Antibody Production Expression But Does Not ...

This information is current as of May 31, 2013.

Protective Intestinal Anti-Rotavirus B Cell Immunity Is Dependent on α4β7 Integrin Expression But Does Not Require IgA Antibody Production Nelly A. Kuklin, Lusijah Rott, Ningguo Feng, Margaret E. Conner, Norbert Wagner, Werner Müller and Harry B. Greenberg J Immunol 2001; 166:1894-1902; ; http://www.jimmunol.org/content/166/3/1894

Subscriptions Permissions Email Alerts

This article cites 29 articles, 19 of which you can access for free at: http://www.jimmunol.org/content/166/3/1894.full#ref-list-1 Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscriptions Submit copyright permission requests at: http://www.aai.org/ji/copyright.html Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/cgi/alerts/etoc

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 9650 Rockville Pike, Bethesda, MD 20814-3994. Copyright © 2001 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606.

Downloaded from http://www.jimmunol.org/ by guest on May 31, 2013

References

Protective Intestinal Anti-Rotavirus B Cell Immunity Is Dependent on ␣4␤7 Integrin Expression But Does Not Require IgA Antibody Production1 Nelly A. Kuklin,2* Lusijah Rott,† Ningguo Feng,* Margaret E. Conner,‡ Norbert Wagner,§ Werner Mu¨ller,¶ and Harry B. Greenberg*†

R

otavirus (RV)3 infection is the most common cause of severe gastroenteritis in young children and results in high mortality in less developed countries. Viral replication in humans, as well as in mice, generally remains localized to the villus epithelial cells in the small intestine. The intestinal localization suggests that anti-RV immunity is probably mediated by effector cells in the gut. In fact, oral RV infection generates a large amount of anti-RV-specific IgA-producing cells in the lamina propria (LP), and virus-specific IgA Abs in the stool have been correlated with protection in humans (1, 2) and mice (3). However, the direct role of intestinal IgA Abs in protection from RV infection is not fully understood. For example, studies in rabbits have correlated anti-RV protection with intestinal IgG rather than IgA (4). Additionally, IgA-deficient (IgA⫺/⫺) mice can resolve RV and are protected from reinfection (5).

*Department of Microbiology and Immunology, and †Veterans Affairs Hospital, Palo Alto Health Care System, Palo Alto, CA 94305; ‡Department of Molecular Virology and Microbiology, Baylor College of Medicine, Houston, TX 77030; §Institute of Genetics, University of Cologne, Cologne, Germany; and ¶Department of Pediatrics, University of Bonn, Bonn, Germany Received for publication August 30, 2000. Accepted for publication November 1, 2000. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was supported by National Institutes of Health Grant R37AI21362, a Veterans Affairs Merit Review Grant (to H.B.G.), and National Institutes of Health Training Grant 5 T32 AI07328-12 (to N.K.). 2 Address correspondence and reprint requests to Dr. Nelly A. Kuklin, WP 16-214C P.O. Box 4, Merck Research Laboratories, West Point, PA 19486. E-mail address: [email protected] 3 Abbreviations used in this paper: RV, rotavirus; RRV, rhesus RV; mRV, murine RV; MLN, mesenteric lymph node(s); LP, lamina propria; SFC, spot forming cell(s); EC, epizootic diarrhea of infant mice cambridge.

Copyright © 2001 by The American Association of Immunologists

Local IgG or IgA Ab production depends in large part on the ability of B cells to migrate to and remain in the intestine. Lymphocyte homing to the gut is facilitated by interaction of ␣4␤7 integrin with vascular cell adhesion molecule-1 (Mad-CAM-1) (6 – 8). The majority of RV-specific, as well as total, IgA-producing cells in the spleen express high levels of ␣4␤7 integrin. Our recent observations suggested that thymic epithelial CC chemokine (TECK), which is selectively expressed by epithelial cells in the small intestine, is also involved in intestinal trafficking of IgAproducing cells (E. Bowman, N. A. Kukli, K. R. Youngman, N. Lazarus, E. J. Kunkel, J. Pan, H. Greenberg, and E. C. Butcher, manuscript in preparation). Past studies have demonstrated the involvement of ␣4␤7 integrin expression in lymphocyte homing to mucosal lymphoid tissues (9) and lymphocyte recirculation through the gastrointestinal tract and presumably the LP (6, 10, 11). However, the role of lymphocyte targeting signals in mediating immune effector function against intestinal pathogens has not been fully characterized. The mouse model of RV infection is a particularly useful system in this regard because virus replication is restricted to small intestinal enterocytes. Our previous studies, using RV mouse model and adoptive transfer experiments, provided initial evidence for the importance of ␣4␤7 integrin expression in B cell-mediated anti-RV immunity (12). However, interpretation of these studies was limited by the fact that ␣4␤7 could be up-regulated following adoptive transfer. Hence, the absolute requirement for ␣4␤7 in mediating B cell function could not be directly measured. To determine directly whether ␣4␤7 is necessary in mediating B cell anti-RV functions we used ␤7-deficient mice (␤7⫺/⫺) (unable to express ␣4␤7 integrin). Specifically, we tested whether ␣4␤7⫺/⫺ B cells could migrate to the intestine and resolve RV infection. This work extends our knowledge based on in vitro and short-term in vivo studies by testing the role of ␣4␤7 integrin expression during an ongoing infection in the 0022-1767/01/$02.00


Rotavirus (RV) is the main cause of severe gastroenteritis in young children; protection has been correlated with intestinal Ab responses. Using a mouse model of RV infection and ␤7-deficient (␤7ⴚ/ⴚ) mice, which do not express ␣4␤7 integrin, we demonstrated the importance of ␣4␤7 integrin in B cell-mediated anti-RV immunity. ␤7ⴚ/ⴚ mice acutely infected with murine RV resolved infection and developed normal serum IgG Abs but had diminished intestinal IgA responses. ␣4␤7ⴚ/ⴚ immune B cells did not resolve RV infection when adoptively transferred into RV-infected Rag-2-deficient mice. Fewer RV-specific B cells were found in the intestine of Rag-2-deficient mice transferred with ␤7ⴚ/ⴚ B cells compared with wild type. The absence of ␣4␤7 expression and/or a lower frequency of IgA-producing cells among transferred ␤7ⴚ/ⴚ B cells could have accounted for the inability of these cells to resolve RV infection following passive transfer. To distinguish between these possibilities, we studied the importance of IgA production in RV infection using IgA-deficient (IgAⴚ/ⴚ) mice. IgAⴚ/ⴚ mice depleted of CD8ⴙ T cells were able to clear primary RV infection. Similarly, adoptive transfer of immune IgAⴚ/ⴚ B cells into chronically infected Rag-2-deficient mice resolved RV infection. We further demonstrated in both wild-type and IgAⴚ/ⴚ mice that, following oral RV infection, protective B cells reside in the ␣4␤7high population. Our findings suggest that ␣4␤7 integrin expression is necessary for B cell-mediated immunity to RV independent of the presence of IgA. The Journal of Immunology, 2001, 166: 1894 –1902.

