Actin fence therapy with exogenous V12Rac1 protects against acute lung injury

High mortality in Acute Lung Injury (ALI) results from sustained proinflammatory signaling by alveolar receptors, such as TNFα receptor type 1 (TNFR1). Factors that determine the sustained signaling are not known. Unexpectedly, optical imaging of live alveoli revealed a major TNFα-induced surge of alveolar TNFR1 due to a Ca2+-dependent mechanism that decreased the cortical actin fence. Mouse mortality due to inhaled LPS was associated with cofilin activation, actin loss and the TNFR1 surge. The constitutively active form of the GTPase, Rac1 (V12Rac1), given intranasally as a non-covalent construct with a cell-permeable peptide, enhanced alveolar F-actin and blocked the TNFR1 surge. V12Rac1 also protected against ALI-induced mortality resulting from intranasal (i.n.) instillation of LPS, or of Pseudomonas aeruginosa. We propose a new therapeutic paradigm in which actin enhancement by exogenous Rac1 strengthens the alveolar actin fence, protecting against proinflammatory receptor hyperexpression, hence blocking ALI.


Introduction
Life-threatening tissue injury to critical organs occurs as a result of host-pathogen interactions involving proinflammatory receptors. In lung, the resulting inflammation underlies acute lung injury (ALI), which can lead to the acute respiratory distress syndrome (ARDS), a condition that associates with high mortality (1). Pharmacological therapies are not available for ALI, but are required in order to stem disease progression.
ALI due to inhaled Gram-negative bacteria occurs through initiating and progressive phases. In the initiating phase, inhaled pathogens ligate toll-like receptors on macrophages and the alveolar epithelium (2, 3) causing release of proinflammatory cytokines, such as TNFα that ligate alveolar epithelial receptors (4). Crosstalk with the endothelium follows (5) and chemoattractants, such as IL-8 activate inflammatory cell recruitment (6). In the progressive phase, recruited and resident immune cells continue to secrete cytokines thereby sustaining the inflammatory response (7,8).
The extent to which the initiating mechanisms continue to enhance the progressive phase of the response remains unclear. Here we addressed this question in the context of the interactions of TNFα with its alveolar receptor, TNFR1.
Studies from multiple cell types indicate that following ligation, TNFR1, a transmembrane protein, sheds its ectodomains (9)(10)(11). The sheddase, ADAM-17 mediates the shedding (10), which is protective since shedding inhibition augments lung inflammation (9). The shed domains are not recycled by the cell from which the shedding occurred (12). Brefeldin A, the inhibitor of protein trafficking to the Golgi, abrogated TNFR1 receptor mobilization and decreased TNFR1 abundance on the cell surface, indicating that membrane replenishment of TNFR1 occurs by receptor trafficking from storage pools in the Golgi (11,(13)(14)(15). The replenishment is translation independent (13,16). However, the extent and the time course of TNFR1 replenishment in the alveolar epithelium remain unknown.
In this regard, the regulatory role of cortical actin remains unclear. Cortical actin is the layer of filamentous actin (F-actin) that forms a network adjacent to the plasma membrane (PM).
Although cortical actin is known to form a "fence" against trafficking of vesicles and receptors to the PM (17,18), dynamic data from live alveoli are lacking detailing the role of the actin fence in TNFR1 expression. Here, we addressed these issues by application of confocal microscopy of the live alveolar epithelium. Our goals were to determine strategies by which fence enhancement might impede alveolar receptor display, hence ALI pathogenesis. We show for the first time, to our knowledge, that loss of the fence causes a proinflammatory surge of TNFR1 expression, and that the fence is a druggable target for mitigating ALI.

Results
F-actin determines alveolar TNFR1 display. Optical access to the live alveolar epithelium of mouse lung provided an opportunity for dynamic evaluation of the role of the F-actin fence as a determinant of epithelial TNFR1 expression. By real-time confocal microscopy of pulmonary alveoli, we determined fluorescence expressions of TNFR1 in the alveolar epithelium in terms of a fluorescent mAb that detects TNFR1 ectodomains (9) ( Figure 1A). We confirmed that the TNFR1 immunofluorescence was on the cell surface, since it was eliminated by alveolar injection of a cell-impermeable fluorescence quenching agent. Our procedures did not label TNFR1 on the capillary endothelium adjacent to the epithelium (Supplemental Figure 1A).
