ARTHRITIS & RHEUMATISM Vol. 56, No. 6, June 2007, pp 1798–1805 DOI 10.1002/art.22715 © 2007, American College of Rheumatology
Female Mice Are More Susceptible to Developing Inflammatory Disorders due to Impaired Transforming Growth Factor  Signaling in Salivary Glands Seshagiri R. Nandula,1 Shoba Amarnath,1 Alfredo Molinolo,1 Bidhan C. Bandyopadhyay,1 Bradford Hall,1 Corinne M. Goldsmith,1 Changyu Zheng,1 Jonas Larsson,2 Taduru Sreenath,1 WanJun Chen,1 Indu S. Ambudkar,1 Stefan Karlsson,2 Bruce J. Baum,1 and Ashok B. Kulkarni1 Objective. Transforming growth factor  (TGF) plays a key role in the onset and resolution of autoimmune diseases and chronic inflammation. The aim of this study was to delineate the precise function of TGF signaling in salivary gland inflammation. Methods. We impaired TGF signaling in mouse salivary glands by conditionally inactivating expression of TGF receptor type I (TGFRI), either by using mouse mammary tumor virus–Cre mice or by delivering adenoviral vector containing Cre to mouse salivary glands via retrograde infusion of the cannulated main excretory ducts of submandibular glands. Results. TGF  RI–conditional knockout (TGFRI-coko) mice were born normal; however, female TGFRI-coko mice developed severe multifocal inflammation in salivary and mammary glands and in the heart. The inflammatory disorder affected normal growth and resulted in the death of the mice at ages 4–5 weeks. Interestingly, male TGFRI-coko mice did not
exhibit any signs of inflammation. The female TGFRIcoko mice also showed an increase in Th1 proinflammatory cytokines in salivary glands and exhibited an up-regulation of peripheral T cells. In addition, these mice showed an atypical distribution of aquaporin 5 in their salivary glands, suggesting likely secretory impairment. Administration of an adenoviral vector encoding Cre recombinase into the salivary glands resulted in inflammatory foci only in the glands of female TGFRI– loxP-flanked (floxed) mice (TGFRI-f/f mice), but not in those of male and female wild-type mice or male TGFRI-f/f mice. Conclusion. These results suggest that female mice are uniquely more susceptible to developing inflammatory disorders due to impaired TGF signaling in their salivary glands. Transforming growth factor  (TGF), a multifunctional cytokine, regulates many important biologic processes such as embryonic development, cell proliferation and differentiation, extracellular matrix synthesis, immune response, inflammation, and apoptosis (1). Three different isoforms of TGF (TGF1, TGF2, and TGF3) are expressed in mammals, and they display overlapping expression patterns in vivo and exhibit functional similarities in vitro (2). TGF1, the prototype for the superfamily, is secreted in a latent form and undergoes proteolytic cleavage to generate an active form. The active TGF1 elicits cellular responses by first binding to TGF receptor type II (TGFRII), which in turn initiates phosphorylation of TGFRI. Subsequent phosphorylation and translocation of Smad proteins into the nucleus trigger activation of target genes (1). TGFs are ubiquitously expressed, and experiments with mice with mutations of various TGF family
Supported by the Intramural Division of the National Institute of Dental and Craniofacial Research, NIH. 1 Seshagiri R. Nandula, PhD (current address: School of Medicine, University of Virginia, Charlottesville), Shoba Amarnath, PhD, Alfredo Molinolo, MD, PhD, Bidhan C. Bandyopadhyay, PhD, Bradford Hall, MS, Corinne M. Goldsmith, BS, Changyu Zheng, MD, PhD, Taduru Sreenath, PhD (current address: Uniformed Services University of the Health Sciences, Bethesda, Maryland), WanJun Chen, MD, PhD, Indu S. Ambudkar, PhD, Bruce J. Baum, DMD, PhD, Ashok B. Kulkarni, PhD: National Institute of Dental and Craniofacial Research, NIH, Bethesda, Maryland; 2Jonas Larsson, PhD, Stefan Karlsson, MD, PhD: Lund University Hospital, Lund, Sweden. Address correspondence and reprint requests to Ashok B. Kulkarni, PhD, Functional Genomics Section, Laboratory of Cell and Developmental Biology, National Institute of Dental and Craniofacial Research, NIH, 30 Convent Drive, MSC 4395, Bethesda, MD 20892. E-mail:
[email protected]. Submitted for publication August 31, 2006; accepted in revised form March 7, 2007. 1798
IMPAIRED TGF SIGNALING IN FEMALE MICE
