|Home | About | Journals | Submit | Contact Us | Français|
Periostin is a 90kD member of the fasciclin-containing family and functions as part of the extracellular matrix. Periostin is expressed in a variety of tissues and expression is increased in airway epithelial cells from asthmatic patients. Recent studies have implicated a role for periostin in allergic eosinophilic esophagitis. To further define a role for periostin in Th2-mediated inflammatory diseases such as asthma, we studied the development of allergic pulmonary inflammation in periostin-deficient mice. Sensitization and challenge of periostin-deficient mice with ovalbumin resulted in increased peripheral Th2 responses compared to control mice. In the lungs, periostin-deficiency resulted in increased airway resistance and significantly enhanced mucus production by goblet cells concomitant with increased expression of Gob5 and Muc5ac, compared to wild type littermates. Periostin also inhibited the expression of Gob5, a putative calcium-activated chloride channel involved in the regulation of mucus production, in primary murine airway epithelial cells. Our studies suggest that periostin may be part of a negative-feedback loop regulating allergic inflammation that could serve as a therapeutic target in the treatment of atopic disease.
Allergic asthma is a Th2-mediated inflammatory disease characterized by pulmonary eosinophilia, airway hyperresponsiveness (AHR) and mucus overproduction by goblet cells (1, 2). With subsequent progression of this disease, airways undergo structural and phenotypic changes resulting in airway remodeling (3) including damage to epithelial cells, goblet cell metaplasia (GCM), subepithelial fibrosis, and smooth muscle hyperplasia and hypertrophy (4). One of the key features of airway remodeling is the development of GCM which leads to airflow obstruction in asthma. Mortality of asthma patients reveal mucus plugging on postmortem examination. Several studies have implicated that the Th2-cytokines IL-4/IL-9/IL-13 play critical roles in inducing mucus production in asthma (5–9). However, the mechanism underlying how Th2 cytokines induce mucus production is poorly understood. Mouse calcium-activated chloride channel-3 (mCLCA-3; gob-5) / human CLCA-1 acting as a downstream molecule of the Th2 cytokines, IL-4/IL-9/IL-13 has been suggested to be important in mucus production (10). Regulatory pathways controlling mucus production need to be explored further in order to alleviate excess mucus production in respiratory diseases including asthma.
Periostin, an extracellular matrix protein originally isolated from an osteoblast cell line, is induced by IL-4 and IL-13 in airway epithelial cells (11, 12). Periostin is secreted by fibroblasts and composed of a cysteine-rich domain in its N-terminal end, 4 tandem fasciclin I (FASI) domains, and an alternatively spliced domain at the C-terminal end (13). Periostin shares structural homology to fasciclin I (FASI) identified in insects, and other mammalian proteins including βig-h3, stabilin I and II, MBP-70, Algal-CAM and Periostin-like-factor (PLF) (13–19). IL-13 and TGF-β induce the expression of periostin in esophageal fibroblasts, and there is increased periostin expression in eosinophilic esophagitis (EE) biopsy samples (20). Periostin-deficient (Postn−/−) mice subjected to intranasal Aspergillus infection had decreased eosinophilic inflammation in the lungs and esophagus (20).
In this report, we examine the development of allergen-induced pulmonary inflammation, examining the effects of periostin-deficiency on peripheral Th2 responses, lung function and mucus production in mice. Our study demonstrates an important role for periostin in the regulation of mucus production, as deficiency of periostin significantly increased mucus expression by goblet cells and augmented airway resistance. However, periostin deficiency did not affect allergic airway inflammation, indicating that periostin function in allergic airway disease is restricted to a subset of features including GCM.
Postn−/− mice on a mixed genetic background (B6/129) were generated as described previously (21). Briefly, the Postn gene was targeted in embryonic stem cells, by replacing exons 4 to 10 encoding 3 of the 4 fasciclin domains, which produced a null allele. Control mice were wild-type littermates. All mice were maintained in specific pathogen-free conditions and experiments were approved by the Indiana University Institutional Animal Care and Use Committee.