The Journal of Immunology intestine. In addition, we used IgA-deficient (IgA⫺/⫺) mice to distinguish between the importance of intestinal homing integrin ␣4␤7 expression and IgA production in B cell-mediated anti-RV immunity. Our results demonstrate that intestinal anti-RV B cell immunity is dependent on ␣4␤7 integrin expression, whereas IgA Ab production is not an absolute requirement for effector function.

Materials and Methods Viruses Stocks of wild-type (wt) murine RV (mRV) strain epizotic diarrhea of infant mice cambridge (EC) were prepared as intestinal homogenates and the infectious titer (diarrhea dose 50) was determined by infecting suckling mice as previously described (13).

Mice

Oral virus inoculation Oral immunizations were performed as follows: 3- to 5-wk-old ␤7⫺/⫺, Rag-2, IgA⫺/⫺, and C56BL/6 mice were orally gavaged with 5 ⫻ 105 diarrhea dose 50 of mRV strain EC after receiving 100 ␮l of 1.33% sodium bicarbonate to neutralize stomach acid. Rag-2 mice (used as recipients for adoptive transfer) were infected 1– 4 mo before use in the cell transfer studies. Stools were collected 2 wk post viral inoculation of Rag-2-deficient mice to confirm the establishment of chronic infection.

Detection of RV Ag For detection of virus, Ag sandwich ELISA was conducted as described previously (13). Briefly, microtiter plates (Dynatech, McLean, VA) were coated with guinea pig anti-rhesus RV (anti-RRV) serum and subsequently blocked with 5% nonfat dry milk. Stool samples were suspended as a 10% suspension (weight/volume) (Tris, 10 mM NaCl, 0.5 mM CaCl2 containing 5% FC, 0.05% Tween 20, 0.02% sodium azide, 1% protease inhibitors), added to the plates, and incubated overnight at 4°C. Ag was detected with rabbit anti-RRV serum, followed by HRP-conjugated goat anti-rabbit serum (Kirkegaard & Perry Laboratories, Gaithersburg, MD). ABTS substrate (Kirkegaard & Perry Laboratories) was used for color development, and the reaction was stopped with 10% SDS. Plates were read at 405 nm using an Autoreader (Bio-Tek, Burlington, VT). The total fecal virus Ag shedding was calculated as previously described (3). RV Ag shedding in the stool of infected mice was always analyzed together with specimens from uninfected, naive mice as negative controls. The mean OD from stool samples of naive mice was ⫽ 0.070. Therefore, we determined that samples with OD values ⱖ0.150 (twice that of naive mice) were considered as positive for RV Ag.

Detection of anti RV Abs Virus-specific Abs were detected using a standard ELISA. Plates were first coated as described for Ag detection and then incubated overnight at 4°C with a 1:5 dilution of RRV stock. After washing, 10% stool suspensions or specific serum dilutions as indicated were added to the plates and incubated overnight at 4°C. Ab was detected with HRP-conjugated anti-mouse IgA or IgG (Kirkegaard & Perry Laboratories). Stools or serum from noninfected mice were used as negative controls. The concentration of anti-RV Abs in serum or feces was determined by running an IgA standard in each individual plate as described previously (17). Briefly, three rows per plate were coated with purified goat anti-mouse IgA (Kirkegaard & Perry Laboratories) followed by blocking with 5% dry nonfat milk and washing with PBS 0.05% Tween 20. A standard of 250 ng/ml of purified mouse IgA isotype (PharMingen, La Jolla, CA) was serially diluted and added to the plate. The Abs were detected by using anti-mouse IgA conjugated to HRP as described above.

B cell purification, FACS sorting, and adoptive transfer experiments RV-immune ␤7⫺/⫺, IgA⫺/⫺, and wt mice were used as donors for adoptive transfer of B cells into chronically infected Rag-2 mice. Thirty days following oral infection with mRV strain EC, spleens from the immune mice were harvested and cell suspensions were made using a sterile cell strainer (40 ␮m; Fisher Scientific, Pittsburgh, PA). The splenocytes were washed with DMEM supplemented with 10% FBS (DMEM-10) and the RBC were lysed using lysing buffer (8.3 g/L ammonium chloride in 0.01 M Tris-HCL buffer, pH 7.5). Splenocytes were depleted of CD8⫹ T cells using antiCD8-conjugated beads (Dynabead, Dynabead, NY) and subsequently stained with FITC-labeled anti-IgD (PharMingen) and PE-labeled antiB220 (PharMingen). Gates for cell sorting were set on small PE-labeled lymphocytes, and FITC-labeled cells were gated out. The cells were sorted twice using a modified FACStar (Becton Dickinson, San Jose, CA) with a single 488-mm argon laser and three fluorescence detectors. The purity was 98.8 ⫾ 0.4% after the first sort and ⱖ99.9% after the second sort. Sorted cells were resuspended in sterile HBSS, and 1 ⫻ 106 ␤7⫺/⫺, IgA⫺/⫺, or wt B220⫹ cells were injected i.p. into chronically infected Rag-2 mice. Additionally, in some experiments 30,000 immune CD4⫹ T cells were double-sorted by FACS and coinjected with the wt or IgA⫺/⫺ B cells. Chronically infected Rag-2-deficient mice were also adoptively transferred with only purified CD4⫹ T cells as a negative control. The ability of the transferred cells to resolve chronic RV infection in Rag-2-deficient mice was determined by measuring viral shedding in the stools of the recipients. Additional experiments were designed to evaluate the anti-RV function of ␣4␤7high B220⫹ IgD⫺ and ␣4␤7low B220⫹ IgD⫺ cell populations purified from RV-immune mice. Splenocytes from 30 day immune ␤7⫺/⫺, IgA⫺/⫺, or wt mice were depleted of CD8⫹ T cells by incubation with anti-CD8 beads (Dynabead) following the manufacturer’s instructions. Subsequently the cells were stained with FITC-labeled anti-IgD, antiGR-1, anti-Mac-1, anti-CD3 (PharMingen), APC-labeled anti-␣4␤7 (DAKT 32), and PE-labeled anti-B220 (PharMingen). Gates were set on small PE-labeled lymphocytes, and FITC-labeled cells were gated out. The ␣4␤7high B220⫹ IgD⫺ and ␣4␤7low B220⫹ IgD⫺ cells were sorted by threecolor sorting using a modified FACSstar with filters for FITC detection (530/30), for PE detection (585/42), and for Red 613 detection (630/22). Some of the cell sorting experiments were performed using a FACSVantage equipped with an argon and Helium Neon laser and board pass filters. The purity of the single-sorted ␣4␤7high B220⫹ IgD⫺ cells was usually ⱖ98.6%. The ␣4␤7low B220⫹ IgD⫺ were sorted twice. Their purity was ⱖ99.5% after the second sort. The sorted cells were resuspended in DMEM-10 and used for Ab enzyme-linked immunospot (ELISPOT) analysis (described in detail in the section below) or were resuspended in sterile saline solution, and 10,000 B220⫹, ␣4␤7high, or B220⫹, ␣4␤7low lymphocytes were injected i.p. into chronically infected Rag-2 mice. The ability of the transferred cells to resolve chronic RV infection was determined by measuring RV shedding in the stools of recipient mice.

