• antibody repertoire development;
  • class switch recombination;
  • IgG subclasses;
  • swine influenza


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Infection of germ-free isolator piglets with swine influenza (S-FLU) that generates dsRNA during replication causes elevation of immunoglobulins in serum and bronchoalveolar lavage, a very weak response to trinitrophenyl conjugates but an immune response to S-FLU. The increased immunoglobulin levels result mainly from the polyclonal activation of B cells during the infection, but model antigen exposure may contribute. The 10-fold increase in local and serum IgG accompanies a 10-fold decrease in the transcription of IgG3 in the tracheal–bronchial lymph nodes and in the ileal Peyer's patches. Infection results in class switch recombination to downstream Cγ genes, which diversify their repertoire; both features are diagnostic of adaptive immunity. Meanwhile the repertoires of IgM and IgG3 remain undiversified suggesting that they encode innate, natural antibodies. Whereas IgG3 may play an initial protective role, antibodies encoded by downstream Cγ genes with diversified repertoires are predicted to be most important in long-term protection against S-FLU.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Effective vaccines depend on many factors, including their adjuvant properties. These may be intrinsic, as in the case of pertussis, or supplied as additives. Using an isolator piglet model in which animals have no previous exposure to maternal IgG, normal flora, or other known immune regulatory factors, we have previously shown that normal gut flora or their pathogen-associated molecular patterns (PAMPs) act as adjuvants through innate immune receptors to allow piglets to respond to T-cell-dependent and TI-2 antigens. As PAMPs and other adjuvants are polyclonal B-cell activators, these also elevate the serum levels of the major immunoglobulin isotypes.[1-3] RNA viruses generate double-stranded RNA (dsRNA), which is considered an adjuvant.[4] Therefore we were keen to know if an RNA virus would have the same effect in stimulating the development of adaptive immunity as gut colonization.

Neonatal vaccines are often administered during the ‘critical window of immunological development’ when only innate immunity, including natural antibodies, and passive antibodies (if provided) are available for protection until the adaptive immune system becomes competent.[5, 6] Piglets are precocial at birth, in contrast to rodents, and obtain no maternal immunoglobulins or known maternal regulators through placental transport, and therefore remain naive.[7, 8] Consistent with their precocial nature, they synthesize IgM, IgG (especially IgG3) and IgA in utero.[9, 10] These pre-immune, or natural antibodies, are considered by some as part of innate immunity since their VH genes show little somatic hypermutation in rodents, infants or swine.[11-14] Fetal pigs recovered by Caesarean surgery or naturally delivered are able to immediately forage in germ-free isolators or in specific pathogen-free auto sows, allowing the effect of environmental factors on a totally naive immune system to be studied in a setting in which these factors are controlled by the experimenter.[5, 15] The primary focus of research using the piglet model is on events within the ‘critical window’ in an effort to better understand developmental immunology during this period.

The genomic potential for antibody production in swine primarily resides in the heavy chain locus encoded on chromosome seven. The relevant genes are organized in a similar way to those of most other mammals; there are < 30 VH genes, two functional DH genes and one JH gene, which is then followed by a relatively monomorphic gene for Cμ (IgM) followed by Cδ (IgD). The B-cell receptor is encoded by Cμ at all stages of B-cell lymphogenesis, which determines the pre-immune B-cell repertoire in all species. In some species IgD can play a compensatory role.[16] These two C-region genes are followed in order downstream by a series of Cγ genes encoding numerous IgG subclasses, Cε (IgE) and finally Cα (IgA). Relevant to this study is Cγ3 (IGHG3), which is the most 5′ Cγ gene among the six known[17] and accounts for a high proportion of fetal IgG transcripts.[18, 19] IgG3 is dominantly expressed in the tracheal bronchial lymph nodes (TBLN; this report), the ileal Peyer's patches (IPP) and mesenteric lymph nodes (MLN), and might be important in the protective immune response of newborns.[10]

Studies reported here used swine influenza virus (S-FLU), which was administered by nasal intubation to test whether a viral infection that generates a PAMP would allow the induction of responses to model antigens, cause class switch recombination, diversification of the antibody repertoire and polyclonally activate pre-immune B cells; all features of adaptive immunity.[2, 3] Whereas S-FLU infection stimulated the production of anti-viral antibodies and those to irrelevant model antigen, these accounted for < 10% of the increase in serum immunoglobulin levels in infected piglets. The impact of S-FLU infection on the response to irrelevant model antigens was modest compared with previous studies using gut colonization with normal gut flora.[2, 3] However, it was nevertheless sufficient to promote class switch recombination to downstream Cγ genes that diversify their repertoire while the repertoires of IgM and IgG3 remain undiversified. Our results predict that IgG antibodies encoded by downstream Cγ genes, not IgG3, will be most important in protection against subsequent S-FLU infections.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Isolator piglets

Pregnant gilts at 112 days of gestation were anaesthetized, their fetuses were recovered by Caesarean aseptic procedures and transferred to germ-free isolators where they were reared on sterile formula as previously described.[2, 15] All animal studies adhered to approved guidelines at the National Animal Disease Center. Four days after birth most animals were intranasally infected with a 1-ml dose (1 × 104 CCID50) of S-FLU (A/Swine/Iowa 1930) that had been propagated in Madin–Darby canine kidney (MDCK) cells. Eight animals (two isolators) were maintained as germ-free (GF) controls (Table 1). On 0 days post-infection (dpi), pigs in isolators 90, 93, 104 and 105 were given intraperitoneal injections of 6 mg trinitrophenyl-keyhole limpet haemocyanin (TNP-KLH) (Table 1). TNP-Ficoll and TNP-KLH were prepared as previously described.[2] Immunization with irrelevant antigens was performed at the time of virus infection (0 dpi), which was 4 days after birth. Blood samples were collected before infection (−4 dpi and 0 dpi) and thereafter at 6, 13, 20, 25 and 28 dpi. Animals that had been immunized with model antigens received a booster immunization of the same dosage at 25 dpi. All piglets were killed at 28 dpi; a spectrum of lymphoid tissues were collected into liquid nitrogen and the lungs were lavaged with minimal essential medium. The bronchoalveolar lavage (BAL) was used for immunoglobulin determination and virus recovery. Lungs were examined for lesions and results were recorded. The periodic blood samples collected were tested for virus, used to determine immunoglobulin concentrations and used for detection of antibodies to S-FLU and TNP by ELISA.