members and components of the TGF signaling pathway have shown their crucial roles in multiple physiologic processes (2). TGF1-null mice develop severe multifocal inflammation primarily in the lungs, heart, and salivary glands and succumb to multiorgan failure by age 3–4 weeks (3,4). Many developmental defects have been noted in TGF2-null mice, which die perinatally (5), and TGF3-null mice display delayed lung development and die shortly after birth (6,7). TGFRII-null mice die at midgestation as a result of defects in yolk sac vasculogenesis and hematopoiesis (8). Mice lacking TGFRI also die at midgestation with severe defects in vascular development of the yolk sac and placenta (9). The lethal phenotypes of these null mice indicate the critical role of TGF signaling pathways in important physiologic processes. TGF regulates immune responses involved in the induction and maintenance of immune tolerance by suppressing lymphocyte proliferation and differentiation, thereby preventing inappropriate autoimmune responses (10). Salivary gland lesions in TGF1-null mice resemble those seen in patients with Sjo ¨gren’s syndrome (SS) (11), which are characterized by focal lymphocytic infiltrations (12). Altered levels of ductal expression of TGF isoforms have been observed in patients with primary SS and in those with an autoimmune disorder in which benign lymphoepithelial lesions occur (13). However, immunohistochemical studies have yielded conflicting results on TGF levels in salivary gland epithelia from SS patients (13). A precise role of TGF signaling in the etiology of autoimmune disorders that affect salivary glands, such as SS and benign lymphoepithelial lesions, has not been characterized. In this study, we evaluated this role by impairing TGFRI expression primarily in mouse salivary glands by using the Cre/lox system. In this system, the Cre recombinase from bacteriophage P1 excises the intervening DNA sequence located between 2 unidirectional lox sites positioned on the same linear DNA segment, leaving the lox site behind, and, in our present studies, this results in the deletion of the TGFRI sequence in tissues where Cre is expressed (9). Body weight loss and early mortality were observed only in female TGFRI–conditional knockout (TGFRI-coko) mice at age 4–5 weeks. Histopathologic analysis of female TGFRI-coko mice showed multifocal inflammation in the salivary glands, mammary glands, and heart. Moreover, flow cytometric analysis revealed T cell infiltration in salivary glands of these mice, which resembled the infiltration seen in SS patients. Increased levels of interleukin-1 (IL-1), IL-2,
1799
IL-12, and interferon- ␥ (IFN ␥ ) messenger RNA (mRNA) expression were noted in the salivary glands of female TGFRI-coko mice, suggesting a trend toward a Th1-specific response rather than a Th2-specific response. Interestingly, administration into the salivary glands of an adenoviral vector encoding Cre recombinase resulted in inflammatory foci in female TGFRI– loxP-flanked (floxed) mice (TGFRI-f/f mice), but not in male and female wild-type (WT) mice or male TGFRI-f/f mice, indicating that female mice are more susceptible to autoimmune disorders due to impaired TGF signaling. MATERIALS AND METHODS Mice. TGFRI-f/f mice and mouse mammary tumor virus (MMTV)–Cre mice were generated as previously described (9,14). Mice were housed in standard cages and maintained in climate- and light-controlled rooms (mean ⫾ SD 22 ⫾ 0.5°C, 12-hour dark/12-hour light cycle) with free access to food and water. Studies were performed in compliance with the National Institutes of Health Guidelines on the Care and Use of Laboratory and Experimental Animals. All experimental procedures were approved by the Animal Care and Use Committee of the National Institute of Dental and Craniofacial Research. Identification of genotypes. Genotyping of TGFRI-f/f mice was carried out by using polymerase chain reaction (PCR) as described previously (9). The mice carrying the Cre transgene were identified by PCR as described by Wagner et al (14). Antibodies. Anti-TGFRI antibody (Santa Cruz Biotechnology, Santa Cruz, CA) was used at 1:200 dilutions for immunostaining. Anti-CD4 and anti-CD8 antibodies (1:100) were purchased from BD Biosciences (San Diego, CA). Anti– aquaporin 5 antibody (see below) was purchased from Calbiochem (San Diego, CA). Secondary horseradish peroxidase– conjugated antibodies (1:500) were obtained from Zymed (South San Francisco, CA). Fluorescein isothiocyanate (FITC)–conjugated anti-CD4, anti-CD8a, anti-CD45RB, and anti-CD19; phycoerythrin (PE)–conjugated anti-CD3; and PE- or FITC-conjugated IgG (BD Biosciences) were used as isotype controls. Histopathology and