression on cDC in spleen compared to the high dose LA and AttHRV only groups (Fig. 4). High dose LA did not have a significant effect on TLR expression in ileum and spleen. In blood, high dose LA group had significantly lower TLR3 expression in cDC (and lower in pDC) compared to the

The most striking dose effect of LA on the cytokine production profile in DCs is the significantly increased IL-6 in the low dose LA group (Fig. 5). The low dose LA group had significantly higher frequencies of IL-6 producing pDC and cDC in all tissues compared to the high dose LA and AttHRV only groups. Interestingly, the low dose LA reduced or significantly reduced the other cytokine TNF-α, IL-10 and IFN-α production in pDC in ileum and spleen. In contrast to ileum and spleen, the low dose LA increased or significantly

**Figure 4.** TLR expression patterns of pDC and cDC in intestinal and systemic lymphoid tissues of Gn pigs vaccinated with AttHRV vaccine with high dose, low dose or no LA at PID 28. MNC were stained freshly without *in vitro* stimulation before flow cytometry analyses. Data are presented as mean frequency ± standard error of the mean (n = 3-5). Different letters on top of bars indicate signicant differences in frequencies among groups for the same TLR in the same tissue (Kruskal–Wallis test, p < 0.05), while shared letters indicate no significant difference.

Therefore, the effects of high versus low dose LA on the frequencies, maturation status and functions of DCs were strikingly different. Low dose LA significantly increased frequencies of both DC subsets, but these DCs were immature because they expressed lower frequencies of activation makers CD80/86 and MHC II and had reduced TNF-α, IL-10 and IFN-α production compared to the high dose LA and AttHRV only groups. Low dose LA promoted a strong IL-6 response in all tissues and increased all the other cytokine TNF-α, IL-10 and IFN-α production in blood for both pDC and cDC. High dose LA did not have such a significant modulating effect on the DC responses compared to the low dose (with a few exceptions). These findings are consistent with the differential effects of low dose versus high dose LA on the adaptive immune responses. The differential modulating effects of high versus low dose LA are intriguing. The biological and immunological implications of these effects and the underlying mechanisms require further investigation. From these data, it is clear that the same probiotic strain at different doses can exert qualitatively different modulating effects on DCs and consequently on adaptive immune responses induced by rotavirus vaccines. It has been reported that the effect of low dose microbe-associated pattern molecular (MAPM), such as lipopolysaccharide, was strikingly different as compared to that of high dose on macrophage cell functions: low dose lipopolysaccharide induced a strong inflammatory response in macrophages (Maitra et al., 2011). It is plausible that a similar interaction occurs between the MAPM from LA and DCs in the gut. Future studies are needed to identify the molecular mechanisms of the dose responses of different MAPM.

Dose Effects of LAB on Modulation of Rotavirus Vaccine Induced Immune Responses 275

(by CCIF and/or ELISA)

Mean duration days \*\*

(0.3) A

(0.0) B

(0.2) B

(0.7) A

Mean peak titer (FFU/ml) \*\*, d

2.0 B

5.6 B

4.9 B

4558 A

 % shedding virus \*

31 B 3.8

36 B 1.0

50 B 1.3

100 A 4.7

and the lowest mean cumulative fecal consistency score among all the treatment groups. The durations of diarrhea in the LoLA+AttHRV pigs were significantly shorter compared to the AttHRV only and the mock-vaccinated control pigs. The durations of virus shedding in the LoLA+AttHRV pigs were significantly shorter compared to the HiLA+AttHRV and the mock control pigs. The mean cumulative fecal consistency scores in all the pigs in the LoLA+AttHRV and AttHRV only groups (8.4 and 9.0, respectively) were significantly lower than the control group, indicating significant protection against the severity of diarrhea. Thus, low dose LA slightly, but clearly improved the protection conferred by the AttHRV vaccine against rotavirus diarrhea. In contrast, high dose LA reduced the protection conferred by the AttHRV vaccine as indicated by the significantly longer mean duration of virus shedding (3.8 versus 1.3 days) and higher mean cumulative fecal scores compared to

Treatments n Clinical signs Fecal virus shedding

Mean cumulative score \*\* , c

12.5 (1.4) AB

8.4 (1.3) B

9.8 (1.4) B

14.4 (1.0) A

<sup>∆</sup> The data was partially presented previously (Wen et al., 2012). a Pigs with daily fecal scores of ≥2 were considered diarrheic. Fecal consistency was scored as follows: 0, normal; 1,

d FFU, fluorescent focus forming units. Geometric mean peak titers were calculated among pigs that shed virus.

 Proportions in the same column with different superscript letters (A, B) differ significantly (Fisher's exact test, <sup>p</sup>≤0.05). \*\* Means in the same column with different superscript letters (A, B, C) differ significantly (Kruskal Wallis Test, p≤0.05).

We reported previously that protection rates against rotavirus diarrhea are correlated with virus-specific intestinal IgA ASC and IFN-γ producing T cell responses at PID 28 in Gn pigs (Yuan et al., 1996; Yuan et al., 2008). A balanced Th1 and Th2 type response is needed for the optimal protective immunity against rotavirus. Although low dose LA further reduced the duration of diarrhea in the AttHRV-vaccinated pigs postchallenge, neither low nor high dose LA significantly altered protection rate against rotavirus challenge (proportions of pigs that were infected and developed diarrhea after challenge). Because virus-specific intestinal

**Table 4.** Clinical signs and rotavirus fecal shedding in Gn pigs after virulent HRV challenge<sup>∆</sup>

Mean duration days \*\*,

(0.7b) AB

(0.7) B

(0.5) A

(0.3) A

Mean cumulative score calculation included all the pigs in each group.

% with diarrhea \*, a

HiLA+AttHRV 13 92 A 4.3

LoLA+AttHRV 8 88 A 2.4

AttHRV only 12 67 A 4.6

Mock control 9 100 A 5.6

pasty; 2, semiliquid; and 3, liquid. b Standard error of the mean.

c

\*

the AttHRV only pigs.

**Figure 5.** Cytokine production profiles of pDC and cDC in intestinal and systemic lymphoid tissues of Gn pigs vaccinated with AttHRV vaccine with high dose, low dose or no LA at PID 28. MNC were stained freshly without *in vitro* stimulation before flow cytometry analyses. Data are presented as mean frequency ± standard error of the mean (n = 3-8). Different letters on top of bars indicate signicant differences in frequencies among groups for the same cytokine in the same tissue (Kruskal–Wallis test, p < 0.05), while shared letters indicate no significant difference.