Table 1. Experimental design and anti-swine influenza virus (S-FLU) and anti-trinitrophenyl (TNP) responses in serum
GroupaInfectedbImmunizationSerum VNcAnti-FLU ELISAHIAnti-TNPd
  1. a

    Each group contained four animals (A–D) housed in the same isolator.

  2. b

    All infected animals had viral RNA recovered by PCR.

  3. c

    Virus neutralization (VN) and haemagglutination inhibition (HI) titres measured at necropsy, 28 days post-infection.

  4. d

    Considered positive when the optical density at 450 nm was 10-fold higher than in germ-free controls after adjustment for immunoglobulin content (see Fig. 1).

  5. e

    Two piglets in this group had IgG and IgM levels that were greater than 10-fold higher than in all other infected piglets.

90YesTNP-FicollND4/465 ± 300/42/4
91eYesNoneND3/375 ± 621/31/3
92NoTNP-Ficoll/TNP-keyhole limpet haemocyanin (KLH)ND0/400/40/4
93YesTNP-KLHND3/372 ± 601/43/3
96NoNone< 100/400/40/4
100YesNone1960 ± 12003/460 ± 230/40/4
103YesNone1280 ± 9054/45 ± 61/40/4
104YesTNP-Ficoll1920 ± 2154/425 ± 354/44/4
105YesTNP-KLH1280 ± 9054/435 ± 344/44/4

Measurement of immunoglobulin levels

The concentration of IgM, IgA and IgG was determined by sandwich ELISA as previously described.[1] These assays have a lower limit of detection of 3·5, 3·2 and 1·4 ng/ml, respectively. In the case of BAL, total protein was also measured using a small volume system (NanoDrop Spectrophotometer, ND 1000; Wilmington, DE).

Measurement of anti-S-FLU antibody

Serum and BAL harvested at necropsy were tested for S-FLU antibodies by ELISA, haemagglutination inhibition and virus neutralization.[20] We modified a commercially available blocking ELISA designed for detecting avian antibodies to influenza A virus.[21] The lower the optical density (expressed as signal/noise) the higher the anti-FLU titre. A signal/noise ratio of less than 0·673 was considered positive. Data are reported as the geometric mean and in Fig. 1, the reciprocal values are plotted so that an increase in responsiveness for all parameters would be in the same direction.


Figure 1. Response profiles for isolator piglets in eight treatment groups. Data on the concentration of IgG and IgM in serum, the titre of IgG and IgM responses to trinitrophenyl (TNP), and the ELISA titre are presented. Split scales were required to present the data comparatively. Values for anti-swine influenza virus (S-FLU) titres are plotted as a reciprocal so increasing values would indicate increasing responsiveness for all parameters. The horizontal bar represents the mean value. All groups are statistically different from control Groups 92 and 96.

Download figure to PowerPoint

Measurement of anti-TNP responses

The IgG and IgM anti-TNP antibodies were detected using the ELISA configuration described previously by using TNP-thyroglobulin as the solid-phase antigen.[2, 3] Because S-FLU-infected piglets did not generate the level of anti-TNP activity needed to produce the complete titration plots needed for quantification in ELISA units or ng/ml of specific antibody,[2, 3][22] we recorded anti-TNP activity as the mean of triplicate values of optical density at 405 nm at a fixed dilution (Fig. 1). In Fig. 1 we express these titres ×10 so that values for various parameters can be shown on the same plot.

Affinity chromatography

Columns were constructed using the Amino Link system according to the manufacturer's instructions (Pierce-Thermal, Rockford, IL). Initially, the S-FLU strain (A/swine/Iowa 1930) was propagated in MDCK cells, cell debris was removed by centrifugation at 20 000 × g for 2 hr and the pelleted virus was re-suspended in a small volume of coupling buffer (0·1 m phosphate, 0·15 m NaCl, pH 7·2). These columns recovered only small amounts of anti-S-FLU so after consultation with other FLU investigators, we purchased the recombinant nuclear protein (NP; IMR-274) from IM Genex (San Diego, CA) and constructed an affinity column using the Amino Link technology. With 65% binding and with corrections for molecular weight and Avogadro number, this column should be able to capture 2 mg anti-FLU, assuming a 1 : 1 molecular configuration. The TNP Amino Link column was made using a saturating amount of picryl sulphonic acid, which yielded a bright orange column that should be able to capture 10 mg anti-TNP.

Sera from piglets in the same treatment group were pooled and applied to the NP affinity column and the bound fraction was eluted with 0·025 citrate buffer, pH 2·3. The affinity peak was concentrated using a Millipore PLGC membrane (10-kb cut-off; Millipore, Chicago, IL). The protein concentration of the eluted fraction was determined by NanoDrop spectrophotometry, converted to IgG concentration using an absorbancy coefficient of 1·36 and compared with the total IgG in the pooled samples applied to the column (determined by sandwich ELISA; see above).

