On B Cell Cell-Stimulating Factor-3

Regulatory Effects of Novel Neurotrophin-1/B Cell-Stimulating Factor-3 (Cardiotrophin-Like Cytokine) on B Cell Function This information is current as of September 26, 2021. Giorgio Senaldi, Marina Stolina, Jane Guo, Raffaella Faggioni, Susan McCabe, Stephen A. Kaufman, Gwyneth Van, Weilong Xu, Frederick A. Fletcher, Thomas Boone, Ming-Shi Chang, Ulla Sarmiento and Russell C. Cattley

J Immunol 2002; 168:5690-5698; ; Downloaded from doi: 10.4049/jimmunol.168.11.5690 http://www.jimmunol.org/content/168/11/5690

References This article cites 34 articles, 14 of which you can access for free at: http://www.jimmunol.org/ http://www.jimmunol.org/content/168/11/5690.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists by guest on September 26, 2021

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2002 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Regulatory Effects of Novel Neurotrophin-1/B Cell-Stimulating Factor-3 (Cardiotrophin-Like Cytokine) on B Cell Function

Giorgio Senaldi,1 Marina Stolina, Jane Guo, Raffaella Faggioni, Susan McCabe, Stephen A. Kaufman, Gwyneth Van, Weilong Xu, Frederick A. Fletcher, Thomas Boone, Ming-Shi Chang, Ulla Sarmiento, and Russell C. Cattley

We describe regulatory effects that a novel neurotrophin-1/B cell-stimulating factor-3 (NNT-1/BSF-3; also reported as cardiotrophin-like cytokine) has on B cell function. NNT-1/BSF-3 stimulates B cell proliferation and Ig production in vitro. NNT-1/BSF- 3-transgenic mice, engineered to express NNT-1/BSF-3 in the liver under control of the apolipoprotein E promoter, show B cell hyperplasia with particular expansion of the mature follicular B cell subset in the spleen and the prominent presence of plasma cells. NNT-1/BSF-3-transgenic mice show high serum levels of IgM, IgE, IgG2b, IgG3, anti-dsDNA Abs, and serum amyloid A. Downloaded from NNT-1/BSF-3-transgenic mice also show non-amyloid mesangial deposits that contain IgM, IgG, and C3 and are characterized by a distinctive ultrastructure similar to that of immunotactoid glomerulopathy. NNT-1/BSF-3-transgenic mice produce high amounts of Ag-specific IgM, IgA, and IgE and low amounts of IgG2a and IgG3. Normal mice treated with NNT-1/BSF-3 also produce high amounts of Ag-specific IgE. NNT-1/BSF-3 regulates immunity by stimulating B cell function and Ab production, with preference for Th2 over Th1 Ig types. The Journal of Immunology, 2002, 168: 5690Ð5698. http://www.jimmunol.org/ novel neurotrophin-1/B cell-stimulating factor-3 BSF-3 is postulated to be the elusive CNTF II, a developmentally (NNT-1/BSF-3),2 also reported as cardiotrophin-like cy- important second ligand for the receptor for CNTF (10). A tokine (CLC), is the most recently identified member of To gain knowledge of the functional role of NNT-1/BSF-3 in the IL-6 family of cytokines (1, 2). NNT-1/BSF-3 is a 225-aa immunity we have conducted in vitro and in vivo experiments that protein with the highest homology to cardiotrophin-1 and ciliary involved the generation of NNT-1/BSF-3-transgenic mice. The re- neurotrophic factor (CNTF) (1, 2). We cloned NNT-1/BSF-3 sults of these experiments indicate that NNT-1/BSF-3 regulates cDNA from activated human T cells and found that NNT-1/BSF-3 immunity by stimulating B cell function and Ab production, with mRNA is mainly expressed in lymph nodes and spleen (1). NNT- preference for Th2 over Th1 Ig types.

1/BSF-3 shows activities typical of IL-6 family members, proving by guest on September 26, 2021 to be both a neurotrophic factor and a BSF (1). NNT-1/BSF-3 Materials and Methods induces tyrosine phosphorylation of gp130 (1, 2), leukemia inhib- Mice and reagents itory factor receptor-␤ (LIF-R␤) (1), and STAT-3 (1). In vitro, it Transgenic mice with expression targeted to the liver were prepared as supports the survival of chicken and rat embryo neurons (1, 3, 4). previously described (11). Briefly, the sequence encoding human NNT-1/ In mice, it induces serum amyloid A (SAA), potentiates the in- BSF-3 (1) was subcloned into the apolipoprotein E expression vector, duction of corticosterone and IL-6 by IL-1, and causes body which is liver specific (11). The resultant plasmid was microinjected into weight loss and B cell hyperplasia with serum IgG and IgM in- single-cell embryos from BDF1 mice following published procedures (12, crease (1). NNT-1/BSF-3 can be secreted as a heterocomplex with 13). After overnight culture, 15Ð20 two-cell embryos were transferred to the oviducts of pseudopregnant CD1 female mice. Following term preg- cytokine-like factor (CLF) (5), also reported as NR6 (6), and uses nancy, offspring were screened by PCR for the presence of the integrated the receptor complex for CNTF, which consists of CNTF-R␣, transgene and bred to generate transgenic heterozygotes. These heterozy- gp130, and LIF-R␤ (7), to initiate signal transduction (3, 8). Al- gotes (positive for the NNT-1/BSF-3 transgene), males and females, were ternatively, NNT-1/BSF-3 can be secreted as a heterocomplex with studied between 6 and 28 wk of age together with sex- and age-matched ␣ ␤ littermate controls (negative for the NNT-1/BSF-3 transgene). Expression soluble CNTF-R and uses the gp130/LIF-R receptor dimer for of the transgene was assessed by Northern blot of liver RNA and quantified signaling (4). Since mice lacking CLF or CNTF-R␣, unlike those with a PhosphorImager and the ImageQuant program (Molecular Dynam- lacking CNTF, show a lethal motor neuron defect (6, 9), NNT-1/ ics, Sunnyvale, CA). Normal mice (BALB/c females, 9Ð12 wk old) were obtained from Charles River Laboratories (Wilmington, MA). Mice were housed in rooms at constant temperature and humidity under 12-h light/ Amgen, Inc., Thousand Oaks, CA 91320 dark cycles and fed with standard laboratory diet and water ad libitum. Procedures involving animals and their care were conducted observing Received for publication October 12, 2001. Accepted for publication April 2, 2002. institutional guidelines that are in compliance with national and interna- The costs of publication of this article were defrayed in part by the payment of page tional laws and policies. Recombinant human NNT-1/BSF-3 was produced charges. This article must therefore be hereby marked advertisement in accordance in Escherichia coli as previously described (1). Before use in vitro or in with 18 U.S.C. Section 1734 solely to indicate this fact. vivo, all preparations of NNT-1/BSF-3 were checked for the presence of 1 Address correspondence and reprint requests to Dr. Giorgio Senaldi, Amgen, Inc., endotoxin and were always found to contain Ͻ4.4 endotoxin U/mg protein. Amgen Center M/S 15-2-B, 1 Amgen Center Drive, Thousand Oaks, CA 91320. E-mail address: [email protected] B cell proliferation and Ig production in vitro 2 Abbreviations used in this paper: NNT-1/BSF-3, novel neurotrophin-1/B cell-stim- The Raji human Burkitt’s B cell lymphoma cells (American Type Culture ulating factor-3; CLC, cardiotrophin-like cytokine; CLF, cytokine-like factor; CNTF, ciliary neurotrophic factor; KLH, keyhole limpet hemocyanin; MZ, marginal zone; Collection, Manassas, VA) were cultured in complete RPMI 1640 medium PAS, periodic acid-Schiff; SAA, serum amyloid A; T1, type 1 transitional; T2, type (Life Technologies, Grand Island, NY) with 10% FBS (HyClone, Logan, Ϫ3 2 transitional; TEM, transmission electron microscopy; LIF-R␤, leukemia inhibitory UT) and 25 ϫ 10 M HEPES in the presence or the absence of different factor receptor-␤. concentrations of NNT-1/BSF-3, IL-6 (Amgen), and soluble CNTF-R␣