immunohistochemistry. Selected mice were killed, and complete autopsies were performed for detailed pathologic analysis. Tissues from WT and TGFRIcoko mice were fixed in 10% buffered formalin or freshly prepared 4% paraformaldehyde in phosphate buffered saline (PBS) overnight and then embedded in paraffin. Sections (5 m) were stained with hematoxylin and eosin for histopathologic study. For frozen sections, the tissue samples were mounted and embedded in OCT compound (Sakura, Torrance, CA), frozen on a bed of dry ice, and stored at ⫺70°C. Cryostat sections (8 m) were air-dried for 5 minutes at 4°C, fixed in precooled acetone for 2 minutes, air-dried for 1 hour at room temperature, and stored at –70°C. For immunostaining, slides were dewaxed with xylene, rehydrated with descending grades of ethanol, and rinsed with
1800
NANDULA ET AL
distilled water. After washing with PBS, endogenous peroxidase was blocked by incubating sections with 1% H2O2 for 20 minutes at room temperature. Slides were then washed with PBS and blocked with goat serum for 1 hour at 4°C. After washing with PBS, primary antibodies against TGFRI at 1 g/ml, against aquaporin 5 at 2 g/ml, and against zona occludens 1 (ZO-1) at 1 g/ml (Invitrogen, Carlsbad, CA), as well as anti–Na⫹/K⫹/2Cl⫺ cotransporter at 1 g/ml (15), were diluted in PBS and applied to slides, followed by incubation overnight at 4°C. Slides were then washed with PBS and incubated for 1 hour at room temperature with either biotinylated or peroxidase-conjugated antibodies. All slides were then incubated with 3,3⬘-diaminobenzidine (Zymed), washed with distilled water, counterstained with Mayer’s hematoxylin, dehydrated, and mounted. Flow cytometric analysis. Flow cytometric analysis of peripheral T cells was performed as described previously (16). Briefly, spleens were gently minced in complete Dulbecco’s modified Eagle’s medium containing 10% fetal calf serum (FCS) and washed twice in the medium, and the red blood cells were lysed using ACK cell lysis buffer (BioWhittaker, Walkersville, MD). Spleen cells were resuspended in PBS containing 5% FCS. The cells were labeled with the indicated antibody and analyzed with a FACSCalibur instrument using CellQuest software (BD Biosciences, San Jose, CA). Preparation of RNA and reverse transcriptase (RT)– PCR. Total RNA was extracted from WT mouse and female TGFRI-coko mouse salivary glands using TRIzol reagent (Invitrogen, Gaithersburg, MD). RNA (2 g) was reversetranscribed, and the complementary DNA was synthesized at 42°C for 50 minutes using oligo(dT)16 primer and Superscript II RT (Invitrogen). PCR amplification using the primers for IL-1, IL-4, IL-6, IL-12, IFN␥, and GAPDH (17) was performed for 40 cycles consisting of 94°C for 1 minute, 60°C for 45 seconds, and 72°C for 2 minutes. PCR products were size-separated by electrophoresis using 0.9% agarose gels, visualized by ethidium bromide staining, and normalized to GAPDH for quantification. Adenovirus vector. The adenovector-based Cre recombinase expression system, AdcreM1, was purchased from Microbix (Toronto, Ontario, Canada). The virus was plaquepurified, grown in large quantity, and purified by CsCl gradient centrifugation (18). The titer of the virus, determined by limiting dilution plaque assay, was 5.4 ⫻ 1012 plaque-forming units (PFU)/ml. Delivery of AdcreM1 (5 ⫻ 108 PFU in 50 l) to mouse salivary glands was carried out via retrograde infusion of the cannulated main excretory ducts of submandibular glands as described by Baum et al (19). Statistical analysis. All experiments were performed a minimum of 3 times. Statistical evaluation was done with GraphPad Prism software, version 4.0 (GraphPad Software, San Diego, CA). Significant differences between groups were assessed by Student’s unpaired t-test.
RESULTS Generation of TGFRI-coko mice. TGFRIcoko mice were generated by mating female TGFRI-f/f mice with male TGFRI-f/⫹;MMTV-Cre mice. Female
Figure 1. Body weight profile and longevity of transforming growth factor  receptor type I–conditional knockout (TR1coko) mice. A, Body weight profile of wild-type (WT) and TR1coko mice indicating normal body weight gain in female WT mice and male TR1coko mice, but reduced body weight gain in female TR1coko mice starting at age 2 weeks. Values are the mean ⫾ SD. B, Kaplan-Meier analysis of longevity of female TR1coko mice (n ⫽ 14) and WT mice (n ⫽ 30), indicating early mortality of female TR1coko mice. Male TR1coko mice displayed a lifespan similar to that of the WT mice (data not shown).