Moreover, co-staining with the lamellar body marker, Lysotracker Red (LTR) revealed TNFR1 expression on the type 1 epithelial cells (AT1), and no detectable expression of TNFR1 on the type 2 (AT2) (Supplemental Figure 1B).
To determine the dynamics of epithelial TNFR1 expression, in each alveolar field we sequentially microinjected the anti-TNFR1 mAb, TNFα (or PBS), and then repeated the anti-TNFR1 Ab for a second time ( Figure 1A). The first mAb injection marked the baseline TNFR1 expression on the alveolar epithelium. Alveolar microinjection of TNFα rapidly induced a time dependent decrease of TNFR1 immunofluorescence ( Figure 1A), affirming our previous findings that TNFα causes shedding of TNFR1 ectodomains (9). However, in the same alveoli that had undergone receptor shedding, the second mAb injection revealed marked enhancement of TNFR1 re-expression ( Figure 1A).
In separate experiments, we gave the second mAb injection at different time points to further delineate the time course of TNFR1 re-expression. Our findings indicate that within 10 minutes after the TNFα-induced shedding, TNFR1 expression was similar to baseline ( Figure 1B), indicating that the receptor was rapidly replaced. Subsequently, there was a surge of TNFR1 reexpression that on average, reached a peak at 1 hour and returned to baseline at 4 hours ( Figure   1B).
To determine the effects of the cytosolic Ca 2+ on F-actin and TNFR1 expression, we microinfused TNFα in alveoli expressing a transfected F-actin probe (5). At baseline, fluorescence of Ca 2+ and F-actin were steady for at least 20 min ( Figure 1C), ruling out photobleaching as an artefact. Alveolar microinjection of TNFα rapidly increased Ca 2+ , while concomitantly decreasing F-actin ( Figure 1C). To inhibit the Ca 2+ response, we gave alveolar microinjection of the Ca 2+ chelator, BAPTA-AM, then determined responses after 30 minutes.
BAPTA-AM blocked the TNFα-induced cytosolic Ca 2+ increase, the F-actin decrease and the TNFR1 surge ( Figure 1D). PBS pre-treatment had no effect. These first dynamic quantifications of actin in live alveoli indicated a strong effect of the cytosolic Ca 2+ on F-actin and TNFR1 expression.
To depolymerize F-actin by a receptor-independent mechanism, we gave alveolar microinjection of the actin depolymerizing agent, cytochalasin D (cytD) and determined the TNFR1 response after 60 min (19). CytD exposure markedly enhanced TNFR1 expression ( Figure 1E). Thus, even in the absence of TNFα, actin depolymerization with CytD was sufficient to induce the TNFR1 surge. Taken together, these findings affirmed that F-actin constitutively inhibited alveolar TNFR1 expression, and that decrease of F-actin caused the TNFR1 surge.

TNFα causes calcineurin dependent alveolar TNFR1 expression.
To determine mechanisms downstream of the TNFα-induced cytosolic Ca 2+ increase, we considered the role of calcineurin (Cn), which is a Ca 2+ sensor. Cn is a Ca 2+ -calmodulin dependent protein phosphatase that contains the catalytic CnA (α, β and γ isoforms) and the Ca 2+ -sensing CnB subunits (20,21). In CnAβ-null mice, the β isoform of CnA is disrupted, inhibiting Cn's phosphatase activity (22).
Optical imaging revealed that the TNFα-induced F-actin decrease and the accompanying TNFR1 surge were absent in CnAβ-null mice (Figure 2A, Supplemental Figure 2A). Moreover, the calcineurin inhibitor, FK-506 blocked the TNFR1 surge (Supplemental Figure 2B). These findings mechanistically implicated calcineurin in the TNFR1 responses. Calcineurin dephosphorylates, hence activates the actin-severing protein, cofilin (23,24). To evaluate this hypothesis, we carried out immunoblots on lysates of lungs of wild type and CnAβ-null mice.