(R&D Systems, Minneapolis, MN). Raji cell proliferation was estimated and IgE were measured in serum by ELISA. Briefly, plates were coated after 3 days of culture using the method of Mosmann (14). Mouse B cells with KLH in PBS, blocked, and added with dilutions of test samples. Cap- were purified by negative selection from the spleens of normal mice using tured anti-KLH Ab were revealed using biotinylated Ab specific for Ig a mouse B cell recovery column (Cedarlane, Hornby, Canada). Cells pu- classes and subclasses (all from Southern Biotechnology Associates (Bir- rified by this method are Ͼ90% B220ϩ (15). B cells were cultured in RPMI mingham, AL), except anti-IgE, which was from BD PharMingen (San 1640 medium with 5% FBS and 5 ϫ 10Ϫ5 M ME in the presence or the Diego, CA)), neutravidin-conjugated peroxidase (Pierce), and TMB micro- absence of different concentrations of NNT-1/BSF-3 and 2 ␮g/ml affinity- well peroxidase substrate (Kirkegaard & Perry, Gaithersburg, MD). Plates Ј purified goat F(ab )2 anti-mouse IgM (Jackson ImmunoResearch Labora- were also coated with standard preparations (IgM and IgG from Calbio- tories, West Grove, PA). Mouse B cell proliferation was estimated after 4 chem (La Jolla, CA); IgG1, IgG2a, IgG2b, and IgG3 from Southern Bio- days of culture studying [3H]thymidine incorporation, for which [3H]thy- technology Associates; and IgA and IgE from PharMingen) in PBS, midine was added to the cells during the final 18 h of culture. Mouse B cell blocked, and further processed using biotinylated Ab and other reagents as Ig production was estimated after 7 days of culture measuring IgM, IgG, described above. Anti-Pneumovax IgM were also measured in serum by and IgA in culture supernatants. ELISA. Briefly, plates were coated with Pneumovax using poly-L-lysine, blocked, and added with dilutions of test samples. Captured anti- Necropsy, histological examination, and counts of spleen, lymph Pneumovax IgM were revealed using a biotinylated anti-IgM Ab, neutra- node, and peripheral blood cells of NNT-1/BSF-3-transgenic mice vidin-conjugated peroxidase, and peroxidase substrate as described above. Plates were also coated with a standard IgM preparation (Calbiochem) and NNT-1/BSF-3-transgenic mice were sacrificed for necropsy analysis at dif- further processed as described above. OD were quantitated in a Thermo- ferent ages, ranging between 6 and 28 wk. Organs were weighed and fixed max ELISA reader (Molecular Devices, Menlo Park, CA). in formalin, embedded in paraffin, sectioned, and stained with H&E or Ig, anti-dsDNA Ab, and SAA Ig classes and subclasses were measured periodic acid-Schiff (PAS) using standard histochemical techniques. Livers in serum and culture supernatants by ELISA. Briefly, plates were coated and kidneys were also collected in OCT (Miles, Elkhart, IN), snap-frozen with capture Ab specific for Ig classes (all from Southern Biotechnology in liquid nitrogen, and cut in a cryostat. Livers were stained by an indirect Associates, except anti-IgE, which was from BD PharMingen) in PBS, Downloaded from immunoperoxidase technique using mAb against B220 and CD3 (BD blocked, and added with dilutions of test samples. Captured Ig were re- PharMingen, San Diego, CA) and avidin-biotin-peroxidase complexes. vealed using biotinylated Ab specific for Ig classes, neutravidin-conjugated Sections were finally counterstained with hematoxylin. Kidneys were peroxidase, and peroxidase substrate as described above. Serum anti- stained by direct immunofluorescence technique using FITC-conjugated dsDNA IgM and IgG were also measured by ELISA. Briefly, plates were polyclonal Ab (goat anti-mouse IgM and horse anti-mouse IgG from Vec- coated with a DNA coating solution (DNA-3000, Immunovision, Spring- tor (Burlingame, CA) and goat anti-mouse C3 from Cappel (Aurora, OH)). dale, AZ) diluted in PBS, blocked, and added with dilutions of test sam-

