Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Surfactant Protein-D (SP-D), a member of the collectin family, is expressed in the lung and in lower abundance in many other organs (1). SP-D binds microbial cell-surface carbohydrates by a lectin-mediated mechanism (2). In some cases such binding has been shown to facilitate phagocytosis; in other cases, agglutination of microorganisms has been observed (3). SP-D is also chemotactic for neutrophils and mononuclear phagocytes and has been shown to stimulate an oxidant burst from macrophages in vitro (4, 5). Such observations have led to the general notion that SP-D recognizes non-host carbohydrates, both altering inflammatory cell responses to bound pathogens and facilitating their physical clearance. Thus, SP-D is believed to play an important role in intrinsic immunity in the lung.

SP-D has also been shown to reduce the in vitro proliferation of T lymphocytes in response to Con A, PHA, and anti-CD3 antibodies by a non-lectin-mediated mechanism (6). Purified SP-D reduces the proliferation of peripheral blood mononuclear cells isolated from patients with asthma when exposed to phytohemagglutinin or dust mite allergen (7). Reduced proliferation appeared to be mediated by interaction of SP-D with inflammatory cells. SP-D was also shown to reduce mite allergen-induced histamine release from whole blood from mite-sensitive individuals with asthma. The mechanism for inhibition of histamine release appeared to be dependent upon interaction of SP-D with allergen and was mediated by interaction with antigen carbohydrate (7). Therefore, SP-D may not only have direct effects on lymphocytes, inhibiting their activation and proliferation, but also may prevent inflammatory cell responses to certain antigens. Such observations have led to the hypothesis that SP-D may in part account for the induction and maintenance of a state of reduced immunologic response observed in lung leukocytes (6, 7). Thus, SP-D may have a role in innate immunity and also in regulating clonal immunity in the lung.

The availability of transgenic mice that do not express SP-D or alternatively express supraphysiologic levels of SP-D has allowed in vivo assessment of the role of SP-D (8). In the absence of any apparent infectious process, transgenic mice that do not express SP-D accumulate approximately a 6-fold increase in alveolar surfactant (8, 9). In SP-D null (SP-D^/) mice, alveolar macrophages are also markedly increased by 10 wk of age (8, 9). Many alveolar macrophages are enlarged and contain inclusions that stain for either phospholipid or neutral lipid, but most alveolar macrophages have normal morphology (9). Alveolar macrophages from SP-D^/ mice are also activated because they produce increased concentrations of reactive oxygen species and also secrete matrix metalloproteinases (11). Another pathologic finding is the appearance of peribronchial and perivascular aggregates of lymphocytes without evidence of acute inflammation (8, 9).

Because SP-D^/ mice display increased peribronchial lymphoid aggregates, we sought to determine if there were increased numbers of lymphocytes in the lungs of SP-D^/ mice, if their distribution between lymphocyte classes was altered, and if such lymphocytes displayed markers of activation.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Transgenic Mice

All mice used for this protocol were housed in the animal care facility at Denver Health Medical Center under standard vivarium conditions. SP-D^/ mice were generated using a neomycin replacement strategy with deletion of the second exon of the SP-D gene by homologous recombination. The second exon contains the initiating AUG codon. These mice have previously been described and do not make SP-D protein (9). All experimental mice resulted from 8 sequential outbreedings of the founding strain (129 Ola/Swiss Black hybrids) into a commercially available NIH Black Swiss outbred strain (Taconic, Germantown, NY). Swiss Black outbred mice were chosen for evaluation because the phenotype was first observed in the founding hybrids and persisted through successive outbreedings. However, similar collections of lymphocytes have been observed in SP-D^/ mice derived from different parental strains (8). Both SP-D^+/+ and SP-D^/ mice were pathogen-free by serologic evaluation. Repeated attempts to culture bacterial, fungal, and mycobacterial organisms from the lungs of SP-D^/ mice have been negative. Likewise, attempts to identify Mycoplasma and Chlamydia by ribosomal reverse transcriptase/polymerase chain reaction (RT-PCR) have been negative. All SP-D transgenic mice have normal life spans and gain weight normally, suggesting that SP-D mice do not develop intermittent or chronic lung infections. The University of Colorado Health Sciences Center animal care committee approved all protocols. For these experiments, SP-D^/ and SP-D^+/+ mice were littermates from heterozygous crosses and all were apparently healthy and of equal weight. Every mouse used in these experiments was genotyped by PCR and DNA blot analysis as previously described (9). Mice were killed by lethal pentobarbital injection followed by exsanguination at 6-12 wk of age. Lungs and spleens were removed and placed in iced phosphate-buffered saline (PBS) until analysis was performed (within 2 h).