Our findings indicated that levels of phosphorylated (inactive) cofilin were higher in CnAβ-null than WT mice ( Figure 2B). Moreover, 4 hours after intranasal (i.n.) instillation of the TNFα, 7 cofilin dephosphorylation occurred in wild type (WT), but not CnAβ-null mice ( Figure 2B). We interpret from these findings that TNFα-induced calcineurin activation, led to cofilin dephosphorylation.
To further explore the role of cofilin, we transfected the alveolar epithelium with plasmids to express wild type cofilin (pWT), or cofilin mutants that cannot be phosphorylated (pS3A), or are constitutively phosphorylated (pS3E). Hence these mutants are respectively, constitutively active, or inactive (Supplemental Figure 3) (25)(26)(27). As compared to pWT-expressing epithelium, baseline F-actin was lower in pS3A-, but higher in pS3E-expressing epithelium ( Figure 2C). Alveolar TNFα microinjection decreased F-actin in pS3A-, but not in pS3Eexpressing epithelium ( Figure 2C). This lack of effect was not due to detection failure, since the actin depolymerizing agent, latrunculin B decreased F-actin in pS3E-expressing epithelium ( Figure 2C). The TNFα-induced TNFR1 surge was greater in pS3A-than WT-expressing epithelium, but blocked in pS3E-expressing epithelium ( Figure 2D). These findings indicate that active, namely dephosphorylated cofilin was required for the TNFα-induced TNFR1 surge.
Taken together, our findings indicate that a major effect of TNFα was to activate calcineurin, which then lead to cofilin dephosphorylation. As a consequence, epithelial F-actin decreased, resulting in the TNFR1 surge.
Delivery of exogenous of Rac1 mutants modifies the actin cytoskeleton in alveolar epithelium. The GTPase, Rac1 phosphorylates p21-activated kinase, leading to LIM-kinase dependent cofilin phosphorylation, hence F-actin stabilization (28)(29)(30)(31). To evaluate the therapeutic potential of these mechanisms, we developed non-covalent TAT-linked conjugates with His-V12Rac1 and His-N17Rac1, which are constitutively active and inactive mutants of Rac1, respectively (32). We engineered the constructs to be unstable at low pH to enable intracellular separation of TAT from the cargo protein, hence to retain the protein in the cytosol.
Accordingly, when we labelled TAT and V12Rac1 with different fluorophores and injected the construct by alveolar microinjection, we could detect rapid entry of TAT-V12Rac1 in the alveolar epithelium ( Figure 3A). Subsequently, TAT fluorescence progressively decreased ( Figure 3A), indicating TAT exited from the epithelial cytosol, while V12Rac1 remained. We affirmed that the spatial distribution of V12Rac1 fluorescence in the epithelium matched that of the cytosolic dye, calcein red (Supplemental Figure 4A), and that the cell-impermeable fluorescence quencher, trypan blue (TB) failed to diminish V12Rac1 fluorescence (Supplemental To determine whether the TAT-protein constructs entered the endothelium of adjoining capillaries, we gave alveolar microinjections of TAT-V12Rac1 in which V12Rac1 was fluorophore tagged. Then, we loaded the alveolar epithelium and the endothelium with calcein red. Alveolar microinjection of the detergent, saponin eliminated epithelial, but not endothelial cytosolic fluorescence. Thus, the endothelial PM was intact and no V12Rac1 fluorescence was evident in the endothelial cytosol (Supplemental Figure 4 A, B). We conclude that the construct did not cross the alveolar barrier to enter the endothelium.