IgG3 and total IgG transcript analysis

Total RNA was prepared from various lymphoid tissues and used to prepare cDNA as described previously.[13, 23, 24] The IgG transcripts examined contained the VDJ region downstream through the CH2 domain and were recovered using the hemi-nested primer sets described in the Supplementary material, Table S1. Briefly, the PCR product was cloned into pCR4TOPO and insert-containing clones were identified by blue/white selection. Positive clones were transferred to individual microtitre wells and grown for 18 hr at 37°. One-quarter of the growth was transferred to a second microtitre plate and stored so that individual clones could later be propagated and their inserts could be sequenced. Plasmid DNA was recovered by alkaline hydrolysis from the original plate and transferred to nylon membranes.[10, 18] Membranes were then hybridized with a Cγ pan-specific probe to identify all Cγ-containing clones and subsequently with a probe specific for IgG3 (Cγ3; see Supplementary material, Table S1). Clones from each tissue (n = 40 to n = 50) were analysed and the number of IgG3 clones was subtracted from the number of total Cγ clones identified to generate the proportional data on IgG3 transcription versus other IgG subclass transcript expression. As swine have six expressed IgG subclasses, ‘other’ IgG refers to IgGs encoded by the five Cγ genes encoded downstream from IgG3.

Recovery of IgM and IgA transcripts

The same procedure described above was employed by using the primers sets given in Supplementary material, Table S1, which allow the recovery of VDJ-containing Cμ and Cα transcripts. Cα transcripts were only recovered from a small number of samples from animals that showed an increase in serum IgA levels on 28 dpi (Fig. 2). The Cμ and Cα transcripts were cloned and hybridized as described for Cγ clones except that the appropriate probes that could identify Cμ and Cα clones were employed (see Supplementary material, Table S1).


Figure 2. Serum immunoglobulin levels in periodically collected samples from isolator-reared germ-free piglets, those infected with swine influenza virus (S-FLU) and S-FLU-infected piglets that were also immunized with model antigens. Mean and standard deviations are shown. * Indicates values for immunized/infected piglets that are significantly higher than in non-immunized S-FLU-infected piglets.

Download figure to PowerPoint

Measurement of repertoire diversification index

Measurement of the repertoire diversification index (RDI) is based on successive hybridization of individual clones with two probe cocktails. One contains probes specific for the CDR1 region of porcine VHA, VHB, VHB*, VHC, VHA*, VHE and VHF and a second cocktail containing probes for the CDR2 regions of the same seven genes.[14] We have shown that the use of cocktails gave the same results as sequential hybridization with individual VH gene-specific probes.[14] The seven VH genes analysed comprise > 90% of the pre-immune repertoire in swine.[13, 14] The corresponding ImMunoGeneTics nomenclature for these VH genes has been previously published.[14] We have shown that an increase in the RDI is primarily a result of somatic hypermutation, which is concentrated in the CDR regions so that VDJ transcripts with mutated CDRs fail to hybridize with the cocktail of probes that bind to the CDR regions of germline VH genes.[13, 14] The number of non-hybridizing VH gene clones (designated UNK), partial hybridizing clones (only one CDR cocktail binds) or full hybridizing clones are then recorded and used in the equation below to calculate the RDI. Data are presented here for > 4000 VDJ clones. We also sequenced 140 VDJ clones and determined that the misidentification frequency by hybridization was 2·6–7·5% which has a negligible effect on RDI calculations.[14]

  • display math

Statistical analysis

Mean differences were tested using analysis of variance and Student's t-test programs embedded within the Graphpad Prism data analysis program (Graphpad Software, San Diego, CA).


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

S-FLU-infected piglets developed minimal disease

S-FLU-infected piglets displayed few clinical signs, which consisted mainly of intermittent lack of appetite, which was variable among piglets within or between isolators. No other discernible clinical signs were observed in the infected piglets. Upon necropsy at 28 dpi, all pigs appeared normal and no significant lung lesions were seen. Infected piglets displayed enlargement of the lymph nodes associated with the respiratory tract.

S-FLU-infected piglets have virus-specific antibodies

Control piglets (Groups 92 and 96) remained seronegative and their total IgG and IgM levels remained the same as in newborns (Figs. 1 and 2). At the time of necropsy, ELISA, haemagglutination inhibition and virus neutralization antibodies were detected in the serum of all S-FLU-inoculated piglets (Table 1; Fig. 1). In BAL, signal/noise ELISA values ranged from 0·69 (piglet 100A) to 1·2 (piglets 96A) meaning that no anti- S-FLU IgG was detected in BAL of infected or control piglets. Virus neutralization titres < 10 were the norm for half of the BAL samples whereas the remainder had virus neutralization titres from 1 : 20 to 1 : 160 (data not shown).

Immunization of S-FLU-infected piglets favours an IgG anti-TNP response

Table 1 and Fig. 1 show that among the 15 S-FLU-infected and immunized piglets, nine had positive IgM anti-TNP titres but 13 had positive IgG anti-TNP titres. Uninfected controls gave no anti-TNP response (Groups 92 and 96; Table 1; Fig. 1). Responsiveness was not correlated with the type of TNP conjugate used for immunization (Table 1). We also found that two of eleven non-immunized but S-FLU-infected animals (Groups 91, 100, 103) had IgM anti-TNP titres and one of eleven had an IgG anti-TNP titre.

