WT (C57BL/6), SCID, and Fyn^−/− [2] were purchased from Jackson Laboratories. Fyn^−/− mice were bred to CARD1 [13] (here referred to as CAR⁺ mice) in our animal facility to develop Fyn^−/−CAR⁺ mice. All mice, including SCID recipients, were kept in specified pathogen-free conditions prior to adoptive cell transfers. All experiments were approved by Sunnybrook Research Institute Animal Care Committee and followed guidelines set by the Canadian Council on Animal Care.

Five week old (20 grams) male mice were infected with the influenza virus A/PR8 (PR8, H1N1) through intranasal administration with either 10⁵ or 10⁴ TCID₅₀ in PBS as indicated. For survival experiments, following the procedure the mice were monitored daily and sacrificed when the body temperature had dropped below 32 degrees C, they had lost >30% of total body weight, or appeared moribund. Otherwise, the mice were sacrificed at specific time points post infection.

For viral clearance assays, lungs were excised from the animal at 3, 6, and 9 days post intranasal infection with influenza A/PR8. After weighing, lungs were homogenized in RPMI 1640 medium (1 g of lung tissue/10 ml). Supernatant was obtained and stored at −70˚C after centrifugation at 1200 × g for 20 min at 4˚C. TCID₅₀ was determined by the MDCK assay with the Reed and Muench technique as previously described [14] . The infected lungs and sacrificed mice were disposed of in accordance with Sunnybrook Research Institute Animal Care Committee and followed guidelines set by the Canadian Council on Animal Care.

Polyclonal anti-Lck and anti-Fyn antibodies have been previously described [5] . Anti-phosphotyrosine (4G10) and anti-actin antibodies were purchased from Upstate, anti-Fyb from Lifespan Bio, anti-SLP-76 (clone AS55) from Millipore, anti-CD3ε (145-2C11), anti-CD4 (GK1.5), anti-TcRCβ (H57), anti-CD69 (H1.2F3), anti- CD45RB (C363.16A), anti-IL-4 (11B11), anti-IFN-γ (R4-6A2), mouse recombinant IL-6 and recombinant human TGF-β1 from eBiocience. Murine recombinant IL-7 was purchased from Peprotech. Anti-CAR (RmcB), anti-IL-2 (S4B6), and anti-IL-2 isotype control (rat IgG2) were isolated from their respective hybridomas cultured in our laboratory. Brij-58, Cholera toxin B-HRP (CT-HRP) and Streptavidin were purchased from Sigma- Aldrich.

Splenic CAR⁺ CD4⁺ CD45RB^hi or CD45RB^lo T cells [11] were sorted using BDAria cell sorter and 2.5 × 10⁶ cells were injected iv into each SCID recipient. After cell transfer, SCID recipients were fed non-sterile food and water and kept in cages in the absence of filter tops. Recipients were monitored and sacrificed when bloody diarrhea and/or rectal prolapse was evident. To determine the number of infiltrating donor cells, Lamina Propria lymphocytes were isolated from the large gut of recipients as described [15] , stained using fluorochrome labeled anti-CD4 and anti-CAR and the number of donor cells was determined using percentage of CD4⁺CAR⁺ assessed flow cytometrically using a BD FACSCalibur.

Sorted splenic CAR⁺ CD4⁺ CD45RB^hi T cells were CFSE labeled [16] and 2.5 × 10⁶ cells were injected iv into each SCID recipient on day zero. Recipients also received daily intraperitoneal injections of 0.5 mg of S4B6, 0.5 mg of isotype control, or PBS in a total volume of 200 μL, starting on day zero, for seven days. Recipients were sacrificed on day seven, and donor CAR⁺CD4⁺ T cells in spleen and lymph nodes were analyzed for CFSE expression flow cytometrically [16] .

The shuttle vector (pENTR-UbC) was designed by cloning human Ubiquitin C promoter [13] followed by a multiple cloning site (MCS) and a region containing the simian virus 40 late polyadenylation sequence [13] using pENTR (Invitrogen) as the vector backbone. emGFP (Invitrogen) and wild type murine FynT (WT-Fyn) [5] cDNAs were cloned into the MCS of pENTR-UbC.

Adenoviral constructs were created using Clonase-mediated recombination between pENTR-UbC-emGFP or pENTR-UbC-WT-Fyn with pAd/PL-DEST (Invitrogen), followed by preparation of amplified adenoviral stocks according to the manufacturer’s recommendations. Adenoviral stocks were purified using Adeno-X Maxi purification kits from Clontech and titered using plaque formation assays in 293A cells (Invitrogen) and Quantum’s AdEasy TCID₅₀ protocol.