Flow Cytometric Analysis of Lung and Spleen Cells

Lymphocyte analysis was performed as previously described (12). Lung lobes were minced with scissors in a 35-mm tissue culture dish and then digested in a solution of 1.9 mg/ml collagenase, DNase at 10 U/ml, and soybean-trypsin inhibitor at 10 µg/ml in tissue culture medium with 5% fetal bovine serum (all reagents were from Sigma Chemical Co., St Louis, MO) in the same dish. Lung tissues were incubated in the digestion solution for 1 h at 37°C, then triturated through a #20 needle. The entire cell suspension was then passed through a 70-µM centrifugal filter (BD Falcon, Bedford, MA). Mononuclear cells were isolated from the filtrate by Ficoll density gradient centrifugation and the total number of viable mononuclear cells recovered from each lung was determined by manual counting. Spleen cells were obtained after mechanical disruption and erythrocytes were lysed by ammonium chloride. A second set of experiments was performed to eliminate the potential for erroneous assessment of lymphocyte lineages from lymphocytes in the pulmonary circulation. After bronchoalveolar lavage (BAL), the lungs were perfused free of blood using 10 ml of PBS, 5 mM EDTA. Lymphocytes from disassociated lung tissue were analyzed by fluorescence-activated cell sorter (FACS).

Cells were analyzed using a FACSCalibur flow cytometer (Becton Dickinson, Palo Alto, CA), with analysis gates set by gating on unstained spleen lymphocytes. Between 10,000 and 30,000 gated events were analyzed for each cell population. Isotype controls have been performed on previous occasions for each antibody used, and background has not been significant; therefore isotype controls were not duplicated for these experiments (12). Data was analyzed using Repromac software (True Facts Software, Seattle, WA) and the percentage of each cell type in the total lung and spleen mononuclear population was determined.

For analysis of T cell activation, cells were immunostained with anti- TCR antibody (biotin H57.597; Pharmingen, San Diego, CA) and with antibodies to either CD4 (FITC RM4-5; Pharmingen) or CD8 (FITC 53-6.7; Pharmingen), and with anti-CD69 (Very Early Activation Antigen) (PE, H1.2F3; Pharmingen). NK cells were labeled using a pan NK cell marker (biotin DX5; Pharmingen) and CD3 (FITC 2C11) and anti-CD69. B cells were evaluated using anti-B220 (biotin CD45R; Pharmingen), anti-IA (FITC I-A^d/I-E^d, 2G9; Pharmingen), and anti-CD69. For T cells, CD69 expression by CD8⁺//-TCR⁺ or CD4⁺//-TCR⁺ populations was assessed. For NK cells, CD69 expression by NK⁺/CD-3⁺ cell populations was analyzed. For B cells, CD69 expression by B220⁺/IA⁺ cells was analyzed. Biotin-labeled primary antibodies were detected using either Streptavidin-Cychrome or Streptavidin-APD (Pharmingen).

The T cell memory population was assessed by quantifying the level of expression of CD62 ligand (L-selectin) on CD4⁺ T cells that also express high levels of CD44. T cells that express low levels of CD62L and high levels of CD44 are considered to represent the memory cell population (13). Cells were labeled with antibodies to CD4 or CD8, CD44 (PE: 1M7; Pharmingen), and CD62L (biotin MEL-14; Pharmingen).

To determine if BAL lymphocytes were activated, SP-D^+/+ and SP-D^/ mice 6 wk of age were killed and underwent BAL using PBS, 5 mM EDTA (pH 7.4) as previously described (9). BAL cells were pelleted at 1,000 × g for 10 min, resuspended in PBS, and subjected to FACS analysis.

Analysis of T Cell Receptor V Repertoire of Intrapulmonary T Cells

Intrapulmonary mononuclear cells were stained with antibodies to CD4 or CD8 and biotinylated antibodies to the following T cell receptor V chains: Vb2, 3, 8, 11, and 14; all antibodies from Pharmingen). Cells were washed and stained with streptavidin-PE and then analyzed by flow cytometry.

Enzyme-Linked Immunosorbent Assays for Spontaneous Cytokine Release

The enzyme-linked immunosorbent assays (ELISA) for murine IL-2, IL-4, and IL-10 were purchased from Pharmingen and were performed according to the manufacturer's directions. The ELISAs for total IL-12 and tumor necrosis factor (TNF)- were purchased for Genzyme (Boston MA). The ELISA for murine interferon (IFN)- was performed using the XMG1.2 mAB as a capture antibody and biotinylated R4GA2 mAb as the detecting antibody. Concentrations of the IFN- were determined by comparison to a standard curve generated with recombinant murine IFN- (R&D Systems, Minneapolis, MN). Cytokine concentrations were determined in tissue culture supernatants obtained from unstimulated cell cultures after 18 h in vitro incubation. Cells from lungs and spleen tissues were cultured in complete medium at a final concentration of 2.5 × 10⁵/ml (lungs) and 1 × 10⁶/ml (spleens). The supernatants were harvested and stored at 4°C before analysis. Each sample was assayed in duplicate, with four mice per assay.