Splenocytes or bronchoalveolar lavage (BAL) cells (0.5–1.0 × 106 per sample) were washed and stained in PBS with 2% BSA and 0.1% NaN3 (FACS buffer). Cells were first incubated with anti-FcγR Abs, clone 2.4G2 (BD Pharmingen), for 10 min. All staining was done at 4°C, followed by one wash with FACS buffer, then fixed in PBS with 2% BSA, 0.1% NaN3, and 0.5% formaldehyde. Samples were stained with Abs directly conjugated to FITC, PE, CyChrome, or allophycocyanin (BD Pharmingen) for CD4, CD8, CD3, B220, CD44, CD62L, CD11c, CD3, B220, CCR3 and MHC class II (BD Pharmingen). Analysis of lymphocyte populations was performed using a forward scatter (size) gate, and the expression of CD44 and CD62L were determined on CD4+ cells. The cellular composition of BAL for the populations of eosinophils, neutrophils, T cells, B cells, dendritic cells, and macrophages was performed as described by a flow cytometric method (22). Eosinophils, neutrophils, T cells, B cells, and mononuclear cells were distinguished by cell size and the expression of CD3, B220, CCR3, CD11c, and MHC class II.
Total CD4+ T cells were isolated from spleens of periostin-deficient mice and wild-type littermates on a mixed background (MACS isolation system; Miltenyi Biotec, Auburn, CA). T cells were activated with plate-bound anti-CD3 (4 µg/ml 145-2C11) and soluble anti-CD28 (1 µg/ml; BD Pharmingen) and cultured under conditions that prime Th1 [IL-12 at 10 ng/ml and anti-IL-4 at 10 µg/ml], Th2 [IL-4 at 10 ng/ml (Peprotech, Rocky Hill, NJ) and anti-IFNγ at 10 µg/ml], and Th17 [TGF-β1 (5 ng/ml) and IL-6 (100 ng/ml; Peprotech) + IL-1β (10 ng/ml) + anti-IL-4 (10 µg/ml)+ anti-IFN-γ (10 µg/ml)] conditions. After 5 days in culture, cells were restimulated with plate-bound anti-CD3 (4µg/ml) for 24 hours and cell-free supernatants were analyzed for IFN-γ, IL-4, and IL-17 using ELISA (reagents from BD Pharmingen) (23).
Mice were sensitized by intraperitoneal injections of ovalbumin (OVA) adsorbed with alum (OVA and alum from Sigma, St. Louis, MO) at a dose of 20 µg OVA / 2 mg alum on days 0 and 7. On day 14, mice were exposed to intranasal OVA (100 µg) per day for 6 consecutive days. Mice were euthanized by intraperitoneal injection of pentobarbital (5 mg/mouse), 48 h after the last intranasal challenge. The trachea was cannulated and lungs lavaged 3 times with 1 ml of PBS. The cells recovered in bronchoalveolar lavage (BAL) fluid were counted with a hemocytometer. Eosinophils, neutrophils, T cells, B cells, and mononuclear cells in the BAL were distinguished by cell size and the expression of CD3, B220, CCR3, CD11c, and MHC class II, as described by a flow cytometric method (22). Cytokine levels in cell-free BAL fluid were measured using ELISA. Total BAL cells were stimulated with 4 µg/ml anti-CD3 or 100 µg/ml OVA for 72 h before supernatants were tested for cytokine levels using ELISA. For qPCR measurements, lung tissues were homogenized in a tissue lyser (Qiagen) and RNA isolated with RNeasy kit (Qiagen) was used to synthesize cDNA for subsequent analysis. Peripheral immune responses in sensitized and challenged wild type or Postn−/− mice were assessed by culturing splenocytes with anti-CD3 (4 µg/ml) or OVA (100 µg/ml) for 72 h before cell-free supernatants were tested for cytokines using ELISA.
AHR to methacholine challenge (dose range from 10–200 mg/ml) was determined 24 h after the last intranasal challenge. Non-invasive unrestrained whole body plethysmography (Buxco System, NC) was used to record airway responsiveness via a dimensionless parameter termed enhanced pause (Penh) to estimate the total pulmonary resistance, an indicator of bronchoconstriction. In separate animals, AHR was measured by invasive means of determining airway resistance (Buxco Systems). Mice were anesthetized, trachea intubated and mice ventilated before intratracheally delivering nebulized saline followed by methacholine doses of 5–200 mg/ml.