For transmission electron microscopy (TEM), kidneys were fixed by ples. Captured anti-dsDNA IgM and IgG were revealed as described above http://www.jimmunol.org/ perfusion or immersion. For perfusion fixation, a transcardiac catheter was for Ig. For the measurement of both Ig and anti-dsDNA IgM and IgG, employed. After a brief initial flush with saline, the vasculature was per- plates were also coated with standard preparations and further processed as fused for 20 min with a fixative consisting of 2.5% glutaraldehyde and 1% described above. SAA was measured in serum by ELISA using commer- paraformaldehyde in 0.1 M sodium cacodylate buffer, pH 7.3. Immediately cially available kits (BioSource, Camarillo, CA). OD was quantitated as after perfusion, the kidneys were removed, thinly sliced, and immersion described above. fixed at 4¡C overnight. For immersion fixation, blocks approximately 1 mm3 were placed directly into fixative overnight. The specimens were trimmed into blocks and rinsed for 24 h in 0.1 M sodium cacodylate buffer at 4¡C. The kidneys were then postfixed in 1% aqueous osmium tetroxide for 1 h, rinsed in distilled water, and dehydrated in ethanol. Following a transition to propylene oxide, the tissue blocks were infiltrated and em- by guest on September 26, 2021 bedded with an Eponate-12 epoxy resin mixture (Ted Pella, Redding, CA) and polymerized in gelatin capsules for 48 h at 60¡C. Multiple samples from each animal were semithin sectioned and stained with toluidine blue. Following light microscopic examination, the block faces were trimmed and ultra-thin-sectioned. The sections were collected onto 200-mesh cop- per grids and stained with uranyl acetate and lead citrate. The grids were examined and photographed on a CM120 transmission electron microscope (Phillips, Eindhoven, The Netherlands) operated at 60 kV. Spleens and lymph nodes (two axillary and one inguinal from each side of the body) were also collected in saline and gently disrupted by manual homogenization to yield a cell suspension. The total cell number was obtained with an H1E counter (Technicon, Tarrytown, NY). The percentages of B and T cells were derived by immunofluorescence double or triple staining and flow cytometry using FITC-, PE-, and biotin-conjugated mAb against CD3, B220, IgM, IgD, CD5, CD21, and CD23 (BD PharMingen), RED670-conjugated streptavidin (Life Technologies), and a FACScan an- alyzer (BD Biosciences, Mountain View, CA). Double staining was used for CD3 and B220 to identify T cells (CD3ϩ) and B cells (CD3ϪB220ϩ); for IgM and IgD to identify B cells (IgMϩ), B1 plus MZ plus T1 B cells (IgMhighIgDlow), T2 B cells (IgMhighIgDhigh), and B2 (mature follicular) B cells (IgMlowIgDhigh); and for IgM and CD5 to identify B1a B cells (IgMϩCD5ϩ) (16). Triple staining was used for IgM, CD21, and CD23 to identify MZ B cells (IgMϩCD21ϩCD23Ϫ) (16). Peripheral blood was collected to count cells and derive serum. Blood cells were counted with an H1E counter and by immunofluorescence and flow cytometry as described above. FIGURE 1. NNT-1/BSF-3 stimulates B cell proliferation and Ig production in vitro. a, NNT-1/BSF-3 stimulates the proliferation of Raji hu- Induction and measurement of Ag-specific Ab man Burkitt’s B cell lymphoma cells with effects similar to those of IL-6. The experiment was performed in triplicate. b, NNT-1/BSF-3 stimulation NNT-1/BSF-3-transgenic mice were immunized on day 0 by s.c. injection of Raji cells is potentiated by soluble (s) CNTF-R␣. The experiment was ␮ of 100 g keyhole limpet hemocyanin (KLH; Pierce, Rockford, IL) in CFA performed in triplicate. c, NNT-1/BSF-3 costimulates with anti-IgM Ab the along the flanks of the abdomen or by i.p. injection of 115 ␮g Pneumovax (Merck, West Point, PA). Mice were bled to detect Ag-specific Ab imme- proliferation of purified mouse spleen B cells. The experiment was per- diately before and 7, 14, and 21 days after immunization. Normal mice formed in triplicate. d, NNT-1/BSF-3 stimulates purified mouse spleen B were immunized and bled as described above and also given NNT-1/BSF-3 cells to produce IgM, IgG, and IgA. NNT-1/BSF-3 was used at 15 ␮g/ml (3 mg/Kg) or NNT-1/BSF-3 vehicle as a control i.p. for 7 consecutive days in the absence of anti-IgM Ab. Medium alone was used as a control. Su- starting on day 0. Anti-KLH IgM, IgG, IgG1, IgG2a, IgG2b, IgG3, IgA, pernatants from three culture wells were pooled for Ig measurement. 5692 REGULATORY EFFECTS OF NNT-1/BSF-3 ON B CELLS

Table II. Counts of spleen B cell subsets in NNT-1/BSF-3-transgenic micea

NNT-1/BSF-3 TG Control LM p

Total cells (no.) 226.4 Ϯ 27.2 107.2 Ϯ 13.4 Ͻ0.01 IgMhighIgDlow cells (%) 2.5 Ϯ 0.2 2.7 Ϯ 0.4 NS IgMhighIgDlow cells (no.) 5.7 Ϯ 0.9 3.0 Ϯ 0.5 Ͻ0.04 IgMhighIgDhigh cells (%) 6.0 Ϯ 0.5 6.5 Ϯ 0.7 NS IgMhighIgDhigh cells (no.) 13.8 Ϯ 2.3 6.9 Ϯ 1.1 Ͻ0.03 IgMlowIgDhigh cells (%) 39.8 Ϯ 3.0 31.1 Ϯ 1.4 Ͻ0.04 IgMlowIgDhigh cells (no.) 91.1 Ϯ 16.4 33.7 Ϯ 5.0 Ͻ0.02 IgMϩCD5ϩ cells (%) 6.9 Ϯ 0.9 6.9 Ϯ 0.7 NS IgMϩCD5ϩ cells (no.) 10.3 Ϯ 1.6 5.2 Ϯ 1.1 Ͻ0.02 IgMϩCD21ϩCD23Ϫ cells (%) 9.4 Ϯ 1.2 8.7 Ϯ 1.0 NS IgMϩCD21ϩCD23Ϫ cells (no.) 13.8 Ϯ 2.3 6.1 Ϯ 0.8 Ͻ0.01

a TG, transgenic mice; LM, littermate controls; %, percentage of total cells; no., count in millions of cells; n ϭ 7.

NNT-1/BSF-3 also costimulates the proliferation of murine pri- Downloaded from mary B cells by potentiating the proliferation-inducing effect of an anti-IgM Ab (Fig. 1c). NNT-1/BSF-3 does not stimulate the proliferation of primary B cells either alone or in conjunction with LPS or IL-4 or an anti-CD40 Ab (data not shown). Alone, however, NNT-1/BSF-3 is able to stimulate primary B cells to produce

IgM, IgG, and IgA (Fig. 1d). NNT-1/BSF-3 does not appear to http://www.jimmunol.org/ affect the proliferation of murine primary T cells either alone or in conjunction with an anti-CD3 Ab (data not shown).