Immunostaining

Formalin-fixed paraffin embedded sections were deparaffinized and rehydrated. Endogenous peroxidase was blocked by incubation in 3% H₂O₂ in methanol for 10 min. Antigen retrieval was done using Retrievagen A solution from BD Biosciences (San Diego, CA). Slides were blocked with 3% goat serum and 0.2% Tween in PBS for 30 min and washed with three changes of PBS. Primary antibody, rat anti-CD45 BD (BD Biosciences), was diluted 1:50 in PBS 0.2% Tween 3% goat serum and incubated overnight at 4°C and washed with three changes of PBS. Secondary antibody, biotinylated mouse anti-rat IgG (BD Biosciences), was applied in PBS 3% goat serum for 30 min followed by washing three times with PBS. A Vectastain ABC kit from Vector Labs (Burlingame, CA) was used according to the manufacturers directions using Vector Red staining and haematoxylin counterstain.

Morphometric Analysis

Both lungs from each of six SP-D^/ and six SP-D^+/+ mice were stained with CD45 and used to estimate lymphocyte abundance in three compartments. Interstitial lymphocytes were defined as lymphocytes present in the interstitium of alveoli. Intra-alveolar cells were not scored. Airway-associated lymphocytes were defined as lymphocytes contiguous with the airway epithelium, airway adventitial tissues, or mononuclear cells which were contiguous with the airway tissues. Vessel-associated lymphocytes were defined as lymphocytes contiguous with vessels, vascular adventitial tissues, or vessel-associated mononuclear cells. Twenty randomly chosen fields of 16 µM² were scored for each section with one section for each lung. Because the data was not normally distributed, the Kruskal-Wallis test was used to evaluate differences in lymphocyte density in each of the three compartments between SP-D^/ and SP-D^+/+ mice.

Assessment of Lymphocyte Proliferation

Age-matched mice were given 600 µl BrdU solution (Amersham, Piscataway, NJ) intraperitoneally and killed 24 h later. BAL was performed and the lungs were perfused free of blood, after which they were minced, subjected to enzymatic digestion, and collected on a density step gradient (12). Cytospins were performed from BAL and lung mononuclear cells using at least 50,000 cells. Cytospin slides were stained for PCNA and BrdU using kits obtained from Zymed Laboratories (South San Francisco, CA) following the directions of the manufacturer. Proliferation of lymphocytes was determined by assessing the number of lymphocytes that stained positive for either marker. Six mice of each genotype were analyzed, and 100 lymphocytes per mouse counted. Significance was assessed using an unpaired t test.

RNase Protection Assays

Assays were performed using RiboQuant Multi-Probe RNase Protection Assay kits supplied by Pharmingen. Total lung RNA was purified using the acid phenol extraction method. Templates were labeled following recommendations of the kit manufacturer. Ten micrograms of total lung RNA was hybridized with labeled templates for 16 h, treated with RNase, extracted with phenol/chloroform:isoamyl alcohol and precipitated with ethanol. Protected fragments were resolved on Life Technologies S2 vertical slab gel electrophoresis apparatus (Gibco, Gaithersburg, MD) using a 5% polyacrylamide gel. Autoradiography was performed for various times to evaluate both abundant and rare transcripts. Semiquantitative analysis was performed with scanning densitometry using a Bio-Rad scanning densitometer (Hercules, CA). All fragment intensities were normalized for glyceraldehyde phosphate dehydrogenase (GAPDH) mRNA abundance. For comparisons, four or five mice of each genotype were evaluated twice and normalized band intensity averaged for each mouse. Normal mouse signal was arbitrarily assigned a value of 100. SP-D^/ mice values were expressed as a fraction of the normal mouse value. Not all cytokines assayed by the kits were sufficiently abundant to be detected by the assay and were not reported. Differences in cytokine mRNA abundance were analyzed with GraphPad In-Stat statistics package (San Diego, CA) using one-way ANOVA. Because there were more than five groups, the Tukey post hoc test was applied. Statistical significance was a P value of < 0.05.

IL-5 RT-PCR

To estimate IL-5 transcript abundance, RT-PCR was performed as follows: after DNase treatment, 200 ng each total RNA from four SP-D^/ and four SP-D^+/+ mice was subjected to reverse transcription, using 25 U MuLV reverse transcriptase (Applied Biosystems, Foster City, CA). Following RT, 10% of the RT mix was subjected to 40-cycle PCR using a murine IL-5 Amplifluor Cytokine Direct Gene Systems fluorescence assay kit (Intergen, Purchase, NY) according to the recommendations of the supplier. These methods were sufficient to reliably detect a 25% difference in input RNA. Similar efficiencies of RT and PCR were demonstrated using primers for -actin and results were normalized for -actin abundance after RT-PCR.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Ten-week-old female SP-D^/ and SP-D^+/+ littermates from a heterozygous cross were caged together from birth. Both DNA blot analysis and PCR of tail-clip DNA were used to confirm genotypes. Representative sections are shown in Figure 1. In SP-D^/ mice, lymphocytes are visible near airways, small and large pulmonary vessels adjacent to airways, and also pulmonary veins adjacent to lobar septae. Similar collections of lymphocytes were not seen in SP-D^+/+ littermates using H&E staining.