Alveolar microinjection of TAT-V12Rac1 induced rapid increase of epithelial F-actin that was sustained for at least 4 hours ( Figure 3B). Although baseline F-actin was higher in AT2 than AT1 (p<0.01) (33), V12Rac1 induced a greater relative increase of F-actin in AT1 (Supplemental Figure 5). By contrast, the constitutively inactive form of Rac1, N17Rac1, which was also internalized by the epithelium, failed to increase actin ( Figure 3B). The TNFα-induced TNFR1 surge was absent in TAT-V12Rac1 loaded epithelium, but present in epithelium loaded with TAT-N17Rac1 ( Figure 3B). These findings indicated that epithelial loading with V12Rac1 increased F-actin, inhibiting the TNFα-induced TNFR1 surge.
To determine the extent to which these responses occurred at a whole-organ level, we carried out immunoblots on lung lysates 4 hours after i.n. instillation of the TAT-linked constructs. Our findings indicated that within 4 hours, TAT-V12Rac1, but not TAT-N17Rac1, increased F-actin, while decreasing G-actin ( Figure 3C), and concomitantly increasing cofilin phosphorylation ( Figure 3D). To determine longer term effects, we gave i.n. instillations of the TAT constructs.
Then after 24 hours, we immunoprecipitated the Rac1 mutants from lung lysates using anti-His antibody. Both His-tagged proteins were detectable in immunoblots ( Figure 3E), indicating that exogenous Rac1 mutants remained in the lung for at least 24 hours after cell internalization.
Taking the imaging and the immunoblot data together, our findings indicated that epithelial loading with V12Rac1 increased F-actin, inhibiting the TNFα-induced TNFR1 surge.
Protective effects of alveolar F-actin enhancement on ALI outcomes. To determine the role of alveolar F-actin in a mouse model of ALI, we gave i.n. LPS at lethal-dose 0 (LD0), at which there is no mouse mortality (7,34), then determined responses after 24 hours. In optically imaged alveoli, LPS increased epithelial TNFR1 expression ( Figure 4A). Concomitantly, there was alveolar inflammation, as indicated by alveolar neutrophil entry ( Figure 4A). By contrast, i.n. instillation of TAT-V12Rac1 30 min prior to LPS, but not of TAT-N17Rac1, blocked the increase of TNFR1 expression, as well as the alveolar inflammation ( Figure 4A). LPS induced the expected ALI responses after 24 hours, namely increased leukocytes in the bronchoalveolar lavage (BAL) ( Figure 4B), increased alveolar permeability to intravascularly injected albumin ( Figure 4C).
To determine global lung responses, we carried out in situ biotinylation assays to determine cellsurface expression of TNFR1, as also assays of whole lung lysates for cofilin phosphorylation, F-actin and IκB. These studies indicated that LPS decreased cofilin phosphorylation ( Figure 4D) as well as F-actin in 24 hours ( Figure 4E) while increasing surface TNFR1 expression ( Figure   4F). By contrast, pre-LPS i.n. instillation of TAT-V12Rac1, but not of TAT-N17Rac1, stabilized F-actin for 24 hours ( Figure 4E) and markedly abrogated the LPS-induced enhancement of TNFR1 expression. Since NFkB activation causes alveolar inflammation, we confirmed that LPS induced IkB degradation ( Figure 4G). TAT-V12Rac1, but not TAT-N17Rac1, blocked IkB degradation ( Figure 4G). Since inhibition of IkB degradation inhibits NFkB activation, we interpret V12 Rac1-induced actin enhancement inhibited NFkB activation.
Taking the imaging and global data together, our findings indicate that the LPS-induced proinflammatory effect of epithelial TNFR1 hyperexpression was sustained for at least 24 hours, and that these responses were inhibited by epithelial incorporation of V12Rac1.