Influenza infection elevates all immunoglobulin levels in serum

Figure 2 shows that IgG, IgA and IgM remained near newborn levels in non-infected control piglets after 5 weeks (28 dpi), and as values for immunized control piglets (Group 92) were indistinguishable those for from non-immunized control piglets (Group 96) (Fig. 1) we pooled the data in preparing Fig. 2. Hence, exposure to dietary protein had no effect on immunoglobulin levels. The increase in IgA levels between 25 dpi and 28 dpi in GF controls was not significant (Fig. 2). The effect of S-FLU infection on serum IgG and IgM levels was already apparent at 13 dpi (> 10-fold) and on IgA (twofold; Figure 2). Serum IgM declined after 13 dpi, but IgG and IgA levels were sustained. The decline in IgM could reflect a primary response coupled with the short serum half-life (2·8 days) of IgM.[25]

S-FLU infection preferentially elevates IgG in BAL

Only IgA was detectable in the BAL of uninfected controls after 5 weeks and this comprised < 0·1% of BAL fluid protein (Table 2). At the time of necropsy in infected animals, IgA was elevated 5-fold to 30-fold. IgG first appeared after infection and levels were fourfold higher than those for IgA. IgM could also be detected in BAL of infected piglets but levels were fourfold lower than IgA levels and 5-fold to 17-fold lower than IgG levels. In all cases, the proportion of immunoglobulin in the total protein recovered was 10-fold higher than in GF controls (Table 2). Apart from virus neutralization titres, neither anti-S-FLU nor anti-TNP could be detected in BAL.

Table 2. Immunoglobulins (µg/ml) and total protein and immunoglobulin (Ig) ratios in bronchoalveolar lavage
GroupaInfectedbImmunizationIgAIgGIgMTotal protein% Ig
  1. a

    Each group contained four piglets reared in the same isolator (see Table 1).

  2. b

    Anti-FLU was not detected by ELISA and all animals were haemagglutination inhibition negative.

  3. c

    The ratio of IgA : IgM when IgM is assigned a value of 1.

  4. d

    The ratio of IgG : IgM when IgM is assigned a value of 1.


0·6 ± 0·2


0·0 ± 0·0


0·0 ± 0·0


605 ± 265


100YesNone1·1 ± 0·61·3c4·4 ± 1·4 4·92d0·9 ± 0·61·0

804 ± 427



6·0 ± 2·8


24·8 ± 15·1


1·4 ± 1·2


444 ± 140



6·1 ± 2·0


16·2 ± 9·1


1·4 ± 0·6


504 ± 7



2·3 ± 1·3


7·0 ± 2·8


0·4 ± 0·2


590 ± 98



Immunization with irrelevant antigens contributes to increased serum immunoglobulin levels

Figure 2 shows that the exposure to model immunogens of S-FLU-infected piglets further increased IgG and IgA levels throughout the study period but increased IgM only at 13 dpi. IgG and IgA were elevated as much as 400 μg/ml and 3 μg/ml, respectively, in TNP-immunized versus non-immunized S-FLU-infected piglets. IgA levels increased after booster immunization on 25 dpi, but this small increase was also seen in controls; no IgA responses to S-FLU or TNP were detected (see above).

IgG anti-S-FLU levels increase after resolution of infection

We followed the response kinetics of 12 infected piglets throughout the study and found that the number of seropositive piglets progressively increased with time after infection (Fig. 3). Only one piglet was seropositive at 13 dpi, about half at 20 dpi and all but two at 25 dpi. All were seropositive at necropsy. Hence, both serum IgG levels and anti-FLU increased after resolution of the infection at 7–10 dpi. This suggests that resolution of the primary infection is most likely cell-mediated and independent of adaptive humoral immune responses.


Figure 3. The relationship between serum IgG levels and IgG anti-swine influenza virus (S-FLU) antibodies among a panel of 12 S-FLU-infected piglets, some immunized and some not. The type of immunogen used is indicated. Solid symbols indicate that anti-S-FLU was detected. Except for 25 and 28 days post-infection, mean IgG levels did not differ among immunized and non-immunized piglets (see Fig. 2).

Download figure to PowerPoint

Most IgG in S-FLU-infected and immunized piglets is not specific to S-FLU or TNP

Table 3 summarizes the amount of anti-NP recovered from pooled sera from different treatment groups and anti-TNP when sufficient sample was available. This study was limited because an average of 0·5–1·0 ml serum was collected from each animal and most of this had been used in other assays. When volume was limited, samples were not applied to the TNP column (Table 3). These results show that with the exception of Group 93, the amount of antibody recovered was equal to or less than the saturation limit. In the few cases in which the pooled samples were applied to both columns, the total of the amount of anti-TNP and anti-NP was < 10% of the amount of IgG applied.

Table 3. Estimation of anti-trinitrophenyl (TNP) and anti-nuclear protein (NP) in serum using affinity chromatography
GroupaIgG added(mg)Amount recovered (mg)bProportion of IgG
  1. a

    Sera were pooled to provide adequate volume for application to affinity column.

  2. b

    Measured as total protein at optical density at 280 nm.

  3. c

    I.S. = Insufficient volume to apply to both columns.

1000·63I.S.?  I.S.?

Influenza infection significantly effects IgG3 transcription

IgG3 transcripts account for the vast majority of IgG transcripts in peripheral lymphoid tissues from fetal and newborn piglets but not in spleen or peripheral blood mononuclear cells.[10, 19] Influenza virus infection resulted in a highly significant decrease in relative IgG3 expression in the TBLN and surprisingly also in the IPP, whereas the decrease in the MLN was not significantly different from that seen in GF control piglets (Fig. 4). There is little IgG3 expression in the spleen at birth and this does not change in S-FLU infection. Unrelated to S-FLU, relative IgG3 levels increased slightly in the spleen of GF piglets compared with newborns (Fig. 4).