View larger version (92K): [in this window] [in a new window]   Figure 1.   (A) Photomicrograph of an H and E stained section from a 12-wk-old SP-D/ mouse showing abnormal accumulation of lymphocytes near a large airway (B) and its accompanying pulmonary artery (A). Lymphocytes are associated with large foamy macrophages in the adjacent alveoli. (B) H&E-stained photomicrographs at similar magnification from 12-wk-old SP-D+/+ and SP-D/ littermates that were caged together since birth. Both sections demonstrate small conducting airways and alveolar ducts with associated small vessels. In the SP-D/ mouse, there is a local accumulation of lymphocytes between the vessel and bronchiole. Most of the cells appear to be lymphocytes but there are also increased numbers of macrophages. Interstitial accumulations of lymphocytes are not apparent.

CD45 staining was conducted to identify collections of marrow-derived cells. In Figure 2, CD45-stained lung sections from SP-D^/ and SP-D^+/+ mice are compared. There are perivascular and peribronchial collections of mononuclear cells which have the morphologic appearance of lymphocytes. Lymphoid aggregates are not prominent in the interstitium, in airway lumens, or in the alveolar airspace. Collections of lymphocytes are often seen in close proximity to enlarged alveolar macrophages. Similar lymphocytic aggregates have been seen in 4/4 SP-D^/ mice. Collections of lymphocytes were apparent in (4/4) matched SP-D^+/+ littermates in similar locations, but they were markedly smaller and also less abundant. In total, mice from four separate litters were evaluated morphologically.

View larger version (135K): [in this window] [in a new window]   View larger version (135K): [in this window] [in a new window]   View larger version (131K): [in this window] [in a new window]   View larger version (113K): [in this window] [in a new window]   Figure 2.   CD45 immunoperoxidase sections from a SP-D+/+ mouse (A) and from SP-D/ mice from separate litters (B, C, and D). CD45 is a membrane protein identifying marrow-derived cells. CD45 stains red; haematoxylin was used as a counterstain. (A) Section from a SP-D+/+ mouse at low and high power. The section demonstrates a large airway, pulmonary artery, and pleural surface. Alveolar macrophages and cells associated with the airway having morphologic features consistent with dendritic cells stain for CD45. There is a small focal collection of lymphocytes visible adjacent to the large airway in the low power section. (B) Section from a SP-D/ mouse at low and high power, showing abundant staining of alveolar macrophages. There are several terminal airways and a pulmonary vessel, probably a pulmonary vein, with a focal accumulation of mononuclear cells. High power shows that these cells are morphologically consistent with lymphocytes. (C) Section from an SP-D/ mouse demonstrating a large airway and pulmonary artery at high and low power, demonstrating staining of alveolar macrophages and prominent perivascular accumulation of lymphocytes. (D) Section from a 12-wk-old SP-D/ mouse that was not perfused free of blood. There is a perivascular accumulation of lymphocytes as well as accumulation around a large bronchus and a respiratory bronchiole. High power demonstrates an increased abundance of cells with morphologic characteristics of lymphocytes near the bifurcation of a small bronchiole surrounding a small blood vessel.

To determine the total number of lung mononuclear cells and their distribution between subtypes in SP-D^/ and SP-D^+/+ littermates, whole lung and spleen were minced, enzymatically disassociated, and filtered through a 70-µM centrifugal filter. Viable cells were counted before and after sedimentation through a Ficoll gradient and efficiency of mononuclear cell isolation was identical between SP-D^/ and SP-D^+/+ mice. Total numbers of lung mononuclear cells exclusive of pulmonary macrophages were not significantly different between null and normal mice, as shown in Figure 3A. Because the numbers of lymphocytes appeared to be increased by both H&E staining and CD45 immunohistochemical staining, we evaluated our lung disassociation protocol. Cytospin analysis of disassociated lung before filtration demonstrated aggregates of undigested airway epithelium, which in the case of SP-D^/ mice was associated with lymphocytes by CD45 staining. Aggregates could be eluted from the 70-µM filter retentate and were therefore not subjected to density gradient (data not shown). Such aggregates of airway cells and lymphocytes therefore may not be well represented in lymphocytes obtained from disassociated lung. Total lung lymphocytes could therefore be underestimated. The distribution of lymphocytes between CD4⁺, CD8⁺, NK, and B cell lineages is shown in Figure 3B and was likewise not significantly different between genotypes (P > 0.05).