Necropsy ﬁndings in NNT-1/BSF-3-transgenic mice The transgenic mice presented in this paper were engineered to express NNT-1/BSF-3 in the liver under control of the apolipoprotein E promoter and secrete it into the circulation together with plasma proteins (11, 13). Transgenic mice, i.e., positive by PCR FIGURE 2. NNT-1/BSF-3-transgenic mice (TG) show liver expression by guest on September 26, 2021 of NNT-1/BSF-3 mRNA, which is increased compared with that in litter- for the NNT-1/BSF-3 transgene, were studied together with sex- mate controls (LM) and is accompanied by high counts of spleen B cells and age-matched littermate controls, i.e., negative for the NNT-1/ and high levels of serum IgM. a, Northern blot of liver RNA revealing BSF-3 transgene. Transgenic mice show liver expression of NNT- NNT-1/BSF-3 mRNA. b, Levels of NNT-1/BSF-3 mRNA quantiﬁed with 1/BSF-3 mRNA (Fig. 2, a and b). This expression is variable, but a PhosphorImager and the ImageQuant program. c, Counts of spleen B much increased compared with that in littermate controls (Fig. 2, cells (IgMϩ). d, Levels of serum IgM. a and b) and is accompanied by the expression of the phenotype, as represented by high spleen B cell counts and high serum IgM levels (see below; Fig. 2, c and d). Statistical analysis NNT-1/BSF-3-transgenic mice show B cell hyperplasia. Results are expressed as the mean Ϯ SE. Differences between groups were Spleens from NNT-1/BSF-3-transgenic mice are larger and have analyzed by Student’s t test. higher absolute counts of total cells and B cells, but not T cells, than spleens from littermate controls (Tables I and II). Compared Results B cell proliferation and Ig production in vitro NNT-1/BSF-3 stimulates the proliferation of the Raji human Bur- Table III. Lymph node and peripheral blood cell counts in NNT-1/BSF-3- kitt’s B cell lymphoma cells in culture dose-dependently and sim- transgenic micea ilarly to IL-6 (Fig. 1a). This stimulation is potentiated by soluble CNTF-R␣ dose-dependently and with saturation effect (Fig. 1b). NNT-1/BSF-3 TG Control LM p Lymph nodes Total cells (no.) 8.6 Ϯ 2.3 9.7 Ϯ 1.8 NS Table I. Spleen weight and cell counts in NNT-1/BSF-3-transgenic CD3ϩ cells (%) 46.6 Ϯ 5.8 66.0 Ϯ 5.5 Ͻ0.05 a mice CD3ϩ cells (no.) 3.5 Ϯ 0.6 6.1 Ϯ 0.8 Ͻ0.04 CD3ϪB220ϩ cells (%) 50.0 Ϯ 5.5 32.0 Ϯ 5.4 Ͻ0.05 Ϫ ϩ NNT-1/BSF-3 TG Control LM p CD3 B220 cells (no.) 4.8 Ϯ 1.7 3.4 Ϯ 1.1 NS

Weight (mg) 275 Ϯ 22 179 Ϯ 17 Ͻ0.001 Peripheral blood Total cells (no.) 243.5 Ϯ 30.2 134.6 Ϯ 16.1 Ͻ0.02 Lymphocytes (no.) 10.9 Ϯ 1.7 10.9 Ϯ 0.7 NS CD3ϩ cells (%) 13.4 Ϯ 2.1 20.8 Ϯ 3.4 NS CD3ϩ cells (%) 14.0 Ϯ 2.1 21.2 Ϯ 0.7 Ͻ0.03 CD3ϩ cells (no.) 29.3 Ϯ 3.8 26.6 Ϯ 5.4 NS CD3ϩ cells (no.) 1.5 Ϯ 0.3 2.3 Ϯ 0.2 Ͻ0.05 CD3ϪB220ϩ cells (%) 68.8 Ϯ 2.7 62.8 Ϯ 3.1 NS CD3ϪB220ϩ cells (%) 80.0 Ϯ 2.1 72.8 Ϯ 1.3 Ͻ0.03 CD3ϪB220ϩ cells (no.) 163.5 Ϯ 25.1 81.1 Ϯ 8.1 Ͻ0.03 CD3ϪB220ϩ cells (no.) 8.8 Ϯ 1.4 8.0 Ϯ 0.6 NS

a TG, transgenic mice; LM, littermate controls; %, percentage of total cells; no., a TG, transgenic mice; LM, littermate controls; %, percentage of total cells (lymph count in millions of cells; n ϭ 5. nodes) or lymphocytes (blood); no., count in millions of cells; n ϭ 5. The Journal of Immunology 5693

FIGURE 3. NNT-1/BSF-3-transgenic mice show B cell hyperplasia. a and b, Spleens from an NNT-1/BSF- 3-transgenic mouse and a littermate control. The spleen from the transgenic mouse shows follicular hyperplasia (arrows, a). cÐf, Lymph nodes from an NNT-1/BSF-3- transgenic mouse (c) and a littermate control (d). The lymph node from the NNT-1/BSF-3-transgenic mouse shows sinusoidal hyperplasia (arrows, c) and medullary cord expansion with prominent presence of plasma Downloaded from cells (arrows, e), some of which contain eosinophilic cytoplasmic globules, also called Russell bodies (arrows, f). H&E staining. The microbar length is indicated in micrometers. http://www.jimmunol.org/ by guest on September 26, 2021

FIGURE 4. NNT-1/BSF-3-transgenic mice show anomalous lymphoid aggregates. aÐc, Kidney (a), lung (b), and liver (c) from NNT-1/BSF-3-transgenic mice showing anomalous lymphoid aggregates. d, Liver from an NNT-1/BSF-3-transgenic mouse with a lymphoid aggregate containing numerous plasma cells (arrows). e and f, Liver from an NNT-1/BSF-3-transgenic mouse showing a lymphoid aggregate largely consisting of B cells (e) and only few T cells (f). aÐd, H&E staining. e and f, Immunoperoxidase staining for B220 (e) and CD3 (f). The microbar length is indicated in micrometers. 5694 REGULATORY EFFECTS OF NNT-1/BSF-3 ON B CELLS

cÐf), consist of B cells with only a few T cells (Fig. 4, e and f), and contain a considerable number of plasma cells (Fig. 4d). NNT-1/BSF-3-transgenic mice show hypergammaglobulinemia. NNT-1/BSF-3-transgenic mice have higher serum levels of IgM and IgE, but not total IgG and IgA, than controls (Fig. 5a). Considering the individual IgG subclasses, however, NNT-1/BSF- 3-transgenic mice have higher serum levels of IgG2b and IgG3 than controls. Notably, in the NNT-1/BSF-3-transgenic mice the levels of IgM show a strong tendency to increase with age (Fig. 5b). NNT-1/BSF-3-transgenic mice also have higher serum levels of anti-dsDNA IgM and IgG than controls (Fig. 5c). NNT-1/BSF-3-transgenic mice show conspicuous deposits of amorphous material in the kidney glomeruli, which are collected within the mesangium (Fig. 6, a and b). These deposits contain glycoproteins, as indicated by their eosinophilic nature (Fig. 6, a and b) and by staining with PAS (Fig. 6, c and d). IgM, IgG, and the complement component C3 accumulate within these deposits,