Figure 4. Scattergram showing the proportional recovery of IgG3 transcripts among all IgG transcripts in newborn (NB) piglets, isolator-reared versus germ-free piglets (GF) and swine influenza virus (S-FLU)-infected isolator piglets. Each point represents one animal and the horizontal bar indicates the mean. Mean relative transcript levels in S-FLU-infected animals were compared with both NB and GF animals. *Significant; **highly significant; ***very highly significant.

Download figure to PowerPoint

Diversification of the VH repertoire is confined to downstream Cγ genes

For this analysis, we selected tissues from piglets in the GF and the infected groups. Table 4 shows that the RDI for IgM, IgG3 and other IgG remained low in all tissues of GF piglets (0·5–2·0) except for ‘other IgG’ in the IPP. The degree of diversification of VH genes transcribed with IgM and IgG3 did not significantly increase in the tissues of S-FLU-infected piglets (P = 0·16–0·67; Table 4). IgG3 transcription was too low in the spleen of infected animals to allow successful recovery and cloning. In contrast, the VH repertoire associated with downstream Cγ genes (called ‘other’) was elevated fivefold in the TBLN, spleen and MLN (P = 0·037 to P = 0·005) but not in the IPP (P = 0·15). The RDI in the IPP was already elevated in GF animals. Because of the slight elevation of serum IgA levels and up to a 10-fold increase in IgA on 28 dpi, we examined ~ 196 IgA VH clones from IPP and TBLN. The RDI values were similar to those for IgM in GF piglets (Table 4). Because of this outcome we did not pursue analysis of additional IgA transcripts.

Table 4. Repertoire Diversification Index (RDI) at 28 days post-infectiona
Treatment GroupTissueIgAIgMIgG3Other IgG
  1. GF, germ-free; IPP, ileal Peyer's patches; MLN, mesenteric lymph nodes; S-FLU, swine influenza virus; Spl, spleen; TBLN, tracheal bronchial lymph nodes.

  2. a

    The P values comparing S-FLU and GF piglets are presented in the bottom portion of this table.

  3. b

    Number of VDJ clones examined for each isotype is given in parenthesis.

Germ-freeIPP 0·46 (196)b1·39 (141)4·40 (119)
MLN 0·63 (115)1·30 (222)0·63 (159)
Spl 0·43 (165)1·62 (135)1·04 (267)
TBLN 0·35 (227)2·32 (235)1·87 (197)
(mean; total clones)b  0·47 (703)1·66 (733)1·98 (742)
S-FLUIPP0·5 (196)1·02 (127)1·30 (84)4·92 (151)
MLN 0·52 (227)2·46 7(175)2·25 (149)
Spl 0·26 (171) 6·77 (279)
TBLN0·39 (196)0·55 (218)1·50 (82)8·88 (276)
(mean; total clones)b 0·59 (743)1·75 (341)5·70 (855) 
P value
Germ-free/S-FLUIPP 0·1610·15
MLN 0·670·340·005
Spl 0·233 0·0088
TBLN 0·1240·780·037


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Immunoresponsiveness of newborn piglets has been shown to be dependent on contact with PAMPs, either provided by colonizing bacteria or administered as purified PAMPs.[2, 3] The adjuvant property of bacteria has been known and used for > 50 years to boost immune responses.[26, 27] In this study we tested whether infection with S-FLU that generates dsRNA, another intrinsic adjuvant,[4] would have the same effect. Compared with the responsiveness of colonized piglets to the same model antigens, responses were extraordinarily weak although 13/15 immunized and infected piglets made IgG antibodies to these irrelevant model antigens compared with 2/11 non-immunized piglets (Table 1). Neither IgM nor IgG anti-TNP was detected in immunized control Group 92 and the serum immunoglobulin levels in piglets from non-immunized controls (Grp 96) remained near newborn levels throughout the study (Figs. 1 and 2). This confirms that immune competence for adaptive responses to TNP requires adjuvant i.e. PAMPs, that are not provided by dietary antigen.

The adjuvant effect of S-FLU infection measured in terms of an antibody response was weak compared with colonized piglets immunized in the same manner,[2, 3] yet it was sufficient to cause class switch recombination to IgG encoded by downstream Cγ genes in the heavy chain locus.[17] Furthermore, these IgGs had a diversified repertoire (Fig. 4). Both are features of adaptive immunity and processes generally believed to involve germinal centre formation.[28]

The adaptive immune response to TNP and S-FLU occurred within the backdrop of a very large increase in all immunoglobulins in immunized and non-immunized S-FLU-infected piglets. However, < 10% of the IgG could be accounted for as anti-TNP and anti-S-FLU (Table 3). We interpret the remaining 90% to be the result of polyclonal B-cell activation, a phenomenon known to occur with viral infections.[29-32] Adjuvants like complete Freund's, purified PAMPs like CpG-ODN (bacterial DNA surrogate) or lipopolysaccharide can alone produce a substantial increase in immunoglobulin levels.[3, 27, 32] CpG-ODN administered to isolator piglets can raises IgM levels 10-fold.[3] Repeated immunization of young rabbits can increase IgG levels to > 20 mg/ml yet < 1–2 mg/ml (10%) are antigen-specific and cattle hyperimmunized with killed Pasteurella can have > 50 mg/ml serum IgG whereas < 5 mg/ml is antigen-specific (Butler et al. unpublished data). As a viral example, IgG levels can be elevated 100-fold by porcine-reproductive-and-respiratory-syndrome-virus-infected isolator piglets but < 1% is virus-specific.[30] These polyclonal activators are also associated with the appearance of antibodies that can recognize autoantigens and ubiquitous environmental chemicals such as analogues of TNP.[31] Presumably this explains the presence of anti-TNP in non-immunized piglets (Table 1; Fig. 1). We believe that polyclonal B-cell activation is a means to non-specifically expand the pre-immune repertoire, which can then display a broad array of B-cell receptors and antibodies for selection and further diversification as part of adaptive immunity.[31, 33] The non-IgG3 population we described fits this category. Figure 2 suggests that the hapten carriers Ficoll and KLH also exert a polyclonal effect (Fig. 2). This is well known for Ficoll, which like lipopolysaccharide has been used for in vitro proliferation.[27] Polyclonal B-cell activation may also result in antibodies that may recognize S-FLU and so aid in early resolution of the infection.

