The authors have declared that no competing interests exist.
Surfactant protein A (SP-A) plays a number of roles in lung host defense and innate immunity. There are two human genes,
Surfactant protein-A (SP-A) is a member of the collectin family with an N-terminal collagen-like domain and a C-terminal carbohydrate recognition domain (CRD) that binds to many macromolecules, pathogens, and allergens
SP-A influences host defense function in a variety of ways. It recognizes and binds to pathogen-associated molecular patterns. These interactions are complex and may involve binding sites in addition to the CRD. Direct interaction with pathogens constitutes one aspect of its host defense function, but SP-A also aids in the clearance of particulates, allergens, and debris
Several studies have identified functional differences between SP-A1 and SP-A2 in innate immune functions (including many of those mentioned above), and in several surfactant-related functions. These included cytokine production
The AM, the primary effector cell for lung innate immunity, exhibits a unique phenotype
For this study, we used wild-type C57BL/6 mice obtained from the Jackson Laboratory (Bar Harbor, ME). SP-A KO mice were propagated in the animal core facility of the Pennsylvania State University College of Medicine. Humanized TG SP-A1 and SP-A2 mice that each carried an SP-A1 or SP-A2 variant were generated on the SP-A KO C57BL/6 background as previously described
All procedures involving animals used protocols that were approved by the Institutional Animal Care and Use Committee at the Pennsylvania State University College of Medicine. All mice were maintained in facilities under pathogen-free conditions or in barrier containment facilities.
AM were obtained by BAL
To study the copy number of the transgenes in the SP-A1 and SP-A2 hTG mice, total DNA was extracted and purified from the mouse tails and human lung tissue using the DNeasy Blood & Tissue Kit (Qiagen, Valencia, CA). Eight μg of pure DNA from each sample were digested overnight at 37°C with the restriction enzyme EcoRI and the digestions were subjected to 0.7% agarose gel electrophoresis and transferred to a Nylon membrane for Southern blot analysis. Two DNA probes (0.8 and 1.3 kb) amplified from human SP-A cDNA of SP-A2 (1A3) were labeled with digoxigenin (PCR DIG Probe Synthesis Kit, Roche, Indianapolis, IN) and used for hybridization to detect the SP-A transgenes by Southern blot. The 0.8kb probe was generated using primers 2058/2059 and contains the coding region of exons I-IV, and the 1.3kb probe was generated with primers 2058 and 2060 and contains the coding region (exons I-IV) plus a partial 3’UTR sequence which is part of exon IV. The sequences of the probes are listed in Supplementary
Quantitative PCR (qPCR) was conducted in 10 ml reactions (384-well plates) consisting of 10 ng of mouse genomic DNA, TaqMan SP-A1, SP-A2, or mouse reference assays, and TaqMan master mix (Applied Biosystems, Foster City, CA, USA). Custom-designed TaqMan® Assays were purchased from Applied Biosystems (Foster City, CA, USA) for human SP-A1 (assay Hs01921510) and human SP-A2 (assay Hs00359837). The mouse Trfc TaqMan® Copy Number Reference Assay (Applied Biosystems, Foster City, CA, USA) was used as reference. Gene amplification was achieved using the following protocol: denaturing for 10 min at 94°C, and 40 cycles of 94°C for 15 sec and 60°C for 60 sec. Four replicates were run per reaction. These were monitored with the 7700 Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). A standard curve was also amplified using serial dilutions of the vector used to generate the transgenic mice, ranging from 0.1 to 1000 copies of the transgene. The mass of one copy of the plasmid was calculated by the formula described by Joshi et al
Transgene copy number was calculated by the standard curve method
Endpoint PCR was performed in 400 ng of mouse genomic DNA with the Fast Start High Fidelity PCR system (Roche Diagnostics, Indianapolis, IN) and specific primers listed in
Initial characterization of BAL fluid was done to confirm the presence of the expected SP-A protein. Blots were prepared as described subsequently and immunostained with either an antibody specific to SP-A1
To determine the amount of secreted hSP-A in the lungs of hTG mice, lungs were lavaged and lavage samples (25 µl of BAL/sample) were subjected to gel electrophoresis (12.5% SDS-PAGE) under reducing conditions. Human SP-A in the BAL fluid was detected by Western blot analysis using a specific SP-A antibody that recognizes both SP-A1 and SP-A2, as described previously
For the proteomic study we used humanized transgenic mice expressing SP-A1 and SP-A2 at high levels and two mouse lines that expressed SP-A1 and SP-A2 at low levels, in addition to KO mice and WT mice. The SP-A1 mouse lines were 6A2 T1 (subsequently referred to as 6A2) that expressed high levels of SP-A1, and a mouse expressing low levels of SP-A1 that is designated 6A4-LE. The SP-A2 mice consisted of two high expressing mouse lines, 1A0-T10 and 1A0 T13, subsequently called 1A0 mice, and a low expressing mouse line designated 1A0-LE. The T stands for transgene and the number next to it indicates the particular mouse line expressing a given SP-A variant.