Lung tissue derived from mice 48 h after the last intranasal challenge was fixed in 10% formalin and paraffin-embedded sections were stained with hematoxylin & eosin (H&E) or periodic acid-Schiff (PAS) in order to evaluate the infiltration of inflammatory cells and mucus secretion. Peribronchial and perivascular inflammation was assessed using light microscopy with 400× magnification. Mucus production in histological sections was assessed by measuring color intensity of PAS staining of goblet cells on histological sections, image capture with 10× magnification, in a blinded fashion, and quantitative intensity (expression) measurement of PAS staining exclusively in the large airways, using a macro in the Metamorph Imaging software (Universal), as previously described (24). Immunostaining of periostin (1:10,000 dilution) using the ABC kit (Vector) following manufacturer’s directions, was performed as described (21). Negative controls that lacked either the primary or secondary or both antibodies, or Postn−/− mice, did not produce any DAB staining.
Lung tissue sections from non-sensitized (WT NS) and sensitized and challenged wild type mice (WT) were snap frozen in liquid nitrogen. Tissues were homogenized as described (25). Western blot analysis was performed as previously described (26) using specific Abs against periostin, or control Ab β-actin (Santa Cruz Biotechnology, Santa Cruz, CA). For densitometric analysis, the scanned images of immunoblots were quantified with Image J software.
Periostin-containing and control supernatants were collected from HEK 293 cells transfected with full-length murine periostin cDNA (PS and SJC, unpublished) or empty vector. Briefly, the CD4+ T cells from the spleens and lymph nodes of wild type mice were cultured under Th2-priming conditions in the presence of periostin-containing or control media (10% v/v). After 5 days in culture, cells were restimulated with plate-bound anti-CD3 (4µg/ml) for 24 hours and cell-free supernatants were analyzed for IL-4, IL-5 and IL-13 using ELISA (reagents from BD Pharmingen). The effect of periostin-containing or control media was also tested on CD11c+ DCs. Splenic DCs were isolated by positive selection with anti-CD11c magnetic beads following manufacturer instructions (Miltenyi Biotec). Isolated CD11c+ DCs were stimulated with LPS (1µg/ml) in the presence of periostin-containing or control media, or recombinant periostin (R&D Systems) for 72 h. Cell-free culture supernatants were tested for IFN-γ by ELISA. Next, we also stimulated CD11c+ DCs with LPS (10ng/ml) in the presence of recombinant periostin (0–100 ng/ml) and determined IFN-γ levels in 72 h culture supernatants by ELISA.
MTEC from C57BL/6 mice were cultured as previously described (27). Briefly, tracheas were resected, opened longitudinally, incubated in Ham’s F-12 pen-strep containing pronase followed by gentle inversion 3 times, to release the cells. Cells were then incubated with pancreatic DNase I and finally resuspended in MTEC basic media with 10% FBS. Fibroblasts were removed by incubation of the cells in tissue culture plates for 3–4 h at 37°C. Non-adherent cells were cultured for 48 h in 24-well plates. Once the cells reached confluence, cells were treated with varying concentrations of recombinant mouse IL-13 (R & D Systems, Minneapolis, MN) for 6 h to analyze gene expression. To test the effect of periostin on GCM-associated gene expression, MTEC cells were stimulated with varying concentrations of periostin for 6 h in the presence of 1 ng/ml of IL-13. After 6 h, mucus gene expression was analyzed by Taqman real time quantitative PCR.
Although Postn−/− mice exhibit teeth, bone, and cardiac defects, and while they can be decreased in size, mice develop normally on wet feed (21). Postn−/− mice used for our studies were comparable in size to their wild-type littermates. To determine if Postn−/− mice have altered lymphocyte development, splenocytes were analyzed for surface expression of CD3, B220, CD4, CD8, CD44 and CD62L by flow cytometry. B and T cell populations, as well as populations of memory and naïve CD4+ T cells remain unaltered in Postn−/− mice (Fig. 1A).
To determine if periostin-deficiency had any effect on the ability of naïve T cells to respond to cytokines and differentiate into effector subsets, purified CD4+ T cells from splenocytes were stimulated with anti-CD3 and anti-CD28 under Th1, Th2 and Th17 conditions. Analysis of cytokine production in cell-free culture supernatants revealed that Postn−/− T cells had comparable levels of IFN-γ, IL-4, IL-5, IL-13 and IL-17, to wild-type littermates under Th1, Th2 and Th17 conditions respectively (Fig. 1B). Thus, Postn−/− mice have normal lymphocyte development in the periphery and T cells are unimpaired in their ability to acquire Th effector phenotypes.