TGFRI-coko mice were born normal and did not display any phenotype until age 2 weeks, when they started showing a decline in body weight gain. At the end of 4 weeks, they weighed significantly less than controls (Figure 1A). However, male TGFRI-coko mice and female TGFRI-f/⫹;MMTV-Cre mice did not show any loss in body weight and appeared outwardly normal. Female TGFRI-coko mice died as early as age 4–5 weeks. Kaplan-Meier analysis showed significant reduction in longevity of female TGFRI-coko mice compared with WT controls (P ⫽ 0.0001) (Figure 1B). The longevity of male TGFRI-coko mice was similar to that of WT controls (data not shown). Inflammation in the salivary glands of female TGFRI-coko mice. Immunostaining for TGFRI showed fairly good expression of TGFRI in the ductal
IMPAIRED TGF SIGNALING IN FEMALE MICE
Figure 2. Inflammation in salivary glands, mammary glands, and heart of female TR1coko mice. A and B, Reduced immunostaining of transforming growth factor  receptor type I in ductal cells of salivary glands of female TR1coko mice (arrow in B) compared with that in those of WT mice (arrow in A). C–H, Representative hematoxylin and eosin–stained tissue sections from WT mice (C, E, and G) and female TR1coko mice (D, F, and H). Inflammatory foci are clearly visible in salivary glands (arrows in D), mammary glands (arrows in F), and heart (arrow in H) of female TR1coko mice. (Original magnification ⫻ 20.) See Figure 1 for definitions.
cells of salivary glands of female WT mice (Figure 2A), but not in those of TGFRI-coko mice (Figure 2B). A total of 14 female TGFRI-coko mice were subjected to a detailed histopathologic examination (Table 1). The salivary glands of female TGFRI-coko mice showed significant focal chronic lymphocytic infiltrates composed mainly of lymphocytes as well as histiocytes and scattered plasma cells (Figure 2D). No inflammatory cell infiltration was observed either in WT mice (Figure 2C) or in male TGFRI-coko mice (results not shown). In
1801
addition to salivary glands, inflammatory cell infiltration was observed in the mammary glands and heart of female TGFRI-coko mice (Figures 2F and H). The described chronic infiltration was associated with tissue damage in salivary and mammary glands and in the heart (Figures 2D, F, and H). The resulting endomyocarditis possibly contributed to early mortality in these mice. Immunostaining of the salivary gland sections from the female TGFRI-coko mice showed the presence of CD4⫹ T cells and macrophages in the focal infiltrates (results not shown). These infiltrates were mainly localized near the ducts, but some were scattered in the perivascular space. In order to characterize the type of immune response elicited in the salivary glands of female TGFRI-coko mice, we determined cytokine levels by RT-PCR on total RNA extracted from salivary glands of the WT controls and TGFRI-coko mice, using GAPDH as an internal control. Levels of mRNA for IL-1, IL-6, IL-12, and IFN␥, but not for IL-4, were increased in salivary glands of female TGFRI-coko mice (data not shown), indicating a Th1 immune response. In order to assess whether the focal inflammation seen around the salivary gland ducts of female TGFRIcoko mice affected a key acinar cell membrane protein essential for salivary fluid secretion (20,21), we performed immunostaining for aquaporin 5. As shown in Figure 3A, salivary gland sections from female WT mice displayed the typical abundant and polarized pattern of aquaporin 5 staining around the apical plasma membranes of acinar cells (15). However, salivary gland sections from female TGFRI-coko mice did not show this typical aquaporin 5 localization. Staining was considerably reduced and nonpolarized, with substantial intracellular staining evident (Figure 3B). Given the critical role that aquaporin 5 plays in salivary fluid secretion (20,21), this finding suggests the potential for abnormalities in fluid secretion in female TGFRI-coko mice. Aquaporin 5 staining of the lung sections did not reveal abnormal distribution in female TGFRI-coko mice (results not shown). We also stained salivary gland sections for a basolateral marker of acinar cells using anti–Na⫹/K⫹/2Cl⫺ cotransporter antibody and for a tight junction marker using anti–ZO-1 antibody. Interestingly, staining for ZO-1 and Na⫹/K⫹/2Cl⫺ cotransporter did not show any abnormal distribution in female TGFRI-coko mice (results not shown). Analysis of peripheral T cells in female TGFRIcoko mice. The apparent increase in the Th1 type of proinflammatory infiltrates in salivary glands of female
1802
NANDULA ET AL
Table 1. Summary of pathologic findings in female transforming growth factor  receptor type I–conditional knockout mice* Mouse
Age, weeks
Salivary gland
Mammary gland
Heart
Lung
Liver
Spleen
Kidney
1 2 3 4 5 6 7 8 9 10 11 12 13 14
3 3 3 3 3 3 3 4 4 4 4 4 4 4
⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹
⫹⫹⫹ ⫹⫹⫹ NE ⫹⫹⫹ NE NE NE ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ NE ⫹⫹⫹
⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹
⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺
⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺
⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺
⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺
* Inflammation was graded as none (⫺), slight (⫹), or severe (⫹⫹⫹). NE ⫽ not evaluated.