Our use of affinity chromatography to test this hypothesis has some caveats. As the NP for the S-FLU used is not available, we used a recombinant from A/Puerto Rico/8/34/Mount Sinai H1N1. Whereas NP is conserved, one could argue that this cross-reactive NP did not capture all the anti-S-FLU anti-NP. Furthermore, it would not capture antibodies to the haemagglutinins. Our lack of success using a ‘whole FLU’ affinity column is probably because the labile haemagglutinins were denatured in the process and too little NP remained. Assuming that we recovered only one-quarter to one-third of the anti-S-FLU, this still leaves 80% of the increase in IgG unaccounted for, consistent with our interpretation that it arises from polyclonal B-cell activation.

The IgG and IgA in BAL could not be explained as anti-S-FLU or anti-TNP. If < 5% of the IgG is anti-S-FLU and IgG levels in BAL are 45-fold lower (Table 2; Fig. 2), this could explain our failure to detect these antibodies. These immunoglobulins could be derived from the TBLN, which are enlarged in S-FLU-infected piglets, by polyclonal activation or be the result of inflammation-induced transudation from serum.[35, 36] The infection is resolved by 7–10 dpi, but lesions may persist for 2 weeks, making transudation a possibility. The IgA produced in the lung/bronchi can also contribute to serum IgA levels,[37] which might explain the rise in serum IgA levels on 28 dpi in infected animals. However, a small rise was also seen in GF controls (Fig. 2).

Of particular interest in this study is the switch to expression of downstream Cγ gene expression in the TBLN and IPP of S-FLU-infected piglets. This switch is apparent by the significantly reduced proportional transcription of IgG3 in the IPP and TBLN (Fig. 4). Our data show that pre-immune IgG3 cells do not expand and diversify their repertoire but rather infection results in their relative disappearance (Fig. 4). This seems to suggest that IgG3, which is encoded immediately downstream from the Cμ–Cδ gene complex,[17] acts in a similar manner to IgM in providing ‘natural antibodies’. The shift away from IgG3 expression in the TBLN suggests that IgG3 does not play a dominant role in the IgG secreted by plasma cells, which accounts for the increase seen in the serum and BAL of S-FLU piglets at necropsy (Fig. 2; Table 2). This IgG, we believe, is encoded by downstream Cγ subclass genes such as IgG1.[10] Proving that the IgG response to S-FLU in serum and BAL is indeed IgG1, or some other downstream Cγ gene, requires reliable subclass-specific antibodies, which are not currently available; the anti-IgG1 currently marketed is both pan-specific and allotype-biased.[38]

It is surprising that the IPP are also affected during S-FLU infection because S-FLU is not known to infect the gastrointestinal tract. Much aerosol-administered antigen in rabbits is ingested[39] so we suspect that during S-FLU infection, virus is shed from the nasal cavity and ingested. Surprisingly, we did not see an effect in the MLN although many B cells in the MLN are believed to be immigrants from gut-associated lymphoid tissues like the IPP.[40] Perhaps ingested virus is sufficient to initiate early steps in adaptive immunity such as class switch recombination in B cells in the IPP (Fig. 4) but insufficient to cause B-cell migration to the MLN. Alternatively, events in the IPP may be a type of antigen-independent development. However, our studies indicate that development of the B-cell repertoire in the IPP is antigen dependent.[2, 41]

Resolution of primary viral infections like S-FLU is generally believed to be dependent on cytotoxic T cells, which eliminate virus-infected cells.[42-45] This study does not challenge that view, especially since virus-specific IgG antibodies appear after resolution of the infection (Fig. 3). Of course, natural antibodies such as IgG3 could be involved in the resolution phase of the infection at 7–10 dpi. Humoral immunity can aid in termination of the infection[46] and can provide long-lasting sterilizing immunity in mice and humans.[46] We believe that antibodies that emerge after resolution of the infection are likely to be important in viral neutralization during secondary infections. Based on relative transcription and repertoire diversification, these IgG antibodies are encoded by downstream Cγ genes, not by IgG3.

The variation in responsiveness to S-FLU and TNP among immunized and non-immunized piglets (Figs. 1 and 3) serves as further evidence that outbred farm pigs, even when reared in a highly controlled setting, give different responses to viral pathogens and irrelevant model antigens. It also suggests that choosing a model antigen for studies in fetal piglets and newborns, for which natural antibodies are rare, may be difficult.