We next determined the role of periostin in allergic airway inflammation by employing an OVA-induced model of sensitization and intranasal challenge. Postn−/− mice and wild-type littermates were subjected to sensitization and challenge with OVA. To determine if periostin is induced in the lungs of sensitized mice, we performed immunohistochemistry of paraffin-embedded tissue sections. Immunohistochemistry revealed that periostin is normally present at low levels within control wild type non-sensitized lung sections (NS), and staining is significantly increased in lung tissue from sensitized and challenged wild type mice (Fig. 2A, middle). Periostin extracellular matrix deposition is predominantly observed within the submucosa and lung parenchyma surrounding the bronchus and bronchioles. Staining was absent in sections from Postn−/− sensitized and challenged mice (Fig. 2A, bottom).
Western blot analysis was used to confirm the induction of periostin expression. We observed a significant increase in periostin expression in the lung protein extracts from sensitized and challenged mice as compared to non-sensitized mice. Expression assessed by immunoblot was averaged from four mice and showed a 2.5-fold increase in the amount of periostin expression (Fig. 2B).
To determine if periostin-deficiency affected pulmonary inflammation following sensitization and challenge, we next analyzed cells in the BAL for populations of T cells, B cells, eosinophils, neutrophils and dendritic cells by flow cytometry. Deficiency of periostin did not result in significant differences in the predominant T cell and eosinophil populations augmented in allergic airway inflammation over non-sensitized mice which did not show any inflammation (Fig. 3A). Histopathological analysis of lung sections did not reveal differences in the amount of inflammation in the lungs of Postn−/− mice compared to wild-type littermates (Fig. 3B). Similarly cytokines in the BAL fluid also remained unaltered in mice with a deficiency of periostin (Fig. 3C, left panel). However, when BAL cells were stimulated ex-vivo, a significant increase in IL-5 and IL-13 production following anti-CD3 stimulation and IL-5 production following OVA stimulation was observed in cultures from Postn−/− mice (Fig. 3C, middle and right panels). Moreover, there were increased amounts of Th2 cytokine mRNA isolated from Postn−/− lung tissue, compared to littermate controls (Fig. 3D). These results indicate that periostin-deficiency does not alter the inflammatory cell populations in the lung but that the cells present in the BAL and lung have modestly increased Th2 cytokine production, compared to littermate controls.
To determine the effects of periostin-deficiency on lung function, we assessed lung function using both non-invasive whole body plethysmography (measured as Penh) and an invasive method of determining airway resistance. In non-sensitized and challenged mice, there was no significant difference in Penh or airway resistance between wild type and Postn−/− mice (data not shown). However, following sensitization and challenge, a significant increase in Penh was observed in Postn−/− mice at higher concentrations of methacholine (Fig. 4A). Airway resistance using tracheal intubation was increased at baseline, and was also significantly increased at the highest dose of methacholine (Fig. 4B) concomitant with decreased compliance (Fig. 4C). Thus in sensitized and challenged Postn−/− mice there is increased airway resistance and decreased compliance, compared to control mice, despite a similar recruitment of inflammatory cells to the lungs in mice of both genotypes.
To determine if increased Th2 cytokines in BAL and lungs was a result of increased Th2 responses in the periphery, we examined the cytokine production from cells in the lung draining lymph nodes and spleens. The production of Th2 cytokines IL-4, IL-5 and IL-13 in both splenocyte and mediastinal lymph node cultures was increased following stimulation with anti-CD3 or OVA from cultures of Postn−/− cells, compared to wild type cells (Fig. 5A & 5B). In contrast, IFN-γ or IL-17 remained unaltered in sensitized and challenged Postn−/− (data not shown). Serum OVA-specific IgE levels in Postn−/− deficient mice were also significantly increased compared to controls (data not shown). These results suggest that periostin-deficiency results in enhanced Th2 responses in vivo.