as shown by immunoﬂuorescence staining (Fig. 6, eÐj). Surpris- Downloaded from ingly, these deposits do not seem to contain amyloid, as indicated by the fact that are negative for Congo Red-associated birefrin- gence (data not shown), even though NNT-1/BSF-3-transgenic mice have high levels of serum SAA (5.77 Ϯ 0.70 ␮g/ml in NNT- 1/BSF-3-transgenic mice (n ϭ 59) vs 1.12 Ϯ 0.15 ␮g/ml in con- ϭ Ͻ

trols (n 22); p 0.001). TEM provided further evidence that http://www.jimmunol.org/ these deposits do not contain amyloid. The deposits were found to consist of arrays of fibrils and tubular structures located in the mesangial and subendothelial sites within affected glomeruli (Fig. 7, aÐc). Where present, the tubular structures were approximately 30Ð50 nm in diameter (Fig. 7d). The regular arrangement and approximate diameter of these structures contrast with the ran- domly arranged 10- to 12-nm diameter fibrils described for amyloid (17). Rather, the findings were reminiscent of a collection of rare diseases reported as fibrillary and immunotactoid glomeru- by guest on September 26, 2021 lopathies (18Ð20). The presence of these unusual glomerular deposits was especially prominent in NNT-1/BSF-3-transgenic mice Ͼ22 wk of age and with marked elevations of circulating IgM. FIGURE 5. NNT-1/BSF-3-transgenic mice show high serum levels of NNT-1/BSF-3-transgenic mice show no other abnormalities other IgM, IgE, IgG2b, IgG3, and anti-dsDNA Abs. a, NNT-1/BSF-3-transgenic than those mentioned above. mice (TG) have higher serum levels of IgM and IgE, but not IgG and IgA, than littermate controls (LM). Mice were bled at the age of 6Ð28 wk (n ϭ 15). b, NNT-1/BSF-3-transgenic mice (TG) have serum IgM levels showing a stronger tendency to increase with age than littermate controls (LM; Ag-specific Ab production in NNT-1/BSF-3-transgenic mice and n ϭ 5). c, NNT-1/BSF-3-transgenic mice (TG; n ϭ 65) have higher serum in normal mice treated with NNT-1/BSF-3 levels of anti-dsDNA IgM and IgG than littermate controls (LM; n ϭ 22). Compared with littermate controls, NNT-1/BSF-3-transgenic mice Ͻ ءءء Ͻ ءء Ͻ ء aÐc, , p 0.05; , p 0.01; , p 0.001. showed increased production of Ag-specific Ab of the IgM, IgA, and IgE classes in response to immunization with the T cell-dependent Ag keyhole limpet hemocyanin (KLH; Fig. 8a). On the with controls, spleens from the transgenic mice have higher abso- other hand, compared with controls, NNT-1/BSF-3-transgenic lute counts of all B cell subsets identified (B1 ϩ MZ ϩ T1, T2, B2, mice showed decreased production of anti-KLH IgG2a and IgG3 B1a, and MZ), higher percentages of B2 cells, and unchanged percentages of the other subsets (Table II). Thus, NNT-1/BSF-3- (Fig. 8a). The production of anti-KLH total IgG, IgG1, and IgG2b transgenic mice have an enlarged population of B2 B cells in both in NNT-1/BSF-3-transgenic mice was not significantly different absolute and relative terms. Lymph nodes and peripheral blood from that in controls (Fig. 8a). Compared with controls, NNT-1/ from NNT-1/BSF-3-transgenic mice have higher percentages of B BSF-3-transgenic mice also showed increased production of Ag- cells and lower percentages of T cells than littermate controls (Ta- specific IgM in response to immunization with the T cell-indepen- ble III). Spleen and lymph nodes from NNT-1/BSF-3-transgenic dent Ag Pneumovax (Fig. 8b). Normal mice treated with NNT-1/ mice show, respectively, follicular and sinusoidal lymphoid hy- BSF-3 had increased production of anti-KLH IgE compared with perplasia with the prominent presence of plasma cells (Fig. 3, a, c, control mice treated with NNT-1/BSF-3 vehicle (Fig. 8c). Mice e, and f). Some of these plasma cells contain eosinophilic cyto- were treated for 7 days starting on the day of immunization and plasmic globules, also called Russell bodies (Fig. 3f). Various or- showed increased anti-KLH IgE production on day 14, but not on gans from NNT-1/BSF-3-transgenic mice, especially liver, kid- day 21. The response to KLH, either proliferation or production of neys, and lungs, show anomalous lymphoid aggregates (Fig. 4). IL-2, IFN-␥, IL-4, and IL-10, of the lymph node cells from NNT- These aggregates are found in the interstitium of kidneys (Fig. 4a) 1/BSF-3-transgenic mice was not significantly different from that and lungs (Fig. 4b) and mostly in the liver periportal area (Fig. 4, of controls (data not shown). The Journal of Immunology 5695 Downloaded from http://www.jimmunol.org/ by guest on September 26, 2021

FIGURE 6. NNT-1/BSF-3-transgenic mice show mesangial deposits of IgM, IgG, and C3. aÐj, Kidney glomeruli from NNT-1/BSF-3-transgenic mice (TG; right column) and littermate controls (LM; left column). aÐd, Compared with the littermate control, the glomerulus from the transgenic mouse shows in the mesangium conspicuous deposits of amorphous eosinophilic material (a and b) that are PAS positive (c and d). eÐj, The glomerulus from the transgenic mouse shows deposits of IgG (f), IgM (h), and C3 (j), while that from the littermate control is devoid of such deposits (e, g, and i). a and b, H&E staining. c and d, PAS staining. eÐj, Immunoﬂuorescence staining for IgG (e and f), IgM (g and h), and C3 (i and j). Microbars are 20 micrometers. 5696 REGULATORY EFFECTS OF NNT-1/BSF-3 ON B CELLS