To prepare AM for 2D-DIGE frozen AM pellets were lyophilized until completely dry and resuspended in 25 µL of standard cell lysis buffer (30 mM Tris-HCl, 2 M thiourea, 7 M urea, 4% CHAPS, pH 8.5). Protein determinations were done using the Bio-Rad Protein Assay (Bio-Rad, Hercules, CA) and the concentration of protein was adjusted to 1mg/ml for labeling with either Cy3 or Cy5. Twenty micrograms of protein was used for each sample. These procedures and the subsequent 2D-DIGE were performed as we have described in detail
Data were normalized between three independent 2D-DIGE experiments using the KO AM samples present in each experiment. The mean KO AM value was determined for each protein for each of the three experiments. These means were then used to calculate a correction factor for each protein and used to normalize the data for all experiments.
Heat maps were generated in Excel (Microsoft). Means of the normalized values for each protein were subject to conditional formatting with the highest value assigned a red color, middle values a yellow color, lowest values a green color, and intermediate values colored with the appropriate shades depending upon where they fall within the range for that protein. All heat maps included values for lines containing the two highly expressed transgenes, the transgenes expressed at low levels, the WT mice, and the KO mice. For each analysis one group of AM was considered the index group and its proteins were arranged so that they went from high to low expression levels for that group.
First, we confirmed the presence of the transgene by Southern Blot (
|
|
---|---|
2 | |
6A4 T1 | 3-4 |
0 | |
0 | |
|
|
1A0 T7 | >10 |
>10 | |
>10 | |
1A3 T2 | 1-2 |
1A3 T4 | 1-2 |
0 | |
0 |
T stands for transgene and the numbers next to the T indicate the particular mouse line that expressed a given SP-A variant. The numbers shown in the right hand column depict the copy caller values. The asterisks (
Next we determined transgene orientation, and concatemer occurrence by end point PCR, with primers described in
No bands were detected in negative controls of one copy (human DNA, plasmid DNA of 1A3-T4 and 6A4-T3) or controls without the transgene (WT, KO). The presence of head to tail concatemers appeared to result in a band(s) of approximately 6.6kb size shown in
We used some of these mice to study the impact of chronic SP-A variant exposure on the AM phenotype using a proteomic approach. SP-A expression in the mouse lines used in the present study was first assessed by antibodies specific to SP-A1 and SP-A2 to confirm the presence and specificity of SP-A1 and SP-A2 expression in each transgenic line, as described in Methods and Materials. Because the affinities of these antibodies were different we could not use these for quantitative determinations. Thus, we measured total SP-A levels in the BAL of each mouse with an SP-A antibody recognizing both SP-A1 and SP-A2. Total SP-A levels were then measured on 25 µl of BAL/sample.
In this study we compared SP-A KO mice or WT mice with several mouse lines containing human SP-A transgenes to determine the effects of endogenous expression of the human SP-A transgenes on the AM proteome. The high-expressing SP-A1 (6A2) and SP-A2 (1A0) lines used in the proteomic study are marked with an asterisk in
|
|
|
---|---|---|
6A4-LE (SP-A1) | 11 | 31 |
1A0-LE (SP-A2) | 8 | 34 |
6A2 (SP-A1) | 26 | 47 |
1A0 (SP-A2) | 19 | 17 |
The table summarizes the specific differences seen in each comparison of transgenic mice to either KO or WT mice.