To define a mechanism for how periostin regulated Th2 development we tested the ability of periostin to alter Th2 cell development in vitro. As noted previously, we did not observe any effects of periostin-deficiency on the acquisition of Th phenotypes in vitro (Fig.1) consistent with a lack of detectable periostin expression in T cells (data not shown). Moreover, addition of periostin-containing media to differentiating T cells did not have an effect on Th2 cytokine production (Fig. 5C top). Since periostin did not have any direct effects on T cells, we determined if periostin could mediate its effects by acting on DCs. Addition of periostin-containing media (Fig. 5C middle) or recombinant periostin (Fig. 5C bottom) to LPS stimulated CD11c+ DC increased the levels of IFN-γ. These results suggest that periostin might enhance Th1 responses and that increased Th2 responses develop in the absence of perisotin.
GCM and mucus overproduction are predominant features of asthma and correlate with AHR resulting in increased severity and mortality of asthmatic patients (28–33). To determine if mucus production is affected following sensitization and challenge of Postn−/− mice, lung tissues were stained with periodic acid-Schiff (PAS) to analyze mucus-producing cells. An increase in mucus production by goblet cells of Postn−/− mice was observed (Fig. 6A). Mucus production was quantified by blinded automated measurement of the intensity of PAS-staining of goblet cells on three random low power fields per mouse, utilizing a macro developed with the Metamorph software. A significant increase in bronchial PAS staining index was observed in Postn−/− mice relative to wild type mice (Fig. 6B). Further analysis of mucin gene expression demonstrated significant enhancement of Muc5ac and Gob5 expression in lung tissues of Postn−/− mice compared to wild-type littermates (Fig. 6C). Thus, periostin-deficiency markedly augmented mucin gene expression and GCM in the lung.
To directly examine the ability of periostin to inhibit GCM, we tested the effect of periostin on MTEC. Tracheal epithelial cells were isolated and cultured as described in Materials & Methods. After 2d, the confluent cells were stimulated with IL-13 to verify its ability to induce GCM. A forty-fold increase in the expression of Gob5 at 1 ng/ml of IL-13 was observed (Fig. 6D). We next determined the effect of varying doses of periostin (10–100 ng/ml) on IL-13-induced Gob5 expression in MTEC. A consistent decrease in the expression of IL-13-induced Gob5 was observed upon treatment with increasing doses of recombinant periostin (Fig. 6D). Thus, periostin may regulate GCM by directly acting on airway epithelial cells.
Mucus hypersecretion, a hallmark of asthma is responsible for the increasing morbidity and mortality associated with this disease. However, the underlying mechanisms of mucus production are still poorly understood. Earlier studies have defined a critical role for IL-13 in the induction of GCM. Periostin, an extracellular matrix protein is induced in epithelial cells by inflammatory cytokines including IL-4 and IL-13 (12) and in lung fibroblasts by the anti-inflammatory cytokine TGF-β (34). Moreover, periostin can also co-localize with other extracellular matrix proteins to trigger subepithelial fibrosis (12). Whether these effects are important in the development of allergic inflammation was not well characterized. Thus, in the present study we determined whether periostin promotes inflammation or has a protective homeostatic effect in regulating airway inflammation. Our study demonstrated that periostin-deficiency led to a marked increase in airway resistance and mucus production upon sensitization and challenge with OVA. Despite the development of eosinophilic airway inflammation to a similar extent in Postn−/− mice as wild-type littermates, deficiency of periostin specifically augmented mucus production as evident from PAS-staining index and expression of Muc5ac and Gob5 in allergen-challenged lung tissues.
Recent studies have implicated periostin as a biomarker for IL-13-dependent corticosteroid responsive asthma (35), expressed in airway epithelial cells of asthma patients. Phenotypic heterogeneity of asthmatics could be classified into “Th2-high” and “Th2-low” asthma based on the epithelial expression of the IL-13-inducible genes POSTN, CLCA1, SERPINB2 (36). Patients with “Th2-high asthma” had significantly increased AHR, eosinophilic inflammation and induction of epithelial mucins MUC5AC and MUC2. There are similar associations of increased periostin expression in biopsies of patients with eosinophilic esophagitis (20, 37). However, these correlations do not determine the biological function of periostin, which could be acting to promote inflammation, or as part of a negative feedback mechanism to control inflammation. Our studies suggest that allergen sensitization and challenge induced periostin expression (Fig. 2), which functioned to attenuate mucus gene expression and GCM. The mechanism of inhibition of GCM by periostin remains to be elucidated. Earlier studies have suggested that periostin acts as a cell adhesion molecule by binding to integrins αvβ3 and αvβ5 in cancer cells (38). Thus, it is possible that periostin binds to integrins alpha4 and beta 1/2, which are involved in asthma (39–48), and may trigger intracellular signaling pathways that repress mucus-inducing transcription factors, including NF-κB, Sp1 and AP-1 (49). Regardless, our data demonstrate that in the absence of periostin, there was increased differentiation of epithelial cells into mucus-producing goblet cells, suggesting that periostin may contribute to the homeostasis of GCM during allergic inflammation.