FIGURE 7. NNT-1/BSF-3-transgenic mice show mesangial fibrils of distinctive ultrastructure. aÐd, TEM of kidney glomeruli of NNT-1/BSF-3-transgenic mice. a, Arrays of fibrils and tubular structures in the mesangial and subendothelial regions of the glomerulus (arrows). b, Deposits, primarily fibrils, located on the endothelial aspect of the glomerular basement membrane (arrows). c, Deposits, primarily tubular structures, located on the endothelial aspect of the glomerular basement membrane (arrows). d, Tubular structures, approximately 40Ð50 nm in diameter, arranged in bundles (arrow). a, c, and d, Immersion fixation; b, perfusion fixation. The microbar in a is 5 ␮m; those in bÐd are 1 ␮m. Downloaded from http://www.jimmunol.org/ Discussion functions within the nervous system, as indicated by its neurotro- This study illustrates the regulatory effects of NNT-1/BSF-3 on phic properties in vitro (1, 3, 4) and by the phenotype of mice immunity, demonstrating that it is an integral molecule of the im- lacking either CLF or CNTF-R␣, two molecules assisting NNT- mune system. NNT-1/BSF-3 has effects on adaptive immunity, 1/BSF-3 secretion and signal transduction, but devoid of neuro- directly stimulating B cells to proliferate and produce Ab with trophic properties per se (3Ð5, 8, 24), which consists of lethal preference of Th2 over Th1 Ig types. motor neuron deficiency (6, 9). The study of NNT-1/BSF-3-defi- NNT-1/BSF-3 is a cytokine of the IL-6 family of very recent cient mice appears at this point to be crucial to prove conclusively identification (1, 2), and little is known of its physiological signif- that NNT-1/BSF-3 is a factor necessary for the development of icance. NNT-1/BSF-3 is a gp130 activator with highest homology motor neurons (10). Studies using neutralizing Ab have the poten- by guest on September 26, 2021 to cardiotrophin-1 (1, 2), hence the denomination CLC (2). NNT- tial to uncover additional functions that NNT-1/BSF-3 may have in 1/BSF-3 has been shown to have general activities typical of IL-6 adult individuals. family members in vivo, such as induction of SAA, potentiation of We noticed that NNT-1/BSF-3, which has a conventional signal induction of corticosterone and IL-6 by IL-1, and induction of peptide (1, 2), is produced, but not readily secreted, in vitro by body weight loss (1, 21). More specifically, NNT-1/BSF-3 has mammalian cells genetically engineered for its recombinant pro- been shown to have neurotrophic properties similar to CNTF in duction (unpublished observations). It seems that NNT-1/BSF-3 vitro (1, 3, 4, 22), which suggested the denomination NNT-1 (1), needs assistance from a concomitantly expressed molecule, such as and to have effects reminiscent of IL-6 on B cells in vivo (1, 23), CLC (3) or CNTF-R␣ (4), to be secreted. The finding in this study which suggested the other denomination, BSF-3 (1). that mice engineered to express an NNT-1/BSF-3 transgene spe- To acquire further information on NNT-1/BSF-3 functions, cifically in the liver and secrete NNT-1/BSF-3 into the general transgenic mice were engineered in this study to aberrantly express circulation show a systemic phenotype consistent with the pheno- NNT-1/BSF-3 in the liver under control of the apolipoprotein E type observed in normal mice given i.p. E. coli-derived NNT-1/ promoter and to secrete it into the circulation together with most BSF-3 (1) indicates the possibility that NNT-1/BSF-3 is secreted, plasma proteins (11). These NNT-1/BSF-3-transgenic mice show a at least by the liver, independently of CLC or CNTF-R␣. In fact, phenotype consisting of B cell hyperplasia, hypergammaglobu- neither CLC nor CNTF-R␣ is expressed in the liver (5, 24). It linemia with anti-dsDNA Ab, and glomerulopathy with mesangial remains to be determined whether this occurs because the expres- Ig deposition. Interestingly, B cell hyperplasia and hypergamma- sion of the transgene abnormally overrides a postulated retention globulinemia are also the main abnormalities seen in normal mice mechanism (4) or because the liver is normally capable of secret- given a daily injection of recombinant NNT-1/BSF-3 for 7 days ing NNT-1/BSF-3, either alone or in conjunction with an as yet (1), indicating that the effects of NNT-1/BSF-3 supplementation, unknown molecule. Interestingly, we have noticed a nonnegligible because of either transgenic expression or pharmacologic admin- amount of NNT-1/BSF-3 expression in normal mouse liver (1). istration, are mainly confined to the immune system. Notably, B cell hyperplasia, hypergammaglobulinemia, and glomerulopa- however, NNT-1/BSF-3-injected mice also demonstrate body thy with mesangial Ig deposition are also the main abnormalities weight loss (1), a phenomenon not shown by NNT-1/BSF-3-trans- seen in IL-6-transgenic mice (25, 26), which significantly rein- genic mice, probably as a consequence of developmental compen- forces the functional connection between NNT-1/BSF-3 and IL-6 satory mechanisms. (1). However, beyond the similarities, there are differences be- The fact that the effects of NNT-1/BSF-3 supplementation tween the phenotypes of NNT-1/BSF-3 and IL-6-transgenic mice. mainly concern the immune system does not exclude the possibil- In NNT-1/BSF-3-transgenic mice B cell hyperplasia also involves ity that this cytokine has functional roles within other systems, cells at less mature stages than plasma cells and particularly con- especially the nervous system. NNT-1/BSF-3 may have important cerns B2 (mature follicular) B cells, while in IL-6-transgenic mice The Journal of Immunology 5697

These observations suggest that NNT-1/BSF-3 and IL-6 stimulate B cells at different developmental stages. In NNT-1/BSF-3-transgenic mice no amyloidosis is found in any organ including kidneys, even when these mice show conspicuous proteinaceous deposits in the mesangium and high levels of circulating SAA that may favor the deposition of AA amyloid. In contrast, in IL-6-transgenic mice, also engineered to express the transgene under control of a liver-specific promoter, i.e., metallothionein I (26), amyloid A amyloidosis is found in spleen, liver, and kidneys, including the mesangium, together with high levels of circulating SAA (27). On the other hand, NNT-1/BSF-3-transgenic mice show the presence in the mesangium of a peculiar type of fibrils of distinctive ultrastructure that has not been observed in IL-6-transgenic mice despite electron microscopic analysis (27). Intriguingly, the accumulation of fibrils and tubular structures in the glomeruli of NNT-1/BSF-3-transgenic mice is reminiscent of a group of rare glomerulopathies of unknown etiology described in patients with pro- teinuria and often the nephrotic syndrome and reported as fibrillary or immunotactoid glomerulopathies (17Ð20, 28, 29). Immunotactoid Downloaded from glomerulopathy is ultrastructurally distinguished from fibrillary glomerulopathy by the presence of tubular structures, although some controversy exists concerning whether these glomerulopathies should be considered distinct entities (20). The biopsy specimens from affected patients are often observed to contain Ig. These patients are