TGFRI-coko mice prompted us to investigate the peripheral immune system. Phenotypic analysis revealed that the percentage of CD3⫹ T cells in spleens was slightly up-regulated, with 17.32% and 20.91% in the WT and female TGFRI-coko mice, respectively (Figures 4A and B). However, there was no change in the CD3⫹ T cell population in male TGFRI-coko mice compared with the WT mice (Figure 4). The percentage of CD4⫹ T cells, but not that of CD8⫹ T cells, was increased in female TGFRI-coko mice compared with that in WT controls (13.69% versus 9.84%) (Figure 4). B cells (CD19⫹) were not changed in female TGFRIcoko mouse spleens (Figure 4). The data indicate an expansion of peripheral CD4⫹ T cells in female TGFRI-coko mice resembling a similar trend in the
Figure 3. Abnormal distribution of aquaporin 5 in salivary glands of a female TR1coko mouse. Representative salivary gland sections from a WT mouse (A) and a female TR1coko mouse (B) immunostained for aquaporin 5 are shown. In the WT mouse, aquaporin 5 staining is restricted to the apical region (arrow in A; arrowhead in inset), while in the female TR1coko mouse, staining indicates an abnormal distribution (arrow in B; arrowhead in inset). (Original magnification ⫻ 20 in A and B; ⫻ 60 in insets.) See Figure 1 for definitions.
CD4⫹ T cell population in TGF1-knockout mice. T cell expansion per se, however, remained unaltered. Focal inflammation resulting from AdcreM1mediated deletion of TGFRI expression in salivary glands of female TGFRI-f/f mice. Since MMTV-Cre– mediated deletion of TGFRI expression is not restricted to salivary glands, we administered AdcreM1 to salivary glands of TGFRI-f/f mice to delete TGFRI expression only in this tissue. For this purpose, we used
Figure 4. Increased T cell populations in female TR1coko mice. Spleen cells from female WT mice (A, D, and G), female TR1coko mice (B, E, and H), and male TR1coko mice (C, F, and I) were stained with anti-CD3, anti-CD4, anti-CD8, and anti-CD19. Values shown are from representative samples from 3 mice of each genotype. See Figure 1 for definitions.
IMPAIRED TGF SIGNALING IN FEMALE MICE
1803
DISCUSSION
Figure 5. Inflammatory response in female transforming growth factor  receptor type I–loxP-flanked (floxed) mice (female TR1f/f mice) elicited by delivery of AdcreM1 to salivary glands (SG). As indicated in the schematic representation (E), administration of AdcreM1 into mouse salivary glands was carried out via retrograde infusion of the cannulated main excretory ducts (see Materials and Methods). Salivary glands were analyzed after 30 and 60 days. PFU ⫽ plaque-forming units. Representative hematoxylin and eosin–stained salivary gland sections from male wild-type (WT) mice (A), female WT mice (B), male TR1f/f mice (C), and female TR1f/f mice (D) are shown. AdcreM1 elicited inflammatory responses in the salivary glands of only female TR1f/f mice after 60 days (arrow). (Original magnification ⫻ 100.)
10-month-old male and female mice of either the WT or TGFRI-f/f genotype. Following the delivery of AdcreM1 (5 ⫻ 108 PFU), the first group of mice was analyzed on day 30, while the second group was analyzed on day 60. All the mice appeared outwardly normal after the vector was administered. Analysis of salivary glands from the first group of mice did not show any significant abnormalities. However, the salivary glands analyzed from the second group of female TGFRI-f/f mice showed focal inflammation, which was not seen in male TGFRI-f/f mice or in WT males or females (Figure 5). Staining for aquaporin 5, Na⫹/K⫹/2Cl⫺ cotransporter, and ZO-1 in the salivary gland sections from the female TGFRI-f/f mice did not show any abnormal distribution (results not shown), indicating milder and highly localized inflammatory foci at this stage.