How perisotin contributes to airway function is still unclear. There were no differences in cellular infiltration, Penh, or airway resistance and compliance, between wild type and Postn−/− mice in the absence of sensitization and challenge, suggesting that the differences observed in Fig. 4 arise from the allergic inflammation and not from a requirement for periostin in lung development. This is consistent with the observation of low perisotin expression in non-challenged lungs (Fig. 2). Analysis of sensitized and challenged mice by whole body plethysmography suggested an increase in airway hyperreactivity. However, use of the more invasive method of testing lung function demonstrated increased airway resistance and decreased compliance in Postn−/− mice compared to control mice, even in the absence of methacholine challenge (Fig. 4). While the airway resistance was significantly increased in Postn−/− mice compared to wild type mice following methacholine challenge, the fold difference between wild type and Postn−/− mice at baseline and after methacholine challenge was similar. These results suggest that periostin-deficiency is not affecting airway hyperreactivity per se, but rather affects baseline lung function following the induction of allergic inflammation. We did not observe obvious changes in epithelial size, smooth muscle content or fibrosis between wild type and Postn−/− lungs, suggesting that the primary effect of perisotin-deficiency is in increased mucus production that results in decreased airflow through the respiratory tract.
The increased Th2 responses in the periphery were somewhat surprising, considering the lack of a direct effect of periostin on Th cell differentiation. As a possible mechanism for this phenomenon, we demonstrated that periostin augments the production of IFNγ, a cytokine that both inhibits Th2 and promotes Th1 responses. We propose a model wherein DCs in the lung would be exposed to periostin induced during allergic inflammation that would alter the amounts of cytokines they secrete, when as mature DC, they migrate to lymph nodes to activate T cells. The ability of periostin to alter Th2 cytokine production might also impact GCM. Thus, the effects of periostin-deficiency on GCM could be through at least two mechanisms; direct effects on airway epithelial cells, and indirect effects of altering APC function and Th2 cytokine production. Although Th2 cytokine production was modestly increased in the lung of Postn−/− compared to wild type mice, neither inflammation nor Tslp expression were appreciably increased (Fig. 3 and data not shown). It is possible that the modest increase in Th2 cytokines was not sufficient to alter the amount of inflammation. It is also possible that, as previously reported (20), periostin plays a role in cell recruitment and this was balanced by increased Th2 cytokines. Further studies can distinguish these possibilities.
In a recent study (20) of periostin-deficient mice, a decrease in eosinophil recruitment to the lungs and increase in blood eosinophilia was observed with no significant differences in mucus production upon intranasal inhalation with allergens from A. fumigatus. There may be several reasons for these differences, including the allergen used to generate airway inflammation. It is also possible that genetic background may contribute to the observed differences; the mice used in our study were on a mixed background, the mice in the previous study, while not stated specifically, were likely C57BL/6 mice (50). Differences in immune pathways activated by these two protocols may result in differential dependence on periostin for regulation of GCM.
In conclusion, our studies have defined a novel previously unrecognized role for periostin in regulation of GCM. Regulation of mucus production is important in minimizing the symptoms and reducing mortality associated with severe asthma attacks (51–53). Our data suggest that although periostin is induced by cytokines associated with allergic inflammation, it functions in the lung to limit mucus production and the resulting effects on airway resistance. This negative feedback mechanism may be an important component of maintaining airway function while the immune system combats pathogens, or in the case of allergens, responds to innocuous proteins. Our results examining the function of periostin in vivo and on primary tracheal epithelial cells suggest that it may be a potential therapeutic in the regulation of mucus production.
This work was supported in part, by the Riley Children's Foundation (SJC & PS), the Indiana University Department of Pediatrics (Cardiology) and Public Health Service Awards U19 AI070448 (MHK, RST), RO1 HL080071 (RST), R01 HL092508 (SJC).