sometimes found to have an underlying lymphoproliferative disorder http://www.jimmunol.org/ possibly with dysproteinemia (18), suggesting a comparable patho- genesis of these types of human glomerulopathies to that of NNT-1/ BSF-3-transgenic mice. NNT-1/BSF-3 has stimulating effects on B cells by virtue of direct action, as shown by the experiments involving the use in vitro of the human B cell line Raji or primary B cells purified from mouse spleens. Two interesting observations were made using the Raji cells, i.e., that NNT-1/BSF-3 has growth-stimulating activity similar to that of IL-6 and that such activity can be increased by by guest on September 26, 2021 soluble CNTF-R␣. That soluble CNTF-R␣ can increase NNT-1/ BSF-3 activity is consistent with the recent reports that in addition to gp130 and LIF-R␤, NNT-1/BSF-3 needs CNTF-R␣ for signaling in either membrane-bound (3) or soluble form (4) and illustrates the possibility of modulating NNT-1/BSF-3 activity with soluble CNTF-R␣, as previously shown for CNTF (30). NNT-1/ BSF-3 does not seem to be able to affect primary B cells independently of stimulation through their Ag receptor. Thus, NNT-1/BSF-3 is likely to act physiologically only as a B cell costimulatory agent. Primary B cells are much less sensitive to NNT-1/BSF-3 than Raji cells. This may depend on various factors, including differences in the expression of NNT-1/BSF-3 receptor components by these cell types. An interesting finding of this study is the ability of NNT-1/ BSF-3 to privilege a Th2 over Th1 type of Ig production. This is FIGURE 8. NNT-1/BSF-3-transgenic mice produce high amounts of already apparent when total levels of circulating Ig of different Ag-specific IgM, IgA, and IgE and low amounts of IgG2a. a, Compared postswitch classes (IgG, IgA, and IgE) are examined, since NNT- with littermate controls (LM), NNT-1/BSF-3-transgenic mice (TG) have 1/BSF-3-transgenic mice, in addition to pre-switch IgM, show higher serum levels of anti-KLH IgM, IgA, and IgE; not significantly dif- higher levels of IgE than controls, but not of IgG or IgA (31). It ferent levels of IgG and IgG1; and lower levels of IgG2a. b, NNT-1/BSF- becomes even more apparent when Ag-specific levels of circulat- 3-transgenic mice (TG) have higher serum levels of anti-Pneumovax IgM ing Ig classes and subclasses are examined, since NNT-1/BSF-3- than littermate controls (LM). c, Normal mice treated with NNT-1/BSF-3 transgenic mice not only show higher levels of Ag-specific IgE have higher serum levels of anti-KLH IgE than controls on day 14 after than controls, but also have lower levels of IgG2a. In fact, IgG2a p Ͻ ,ءء ;p Ͻ 0.05 ,ء ,immunization. Mice were treated from days 0Ð6. aÐc are considered to be the expression of a Th1-type immune re- .p Ͻ 0.001. n ϭ 10 ,ءءء ;0.01 sponse and IgE of a Th2-type response (32, 33). Since lymph node cells from NNT-1/BSF-3-transgenic mice were not found to proliferate or produce Ig class switch-modulating cytokines, such as it involves virtually only plasma cells (25, 26). In NNT-1/BSF-3- IFN-␥ and IL-4 (34), under Ag-specific stimulation differently transgenic mice hypergammaglobulinemia is limited to a 4- to 6-fold from controls, it seems that the ability of NNT-1/BSF-3 to privi- increase in IgM and, interestingly, IgE, while in IL-6-transgenic mice lege the Ig class switch of Th2 over Th1 type is independent of it involves a characteristic increase in IgG1, up to 400-fold (25, 26). effects on T cells and, rather, reflects a direct action on B cells. 5698 REGULATORY EFFECTS OF NNT-1/BSF-3 ON B CELLS

This is supported by the observation that NNT-1/BSF-3 stimulates 13. Simonet, W. S., D. L. Lacey, C. R. Dunstan, M. Kelley, M. S. Chang, R. Luethy, B cell proliferation and Ig production in vitro. Other interesting H. Q. Nguyen, S. Wooden, L. Bennett, T. Boone, et al. 1997. Osteoprotegerin: a novel secreted protein involved in the regulation of bone density. Cell 89:309. findings of this study are that NNT-1/BSF-3 can stimulate the pro- 14. Mosmann, T. 1983. Rapid colorimetric assay for cellular growth and survival: duction of Ag-specific Ab of the IgA class and that it can stimulate application to proliferation and cytotoxicity assays. J. Immunol. Methods 65:55. 15. Yu, G., T. Boone, J. Delaney, N. Hawkins, M. Kelley, M. Ramakrishnan, Ab production against both T cell-dependent (like KLH) and in- S. McCabe, W. Qiu, M. Kornuc, X. C. Xia, et al. 2000. APRIL and TALL-1 and dependent Ag (such as Pneumovax) (35). receptors BCMA and TACI: system for regulating humoral immunity. Nat. Im- In conclusion, NNT-1/BSF-3-transgenic mice show a phenotype munol. 1:252. 16. Hardy, R. R., and K. Hayakawa. 2001. B cell development pathways. Annu. Rev. of immunological relevance. Since NNT-1/BSF-3 is strongly ex- Immunol. 19:595. pressed in lymphoid tissues (1, 2), and since NNT-1/BSF-3-treated 17. Alpers, C. E. 1994. Glomerulopathies of dysproteinemias, abnormal immuno- mice mainly show immune alterations (1), this new observation globulin deposition, and lymphoproliferative disorders. Curr. Opin. Nephrol. Hy- pertension 3:349. indicates that NNT-1/BSF-3 is an integral molecule of the immune 18. Alpers, C. E. 1992. Immunotactoid (microtubular) glomerulopathy: an entity dis- system. Future studies, possibly involving deficient mice and neu- tinct from fibrillary glomerulonephritis? Am. J. Kidney Dis. 19:185. 19. Korbet, S. E., M. M. Schwartz, and E. J. Lewis. 1994. The fibrillary glomeru- tralizing Abs, may reveal additional physiological roles. lopathies. Am. J. Kidney Dis. 23:751. 20. Nickeleit, V., S. Moll, M. Schmid, and M. J. Mihatsch. 1999. Nephrotic syndrome in an adult: the ongoing saga of fibrils versus microtubules. References Am. J. Kidney Dis. 34:1146. 21. Benigni, F., G. Fantuzzi, S. Sacco, M. Sironi, P. Pozzi, C. A. Dinarello, 1. Senaldi, G., B. C. Varnum, U. Sarmiento, C. Starnes, J. Lile, S. Scully, J. Guo, J. D. Sipe, V. Poli, M. Cappelletti, G. Paonessa, et al. 1996. Six different cyto- G. Elliott, J. McNinch, C. L. Shaklee, et al. 1999. Novel neurotrophin-1/B cell kines that share gp130 as a receptor subunit, induce serum amyloid A and po- stimulating factor-3: a cytokine of the IL-6 family. Proc. Natl. Acad. Sci. USA tentiate the induction of interleukin-6 and the activation of the hypothalamus- 96:11458. Downloaded from pituitary-adrenal axis by interleukin-1. Blood 87:1851. 2. Shi, Y., W. Wang, P. A. Yourey, S. Gohari, D. Zukauskas, J. Zhang, S. Ruben, 22. Stoekli, K. A., F. Lottspeich, M. Sendtner, P. Masiakowski, P. Carroll, R. Goetz, and R. F. Alderson. 1999. Computational EST database analysis identifies a novel D. Lindholm, and H. Thoenen. 1989. Molecular cloning, expression and regional member of the neuropoietic cytokine family. Biochem. Biophys. Res. Commun. distribution of rat ciliary neurotrophic factor. Nature 342:920. 262:132. 23. Hirano, T., T. Taga, N. Nakano, K. Yasukawa, S. Kashiwamura, K. Shimizu, 3. Elson, G. C. A., E. Lelievre, C. Guillet, S. Chevalier, H. Plun-Favreau, J. Froger, K. Nakajima, K. H. Pyun, and T. Kishimoto. 1985. Purification to homogeneity I. Suard, A. B. de Coignac, Y. Delneste, J. Y. Bonnefoy, et al. 2000. CLF asso- and characterization of human B-cell differentiation factor (BCDF or BSF-2). ciates with CLC to form a functional heteromeric ligand for the CNTF receptor Proc. Natl. Acad. Sci. USA 82:5490.