Determination of the frequency of RV-specific Ab-producing cells by ELISPOT The method used was described in detail previously (17). Splenocytes from orally immunized animals were depleted of CD8⫹ T cells using anti-CD8conjugated beads (Dynabead) and subsequently stained with FITC-labeled anti-IgD, anti-GR-1, anti-Mac-1, anti-CD3 (PharMingen), APC-labeled anti␣4␤7 (DAKT 32), and PE-labeled anti-B220 (PharMingen). Gates were set on small PE-labeled lymphocytes, and FITC-labeled cells were gated out. The ␣4␤7high B220⫹ IgD⫺ and ␣4␤7low B220⫹ IgD⫺ were sorted by threecolor sorting as described above. Millipore 96-well filtration plates with Imobilon-p membranes (Millipore, Bedford, MA) were coated overnight at 4°C with anti-mouse IgG, IgM, or IgA (Kirkegaard & Perry Laboratories) at a concentration of 2 ␮g/ml in 100 ␮l of carbonate buffer for detection of total Ab-secreting cells (ASC). For detection of anti-RV ASC, ELISPOT plates were coated with VP2 and VP6 virus-like particles (VLPs) made from baculovirus recombinants expressing heterologous bovine RV VP6 and VP2 (provided by J. Cohen, Cedex, France) (18) at a concentration of 5 ␮g/ml in 50 ␮l of TNC and incubated at 4°C overnight. The following day the plates were washed with PBS and blocked with DMEM-10% serum for 1 h at 37°C. Two hundred microliters of 2 ⫻ 105/ml sorted ␣4␤7high B220⫹ IgD⫺ or ␣4␤7low B220⫹ IgD⫺ cells were added and were serially (2-fold) diluted into the ELISPOT plates. After 24 h of incubation in a vibration-free incubator, the ELISPOT plates were washed and anti-mouse IgA, IgG, or IgM conjugated to HRP (Kirkegaard & Perry Laboratories) diluted 1:10,000 in PBS 1% FBS was added to the plates. After a 1-h


C57BL/6 mice (wt) were obtained from Charles River Breeding Laboratories (Hollister, CA), ␤7 knockout (␤7⫺/⫺) mice (C57BL/6 background) were produced by Norbert Wagner (Institute for Genetics, University of Cologne, Cologne, Germany) as previously described (14), and Rag-2deficient (Rag-2) mice were obtained from Taconic Farms (Germantown, NY). IgA-deficient (IgA⫺/⫺) mice were produced as previously described (15, 16). All mice were bred in the Palo Alto Veterans Administration vivarium. Mice were routinely tested for RV Abs (or RV shedding for Rag-2-deficient mice) before infection and were negative.

1895

1896 incubation at 37°C, the plates were washed and visualized with the precipitating substrate, 3-amino-9-ethylcarbazole. Spots were counted using a dissecting microscope (Stereomaster; Fisher).

Determination of the frequency of RV-specific memory B cells using FACS analysis Splenocytes from RV-immune ␤7⫺/⫺ or wt mice were stained with FITClabeled anti-IgD, anti-GR-1, anti-Mac-1, anti-CD3, and anti-IgM, APClabeled anti-␣4␤7 (DAKT 32); and PE-labeled anti-B220. Biotin-conjugated VLPs incubated with streptavidin PerCP were used to detect RVspecific cells. VLPs were produced as previously described (18) and conjugated with biotin (Chomoprobe, Mountain View, CA). Splenocytes from naive mice and second stage controls (no VLP added) were used as negative controls. Gates were set on small PE-labeled lymphocytes, and FITC-labeled cells were gated out.

Isolation of lymphocytes from the spleen, mesenteric lymph node (MLN), and LP Lymphocytes from spleen and MLN were isolated as previously described (12). LP lymphocytes were isolated from the ␤7⫺/⫺, wt, and recipient Rag2-deficient mice by dispase digestion of intestine from which Peyer’s patches had been removed (12).

C57BL/6 (wt) or ␤7⫺/⫺ mice were treated with ascites fluid containing the rat anti-mouse CD8 mAb 2.43 as previously described (19). In brief, each mouse received 0.5 ml of ascites fluid i.p. 5, 4, 3, and 2 days before RV infection, on the day of RV infection, and on days 3, 6, and 9 after infection. Successful depletion was observed in all experiments and ⬍0.4% CD8⫹ T cells were detected in lymphocytes purified from spleen, MLN, or intraepithelial lymphocytes (IEL) at day 30 from the time of virus inoculation of the experiment (data not shown). All flow cytometry data were analyzed with CellQuest program (Becton Dickinson).

FIGURE 1. Humoral immune responses in ␤7⫺/⫺ and wt mice. A, Levels of anti-RV intestinal IgA in orally infected ␤7⫺/⫺ and wt mice determined by ELISA (six mice per group were tested at each time point) (OD 405 nm). B, Serum and stool anti-RV IgA and IgG Ab responses in ␤7⫺/⫺ and wt mice measured 25 days following oral infection with RV. Stool samples were diluted 1:4 (weight/volume) and serum samples were diluted 1:1000. Each group consists of six mice. C, Frequency of anti-RV IgA and IgG Ab-producing spleen cells determined by ELISPOT 25 days following RV infection. Data are presented as Ab SFC/1 ⫻ 106 splenocytes. D, Frequency of anti-RV IgA and IgG Ab-producing MLN cells determined by ELISPOT 25 days following RV infection. Data are presented as Ab SFC/1 ⫻ 106 MLN cells. The figures (C and D) represent one of three experiments performed with similar results. The SD values are based on four mice per group.

Results ␤7⫺/⫺ animals have a diminished anti-RV humoral immune response in the intestine We evaluated the ability of ␤7⫺/⫺ mice to generate systemic and mucosal humoral immune responses to oral RV infection. ␤7⫺/⫺ and wt mice were orally infected with RV, and the level of RVspecific IgA in the stool was monitored daily by ELISA (Fig. 1A and Table I). Anti-RV IgA was detected in the feces of wt mice as early as 8 days post virus inoculation (Fig. 1A), which correlates with the time of resolution of primary RV infection in ␤7⫺/⫺ and wt mice (data not shown and Ref. 19). Anti-RV IgA responses in the stool and serum of ␤7⫺/⫺ mice were lower that in wt mice (Fig. 1A and Table I). Despite the impaired intestinal anti-RV IgA responses (Fig. 1 and Table I), serum anti-RV IgG titers measured in ␤7⫺/⫺ mice were comparable to wt mice (measured on day 25, Fig. 1B, and on day 20, Table I) post viral inoculation. We further evaluated the ability of ␤7⫺/⫺ mice to generate humoral immunity to RV by determining the number of anti-RV IgG and IgA spot forming cells (SFC) in the spleen and MLN 25 days after oral RV infection. Although comparable numbers of anti-RV IgG SFC were detected 25 days after viral infection in the spleen and MLN of ␤7⫺/⫺ and wt mice, significantly lower numbers of anti-RV-specific IgA SFC were detected in MLN of ␤7⫺/⫺ mice ( p ⬍ 0.05 measured by Student’s t test) (Fig. 1, C and D). The frequency of anti-RV IgA-producing cells in the spleen of ␤7⫺/⫺ mice was lower than wt control animals, but this difference was not significant ( p ⫽ 0.24 using Student’s t test).