To determine the effects of the Rac1 constructs on LPS-induced mortality, we exposed mice to a lethal LPS dose, which caused robust increases of leukocyte counts and protein concentration in the BAL (Supplemental Figure 7A), depletion of surfactant phospholipids in the BAL ( Figure 5A and Supplemental Figure 7B), loss of lung compliance ( Figure 5B), and ~80% mortality in 3-4 days ( Figure 5C). Pretreatment with i.n.TAT-V12Rac1 30 min prior to LPS instillation protected against mortality (Supplemental Figure 6). To evaluate post-ALI therapeutic efficacy, we gave the constructs 4, or 24 hours after LPS. In the 4-hour group, V12 Rac1 mitigated LPS-induced ALI as indicated by recovery of lung compliance and BAL surfactant phospholipids ( Figure 5A, B), and a 15% mortality in 3 days ( Figure 5C). In the group given TAT-V12Rac1 24 hours after LPS, evaluation of ALI after a further 24 hours indicated marked reduction of lung water and BAL leukocytes ( Figure 5D, E), and of mortality ( Figure 5F). TAT-N17Rac1 was without effect ( Figure 5F). The reduction of mortality between the 4-and 24-hour groups was not statistically significant. These findings indicate that given 30 minutes before, or 4, or 24 hours after LPS, TAT-V12Rac1 protected against ALI.
To determine whether the i.n. delivered constructs entered the capillary endothelium, we gave mice i.n. LPS at a lethal dose, or PBS, then followed 4 hours later with i.n. TAT-V12Rac1-GFP.
After a further hour, we freshly isolated cells from harvested lungs, then we carried out flow cytometry analyses on the cells. These analyses indicated that V12Rac1 was taken up in the epithelium, but not the endothelium (Supplemental Figure 7C). Thus, while the imaging data indicated there was no endothelial uptake of the construct under baseline conditions (Supplemental Figure 4), the flow cytometry analyses affirm this finding and show further that endothelial uptake did not take place after LPS treatment. Taken together, our findings indicated that LPS decreased epithelial F-actin by a cofilin-mediated mechanism, increasing TNFR1 expression at the epithelial surface, augmenting lung inflammation, hence mortality. We conclude, V12Rac1 succeeded in increasing survival in LPS-induced lung injury by both prophylactic and therapeutic strategies.
LPS, an outer coat protein of Gram-negative bacteria, causes ALI by activating TLR4-induced proinflammatory signaling in pulmonary alveoli (7). However, Gram-negative bacteria such as Pseudomonas aeruginosa (PAO1) may induce ALI through additional mechanisms, as for example by production of exotoxin (35). To determine the efficacy of actin enhancement therapy in ALI due to live bacterial infection, we i.n. instilled PA at a dose that causes high mortality. Fifteen hours after instillation mouse mortality was 100% ( Figure 5D). To determine therapeutic efficacy of i.n. TAT-V12Rac1, we administered the construct 4 hours after bacterial instillation. At this time point, lung injury was well developed as indicated by the extravascular lung water, which was two-times above baseline (Supplemental Figure 8). Despite the severe lung injury, i.n. TAT-V12Rac1 decreased mortality to ~20% at 15 h. Thus, a post-injury therapeutic strategy with i.n. TAT V12Rac1 protected against ALI caused by instillation of highly toxic bacteria.

Discussion
We report here the first definitive evidence that TNFR1 ligation induced loss of the F-actin fence in the alveolar epithelium, causing receptor hyperexpression and alveolar injury. Rac1 delivery to the alveolar epithelium enhanced the fence, blocking the receptor hyperexpression and injury.
Of translational significance, the fence enhancement protected survival in LPS-and PA-induced ALI. Together, these findings, to our knowledge, constitute the first evidence that the F-actin fence of the alveolar epithelium is a determinant of lung inflammation and injury.
We propose a sequence of events ( Figure 6) in which inhaled pathogen induces macrophagederived TNFα release. TNFR1 ligation on the alveolar epithelium increases epithelial Ca 2+ , activating the calcineurin-cofilin pathway to depolymerize F-actin. The actin fence is thus disabled, causing a surge of proinflammatory receptor expression. The F-actin depolymerizing agent, cytochalasin D also caused the receptor surge, indicating that loss of F-actin was necessary and sufficient for the effect, and that non-specific receptor-mediated mechanisms were not required. Importantly, a brief TNFα exposure induced sustained F-actin decrease in the alveolar epithelium, leading to the receptor surge, which lasted several hours, possibly corresponding to the time taken for F-actin to return to baseline levels and thereby, re-establish 13 fence conditions. The prolonged duration of the receptor surge suggests receptor hyperexpression might be sustained during the inflammatory process if F-actin fence properties are not re-established. The extent to which this interpretation applies to other proinflammatory receptors expressed on the epithelium (2) require further investigation.