complex. Nat. Neurosci. 3:867. 24. Davis, S., T. H. Aldrich, D. M. Valenzuela, V. V. Wong, M. E. Furth, http://www.jimmunol.org/ 4. Plun-Favreau, H., G. Elson, M. Chabbert, J. Froger, O. deLapeyriere, E. Lelievre, S. P. Squinto, and G. D. Yancopoulos. 1991. The receptor for ciliary neurotrophic C. Guillet, J. Hermann, J. F. Gauchat, H. Gascan, et al. 2001. The ciliary neu- factor. Science 253:59. rotrophic factor receptor ␣ component induces the secretion of and is required for 25. Suematsu, S., T. Matsuda, K. Aozasa, S. Akira, N. Nakano, S. Ohno, J. Miyazaki, functional responses to cardiotrophin-like cytokine. EMBO J. 20:1692. K. Yamamura, T. Hirano, and T. Kishimoto. 1989. IgG1 plasmacytosis in inter- 5. Elson, G. C., P. Graber., C. Losberger, S. Herren, D. Gretener, L. N. Menoud, leukin-6 transgenic mice. Proc. Natl. Acad. Sci. USA 86:7547. T. N. C. Wells, M. H. Kosco-Vilbois, and J. F. Gauchat. 1998. Cytokine-like 26. Fattori, E., C. Della Rocca, P. Costa, M. Giorgio, B. Dente, L. Pozzi, and factor-1, a novel soluble protein, shares homology with members of the cytokine G. Ciliberto. 1994. Development of progressive kidney damage and myeloma type I receptor family. J. Immunol. 161:1371. kidney in interleukin-6 transgenic mice. Blood 83:2570. 6. Alexander, W. S., S. Rakar, L. Robb, A. Farley, T. A. Willson, J. G. Zhang, 27. Solomon, A., D. T. Weiss, M. Schell, R. Hrncic, C. L. Murphy, J. Wall, L. Hartley, Y. Kikuchi, T. Kojima, H. Nomura, et al. 1999. Suckling defect in M. D. McGavin, H. J. Pan, G. W. Kabalka, and M. J. Paulus. 1999. Transgenic mice lacking the soluble haemopoietin receptor NR6. Curr. Biol. 9:605. mouse model of AA amyloidosis. Am. J. Pathol. 154:1267. 7. Davis, S., T. H. Aldrich, N. Stahl, L. Pan, T. Taga, T. Kishimoto, N. Y. Ip, and 28. Fogo, A., N. Qureshi, and R. G. Horn. 1993. Morphologic and clinical features by guest on September 26, 2021 G. D. Yancopoulos. 1993. LIFR␤ and gp130 as heterodimerizing signal trans- of fibrillary glomerulonephritis versus immunotactoid glomerulopathy. ducers of the tripartite CNTF receptor. Science 260:1805. Am. J. Kidney Dis. 22:185. 29. Schwartz, M. M. 1993. Renal amyloidosis and other fibrillar glomerular diseases. 8. Lelievre, E., H. Plun-Favreau, S. Chevalier, J. Froger, C. Guillet, G. C. A. Elson, Curr. Opin. Nephrol. Hypertension 2:238. J. F. Gauchat, and H. Gascan. 2001. Signaling pathways recruited by the car- 30. Davis, S., T. H. Aldrich, N. Y. Ip, N. Stahl, S. Scherer, T. Farruggella, diotrophin-like cytokine/cytokine-like factor-1 composite cytokine. J. Biol. P. S. DiStefano, R. Curtis, N. Panayotatos, H. Gascan, et al. 1993. Released form Chem. 276:22476. of CNTF receptor ␣ component as a soluble mediator of CNTF responses. Sci- 9. DeChiara, T. M., R. Vejsada, W. T. Poueymirou, A. Acheson, C. Suri, ence 259:1736. J. C. Conover, B. Friedman, J. McClain, L. Pan, and N. Stahl. 1995. Mice lacking 31. Stavnezer, J. 2000. A touch of antibody class. Science 288:984. the CNTF receptor, unlike mice lacking CNTF, exhibit profound motor neuron 32. Stevens, T. L., A. Bossie, V. M. Sanders, R. Fernandez-Botran, R. L. Coffman, deficits at birth. Cell 83:313. T. R. Mosmann, and E. S. Vitetta. 1988. Regulation of antibody isotype secretion 10. Lesser, S. S., and D. Lo. 2000. CNTF II, I presume? Nat. Neurosci. 3:851. by subsets of antigen-specific helper T cells. Nature 334:255. 11. Simonet, W. S., T. M. Hughes, H. Q. Nguyen, L. D. Trebasky, D. M. Danilenko, 33. Lebman, D. A., and R. L. Coffman. 1988. Interleukin 4 causes isotype switching and E. S. Medlock. 1994. Long-term impaired neutrophil migration in mice over- to IgE in T cell-stimulated clonal B cell cultures. J. Exp. Med. 168:853. expressing human interleukin-8. J. Clin. Invest. 94:1310. 34. Romagnani, S. 1997. The Th1/Th2 paradigm. Immunol. Today 18:263. 12. Brinster, R. L., H. Y. Chen, M. E. Trumbauer, M. K. Yagle, and R. D. Palmiter. 35. Lamam, J. D., and E. Claassen. 1996. T-cell-independent and T-cell-dependent 1985. Factors affecting the efficiency of introducing foreign DNA into mice by humoral immunity. In Cytokine Regulation of Humoral Immunity: Basic and microinjecting eggs. Proc. Natl. Acad. Sci. USA 82:4438. Clinical Aspects, C. M. Snapper, ed. John Wiley & Sons, New York, p. 23.