AM from mice expressing higher levels of SP-A1 and SP-A2 had many more significant differences from KO, 27 in SP-A1 (6A2) and 19 in SP-A2 (1A0), suggesting some degree of dose dependency as compared to the “low expressing” lines. However, the greater number of significant changes vs KO in the 6A2 mice as compared to the 1A0 mice suggests that SP-A dose dependency (based on the differences in SP-A levels between the 1A0 and 6A2 lines (
We organized the map by arranging all 76 proteins in the WT mice on the basis of highest to lowest levels and then compared WT to the other groups (
The impression obtained from examining the actin-related proteins (
Similar patterns, where the 1A0 resembled that of WT, were also seen among the Nrf-2 regulated proteins (
Both low expressing hTG lines were very different from KO, but very similar to one another
The similarity between these two lines is perhaps best appreciated in heat maps of the different functional groups after stratification of the heat maps based on expression levels in AM from SP-A KO mice
A potential limitation of this study is the fact that the different SP-A variants were expressed at different levels. However, in normal and diseased humans SP-A levels in the BAL are quite variable as well
In the present study we compared KO mice to 1A0 high expressing mice that had been stratified further within this high-expressing group into a higher subgroup and a lower subgroup, as well as the 1A0 low expressors. When this was done, although in some instances protein expression levels varied proportionally with SP-A levels, in most cases it did not, and only two of these cases involved significant changes from the KO values. It was interesting to note that in several of the examples where protein expression varied with 1A0 levels, the protein changes with 6A2 were in the other direction (decrease vs. increase or vice versa). In the only case where protein expression differences from KO in 1A0 and 6A2 mice appeared proportional to SP-A levels, the difference vs KO was significant in both 1A0 and 6A2 mice. This analysis led us to draw the conclusion that under these basal conditions most, if not all, of the effects observed occur after reaching a given threshold level rather than varying in proportion to SP-A levels. However, the possibility remains that in the presence of an insult (i.e oxidative stress) the much higher levels may better maintain the alveolar macrophage phenotype, since some of the SP-A could be oxidized, and this can reduce its functional activity
It should be noted that despite the fact that the 6A2 mice had lower levels of SP-A than 1A0 mice did, they exhibited almost 50% more significant changes vs KO, further reinforcing that the different response patterns reflect primarily functional differences between 1A0 and 6A2 rather than differences between absolute amounts of SP-A. Further evaluation of the expression data for individual proteins in this study did not show that differences in specific protein levels between the 1A0 and 6A2 mice were due to levels of SP-A, but rather due to the specific variant present. This was also supported by the very different response patterns elicited by 1A0 and 6A2. At the other end of the spectrum, with the low expressing mouse lines, while we were unable to measure SP-A using the same gel parameters as with the high expressing group, the protein expression patterns in these mice was very different from that present in the SP-A KO group. This indicated that even very low levels of SP-A have a profound effect on protein expression, and this expression profile is very different from that seen when the proteins are expressed at higher levels.
Another potential limitation is whether the effects of human SP-A on mouse alveolar macrophages provides a good surrogate for the human lung. While many gaps remain in our knowledge of the mechanism of action of SP-A and the specific receptors involved, similarities in the effects of human SP-A on phagocytosis in human and rat alveolar macrophages
In the present study we compared the AM proteome of hTG mice to KO mice and were able to gain insight into changes in protein expression resulting from the presence of the SP-A transgenes. We studied the expression of 76 AM proteins using a 2D-DIGE experimental design. We found that even when SP-A was expressed at low levels the AM proteome differed markedly from that of SP-A KO mice. While there were relatively few significant differences from KO, the similarity in these responses led us to speculate that these low amounts of SP-A might be affecting the AM by acting through a high affinity receptor that was not variant-specific.
On the other hand, when we studied the mice expressing higher levels of SP-A there were more than twice as many significant changes, with greater numbers of changes occurring in the SP-A1 (6A2) mice. The other difference between these mouse lines that became apparent with the study of heat maps was that these two mouse lines produced AM with dramatically different phenotypes. Examining both the overall protein expression pattern as well as the individual functional groups, clearly showed that the patterns of expression in the 1A0 and 6A2 mice were almost the inverse of one another (proteins at their highest levels in 1A0 were at their lowest levels in 6A2, and vice versa). Another very interesting feature of the 1A0 expression pattern was its similarity to the WT pattern. These observations led us to speculate that when high levels of SP-A1 and SP-A2 are present they acted in a variant-specific manner, either through different receptors or by eliciting distinctly different responses through a single receptor.
Further studies are required to determine how these changes in the AM proteome in various humanized SP-A hTG mice affect AM function. On the basis of the similarities between WT and the SP-A2 hTG mice we would predict that SP-A2 transgenics would clear bacteria more effectively from their lungs than SP-A1 mice and that their AM would exhibit increased phagocytic function, as we have previously demonstrated in vitro with AM treated with exogenous SP-A2
1.AM protein expression patterns differed markedly from SP-A KO in hTG mouse lines generated on the SP-A KO background.
2.The AM protein expression patterns from mouse lines expressing low levels of SP-A1 and SP-A2 were similar to each other but different from KO indicating a lack of variant-specific effects at very low levels of SP-A.
3.SP-A1 and SP-A2 at higher levels exhibited very different AM protein expression patterns indicating variant-specific patterns at higher levels of expression of hSP-A.
4.AM protein expression in SP-A2 hTG mice was similar to WT and the pattern from SP-A1 mice was very different from WT.
The authors thank Susan DiAngelo for genotyping. Current address for Guirong Wang: Department of Surgery, SUNY Upstate Medical University, Syracuse, NY 13210.
This study was supported in part by R01 ES009882 from the National Institute of Environmental Health Sciences.