Actin depolymerization also destabilizes cell-junctional barriers (36), causing increased alveolar permeability to plasma proteins. Repetitions of this sequence of events may perpetuate lung injury, progressively increasing its severity and leading to the high mortality of ARDS. The fact that Rac1 delivery to the alveolar epithelium inhibited this injurious sequence indicates that actin enhancement in the epithelium was sufficient for opposing two critical players in the injury response, namely hyperexpression of pro-inflammatory receptors and weakening of the epithelial fluid barrier. Since we did not detect exogenous Rac1 in the capillary endothelium, we interpret that V12Rac1's protective effects were predominantly due to epithelial actin enhancement.
The cell-surface expression of the receptor was AT1 restricted, with no detectable AT2 expression. TNFR1 expression on AT1 related inversely to F-actin levels, indicating that the fence was a determinant of the expression even under unstressed conditions at baseline. Cultured, AT2-like A549 cells express TNFR1 (37), although these cells do not entirely reflect properties of AT2 in situ. Interestingly, as we reported (33), and confirm here, F-actin is threefold higher in AT2 than AT1, suggesting that high F-actin restricts AT2 expression of cellsurface TNFR1.
Our findings reveal novel understanding of Ca 2+ -induced F-actin regulation. An increase of the cytosolic Ca 2+ may affect F-actin differently in different functional contexts. In the context of endothelial barrier regulation, Ca 2+ increases lead to actin polymerization, hence stress fiber formation and barrier loss (38)(39)(40). By contrast, Ca 2+ increases due to T cell receptor engagement at the immunological synapse induces actin depolymerization, thereby promoting the intensity and duration of T cell engagement (41). Here, the TNFα-induced Ca 2+ transient caused actin depolymerization as evident in the rapid F-actin decrease. This response, as well as the ensuing TNFR1 surge were inhibited by blocking the Ca 2+ transient with calcium chelator BAPTA-AM. Hence, the cytosolic Ca 2+ is a major fence regulatory mechanism that determines TNFR1 display on the alveolar epithelium.
The Ca 2+ effect on F-actin was mediated through the calcineurin-cofilin pathway. Several findings support this interpretation. Thus, the TNFα effects on F-actin and the surge were absent after alveolar treatment with FK-506, or in CnAβ-null mice, implicating calcineurin in the responses. We interrogated the mechanistic pathway through epithelial expressions of constitutively active, or inactive mutants of cofilin. The active mutant, which causes F-actin depolymerization, enhanced the TNFα-induced TNFR1 surge. The inactive mutant, which increased F-actin, blocked all TNFα-induced responses including the surge. These results are further evidence that loss of the actin fence was critical for epithelial hyperexpression of the proinflammatory receptor. Taken together, these findings reveal a novel sequence of fast-acting signaling events in which a receptor-mediated Ca 2+ increase activated calcineurin, leading to cofilin-mediated loss of the F-actin fence, resulting in enhanced proinflammatory receptor display on the alveolar surface.
LPS induced the expected lung inflammation and alveolar injury, as indicated by presence of albumin leak, and significant mortality (7,34). LPS induces secretion of multiple cytokines, including TNFα, while also increasing alveolar epithelial Ca 2+ (7). Similar to TNFα, alveolar LPS exposure also dephosphorylated cofilin and decreased F-actin, while increasing TNFR1 expression. However importantly, all of these LPS effects were sustained for 24 hours. This finding indicates that mortality resulted from prolonged loss of the F-actin fence that caused prolonged TNFR1 hyperexpression, exacerbating alveolar inflammation. We tested this possibility by means of a CPP strategy where the plan was to increase the actin fence through intracellular delivery of V12Rac1 in order to phosphorylate, hence inactivate cofilin.