Primary T cells were suspended at a concentration of 1 × 10⁶ cells/100 μL in serum free medium [17] in the presence of adenoviral particles at a multiplicity of infection (MOI) of 10 for one hour at 37˚C. The cells were then washed and cultured in serum free medium in the presence of 10 ng/ml of IL-7 for 24 hours prior to IL-2 and IL-17A assays or for 72 hours prior to cell transfer experiments.

Splenic CD4⁺ T cells were isolated using EasySep negative selection kit from STEMCELL Technologies. 10⁶ CD4⁺ T cells were labeled with biotinylated H57 and/or biotinylated GK1.5 and CD4/TcR coaggregation was achieved using streptavidin as previously described [6] . The cells were lysed and subjected to immunopreciptation (IP) using anti-Lck [6] . The IP samples were subjected to immune complex kinase assay [6] and probed for Lck. For Fyb IPs, 10⁷ CD4⁺ T cells were lysed and subjected to IP using anti-Fyb. The IP samples were equally divided and probed for phosphotyrosine, Fyb, SLP-76, and Fyn.

For IL-2, 5 × 10⁴ CD4⁺ T cells were cultured in 200 μL of serum free medium [17] per well of a 96-well plate (Corning Costar) previously coated with H57 (0.5 μg/ml), GK1.5 (10 μg/ml), or both, in triplicates, for 48 hours. The level of IL-2 was determined using an IL-2 specific ELISA kit (eBioscience). For IL-17A, the cells were cultured in the presence of anti-IL-4 (10 μg/ml), anti-IFN-γ (10 μg/ml), IL-6 (10 ng/ml) and TGF-β1 (5 ng/ml) in 96-well plates (Corning Costar) previously coated with H57 (1.5 μg/ml), GK1.5 (30 μg/ml), or both, in triplicates, for 72 hours. The level of IL-17A was determined using an IL-17A specific ELISA kit (eBioscience).

Lipid rafts (LR) were isolated from 10⁶ cells, as previously described [17] , and 40 μL of each fraction was probed for GM1 ganglioside as a marker for lipid rafts using CT-HRP. LR distributions of Lck and Fyn were determined by immunoblotting 40 μL of each fraction using anti-Lck and anti-Fyn antibodies respectively.

Densitometric analysis was performed using GS-800 densitometer and Quantity One software from Bio-Rad Laboratories on non-saturated signals.

Large intestine from SCID recipients was collected, washed with PBS, cut longitudinally, wrapped to form a roll before being frozen in OCT (Electron Microscopy Sciences), and 8 μm sections were stained with hematoxylin and eosin. Pictures were taken using a DFC300 FX Digital Color Camera and the Application Suite software, both from Leica. For immunofluorescent microscopy the sections were fixed in ice-cold acetone for 10 min and air-dried at room temperature for 30 min, blocked in PBS plus 5% normal goat serum for 30 min followed by incubation in anti-CD3ε diluted in PBS plus 5% normal goat serum. Lastly, the sections were probed using Cy5 conjugated secondary antibody (goat anti-Armenian Hamster IgG (H ⁺ L), Jackson ImmunoResearch Laboratories) as well as 4′,6-Diamidino-2-phenylindole dihydrochloride (DAPI, Sigma-Aldrich). Images were captured using a Zeiss microscope (Carl Zeiss) and analyzed using the AxioVision software (Carl Zeiss). Same time of exposure during the acquisition and same values during processing were applied to each image.

P values among experimental groups were determined by the unpaired Student’s t-test.

An in vitro assay involving antibody-mediated co-aggregation of CD4 and associated Lck with the TcR/CD3 complex was employed. CD4⁺ T cells were cultured in plates pre-coated with anti-CD4, anti-TcRCβ, or both and IL-2 production was assessed at 48 hours. As has been previously described [1] [2] , IL-2 production by Fyn^−/− T cells was profoundly impaired, giving rise to <10% of the IL-2 produced by WT CD4⁺ T cells (Figure 1).