In the present study, we investigated the role of TGF signaling in salivary gland inflammation by conditionally abrogating expression of TGFRI. In the first approach, we generated TGFRI-coko mice by using MMTV-Cre mice. Unexpectedly, only the female TGFRI-coko mice developed a wasting disorder and early mortality reminiscent of findings in TGF1-null mice (3,4). These mice displayed focal inflammation primarily in the salivary glands, mammary glands, and heart. In the second approach, we delivered AdcreM1 to the salivary glands of TGFRI-f/f mice via retrograde infusion of the cannulated main excretory ducts and analyzed the salivary glands after 4 and 8 weeks. AdcreM1 administration elicited a focal inflammatory response only in female TGFRI-f/f mice, similar to the inflammation seen in female TGFRI-coko mice. Interestingly, unlike the case in the TGF1knockout mice, lymphocytic infiltration was restricted to salivary glands, mammary glands, and heart in the female TGFRI-coko mice. Since MMTV-Cre is expressed predominantly in mammary glands and salivary glands (14), the inflammation seen in the heart is intriguing. It is not clear whether this is an indirect effect of primary inflammation in mammary and salivary glands or whether it is due to trace Cre expression in the cardiac tissue. Nonetheless, this inflammation resulted in severe endocarditis with ensuing early mortality in these mice, similar to that seen in TGF1-knockout mice (3). TGF plays an important role in the control of the immune system (22–24). The focal inflammation we observed in the salivary glands presumably can be attributed to the lack of TGF signaling in the ductal cells, since TGF1-null mice also displayed similar inflammation in salivary glands (3). Salivary gland biopsy specimens from SS patients display reduced ductal expression of TGF1 (25). The multifocal inflammation seen in salivary glands of female TGFRI-coko mice was associated with Th1-specific cytokine mRNA expression, which is concordant with the cytokine profile observed by Fox et al in salivary gland biopsy specimens from SS patients (26). Moreover, fluorescence-activated cell sorting analysis revealed an increase in the CD4⫹ T cell population. Since these mice die at an early age, we were not able to compare their saliva secretion rates with those of the controls. However, we did evaluate a key molecule involved in salivary gland secretory function, aquaporin 5. Importantly, we observed an abnormal distribution of aquaporin 5 in salivary glands of female TGFRI-coko
1804
mice. Because of the essential role of aquaporin 5 in salivary fluid secretion (15,20,21), this finding suggests that the saliva secretory system in these mice is potentially defective. Interestingly, Steinfeld et al (27) and Tsubota et al (28) have reported the aberrant sorting of aquaporin 5 in gland biopsy specimens from SS patients (salivary and lacrimal, respectively). Additionally, Konttinen et al (29) reported abnormal distribution of aquaporin 5 in the salivary glands of NOD mice, a widely used SS disease model. Furthermore, Steinfeld et al (30) reported that treatment of SS patients with infliximab, which blocks the proinflammatory cytokine tumor necrosis factor ␣, led to restoration of the normal aquaporin 5 distribution in the apical membranes of minor salivary glands. Clearly, there is an emerging pattern indicating that focal lymphocytic infiltration in exocrine glands leads to altered membrane localization of aquaporin 5, a major component of the fluid secretory system. Interestingly, staining for other key membrane proteins such as Na⫹/K⫹/2Cl⫺ cotransporter, a basolateral marker, and ZO-1, a tight junction marker, did not show any abnormal distribution in the salivary gland sections from female TGFRI-coko mice. It is intriguing to note that only female TGFRIcoko mice developed an inflammatory response, and this sex-specific response was also seen in the female TGFRI-f/f mice that received AdcreM1 in their salivary glands. Many autoimmune diseases, including rheumatoid arthritis, systemic lupus erythematosus, and SS, occur more frequently in women than in men (31,32). There is much evidence suggesting that estrogen is the major cause of this difference in the prevalence of autoimmune diseases between men and women. The functions of estrogens are mediated through their binding to receptors, and Shim et al (33) have recently reported differential expression of estrogen receptors (ER␣ and ER) in secondary lymphoid tissues. Interestingly, TGF has been implicated in regulating ER␣positive glandular epithelial cells (34). It would be of interest to characterize the link between TGF signaling and estrogen functions in sex-specific immune responses. In summary, the present study indicates that female mice are more susceptible to impaired TGF signaling, particularly in their salivary glands, where it resulted in severe multifocal inflammation, increased expression of proinflammatory cytokines, and altered aquaporin 5 sorting. These 3 observations, and their striking sex dependence, were unexpected and are reminiscent of changes that occur in the salivary glands of
NANDULA ET AL
patients with the autoimmune epitheliitis SS. It is not clear whether, in fact, this or other mouse models of impaired TGF signaling are relevant to human SS disease. Indeed, the most widely employed mouse model for studies of SS, the NOD mouse, is neither a perfect model of the human disease nor phenotypically stable (35). However, mouse models in general are important for the study of conditions such as SS in which the etiology is unclear and the patient typically presents with established disease. Unfortunately, there are relatively few studies of possible TGF involvement in SS pathogenesis. Those that exist are equivocal in establishing a role for TGF in SS, with some being consistent (e.g., refs. 13, 36, and 37) and others not (e.g., refs. 38–40). While the findings of the present study may be fortuitous, they suggest that further study of TGF signaling in patients with SS may be useful in obtaining a better understanding of the pathogenesis of this enigmatic disease. Additionally, further studies with the present mouse model should yield valuable insights into sexspecific effects of impaired TGF signaling and inflammation. ACKNOWLEDGMENTS We would like to thank Drs. Jay Chiorini, Matt Hoffman, and Kelly Ten Hagen for critical reading of the manuscript, Dr. Yoshihiko Yamada for helpful discussion, and Harry Grant for editorial assistance. AUTHOR CONTRIBUTIONS Dr. Kulkarni had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study design. Amarnath, Hall, Larsson, Chen, Karlsson, Baum, Kulkarni. Acquisition of data. Nandula, Amarnath, Molinolo, Bandyopadhyay, Goldsmith, Zheng, Sreenath, Kulkarni. Analysis and interpretation of data. Nandula, Amarnath, Molinolo, Bandyopadhyay, Goldsmith, Zheng, Larsson, Sreenath, Chen, Ambudkar, Karlsson, Baum, Kulkarni. Manuscript preparation. Nandula, Amarnath, Ambudkar, Baum, Kulkarni. Statistical analysis. Nandula, Kulkarni.