In vivo CD8⫹ T cell depletion

MECHANISMS OF INTESTINAL IMMUNITY TO ROTAVIRUS

The Journal of Immunology

1897

Table I. Systemic and mucosal Ab titers of RV-infected ␤7⫺/⫺ and wt mice Serum Ab (ng/ml)

Expt. 1 Expt. 2

⫺/⫺

␤7 wt ␤7⫺/⫺ wt

Intestinal Ab (ng/ml)

IgA (day 20)

IgG (day 20)

IgA (day 10)

IgA (day 20)

10 ⫾ 5 443 ⫾ 198 85 ⫾ 15 301 ⫾ 120

5,231 ⫾ 1,817 7,427 ⫾ 579 7,960 ⫾ 2,905 16,860 ⫾ 3,379

5⫾2 23 ⫾ 7 nd nd

11.9 ⫾ 8 108 ⫾ 30 9⫾5 48 ⫾ 7

Identification of RV-specific memory B cells among resting IgD⫺ B220⫹ splenocytes of RV-immune ␤7⫺/⫺ or wt mice using FACS analysis

FIGURE 2. FACS analyses of RV-specific cells among resting IgD⫺ B220⫹ splenocytes from RV-immune ␤7⫺/⫺ or wt mice. Splenocytes from immune animals 30 days post infection were stained with FITC-labeled anti-IgD, anti-GR-1, anti-Mac-1, anti-CD3, and anti-IgM; APC-labeled anti␣4␤7 (DAKT 32); and PE-labeled anti B220. Biotin-conjugated (VP2/VP6) VLP followed by streptavidin PerCP were used to detect RV-specific cells. Gates were set on small PE-labeled lymphocytes, and FITC-labeled cells were gated out. Splenocytes from naive mice were used as negative controls. In second stage negative controls streptavidin PerCP was added to the immune lymphocytes without adding the second stage biotinylated VLPs. Numbers indicate percentage of positive cells in the indicated quadrant. Two experiments with similar results were performed, and five mice were used per group in each experiment.

Immune B cells from wt but not ␤7⫺/⫺ mice are able to resolve RV infection when transferred into chronically infected Rag-2deficient mice We next evaluated the ability of immune splenic ␤7⫺/⫺ B cells to resolve RV infection when transferred into chronically infected Rag-2-deficient mice. Oral RV inoculation of Rag-2-deficient mice, which lack functional immune lymphocytes, results in a chronic infection in which RV Ag is shed in the stool indefinitely. As previously demonstrated, adoptive transfer of wt immune B220⫹ IgD⫺ (memory) cells into RV-infected Rag-2-deficient mice resulted in the resolution of chronic RV infection (Fig. 3A and Ref. 12). In contrast, Rag-2-deficient mice reconstituted with immune ␤7⫺/⫺ B cells continued to chronically shed virus as did the untreated controls (Fig. 3A). Virus was still detected in the stools of the ␤7⫺/⫺-treated Rag-2 mice 30 days following adoptive transfer, the time when the experiments were terminated (data not shown). Resolution of viral infection in Rag-2 mice treated with wt B cells correlated with the appearance of anti-RV IgA in the stool (day 14 post transfer) (Fig. 3B). Significantly lower levels of RVspecific IgA were detected in the stool of Rag-2 mice treated with ␤7⫺/⫺ B cells (Fig. 3B). In contrast, Rag-2-deficient mice transferred with ␤7⫺/⫺ B cells had comparable levels of serum anti-RV IgG (titers of anti-RV serum IgG measured 30 days following adoptive transfer of immune B cells was 1:1800 ⫹ 1006 for Rag2-deficient mice transferred with IgA⫺/⫺ B cells and 1:2000 ⫹ 800 for Rag-2-deficient mice transferred with wt B cells; in both groups n ⫽ 4). Consistent with the low intestinal Ab responses, a low frequency of anti-RV IgA SFC was detected in the LP of Rag-2 recipients of ␤7⫺/⫺ B cells (Fig. 3C). One-tenth as many anti RV IgA SFC were detected in the LP of Rag-2 mice reconstituted with ␤7⫺/⫺ B cells vs Rag-2 mice reconstituted with wt B cells (4.6 vs 55 SFC, respectively, Fig. 3C). Splenocytes from Rag-2-deficient mice reconstituted with ␤7⫺/⫺ immune B cells had ⬃3-fold fewer anti-RV IgA SFC than splenocytes from wt B cell recipients (Fig. 3C). We also measured the frequency of RV-specific B cells in the recipient Rag-2 mice using flow cytometry. The percentages of VLP⫹ B220⫹ IgD⫺ (memory) B cells detected in the spleens of recipients were similar (in both small lymphocyte and in blast cell populations) irrespective of whether the donor was wt or ␤7⫺/⫺ (Fig. 4). However, significantly fewer VLP⫹ IgD⫺ B220⫹ cells were


We have recently developed a FACS-based method to quantify the numbers of RV-specific B cells induced in mice or humans following RV infection (K. R. Youngman, M. Franco, N. A. Kuklin, H. B. Greenberg, and E. C. Butcher, manuscript in preparation). This assay method allowed us to measure the percentage of RVspecific resting IgD⫺ (memory) B220⫹ cells in the spleen and MLN of wt and ␤7⫺/⫺ mice following RV infection (Fig. 2). Because the majority of the humoral anti-RV immune response has been shown to be directed against VP6 (20) we used biotinylated VP2/VP6 VLPs made from baculovirus recombinants expressing heterologous bovine RV proteins (18) to identify anti-RV specific immune memory B cells. We first determined the percentage of VLP⫹ B220⫹ IgD⫺ small lymphocytes in the spleen of ␤7⫺/⫺ and

wt mice 25 days post oral infection with RV (Fig. 2). Splenocytes from naive animals were used as negative controls, and the percentage of VLP⫹ cells in these nonimmune mice was always ⱕ0.3%. Both ␤7⫺/⫺ and wt mice had comparable percentages of VLP-positive cells in the ␣4␤7⫺ cell population (2.4 and 2.5%, respectively). Only splenocytes from wt mice had detectable VLP⫹ B cells in the ␣4␤7⫹ population (1.8% in wt vs 0.2% in ␤7⫺/⫺ mice) (Fig. 2). Hence, RV infection induces similar percentages of ␣4␤7⫺ memory B cells in the spleen of both types of mice, but RV-specific memory B cells lack ␣4␤7 expression in ␤7⫺/⫺ mice.