Our strategy was highly successful. Whether administered by alveolar microinjection, or by the i.n. route, the TAT-V12Rac1 construct was rapidly internalized by the alveolar epithelium. The pH-sensitive, non-covalently conjugated construct hydrolyzed intracellularly, enabling TAT diffusion out of the cell. Hence, although the protein cargo was successfully taken up by the epithelium, TAT, which contains a nuclear-localizing sequence (42) and may therefore induce unwanted transcriptional effects, was eliminated. We consider this TAT elimination from the alveolar epithelium a strength of our CPP strategy, since such an elimination would abrogate the likelihood of transcription-induced long-term toxicity due to TAT. Although reports indicate that extracellular TAT is rapidly removed from the lung (43,44), further studies are required to understand mechanisms underlying the removal process.
The instilled TAT-V12Rac1 construct was not detectable in the capillary endothelium before or after induction of alveolar inflammation. Thus, once dissociated from TAT, non-conjugated V12Rac1 protein remained confined to the epithelial cytosol and did not enter endothelial cells.
We affirmed this result by assaying the membrane permeabilizing effects of saponin on cell fluorescence. Given by alveolar injection, saponin removed epithelial, but not endothelial fluorescence (sFig. 4), indicating that the construct was localized only to the alveolar compartment. This conclusion is further supported by our flow cytometry results showing that even after LPS treatment, a condition that increased alveolar permeability, there was no endothelial uptake of V12Rac1 (SFig. 8C). Therefore, TAT-V12Rac1 did not modify endothelial F-actin. We consider it unlikely that it directly affected endothelial barrier determinants, such as focal adhesions and junctional proteins (45)(46)(47).
Following delivery of the CPP, we could detect V12Rac1 in the epithelium for at least 24 hours, during which epithelial F-actin also remained elevated and the TNFR1 hyperexpression was abrogated. These findings show for the first time, that constitutively active Rac1 increases Factin in vivo, and that the induced F-actin remains stable for sufficient durations to be therapeutically effective. Importantly, our findings indicate that pre-treatment with i.n. TAT-V12Rac1 markedly reduced mortality due to LPS. We point out however, that while our studies support a protective effect of the actin fence in the context of TNFR1 expression, the extent to which actin fence strengthening modifies the effects of other ALI-relevant receptors, such as the proinflammatory interleukin1β receptor, or the anti-inflammatory TNFR2 (48) requires further study. Nevertheless, our findings indicate that pre-treatment strategy with V12Rac1 might be a feasible prophylactic option against ALI.
To determine the therapeutic efficacy of TAT-V12Rac1 as a curative agent, we instilled the construct under conditions of established pathology, namely 4 or 24 hours after instilling a lethal LPS dose that caused major alveolar hyperpermeability and inflammation. These post-LPS V12Rac1instillations were also protective against mortality. An important conclusion is that fence enhancement in the alveolar epithelium mitigates ALI well into the progressive phase of the disease. However, since the mortality protection was less for the delayed intervention group, fence enhancement therapy might be more effective in early than late stages of ALI.
We also considered that infection by live bacteria may induce injury mechanisms that are more extensive than those due to LPS alone. Accordingly, we determined the effects of the construct in the presence of lung infection by a highly lethal dose of PA, an organism that is a major cause of ALI and ARDS (49). In wild-type mice, this dose of PA induced catastrophic mortality in 15 hours. However, when we gave TAT-V12Rac1 after inducing infection, the mortality was markedly abrogated. These findings add to the translational significance of our study, which reveals the therapeutic strength of V12Rac1 in the setting of ongoing lung inflammation.
Antibody inhibition of TNFR1 has been advanced as a possible therapeutic strategy for ALI (50,51). However, antibody therapy was protective for ALI not associated with mortality (50,51).
Hence, these reported findings are not directly comparable to the present mortality-causing inflammatory responses, which could be encountered in clinical ARDS. In severe inflammation, cell-directed therapy as we propose, may be more effective.