REFERENCES 1. Massague J. TGF- signal transduction. Annu Rev Biochem 1998;67:753–91. 2. Letterio JJ, Roberts AB. Transforming growth factor-1-deficient mice: identification of isoform-specific activities in vivo. J Leukoc Biol 1996;59:769–74. 3. Kulkarni AB, Huh CG, Becker D, Geiser A, Lyght M, Flander KC, et al. Transforming growth factor 1 null mutation in mice causes excessive inflammatory response and early death. Proc Natl Acad Sci U S A 1993;90:770–4.
IMPAIRED TGF SIGNALING IN FEMALE MICE
4. Shull MM, Ormsby I, Kier AB, Pawlowski S, Diebold R, Yin M, et al. Targeted disruption of the mouse TGF-1 gene results in multifocal inflammatory disease. Nature 1992;359:693–9. 5. Sanford LP, Ormsby I, Gittenberger-de Groot AC, Sariola H, Friedman R, Boivin GP, et al. TGF  2 knockout mice have multiple developmental defects that are non-overlapping with other TGF  knockout phenotypes. Development 1997;124: 2659–70. 6. Kaartinen V, Voncken JW, Shuler C, Warburton D, Bu D, Heisterkamp N, et al. Abnormal lung development and cleft palate in mice lacking TGF-3 indicates defects of epithelial mesenchymal interaction. Nat Genet 1995;11:415–21. 7. Proetzel G, Pawlowski SA, Wiles MV, Yin M, Boivin GP, Howles PN, et al. Transforming growth factor-3 is required for secondary palate fusion. Nat Genet 1995;11:409–14. 8. Oshima M, Oshima H, Taketo MM. TGF- receptor type II deficiency results in defects of yolk sac hematopoiesis and vasculogenesis. Dev Biol 1996;179:297–302. 9. Larsson J, Goumans M, Sjostrand LJ, van Rooijen MA, Ward D, Leveen P, et al. Abnormal angiogenesis but intact hematopoietic potential in TGF- type I receptor-deficient mice. EMBO J 2001;20:1663–73. 10. Ludviksson BR, Gunnlaugsdottir B. Transforming growth factor- as a regulator of site-specific T-cell inflammatory response. Scand J Immunol 2003;58:129–38. 11. Dang H, Geiser AG, Letterio JJ, Nakabayashi T, Kong L, Fernandes G, et al. SLE-like autoantibodies and Sjogren’s syndromelike lymphoproliferation in TGF- knockout mice. J Immunol 1995;155:3205–12. 12. Daniels TE. Labial salivary gland biopsy in Sjo ¨gren’s syndrome: assessment as diagnostic criterion in 362 suspected cases. Arthritis Rheum 1984;27:147–56. 13. Manson GI, Hamburger J, Bowman S, Matthews JB. Salivary gland expression of transforming growth factor  isoforms in Sjogren’s syndrome and benign lymphoepithelial lesions. Mol Pathol 2003;56:52–9. 14. Wagner KU, McAllister K, Ward T, Davis B, Wisemann R, Hennighausen L. Spatial and temporal expression of the Cre gene under the control of the MMTV-LTR in different lines of transgenic mice. Transgenic Res 2001;10:545–53. 15. He X, Tse CM, Donowitz M, Alper SL, Gabriel SE, Baum BJ. Polarized distribution of key membrane transport proteins in the rat submandibular gland. Pflugers Arch 1997;433:260–8. 16. Chen WJ, Jin W, Hardegen N, Lei KJ, Li L, Marinos N, et al. Conversion of peripheral CD4⫹CD25⫺ naive T cells to CD4⫹CD25⫹ regulatory T cells by TGF- induction of transcription factor Foxp3. J Exp Med 2003;198:1875–86. 17. Gao J, Killedar S, Cornelius JG, Nguyen C, Cha S, Peck AB. Sjo ¨gren’s syndrome in the NOD mouse model is an interleukin-4 time-dependent, antibody isotype-specific autoimmune disease. J Autoimmun 2006;26:90–3. 18. Becker TC, Noel RJ, Coats WS, Gomez-Foix AM, Alam T, Gerard RD, et al. Use of recombinant adenovirus for metabolic engineering of mammalian cells. Methods Cell Biol 1994;43(Pt A):161–89. 19. Baum BJ, Wellner RB, Zheng C. Gene transfer to salivary glands. Int Rev Cytol 2002;213:93–146. 20. Ma T, Song Y, Gillespie A, Carlson EJ, Epstein CJ, Verkman AS. Defective secretion of saliva in transgenic mice lacking aquaporin-5 water channels. J Biol Chem 1999;274:20071–4. 21. Krane CM, Melvin JE, Nguyen HV, Richardson L, Towne JE, Doetschman T, et al. Salivary acinar cells from aquaporin 5-deficient mice have decreased membrane water permeability and altered cell volume regulation. J Biol Chem 2001;276:23413–20.