1898


detected in the MLN of Rag-2-deficient mice treated with ␤7⫺/⫺ B cells (Fig. 4). Hence, the failure of ␤7⫺/⫺ B cells to resolve chronic RV infection correlated both with a decrease in Ab-secreting and memory cells in regional sites (LP and MLN) and lower intestinal IgA but not serum IgG levels in the recipients.

evaluate the role of IgA production in B cell-mediated antiRV immunity using mice deficient in the production of IgA Ab (IgA⫺/⫺). First we treated IgA⫺/⫺ mice with anti-CD8 mAb (because CD8⫹ T cells are known to play a role in the primary resolution of RV; Refs. 19, 21, 22) and tested the ability of the depleted mice (lacking both CD8⫹ T cells and IgA-producing B cells) to resolve primary RV infection. Successful CD8⫹ T cell depletion was confirmed by analyzing the splenocytes and IEL populations of the depleted mice for the presence of CD8⫹ T cells. Percentages of CD8⫹ T cells were always ⬍0.5% in the IEL and ⬍0.2% in the spleen. The CD8⫹ T cell-depleted IgA⫺/⫺ mice were subsequently infected with RV, and RV Ag shedding in the stools was measured. As expected, wt mice, not depleted of CD8 T cells, resolved primary RV infection by day 7 (Fig. 5A and Ref. 22). Untreated IgA⫺/⫺ mice cleared RV with a slight delay compared with wt mice (day 9 for IgA⫺/⫺ mice vs day 7 for wt mice) (Fig. 5, A and B). However, following primary infection, total viral Ag shedding was not different in IgA⫺/⫺ compared with wt mice ( p ⫽ 0.38 using Student’s t test). Previously, it has been shown that B celldeficient (JhD) mice, which lack IgG- and IgM- as well as IgAproducing cells, when depleted of CD8⫹ T cells, became chronically infected (19). In contrast, here we demonstrate that CD8⫹ T cell-depleted IgA⫺/⫺ mice can resolve primary RV infection (Fig. 5B). These findings suggest that, in the IgA⫺/⫺ mice, anti-RV IgG or IgM Ab-producing cells were able to mediate resolution of acute RV infection in the absence of CD8⫹ T cells. In the IgA⫺/⫺ CD8⫹ T cell-depleted mice resolution was complete but delayed (day 14 vs day 11) compared with wt controls (Fig. 5, A and B).

IgA⫺/⫺ mice depleted of CD8⫹ T cells have prolonged viral shedding but still resolve primary RV infection

Comparison of mucosal and systemic humoral anti-RV immune responses between IgA⫺/⫺ and wt mice

We reasoned that the inability of immune ␤7⫺/⫺ B cells to resolve RV infection when transferred into chronically infected Rag-2deficient mice could have been due to the absence of ␣4␤7 integrin expression on these cells and/or insufficient anti-RV IgA-producing B cells in the transferred population. Therefore, we sought to

The fact that CD8⫹ T cell-depleted IgA⫺/⫺ mice cleared RV infection suggested involvement of anti-RV IgG or IgM in the resolution of acute RV infection in these mice. We sought to determine whether compensatory enhancement of anti-RV IgG or IgM developed in the IgA⫺/⫺ mice following RV infection. Slight but

FIGURE 3. RV Ag shedding and IgA Ab responses in the stool of chronically infected Rag-2-deficient mice treated with immune ␤7⫺/⫺ or wt B cells. Continuous shedding in the stool of nontreated RV-infected Rag2-deficient mice is shown as control. A, RV Ag shedding curves measured in the stool of chronically infected Rag-2 mice adoptively transferred with 1 ⫻ 106 B cells from immune ␤7⫺/⫺ (E) or wt (F) Mice not treated with cells are presented as filled diamonds (⽧). There are three mice per group. B, Anti-RV IgA Ab levels in the stool of chronically infected Rag-2-deficient mice treated with 1 ⫻ 106 B cells from immune ␤7⫺/⫺ (E) or wt (F) mice. The experiments presented in A and B were performed three times with similar results. SD is based on three animals per group. C, Frequency of anti-RV IgA Ab-producing spleen and LP cells of chronically infected Rag-2-deficient mice adoptively transferred with immune B cells from ␤7⫺/⫺ (䡺) or wt (f) animals. There were four mice per group.


FIGURE 4. The percentage of VLP⫹ IgD⫺ B220⫹ lymphocytes determined by FACS analysis of spleen and MLN cells from chronically infected Rag-2-deficient mice 30 days following adoptive transfer of 1 ⫻ 106 ␤7⫺/⫺ or wt immune B cells. Cells were stained with FITC-labeled antiIgD, anti-GR-1, anti-Mac-1, anti-CD3, and anti-IgM; APC-labeled DATK 32 (anti-␣4␤7); PE-labeled anti-B220. For detection of RV-specific cells biotinylated VLP followed by streptavidin PerCP was used. Gates were set on PE-labeled small lymphocytes or PE-labeled blast cells. FITC-stained cells were gated out. Splenocytes and MLN cells from naive mice were used as negative controls, and the background was subtracted from the values presented. Data represent one of two experiments performed with similar results, and four mice per group were studied.


1899

FIGURE 5. Stool RV Ag shedding of IgA⫺/⫺ and wt mice depleted of CD8⫹ T cells with mAb. A, Stool RV Ag shedding following primary infection of undepleted (E) and CD8⫹ T cell-depleted (F) wt mice (four mice were used in each group). B, Stool RV Ag shedding following primary infection of untreated (E) and CD8⫹ T cell-depleted (F) IgA⫺/⫺ mice. Four mice were used in each group. CD8⫹ T cells were depleted as previously described (19, 22).

not significant ( p ⬎ 0.05 using a two-tailed Student’s t test) increases in RV-specific IgG or IgM Abs were detected in the serum and the stool of IgA⫺/⫺ mice compared with wt controls (measured 25 days following infection) (Fig. 6). Therefore, we did not detect a significant compensatory enhancement of either anti-RV IgG or IgM Ab in the intestinal secretions or in the serum of the IgA⫺/⫺ mice. Purified B cells from immune IgA⫺/⫺ mice can resolve RV infection when transferred into chronically infected Rag-2deficient mice Because our CD8⫹ T cell depletion studies indicated that IgGand/or IgM-producing cells can mediate resolution of primary RV infection in IgA⫺/⫺ mice, we sought to directly evaluate the ability of immune IgA⫺/⫺ B cells to resolve RV infection following transfer into RV-infected Rag-2-deficient mice. As expected, RV-infected Rag-2-deficient mice transferred with B cells plus CD4⫹ T cells from wt immune donors resolved chronic RV infection (three mice at day 10 and one mouse at day 11) (Fig. 7A and Ref. 12). Immune CD4⫹ T cells alone did not resolve RV infection of Rag-2 recipient mice (data not shown and Ref. 12). Immune wt B cells, without CD4⫹ T cell help, were also able to clear RV following adoptive transfer into RV-infected Rag-2 mice (three of four mice cleared, two on day 16 and one on day 18, Fig. 7B).