From a practical standpoint, the weak response to irrelevant antigens provides insight into events during the ‘critical window’[6] but may be somewhat academic in a conventional setting because newborn piglets given killed vaccines have already been colonized. As reviewed in the Introduction, this study was not undertaken for vaccine purposes. Nevertheless, if nasal or other intubated vaccines are used, such as ‘FluMist’, the addition of a bacterial-based adjuvant to the vaccine such as CpG-ODN, might improve efficacy.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

We are grateful to Zachary Bergman, University of Wisconsin, for assistance with ELISA and to Deb Adolphson, National Animal Disease Center, Ames IA and Marek Sinkora, Czech Academy of Science for assistance with collection and processing of samples from isolator piglets. Mention of trade names or commercial products in this article is solely for the purpose of providing specific information and does not imply recommendation of endorsement by the US Department of Agriculture. The University of Iowa and USDA are equal opportunity providers and employers.

Research Support

This study was supported by NSF-IOS grant 0077237; Biological Mimetics contract NBCHC080090; USDA-AFRI-NIFA/DHS subcontract_#2010-39559-21860 and grant 07-210 from the National Porkboard.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • 1
    Butler JE, Sun J, Weber P, Francis D. Antibody repertoire development in fetal and neonatal piglets III. Colonization of the gastrointestinal tracts results in preferential diversification of the pre-immune mucosal B-cell repertoire. Immunology 2000; 100:11930.
  • 2
    Butler JE, Weber P, Sinkora M, Baker D, Schoenherr A, Mayer B, Francis D. Antibody repertoire development in fetal and neonatal piglets VIII. Colonization is required for newborn piglets to make serum antibodies to T-dependent and type 2 T-independent antigens. J Immunol 2002; 169:682230.
  • 3
    Butler JE, Francis D, Freeling J, Weber P, Sun J, Krieg AM. Antibody repertoire development in fetal and neonatal piglets IX. Three PAMPs act synergistically to allow germfree piglets to respond to TI-2 and TD antigens. J Immunol 2005; 175:677285.
  • 4
    Cunnington PG, Naismith JD. Naturally occurring double-stranded RNA and immune responses III. Immunogenicity and antigenicity in animals. Immunology 1975; 29:100117.
  • 5
    Butler JE, Sinkora M. The isolator piglet: a model for studying the development of adaptive immunity. Immunol Res 2007; 39:3351.
  • 6
    Butler JE, Sinkora M, Wertz N, Holtmeier W, Lemke CD. Development of the neonatal B- and T-cell repertoire in swine: implications for comparative and veterinary immunology. In: Haverson K, Bailey M, Charley B, guest eds. Mucosal Immunity in Domestic Animals. Vet Res 2006; 37: 417441
  • 7
    Brambell FWR. The Transmission of Passive Immunity from Mother to Young. Amsterdam: North-Holland. Publ. Co, 1970
  • 8
    Butler JE. Immunoglobulins of the mammary secretions. In: Larson BL, Smith V, eds. Lactation, a Comprehensive Treatise. New York: Academic Press, 1974:21755.
  • 9
    Butler JE, Sun J, Weber P, Ford SP, Rehakova Z, Sinkora J, Lager K. Antibody repertoire development in fetal and neonatal piglets IV. Switch recombination, primarily in fetal thymus occurs independent of environmental antigen and is only weakly associated with repertoire diversification. J Immunol 2001; 167:323949.
  • 10
    Butler JE, Wertz N. Antibody repertoire development in fetal and neonatal piglets XVII. IgG subclass transcription revisited with emphasis on new IgG3. J Immunol 2006; 177:54809.
  • 11
    Casali P, Schettino EW. Structure and function of natural antibodies. In: Potter M, Rose NR, eds. Immunology of Silicone. Berlin: Springer Verlag, 1996; 16779.
  • 12
    Ochsenbein AF, Zinkernagel R. Natural antibodies and complement link innate and acquired immunity. Immunol Today 2000; 1:62430.
  • 13
    Butler JE, Weber P, Wertz N. Antibody repertoire development in fetal and neonatal pigs XIII. “Hybrid VH genes” and the pre-immune repertoire revisited. J Immunol 2006; 177:545970.
  • 14
    Butler JE, Sun X-Z, Wertz N, Lager KM, Chaloner K, Urban J Jr, Nara P, Tobin G. Antibody repertoire development in fetal and neonatal piglets XXI. VH usage remains constant during development in fetal piglets and postnatally in pigs exposed to environmental antigen. Mol Immunol 2011; 49:48394.
  • 15
    Butler JE, Lager KM, Splichal I et al. The piglet as a model for B cell and immune system development. Vet Immunol Immunopathol 2009; 128:14770.
  • 16
    Lutz C, Ledermann B, Kosco-Vilbois MH, Ochsenbein AF, Zinkernagel RM, Kohler G, Brombacker F. IgD can largely substitute for loss of IgM function in B cells. Nature 1998; 393:797801.
  • 17
    Eguchi-Ogawa T, Sun X-Z, Wertz N et al. Antibody repertoire development in fetal and neonatal piglets XI. The relationship of VDJ usage and the genomic organization of the variable heavy chain locus. J Immunol 2010; 184:373442.
  • 18
    Butler JE, Wertz N, Deschacht N, Kacskovics I. Porcine IgG: structure, genetics and evolution. Immunogenetics 2009; 61:20930.
  • 19