1805
22. Moustaka A, Pardali K, Gaal A, Heldin C. Mechanisms of TGF- signaling in regulation of cell growth and differentiation. Immunol Lett 2002;82:85–91. 23. Boivin GP, O’Toole BA, Orsmby IE, Diebold RJ, Eis MJ, Doetschman T, et al. Onset and progression of pathological lesions in transforming growth factor 1-deficient mice. Am J Pathol 1995;146:276–88. 24. Letterio JJ, Roberts AB. Regulation of immune responses by TGF-. Annu Rev Immunol 1998;16:137–16. 25. Kizu Y, Sakurai S, Katagiri S, Shinozaki N, Ono M, Tsubota K, et al. Immunohistological analysis of tumour growth factor 1 expression on normal and inflamed salivary glands. J Clin Pathol 1996;49:728–32. 26. Fox RI, Kang HI, Ando D, Abrams J, Pisa E. Cytokine mRNA expression in salivary gland biopsies of Sjogren’s syndrome. J Immunol 1994;152:5532–9. 27. Steinfeld S, Cogan E, King LS, Agre P, Kiss R, Delporte C. Abnormal distribution of aquaporin-5 water channel protein in salivary glands from Sjogren’s syndrome patients. Lab Invest 2001;81:143–8. 28. Tsubota K, Hirai S, King LS, Agre P, Ishida N. Defective cellular trafficking of lacrimal gland aquaporin-5 in Sjo ¨gren’s syndrome. Lancet 2001;357:688–9. 29. Konttinen YT, Tensing EK, Laine M, Porola P, Tornwall J, Hukkanen M. Abnormal distribution of aquaporin-5 in salivary glands in the NOD mouse model for Sjogren’s syndrome. J Rheumatol 2005;32:1071–5. 30. Steinfeld SD, Appelboom T, Delporte C. Treatment with infliximab restores normal aquaporin 5 distribution in minor salivary glands of patients with Sjo ¨gren’s syndrome. Arthritis Rheum 2002;46:2249–51. 31. Whitacre CC. Sex differences in autoimmune disease. Nat Immunol 2001;2:777–80. 32. Beeson PB. Age and sex associations of 40 autoimmune diseases. Am J Med 1994;96:457–62. 33. Shim GJ, Gherman D, Kim HJ, Omoto Y, Iwase H, Bouton D, et al. Differential expression of oestrogen receptors in human secondary lymphoid tissues. J Pathol 2006;208:408–14. 34. Ewan KB, Oketch-Rabah HA, Ravani SA, Shyamala G, Moses HL, Barcelos-Hoff MH. Proliferation of estrogen receptor-␣positive mammary epithelial cells is restrained by transforming growth factor-1 in adult mice. Am J Pathol 2005;167:409–17. 35. Lodde BM, Mineshiba F, Kok MR, Wang J, Zheng C, Schmidt M, et al. NOD mouse model for Sjogren’s syndrome: lack of longitudinal stability. Oral Dis 2006;12:566–72. 36. Koski H, Konttingen YT, Gu XH, Hietanen J, Malmstrom M. Transforming growth factor  2 in labial salivary glands in Sjo ¨gren’s syndrome. Ann Rheum Dis 1995;54:744–7. 37. Gottenberg JE, Busson M, Loiseau P, Dourche M, Cohen-Solal J, Lepage V, et al. Association of transforming growth factor 1 and tumor necrosis factor ␣ polymorphisms with anti-SSB/La antibody secretion in patients with primary Sjo ¨gren’s syndrome. Arthritis Rheum 2004;50:570–80. 38. Boumba D, Skopouli FN, Moutsopoulos HM. Cytokine mRNA expression in the labial salivary gland tissues from patients with primary Sjo ¨gren’s syndrome. Br J Rheumatol 1995;34:326–33. 39. Fox PC, Brennan M, Di Sun P. Cytokine expression in human labial minor salivary gland epithelial cells in health and disease. Arch Oral Biol 1999;44 Suppl 1:S49–52. 40. Eriksson P, Andersson C, Ekerfelt C, Ernerudth J, Skogh T. Relationship between serum levels of IL-8 and IgG1 in patients with Sjo ¨gren’s syndrome, rheumatoid arthritis and healthy controls. Clin Exp Immunol 2004;137:617–20.