Next, we evaluated the ability of IgA⫺/⫺ immune B cells to resolve RV infection when transferred into Rag-2 recipient mice. IgA⫺/⫺ B cells coadministered with CD4⫹ T cells also cleared RV infection in recipient Rag-2-deficient mice (four of four mice by day 31 following transfer) (Fig. 7C). IgA⫺/⫺ B cells transferred alone, without CD4⫹ T cells, were less efficient with only two of four recipient mice able to clear RV infection (Fig. 7D). This finding indicates that memory B cells were able to eliminate RV infection without T cell help, although somewhat less efficiently than with CD4⫹ T cells. Our results demonstrated that IgA⫺/⫺ B cells can resolve RV infection of Rag-2 mice (Fig. 7). However, those cells resolved RV infection of the recipients with a significant delay compared with wt B cells (Fig. 7). Additionally, the total amount of fecal RV Ag shedding in Rag-2 mice transferred with IgA⫺/⫺ B cells (Fig. 7A) was much higher than the Ag shedding in Rag-2 mice transferred with wt B cells (Fig. 7C) (7.25 ⫾ 0.89, SEM, and 2.24 ⫾ 0.20, SEM, respectively; p ⬍ 0.001). These results suggest that IgA Ab production is not absolutely required for B cell-mediated resolution of RV infection. Other Ab isotypes (IgG and IgM) are sufficient to resolve RV infection in recipient mice, but those Abs alone are less efficient than IgA, IgG, and IgM together. Oral infection with RV results in generation of protective antiRV SFC primarily in the ␣4␤7high population of B cells The next question we addressed was whether the protective IgM or IgG Ab-producing cells from RV-immune IgA⫺/⫺ animals were expressing ␣4␤7 integrin on their surface. First we used FACS to separate splenocytes from immune wt or IgA⫺/⫺ mice into two cell populations: an IgD⫺ (memory), B220⫹ ␣4␤7high and an IgD⫺ (memory) B220⫹ ␣4␤7low population. Subsequently, the frequency of anti-RV ASC in these two populations was measured by ELISPOT (Table II). Second, we adoptively transferred these sorted B cell populations into RV-infected Rag-2-deficient mice


FIGURE 6. Intestinal and serum RV-specific immune responses in IgA⫺/⫺ and wt mice measured 25 days following primary infection with mRV. A, Intestinal anti-RV IgG and IgA and IgM Ab levels in IgA⫺/⫺ and wt mice (four mice per group). B, Serum anti-RV IgG and IgM Abs levels in IgA⫺/⫺ and wt mice (four mice per group).

1900


and evaluated the ability of the two transferred cell populations to resolve RV infection (Fig. 8). The analysis demonstrated that anti-RV immune effector cells were found in somewhat higher numbers in the ␣4␤7high B cell populations of both IgA⫺/⫺ and wt mice (Table II and Fig. 8). Of note, although some anti-RV specific IgG- and IgM-secreting cells were present in the ␣4␤7low populations of both IgA⫺/⫺ and wt B cells, these cells were not able to mediate resolution of RV infection when transferred into RV-infected Rag-2-deficient mice (Fig. 8 and Ref. 12). Hence, in the absence of IgA production, ␣4␤7high B cells were more efficient than ␣4␤7low B cells in mediating antiRV effect. The efficiency of the immune ␣4␤7high cells from IgA⫺/⫺ donors (3/7) to resolve RV infection appears to be lower than from wt mice (5/6), but the numbers were insufficient to demonstrate statistical difference.

Discussion The humoral immune system plays a critical role in generating protective immunity to many viral infections (20, 23, 24). For infections occurring at mucosal surfaces, such as the respiratory tract or gut, this response has both a local and systemic component with the local component frequently playing the dominant effector role (3, 23, 24). The IgA immune response is the sine quo non of local immunity, and considerable attention has been focused on the role of this Ab isotype in protection. However, recent evidence with RV, HSV, influenza, and Helicobacter pylori indicate that IgA is not required to mediate clearance or protection from infection at mucosal surfaces (5, 7, 25–27). In this study, we have investigated another important determinant of local immunity, the role of ␣4␤7 integrin in anti-RV B cell-mediated immunity. To do this, we have used a murine model of RV infection, passive cell transfer strategies, and several knockout mice deficient in either ␤7, IgA, or Rag-2 genes. We have used the mRV model because, in mice as in humans, RV replicates virtually exclusively in the mature vilus tip cells of the small

bowel, and increased knowledge of the effector mechanisms of mucosal immunity is likely to add to the development of successful vaccine strategies. It has been previously demonstrated that enteric, but not systemic, immunizations generate Ab-producing cells, the great majority of which express ␣4␤7 integrin (28, 29). In another study, the enteric homing potential of circulating lymphocytes was correlated with oral but not parental vaccinations (30). However, none of these studies directly tested the importance of ␣4␤7 integrin expression in B cell-mediated immunity in the intestine. Our previous studies demonstrated that immune splenic B or T cells expressing high, but not low, levels of ␣4␤7 appeared to be the major effectors of RV immunity following passive transfer into chronically infected Rag-2-deficient mice (12). However, passively transferred ␣4␤7low immune B cells had the ability to up-regulate ␣4␤7 integrin expression after transfer, which made definitive assessment of the role of ␣4␤7 difficult. In addition, these studies did not control for possible differences in IgA levels between ␣4␤7high and ␣4␤7low B cell populations. Here, taking advantage of the existence of ␤7-deficient mice, and B cells deficient in ␣4␤7 integrin expression, we were able, for the first time, to directly assess the role of ␣4␤7 expression in RV humoral immunity. Mice deficient in ␤7 (␤7⫺/⫺) have a clearly diminished local IgA response to RV, but a relatively normal systemic humoral response (Fig. 1 and Table I). This finding is consistent with the substantially diminished mucosal inductive sites found in these mice (14). However, despite the defect in mucosal IgA immunity ␤7⫺/⫺ mice resolve primary RV infection normally. Our previous studies indicate that this timely resolution is due to an active class I-restricted T cell response in ␤7-deficient mice.4

4 N. A. Kuklin, L. Rott, J. Darling, J. J. Campbell, M. Franco, N. Feng, W. Mu¨ller, N. Wagner, J. Altman, E. C. Butcher, and H. B. Greenberg. ␣4␤7 independent pathway for CD8⫹ T cell mediated intestinal immunity to rotavirus. Submitted for publication.


FIGURE 7. RV Ag shedding in the stool of Rag2-deficient mice chronically infected with RV. Mice were adoptively transferred with 1 ⫻ 106 B cells with or without 30,000 CD4⫹ T cells from immune IgA⫺/⫺ or wt mice. A, RV Ag shedding in the stool of chronically infected Rag-2-deficient mice adoptively transferred with 1 ⫻ 106 B cells and 30,000 CD4⫹ T cells from wt immune donor mice. B, RV Ag shedding in the stool of chronically infected Rag-2-deficient mice adoptively transferred with 1 ⫻ 106 B cells from wt immune donor mice. C, RV Ag shedding in the stool of chronically infected Rag-2-deficient mice adoptively transferred with 1 ⫻ 106 B cells and 30 000 CD4⫹ T cells from IgA⫺/⫺ immune donor mice. D, RV Ag shedding of chronically infected Rag-2-deficient mice adoptively transferred with 1 ⫻ 106 B cells from IgA⫺/⫺ immune donor mice.