View larger version (14K): [in this window] [in a new window]   Figure 3.   Similar numbers of mononuclear cells are present in lung digests from SP-D/ and SP-D+/+ mice. (A) Lung mononuclear cells were harvested by collagenase digestion from the lungs of age-matched SP-D+/+ and SP-D/ mice (four mice per group) and purified by Ficoll density gradient centrifugation, as described in MATERIALS AND METHODS. (B) Mononuclear cells (excluding lung macrophages) were enumerated by manual counting and the mean (± SE) cell number CD4+, CD8+, NK1.1+ and B220+ cells in the mononuclear cell fraction obtained from lung digests, as described in MATERIALS AND METHODs. Open bars, wild type; solid bars, SP-D/. The mean percentage of each cell type was determined and plotted. Similar results were obtained in two additional experiments.

Because lymphoid aggregates were apparent in SP-D^/ mice, but increased numbers of lymphocytes were not observed by FACS analysis of disassociated lung, we performed morphometric analysis of CD45-stained lung sections (Figure 4). We found significantly increased numbers of airway- and vessel-associated lymphocytes in SP-D^/ mice but did not demonstrate significant differences in interstitial lymphocytes. Therefore redistribution of lymphocytes from interstitial to airway and vascular regions is unlikely. We believe that our data may be explained by inadequate digestion of peribronchial tissues and must be interpreted cautiously because lymphocytes from those regions may have been poorly represented in samples subjected to FACS analysis.

View larger version (12K): [in this window] [in a new window]   Figure 4.   Morphometric analysis of CD45-stained lung slices from six age-matched SP-D/ (solid bars) and six SP-D+/+ (open bars) mice. Lymphocyte density in each of three compartments (peribronchial, perivascular, and interstitial) was estimated as described in MATERIALS AND METHODS. There were significant increases in lymphocyte density in the peribronchial and perivascular compartments in SP-D/ mice, but no significant difference between the interstitial compartments. These findings suggest that there are increased total lymphocytes in SP-D/ mice rather than redistribution between compartments. *P < 0.05, #P > 0.05.

To assess pulmonary lymphocyte activation, expression of CD69 (early activation antigen) and CD25 (IL-2 receptor) was determined flow cytometrically. In Figure 5A, a representative flow cytometric analysis of T cell CD69 expression is presented. CD4⁺ lung lymphocytes from SP-D^/ mice demonstrated a 5-fold increase in the percentage of CD69⁺ cells (P < 0.01), whereas the percentage of CD8⁺ lung lymphocytes positive for CD69 was increased 2-fold (P < 0.05). There was no significant difference in CD69 expression by spleen lymphocytes from SP-D^/ mice compared with normal control mice. There was also approximately a 2-fold increase in the percentage of CD25⁺ intrapulmonary CD4⁺ and CD⁺ T cells in SP-D^/ mice (data not shown), whereas there was no difference in CD25 expression by spleen lymphocytes in either group of mice. Mouse lungs subjected to both BAL and vascular perfusion showed results that were not significantly different from mice analyzed without vascular perfusion and BAL.

View larger version (20K): [in this window] [in a new window]   Figure 5.   Selective activation of T cells in the lungs of SP-D/ mice. Lung mononuclear cells recovered by enzymatic digestion of lungs from SP-D/ and wild-type mice were evaluated for expression of the early activation marker CD69, using flow cytometry. In addition, spleen cells from the same mice were prepared by NH4C1 lysis and also evaluated by flow cytometry. Cells were triple immunostained for expression of / TCR, CD69, and either CD4 or CD8, as described in MATERIALS AND METHODS. A representative histogram of CD69 expression by CD4+ pulmonary parenchymal T cells from a SP-D+/+ and a SP-D/ mouse is shown in the top panels (left and right, respectively). The mean percentage (± SE) of CD4+ or DC8+ T cells positive for CD69 expression was determined for lung mononuclear cells (middle panels) and spleen cells (bottom panels), using cells isolated from four age-matched mice per group. *Significant difference (P < 0.05) in the percentage of CD4+ and CD8+ T cells positive for CD69 expression. NK cells and B cells were also evaluated, but the number of cells expressing CD69 was not significantly different in the lungs of SPD/ mice versus wild-type mice (data not shown).

When BAL lymphocytes were subjected to FACS analysis, there was approximately a 4- to 5-fold increase in the abundance of CD4⁺ and CD8⁺ lymphocytes in BAL from SP-D^/ mice (n = 6 per genotype, P < 0.02 by the Mann-Whitney test). There was a 6-fold increase in the number of SP-D^/ CD4 lymphocytes expressing CD69 (P < 0.02 by the Mann-Whitney test). There was a 4-fold increase in the abundance of CD8 lymphocytes from SP-D^/ that expressed CD69, but the difference was not statistically significant (P > 0.05). Confirming previous observations at the light microscopy level, there was a 6-fold increase in cells expressing Mac1 in BAL from SP-D^/ mice (P < 0.01) even when large alveolar macrophages were excluded by forward scatter.