Since Rac1 plays a role in multiple cellular processes (52-54), we considered whether its sustained epithelial presence affected alveolar function. We approached this issue by quantifying lung surfactant secretion and lung compliance, metrices that report adequacy of overall alveolar homeostasis. Maintenance of surfactant secretion reflects adequacy of several epithelial regulatory parameters, such as cytosolic Ca 2+ regulation (55). The fact that V12Rac1 protected these critical functional responses despite ALI, indicates that V12Rac1 treatment not only did not interfere with major aspects of epithelial function, but that the treatment reinstated alveolar homeostasis in ALI.
Although our study indicates that there were no overall negative consequences attributable to V12Rac1, to support clinical application further data are required to clarify issues including the
To carry out in situ rhodamine-phalloidin and anti-GFP immunofluorescence alveoli were fixed by continuous microinfusion of 4% paraformaldehyde (15 minutes)  In case of an excessive fluorescence response, we reduced gain to prevent fluorescence saturation, then corrected grey levels against the linear relationship between grey levels and gain that we determined for each fluorophore. Evaluation of Acute Lung Injury. We followed reported protocols for quantifications of alveolar permeability (34) and extravascular lung water (EVLW) (56,63). Briefly, alveolar permeability was quantified as the BAL-plasma ratio of Evans Blue-albumin (EB-albumin) concentrations, 4 hours after i.v. EB-albumin. EVLW was the wet-dry ratio of the lung homogenate, corrected for blood water content. To quantify lung compliance (CL), mouse lungs were inflated to 25 cm H2O airway pressure (P). Then, decreases in P were recorded for step decreases in lung volume (V). CL was calculated as the slope of the linear regression of V against P for P=0-10 cmH2O. Data for each lung are means of two replicated V-P plots.
Survival assessment. Animals were randomly divided for the experiments groups. Anesthetized mice were treated with intranasal instillation of TAT-Rac1 conjugates (0.5 mg/kg) 30 minutes before or 4 hours after intranasal instillation of LPS at a lethal dose (10 mg/kg). The mice were scored at frequent intervals following LPS administration, in agreement with the approved Animal Care Protocol. At each assessment, the animals were scores by a blinded investigator applying scoring system which includes measuring of body weight, evaluation of respiration, activity and grooming. Lung fractioning for G-actin and F-actin. A two-step solubilization was carried out by our reported method (36). Briefly, lung tissue cleared from blood and BAL leukocytes were homogenized and solubilized in 400 µl of mild lysis buffer (50 mM NaCl, 10 mM Pipes, pH 6.8, 3 mM MgCl2, 0.05% Triton X-100, 300 mM sucrose) for 20 min at 4 °C on a rocking platform.
The whole lung lysates were clarified by centrifugation at 10,000 g for 10 min. The supernatants, Triton-X 100 soluble fraction, were collected and used to determine G-actin. The resulting pellet was re-suspended in 200 µl of stronger solubilizing buffer (15 mM Tris, pH 7.5, 5 mM EDTA, 2.5 mM EGTA, 1% SDS) and incubated at 95 o C for 10 min to dissolve the pellet and then volume was adjusted to 400 µl with RIPA buffer. The resulting lysates were clarified by centrifugation at 16,000 g for 10 min. Collected supernatant, Triton-X 100 insoluble fraction were used to determine F-actin. Extracted proteins (15µg) were subjected to SDS-PAGE and Western blotting analysis.
Data analysis. Group numbers were designed to enable detection of statistically significant differences with a power of 85%. For imaging experiments carried out with a paired protocol, baseline and test conditions were obtained in the same alveoli and at least 5 determinations were obtained for each lung. These determinations were averaged to obtain a mean for each condition in each lung. There was no statistically significant within-lung variability of effect size. The means for each lung were pooled for the group to obtain mean± SEM, where n represents the number of lungs. The per-lung means are shown in the bar diagrams. Group means were compared by ANOVA with Bonferroni correction. Data for the same alveolar segment were compared by the paired t-test. Survival rates were analyzed by the Kaplan-Meyer log rank test.
We accepted significance at p<0.05.     lungs for each group, *p<0.05 vs baseline using 2-tailed t test. Each dot shows data for a single lung.