1901

Table II. Frequency of anti-RV Ab SFC in ␣4␤7high and ␣4␤7low splenic cell populations of IgA⫺/⫺ or wt mice orally infected with RVa RV SFC/1 ⫻ 104 IgD⫺B220⫹ Cells

␣4␤7high

⫺/⫺

IgA wt

␣4␤7low

IgA

IgM

IgG

IgA

IgM

IgG

0⫾0 9⫾2

8⫾3 10.5

23 ⫾ 8* 16 ⫾ 4#

0⫾0 0.03 ⫾ 0.05

6⫾3 2⫾2

7 ⫾ 5* 9.5 ⫾ 2#

a Results are presented as RV SFC/1 ⫻ 104 memory IgD⫺B220⫹ cells. The data represents of one of three experiments performed with similar results. SD is based on three replicates in one experiment. ⴱ , Statistically significant difference was found comparing ␣4␤7high IgA⫺/⫺ with ␣4␤7low IgA⫺/⫺ group ( p ⬍ 0.05). # , Statistically significant difference was not found when ␣4␤7high wt and ␣4␤7low wt groups were compared ( p ⬎ 0.05) (using Student’s t test Two-Sample Assuming Equal Variances).

FIGURE 8. Resolution of RV infection of chronically infected Rag-2deficient mice transferred with ␣4␤7high expressing IgA⫺/⫺ or ␣4␤7high expressing wt B cells. IgA⫺/⫺ or wt mice were orally immunized with RV. Thirty days following RV infection B cells from the spleen of the immune animals were FACS sorted into memory (IgD-) ␣4␤7high and ␣4␤7low population. Half a million of the sorted cells were adoptively transferred into chronically infected Rag-2-deficient mice. RV Ag was measured in stool samples of recipient Rag-2-deficient mice on days 1, 30, and 35 following adoptive transfer. Neither IgA⫺/⫺ nor wt ␣4␤7low B cells were able to resolve RV infection when transferred into RV-infected Rag-2-deficient mice (data not shown).

mice. These mice were capable of clearing primary RV infection promptly (Fig. 5), as would be expected given their intact T cell immune response (5). Furthermore, removal of CD8⫹ T cells prolonged time to clearance (Fig. 5) but did not induce chronic infection as it does in B cell-deficient mice (22). We interpreted these findings to indicate that B cells producing IgG and/or IgM anti-RV Abs in the IgA⫺/⫺ mice could mediate resolution of primary infection. This conclusion is consistent with the normal serum and mucosal IgG and IgM responses to RV seen in IgAdeficient mice (Fig. 6). We went to directly demonstrate that B cells producing IgG and/or IgM (but not IgA) from RV-immune IgA⫺/⫺ mice could resolve viral shedding when transferred into chronically infected immunodeficient recipients (Fig. 7). Resolution occurred whether or not CD4-immune T cells were cotransferred. However, resolution appears faster and more consistently in the presence of CD4⫹ T cell help. Of note, in these studies using mice on the C57BL/6 background, CD4⫹ T cells alone do not have any effect on chronic shedding (data not shown and Ref. 12). Given the variability of our in vivo model, it is not possible to determine accurately whether B cells from wt mice (which included IgA-producing cells) were more efficient than B cells from IgA-deficient mice in resolving infection, although this seems likely given our initial data (Fig. 7). Further study with a greater number of mice will be required to answer that question definitively. However, it is clear that in the presence and even in the absence of IgA immune response, ␣4␤7 expression is a critical requirement of anti-RV B cell immunity (Figs. 4, 7, and 8). We studied this question by sorting splenic memory B cells from IgA-deficient and wt mice into ␣4␤7high and ␣4␤7low populations and then adoptively transferring these cells into RV-infected Rag-2-deficient mice (Table II and Fig. 8). Only those B cells expressing high levels of ␣4␤7 were capable of resolving chronic RV infection. Additionally, using total and RV-specific Ab ELISPOT, we quantified the frequency of Ag-specific cells in the ␣4␤7high and ␣4␤7low donor cell populations (from immune wt or IgA-deficient mice). In agreement with our in vivo virus clearance data, we found that the majority of the anti-RV memory B cells reside in the ␣4␤7high fraction in the IgA⫺/⫺ mice (Table II). Our findings indicate that IgA production is not an absolute requirement for B cells to resolve RV infection when transferred into chronically infected Rag-2-deficient mice and that immune IgG- and/or IgM-producing cells, if they express ␣4␤7, are capable of mediating anti-RV immunity. It is somewhat surprising that RV immune function is augmented by ␣4␤7 expression in B cells producing IgG or IgM as well as IgA. Presumably, local vs systemic production of these isotypes also enhances movement into the gut by mechanisms that may be either specific or nonspecific.


To directly evaluate the role of ␣4␤7 expression in B cell-mediated immunity we transferred immune splenic B cells from ␤7⫺/⫺ and wt mice into chronically infected Rag-2-deficient mice. ␤7⫺/⫺ mice generate splenic anti-RV memory B cells (Fig. 2) although these cells, as expected, lack ␣4␤7 expression. Of interest, when immune ␤7⫺/⫺ B cells were transferred into chronically infected RAG-2-deficient mice, they were ineffective in clearing infection (Fig. 4). The lack of efficacy correlated with a greatly reduced ability of ␣4␤7-deficient RV-specific B cells to populate the intestinal LP and MLN, but not the spleen, of the Rag-2 recipients (Figs. 3 and 4). This finding stands in contrast to studies of ␤7deficient CD8⫹ T cells, which fully retained their ability to carry out anti-RV functions in a similar passive transfer model.4 The experiments demonstrating that the number of RV-immune B cells in the gut and mesentery was reduced (whereas the total number in the spleen was not) (Fig. 3) supports the notion that efficient effector function of memory B cells requires the ability to traffic to the sites of viral replication. However, these results might be confounded by the finding that ␤7-deficient mice also have a reduced IgA, but not IgG, immune response (Fig. 1). Hence, it was possible that a diminished number of RV-specific IgA memory cells transferred from the ␤7⫺/⫺ mice, rather than the lack of ␣4␤7 on these cells, accounted for the failure of this population to clear chronic infection. To evaluate the respective role of IgA isotype and ␣4␤7 integrin expression in RV B cell-mediated immunity we used IgA⫺/⫺

1902

Acknowledgments We thank Marta Raygoza for helping with the animal work and Sally Morefield for secretarial assistance.