Memory T lymphocytes express high levels of CD44 and low levels of CD62 ligand (L-selectin), indicating previous antigen exposure, whereas naive lymphocytes express low levels of CD44 and higher levels of CD62L. Thus CD4⁺ T cells that are CD44 high and CD62L low represent a T cell memory population. Lymphocytes obtained from SP-D^/ and SP-D^+/+ mouse lung digests were stained for CD44 and CD62L and subjected to flow cytometric analysis (Figure 6). A total of 45 ± 7% of lung-digest CD4 cells from SP-D^+/+ mice and 69 ± 4% of CD4⁺ cells from SP-D^/ mice demonstrated high CD44 fluorescence (P < 0.001), whereas 42 ± 3.8 of CD4 cells from SP-D^+/+ and 23 ± .6% CD4 cells from SP-D^/ mice demonstrated high fluorescence for CD62l (P < 0.001). These results suggest that the pool of memory T cells CD44 (hi), CD62L (lo) was expanded in the SP-D^/ mice relative to their littermate controls.

View larger version (18K): [in this window] [in a new window]   View larger version (29K): [in this window] [in a new window]   Figure 6.   Increase in memory T cells in the lungs of SP-D/ mice. CD4+ T cells isolated from the interstitial tissues of SP-D/ and SP-D+/+ (wild-type) mice were analyzed for expression of cell surface molecules associated with the memory phenotype (CD44 and CD62 ligand). In panel A, a typical histogram of CD62L expression by pulmonary T cells from SP-D/ and SP-D+/+ mice indicates a decrease in the number of CD4 cells expressing CD62L in SP-D/ mice, representing a reduction in the frequency of the naïve phenotype. In panel B, the fraction of CD4 cells expressing CD62L with high fluorescence for CD44 (four mice per group) is plotted. CD4+, CD44 hi T cells from SP-D/ mouse lungs expressed significantly lower levels of CD62L, consistent with an increased number of cells displaying the memory phenotype, compared with SP-D+/+ mice. In panel C, the fraction of CD4+ cells expressing either CD44 (hatched bars) or CD62L (solid bars) is represented. Approximately 30% more CD4+ lymphocytes from SP-D/ mice stain for CD44 (P < 0.001) whereas ~ 50% fewer stain for CD62L (P < 0.001). These findings are consistent with a reduction in naive CD4+ lymphocytes and an increase in memory CD4+ lymphocytes in SP-D/ mice.

We also evaluated pulmonary T cells for evidence of either oligoclonal expansions or deletions. This was done to determine if in the absence of SP-D, there might have been preferential activation and expansion of one or several clones of T cells responding to one or a few dominant antigens. Such a response might be reflected as an increase in the population of T cells expressing a certain T cell receptor V chain. Pulmonary T cells (both CD4⁺ and CD8⁺) were evaluated using antibodies specific for six different T cell receptor V chains. This analysis did not reveal any significant differences in the percentages of pulmonary lymphocytes expressing V2, V3, V6, V8, V11, or V14 in SP-D^/ mice compared with SP-D^+/+ mice (data not shown).

To determine if lymphocytes were polarized to the Th1 or alternatively to the Th2 phenotype, RNase protection assays (RPA) were conducted using RNA from lung homogenate from SP-D^/ and SP-D^+/+ mice. There were statistically significant reductions in Eotaxin and Rantes mRNA (Figure 7). IFN- and IL-4 mRNA was not different but IL-6 expression was increased ~ 3-fold in SP-D^/ mice. (Table 1). IL-5 transcript was not sufficiently abundant to identify expression using our RPA methodology. Therefore, to assess differences in IL-5 expression between SP-D^/ and SP-D^+/+ mice, we performed RT-PCR analysis of samples from four age-matched mice of each genotype. There was no significant difference in the abundance of IL-5 transcript by these methods (data not shown). To assess cytokine production, ELISA of supernatants from cultured-disassociated lung mononuclear cells were assayed for IL-2, IL-4, IL-10, IL-12, TNF-, and IFN- using commercially available kits. There was 2-fold increase in IL-12 expression by lung mononuclear cells from SP-D^/ mice (P < 0.05) but there was no difference in the expression of other cytokines between SP-D^/ and SP-D^+/+ mice by either lung cells or spleen cells. IL-2 production by lung mononuclear cells from SP-D^/ mice was increased 6- to 8-fold but did not reach statistical significance given the small number of samples and high variability (P > 0.07).

View larger version (59K): [in this window] [in a new window]   Figure 7.   Representative RNase protection assay using 4 age-matched mice of each genotype demonstrating differential expression of Eotaxin and Rantes by SP-D/ and SP-D+/+ mouse lung. Expression is normalized for both L32 and GAPDH. There is an approximately 4- to 5-fold reduction in expression of Eotaxin and Rantes in SP-D/ mice.