    Kloep A, Wertz N, Mendicino M, Butler JE. Linkage haplotype for IgG and IgA subclass genes. Immunogenetics 2012; 64:469473.
  • 20
    Gauger PC, Vincent AL, Loving CL, Lager KM, Janke BH, Kehrlie ME Jr, Roth JA. Enhanced pneumonia and disease in pigs vaccinated with an inactivated human-like (d-cluster) H1N2 vaccine and changed with pandemic 2009 H1N1 influenza virus. Vaccine 2011; 29:27129.
  • 21
    Ciacci-Zanella JR, Vincent AL, Prickett JR, Zimmerman SM, Zimmerman JJ. Detection of anti-influenza A nucleoprotein antibodies in pigs using a commercial influenza epitope-blocking enzyme-linked immunosorbent assay developed for avian species. J Vet Diagn Invest 2010; 22:39.
  • 22
    Peterman JH, Butler JE. Application of theoretical considerations to the analysis of ELISA data. Biotechniques 1989; 7:60815.
  • 23
    Sun J, Butler JE. Molecular characteristics of VDJ transcripts from a newborn piglet. Immunology 1996; 88:3319.
  • 24
    Butler JE, Weber P, Sinkora M, Sun J, Ford SJ, Christenson R. Antibody repertoire development in fetal and neonatal piglets II. Characterization of heavy chain CDR3 diversity in the developing fetus. J Immunol 2000; 165:69997011.
  • 25
    Klobasa F, Werhahn E, Butler JE. Regulation of humoral immunity in the piglet by immunoglobulins of maternal origin. Res Vet Sci 1981; 31:195206.
  • 26
    Freund J. Some aspects of active immunization. Annu Rev Microbiol 1947; 1:291308.
  • 27
    Johnson AG, Gaines S, Landy M. Studies on the O antigen of Salmonella typhosa V. Enhancement of antibody response to protein antigens by the purified lipopolysaccharide. J Exp Med 1956; 103:22546.
  • 28
    Bachmann MF, Zinkernagel RM. Neutralizing antiviral B cell responses. Annu Rev Immunol 1997; 15:22570.
  • 29
    Hunziker L, Recher M, Macpherson AJ, Ciurea A, Freigang S, Hengartner H, Zinkernagel RM. Hypergammaglobulinemia and autoantibody induction mechanisms in viral infection. Nat Immunol 2003; 4:3439.
  • 30
    Lemke CD, Haynes JS, Spaete R, Adolphson D, Vorwald A, Lager K, Butler JE. Lymphoid hyperplasia resulting in immune dysregulation is caused by PRRSV infection in pigs. J Immunol 2004; 172:191625.
  • 31
    Bouvet JP, Dighiero G. From natural polyreactive autoantibodies to a la carte monoreactive antibodies to infectious agents: is it a small world after all. Infect Immun 1998; 66:14.
  • 32
    Baumgarth N, Tung JW, Herzenberg LA. Inherent specificities in natural antibodies: a key to immune defense against pathogen invasion. Semin Immunopathol 2005; 26:347362.
  • 33
    Bernasconi NL, Traggial E, Lanzavecchia A. Maintenance of serological memory by polyclonal activation of human memory B cells. Science 2002; 298:2199202.
  • 34
    Tortorella D, Gewurz BE, Furman MH, Schust DJ, Ploegh HL. Viral subversion of the immune system. Annu Rev Immunol 2000; 18:861926.
  • 35
    Butler JE, Kiddy CA, Pierce CS, Rock CA. Quantitative changes associated with calving in the levels of bovine immunoglobulins in selected body fluids I. Changes in the levels of IgA, IgG1 and total protein. Can J Comp Med 1972; 36:23542.
  • 36
    Butler JE, Richerson HB, Kopp WC, Suelzer MJ, Barfknecht C. The influence of muramyl dipeptide on secretory immunity. In: Kunkel H, Mestecky J, McGhee J, eds. The Secretory Immune System. New York: Academy of Science, 1983; 409:66987
  • 37
    Gorin AB, Gould J. Immuoglobulin syntheses in the lung and caudal mediastinal lymph node of sheep. J Immunol 1979; 123:133942.
  • 38
    Butler JE, Wertz N, Sun X-Z, Lunney JK, Myyldermans S. Resolution of an immunodiagnostic dilemma: heavy chain chimeric antibodies for species in which plasmocytomas are unknown. Mol Immunol 2013; 53:1408.
  • 39
    Willoughby JB, Willoughby WF. In vivo responses to inhaled proteins I. Quantitative analysis of antigen uptake, fate and immmunogenicity in a rabbit model system. J Immunol 1977; 119:213746.
  • 40
    Liu L M, MacPherson GG. Lymph-borne (veiled) dendritic cells can acquire and present intestinally administered antigens. Immunology 1991; 73:2816.
  • 41
    Butler JE, Mateo K, Sun X-Z, Wertz N, Sinkora M, Harvey R, Francis DL. Antibody repertoire development in fetal and neonatal piglets XX. The ileal Peyer's patches are not a site of B cell lymphogenesis and are not required for systemic B cell proliferation and Ig synthesis. J Immunol 2011; 187:51419.
  • 42
    Doherty PC, Christensen JP. Accessing complexity: the dynamics of virus-specific T cell responses. Annu Rev Immunol 2000; 18:56182.
  • 43
    Eichelberger M, Allan W, Zijlstra M, Jaenisch R, Doherty PC. Clearance of influenza virus respiratory infection in mice lacking class I histocompatibility complex-restricted CD8+ T cells. J Exp Med 1991; 174:87580.
  • 44
    Stambas J, Huillineau C, Kedzierska K, Mintern JD, Doherty PC, La Gruta NL. Killer T cells in influenza. Pharmacol Ther 2008; 120:18696.
  • 45
    Slifka MK, Antiia R, Whitemore JK, Ahmed R. Humoral immunity due to long-lived plasma cells. Immunity 1998; 8:36372.
  • 46
    Couch RB, Kasel JA. Immunity to influenza in man. Annu Rev Microbiol 1983; 37:52949.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
imm12018-sup-0001-TableS1.docWord document31KTable S1. Primers and probes employed.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.