References

15. Mbawuike, I. N., S. Pacheco, C. L. Acuna, K. C. Switzer, Y. Zhang, and G. R. Harriman. 1999. Mucosal immunity to influenza without IgA: an IgA knockout mouse model. J. Immunol. 162:2530. 16. Harriman, G. R., M. Bogue, P. Rogers, M. Finegold, S. Pacheco, A. Bradley, Y. Zhang, and I. N. Mbawuike. 1999. Targeted deletion of the IgA constant region in mice leads to IgA deficiency with alterations in expression of other Ig isotypes. J. Immunol. 162:2521. 17. Kuklin, N., M. Daheshia, K. Karem, E. Manickan, and B. T. Rouse. 1997. Induction of mucosal immunity against herpes simplex virus by plasmid DNA immunization. J. Virol. 71:3138. 18. Gilbert, J. M., and H. B. Greenberg. 1998. Cleavage of rhesus rotavirus VP4 after arginine 247 is essential for rotavirus-like particle-induced fusion from without. J. Virol. 72:5323. 19. Franco, M. A., C. Tin, L. S. Rott, J. L. VanCott, J. R. McGhee, and H. B. Greenberg. 1997. Evidence for CD8⫹ T-cell immunity to murine rotavirus in the absence of perforin, fas, and ␥ interferon. J. Virol. 71:479. 20. Burns, J. W., M. Siadat-Pajouh, A. A. Krishnaney, and H. B. Greenberg. 1996. Protective effect of rotavirus VP6-specific IgA monoclonal antibodies that lack neutralizing activity. Science 272:104. 21. Dharakul, T., L. Rott, and H. B. Greenberg. 1990. Recovery from chronic rotavirus infection in mice with severe combined immunodeficiency: virus clearance mediated by adoptive transfer of immune CD8⫹ T lymphocytes. J. Virol. 64: 4375. 22. Franco, M. A., C. Tin, and H. B. Greenberg. 1997. CD8⫹ T cells can mediate almost complete short-term and partial long-term immunity to rotavirus in mice. J. Virol. 71:4165. 23. Ahmed, R., and D. Gray. 1996. Immunological memory and protective immunity: understanding their relation. Science 272:54. 24. Murphy, B. R. 1996. Immunization against virus disease. In Virology. B. N. Fields, ed. Lipincott-Raven Publishers, Philadelphia, p. 15. 25. Kuklin, N. A., M. Daheshia, S. Chun, and B. T. Rouse. 1998. Role of mucosal immunity in herpes simplex virus infection. J. Immunol. 160:5998. 26. Blanchard, T. G., J. G. Nedrud, and S. J. Czinn. 1999. Local and systemic antibody responses in humans with Helicobacter pylori infection. Can. J. Gastroenterol. 13:591. 27. Parr, M. B., G. R. Harriman, and E. L. Parr. 1998. Immunity to vaginal HSV-2 infection in immunoglobulin A knockout mice. Immunology 95:208. 28. Quiding-Jabrink, M., I. Nordstrom, G. Granstrom, A. Kilander, M. Jertborn, E. C. Butcher, A. I. Lazarovits, J. Holmgren, and C. Czerkinsky. 1997. Differential expression of tissue-specific adhesion molecules on human circulating antibody-forming cells after systemic, enteric, and nasal immunizations: a molecular basis for the compartmentalization of effector B cell responses. J. Clin. Invest. 99:1281. 29. Qadri, F., P. H. Makela, J. Holmgren, M. J. Albert, K. Mannoor, A. Kantele, D. Saha, M. A. Salam, and J. M. Kantele. 1998. Enteric infections in an endemic area induce a circulating antibody-secreting cell response with homing potentials to both mucosal and systemic tissues. J. Infect. Dis. 177:1594. 30. Kantele, A., J. M. Kantele, E. Savilahti, M. Westerholm, H. Arvilommi, A. Lazarovits, E. C. Butcher, and P. H. Makela. 1997. Homing potentials of circulating lymphocytes in humans depend on the site of activation: oral, but not parenteral, typhoid vaccination induces circulating antibody-secreting cells that all bear homing receptors directing them to the gut. J. Immunol. 158:574.


1. Coulson, B. S., K. Grimwood, I. L. Hudson, G. L. Barnes, and R. F. Bishop. 1992. Role of coproantibody in clinical protection of children during reinfection with rotavirus. J. Clin. Microbiol. 30:1678. 2. Matson, D. O., M. L. O’Ryan, I. Herrera, L. K. Pickering, and M. K. Estes. 1993. Fecal antibody responses to symptomatic and asymptomatic rotavirus infections. J. Infect. Dis. 167:577. 3. Feng, N., J. W. Burns, L. Bracy, and H. B. Greenberg. 1994. Comparison of mucosal and systemic humoral immune responses and subsequent protection in mice orally inoculated with a homologous or a heterologous rotavirus. [Published erratum appears in 1995 J. Virol. 69:3246.] J. Virol. 68:7766. 4. Ciarlet, M., S. E. Crawford, C. Barone, A. Bertolotti-Ciarlet, R. F. Ramig, M. K. Estes, and M. E. Conner. 1998. Subunit rotavirus vaccine administered parenterally to rabbits induces active protective immunity. J. Virol. 72:9233. 5. O’Neal, C. M., G. R. Harriman, and M. E. Conner. 2000. Protection of the villus epithelial cells of the small intestine from rotavirus infection does not require immunoglobulin A. J. Virol. 74:4102. 6. Bargatze, R. F., M. A. Jutila, and E. C. Butcher. 1995. Distinct roles of L-selectin and integrins ␣4␤7 and LFA-1 in lymphocyte homing to Peyer’s patch-HEV in situ: the multistep model confirmed and refined. Immunity 3:99. 7. Hamann, A., D. P. Andrew, D. Jablonski-Westrich, B. Holzmann, and E. C. Butcher. 1994. Role of ␣ 4-integrins in lymphocyte homing to mucosal tissues in vivo. J. Immunol. 152:3282. 8. Picker, L. J. 1992. Mechanisms of lymphocyte homing. Curr. Opin. Immunol. 4:277. 9. Williams, M. B., and E. C. Butcher. 1997. Homing of naive and memory T lymphocyte subsets to Peyer’s patches, lymph nodes, and spleen. J. Immunol. 159:1746. 10. Jalkanen, S., R. A. Reichert, W. M. Gallatin, R. F. Bargatze, I. L. Weissman, and E. C. Butcher. 1986. Homing receptors and the control of lymphocyte migration. Immunol. Rev. 91:39. 11. Mackay, C. R., D. P. Andrew, M. Briskin, D. J. Ringler, and E. C. Butcher. 1996. Phenotype, and migration properties of three major subsets of tissue homing T cells in sheep. Eur. J. Immunol. 26:2433. 12. Williams, M. B., J. R. Rose, L. S. Rott, M. A. Franco, H. B. Greenberg, and E. C. Butcher. 1998. The memory B cell subset responsible for the secretory IgA response and protective humoral immunity to rotavirus expresses the intestinal homing receptor, ␣4␤7. J. Immunol. 161:4227. 13. Burns, J. W., A. A. Krishnaney, P. T. Vo, R. V. Rouse, L. J. Anderson, and H. B. Greenberg. 1995. Analyses of homologous rotavirus infection in the mouse model. Virology 207:143. 14. Wagner, N., J. Lohler, E. J. Kunkel, K. Ley, E. Leung, G. Krissansen, K. Rajewsky, and W. Muller. 1996. Critical role for ␤7 integrins in formation of the gut-associated lymphoid tissue. Nature 382:366.