To assess proliferation of BAL and lung tissue lymphocytes, mice were given BrdU 24 h before BAL and lung digestion. Both PCNA and BrdU immunostaining was done on separate slides. Proliferating lymphocytes were identified by positive staining for either antigen. There was a significant difference in proliferation by PCNA staining with 6.7 ± 1.38% SP-D^/ lung digest lymphocytes positive for PCNA and 4.8 ± 1.07% SP-D^+/+ lymphocytes positive for PCNA (P < 0.02, n = 7/genotype). There was not a significant difference in BAL lymphocyte proliferation between genotypes by PCNA analysis. Incorporation of BrdU into lung and BAL lymphocytes was very low in both genotypes and was also low in bronchial epithelium but incorporation was evident in gastric mucosa. We believe these results can be explained by poor lung penetration of intraperitoneally injected BrdU.

Finally, to determine if activated alveolar macrophages might in turn lead to activation of lung T lymphocytes, equal numbers of BAL alveolar macrophages from SP-D^+/+ or SP-D^/ mice were added to cultured lymphocytes from normal mouse lung and normal mouse spleen for 24 h. Lymphocytes were then subjected to FACS analysis to determine if there were differences in the expression of either CD69 or CD25. There was no difference in expression of activation markers or thymidine incorporation (data not shown).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

SP-D^/ mice from different genetic backgrounds (8, 9) consistently demonstrate abnormal accumulations of lymphocytes around bronchi and pulmonary vessels. Acute inflammation and local tissue destruction are not apparent. Tissue anti-CD45 immunostaining also demonstrates collections of cells that appear to be lymphocytes in perivascular and peribronchial regions. Such collections are not uniform and are often observed in association with enlarged macrophages, suggesting that SP-D^/ macrophages might be responsible for accumulation of lymphocytes. Similar collections of lymphocytes have been described in several models of lung inflammation that are mediated by specific immunoreactivity to inhaled antigens (14). Our colony of SP-D^/ mice is maintained under standard vivarium conditions. We have never been able to document infectious disease in these mice using serologic techniques, bacteriologic and fungal culture, or ribosomal RT-PCR for Chlamydia and Mycoplasma. Therefore it is not likely that such accumulations are due to occult infectious disease.

Contrary to our expectations from examination of lung histology, we did not observe increases in the number of total lung digest lymphocytes in SP-D^/ mice. One likely explanation for this observation is that lung tissue was not adequately digested to release peribronchial and perivascular collections of lymphocytes. The presence of aggregates of airway and vascular tissue with lymphocytes after enzymatic digestion supports this conclusion. Morphometric assessment of lymphocyte density by location confirms that there are increases in lymphocytes around airways and vessels without changes in the abundance in interstitial lymphocytes, suggesting that there is an increase in total lung lymphocytes rather than redistribution from the interstitial note compartment. Because there appears to be an increase in total lung lymphocytes by visual inspection and morphometric analysis, it is likely that our methods are not sufficient to estimate total lung lymphocyte abundance. We also may have underestimated the abundance of specific lymphocyte lineages and their activation in the peribronchial/perivascular compartments. This may be very important because antigen may be encountered in those compartments and antigen presentation may be particularly effective in those regions because of the presence of dendritic cells. Our findings must be interpreted with these limitations in mind.

The relative distribution among lymphocyte classes (CD4⁺, CD8⁺ T cells, NK cells, or B-cells) was not significantly different between normal and SP-D^/ mice. However, if more mice were evaluated, the small apparent differences we observed might have become statistically significant. CD4⁺ and CD8⁺ pulmonary lymphocytes from SP-D^/ mice were activated based on increased expression of both CD69 and CD25, whereas splenic T cells from these same mice were not activated compared with littermate controls. The upregulation of expression of CD25 (the IL-2 receptor) and CD69 are consistent with activation by antigen. Moreover, there was a 2-fold increase in the percentage of CD44(hi)/CD62L(lo)T cells in SP-D^/ mice compared with SP-D^+/+ controls, indicating a significant increase in the proportion of memory CD4⁺ lymphocytes relative to the population of naive CD4⁺ lymphocytes (13). Therefore, not only are T lymphocytes activated in SP-D^/ mice, there is also evidence of previous antigen exposure. The relatively modest alterations in cytokine gene expression may indicate either that effector cells are not abundant, or are not actively producing the cytokines we evaluated.

There were not, however, significant differences in fractions of B-lymphocytes or NK cells expressing CD69 or CD25 between SP-D^/ and normal mice. These findings are consistent with specific activation of CD4 and CD8 lymphocytes by antigen.

We demonstrated increased abundance of both CD4 and CD8 lymphocytes in BAL from SP-D^/ mice compared with normal littermates. BAL CD4 lymphocytes also demonstrated a statistically significant 6-fold increase in expression of CD69, indicating recent and ongoing activation. CD8 lymphocytes had a 4-fold increase in expression of CD69, though this increase was not statistically significant. Thus, increased numbers of activated T lymphocytes enter the airspace in SP-D^/ mice. The entry of activated lymphocytes into the airways of SP-D^/ mice may be the result of nonspecific emigration from a pool of activated lymphocytes in lung interstitial tissues or may instead reflect an alteration in the trafficking of lymphocytes as a consequence of the lack of SP-D.

Lymphocytes obtained by lung disassociation from SP-D^/ mice demonstrated a significant increase in PCNA staining compared with those from SP-D^+/+ mice, consistent with increased in situ proliferation. These findings would be expected if lymphocytes in lung tissue were continuously being exposed to antigen in the presence of effective antigen presenting cells. BAL lymphocytes on the other hand, which may be more mature effector cells, did not demonstrate differences between genotypes in PCNA staining. Taken together, these data suggest that the absence of SP-D in the lungs of mice leads to a state of persistent local activation of pulmonary T cells with in situ proliferation and/or increased longevity for lung-associated memory lymphocytes.

To assess the possibility that one or several dominant antigens might be responsible for activation, we analyzed TCR V chain repertoires expressed by intrapulmonary and BAL T lymphocytes. This analysis did not reveal any differences in the T cell receptor repertoire between SP-D^/ and normal mice. These methods do not exclude stimulation by a small number of antigens, but reduce the likelihood of that mechanism.

Persistent local T cell activation can be driven by antigens or by overproduction of cytokines such as TNF- or type I interferons (12, 17, 18). Analysis of cytokine expression by ELISA and RT-PCR showed slight reductions in Eotaxin and Rantes, no difference in IFN- expression, and increases in IL-6 and IL-12 expression in SP-D^/ mice compared with SP-D^+/+ mice. Expression of IL-4, IL-5, and TNF- was not increased; IFN- and IFN- were not measured. Therefore, there was not an overall increase in T cell-activating cytokines, nor was there a clear polarization toward either a Th1- or Th2-promoting environment or phenotype. Additionally, cytokine expression did not seem sufficient to activate T lymphocytes directly. One explanation for such observations is that activated memory T cells may accumulate around airways and vessels in response to an ongoing antigenic stimulus, but the cytokine environment is not sufficient to induce a marked polarization toward either the Th1 or Th2 phenotype.

Exposure to antigen in the presence of LPS leads to prolonged residence of tissue memory lymphocytes, especially in the lung (22). SP-D^/ mice demonstrate a more intense inflammatory response to both gram-negative and gram-positive organisms. Likewise, intratracheally instilled LPS produced a more pronounced acute inflammatory response in SP-D^/ mice compared with both normal littermates and mice overexpressing SP-D (K. E. Greene, S. Ye, and J. H. Fisher, unpublished observation). It is not known if increased sensitivity to inhaled environmental LPS could account for persistence of memory lymphocytes in the lungs of SP-D^/ mice.

Alveolar macrophages are not efficient antigen-presenting cells under normal circumstances. However, it is possible that macrophages in SP-D^/ mice may develop into more efficient antigen-presenting cells. Because lymphocytes appeared to accumulate near collections of enlarged alveolar macrophages, we sought to demonstrate a direct effect of such alveolar macrophages on lymphocyte activation. Mixing studies using alveolar macrophages from BAL of both SP-D^+/+ and SP-D^/ did not result in activation of spleen lymphocytes. Therefore, alveolar macrophages from SP-D^/ mice alone may not be sufficient to induce T cell activation (at least in vitro), despite the fact that the macrophages are themselves already activated. Activated macrophages in SP-D^/ mice may activate and increase antigen presentation by adjacent antigen-presenting cells such as pulmonary dendritic cells or blood borne monocytes. Additionally, SP-D might alter antigen recognition in regional nodes or the absence of SP-D might increase the number of local dendritic cells that might further facilitate antigen recognition and response. These experiments do not allow conclusions regarding the mechanism by which SP-D regulates lymphocyte proliferation in vivo. For example, it is possible that absence of SP-D may lead to an IL-2-mediated increase in lymphocyte proliferation as has been observed in vitro (6, 7). Taken together with the relatively modest increases in proinflammatory cytokines in the Lungs of SP-D^/ mice, direct lymphocyte activation by cytokines produced by the morphologically abnormal alveolar macrophages that are characteristic of SP-D^/ mice seems unlikely.

Pulmonary leukocytes are less responsive to antigenic stimuli than peripheral blood leukocytes (19). In vivo experiments have linked airway lining material to antigenic hyporesonsiveness (20, 21). Borron and others proposed the hypothesis that SP-D might contribute to inhibition of T cell proliferation in the lung and that this effect might in part account for a relatively antigenic hyporesponsive state in the lung (6). In vivo deficiency of SP-D leads to activation of pulmonary T-lymphocytes in the absence of activation of B-lymphocytes or NK cells. We did demonstrate findings compatible with activation of T-lymphocytes by an antigen or antigens. Whether antigen recognition or another mechanism is responsible for T cell activation in the lungs of SPD^/ mice remains to be clarified.