Formal proof of the role of Fyn underpinning this defect was achieved through ectopic expression of fyn in primary resting T cells using an adenoviral gene delivery system. As mice do not express an adenovirus receptor, WT and Fyn^−/− mice were mated with those transgenic for the coxsackie/adenoviral receptor (CAR) [13] . CAR transgenic mice have been constructed to express the coxsackie/adenoviral receptor using a T cell specific pro-

moter to allow for the expression of CAR specifically on T cells. CAR expression in these transgenic mice is mainly restricted to T cells, and these animals exhibit normal T cell development and T cells expressing CAR allow for the efficient transfer of any gene of interest into primary resting T cells [13] .

The efficacy of adenoviral transduction is illustrated in Figure 2. Specifically, transduction of WT CAR⁺ CD4⁺ T cells with GFP containing virus (Ad/GFP) resulted in ~90% GFP⁺ cells within 24 hours (Figure 2(a)). Figure 2(b) illustrates the kinetics of ectopic expression of Fyn in CD4⁺Fyn^−/−CAR⁺ primary T cells. Further, the levels of ectopic Fyn expression achieved in these circumstances were comparable to WT levels of endogenous Fyn expression, and there were no detectable alterations in the levels of endogenous Lck expression (Figure 2(c)). As illustrated in Figure 2(d), the cellular localization of ectopically expressed Fyn was identical to that observed for endogenous Fyn in WT cells with >90% localizing to LR [6] .

As illustrated in Figure 3, Ad/WT-Fyn transduction of Fyn^−/− CD4⁺ primary T cells rescued levels of IL-2 production upon co-aggregation of TcR and CD4 comparable to those observed in WT CD4⁺ T cells. Ad/GFP transduction of either WT or Fyn^−/− T cells had no significant effect (Figure 3).

Previous reports have demonstrated that the activities of Lck and Fyn are reciprocally regulated [18] [19] . It is therefore possible that the impaired TcR/CD4 induced IL-2 production observed in Fyn^−/− T cells is not due to the absence of Fyn per se, rather, impaired function of Lck in Fyn^−/− T cells. To investigate this possibility, the activity of Lck was assessed subsequent to TcR-CD4 co-aggregation in each of three CAR⁺CD4⁺ T cell populations: WT or Fyn^−/− transduced with Ad/GFP and Fyn^−/− transduced with Ad^/WT-Fyn. As illustrated in Figure 4(a), both aggregation of CD4 and co-aggregation of CD4-TcR resulted in a robust and comparable induction of Lck kinase activity, in each of the three T cell populations, as assessed by levels of phosphorylation of substrate enolase. Hence, the capacity to activate Lck is not impaired in the absence of Fyn, and therefore, taken together, these results formally prove that Fyn predicates TcR induction of IL-2 production in vitro.

We next sought to characterize the signalling defect underpinning compromised IL-2 production in Fyn^−/− T cells. Fyb is a specific Fyn substrate and it has been reported that it is not tyrosine phosphorylated in Fyn^−/− T cells [20] . As illustrated in Figure 4(b), the basal level of tyrosine phosphorylation of Fyb is indeed reduced in Fyn^−/− T cells and is restored in primary resting CD4⁺ T cells transduced with Ad/WT-Fyn (Figure 4(b)). However, and of note, is that notwithstanding the reduced level of PY-Fyb in Fyn^−/− T cells, anti-Fyb co-immuno- precipated SLP-76 from Fyn^−/− T cells as efficiently as it did from WT T cells (Figure 4(b)). Upon TcR-medi- ated cellular activation, levels of P-Y Fyb and P-Y SLP-76 increase in both WT CD4⁺ T cells and Fyn^−/− CD4⁺ T cells transduced with Ad/WT-Fyn, but not in Fyn^−/− CD4⁺ T cells transduced with Ad/GFP (Figure 4(b)). These results confirm that Fyb is a direct Fyn substrate and also reveal the presence of a tri-molecular complex of Fyb-SLP-76-Fyn in WT and Fyn^−/− T cells transduced with Ad/WT-Fyn in resting T cells. Further, that TcR mediated activation results in the parallel and Fyn-dependent increases in P-Y of both Fyb and SLP-76.

Towards mechanistically tethering the observed impairment in IL-2 production in vitro by CD4⁺Fyn^−/− T cells with a phenotype attributable to Fyn deficiency in an in vivo setting, we sought a model system whose endpoint was predicated by T cell expansion. The T cell-induced colitis model is suitable in this regard. Briefly, reconsti-

tution of severe combined immune deficient (SCID) mice with WT CD4⁺ CD45RB^hi T cells has been shown to cause colitis due to expansion of the injected T cells in the large intestine [8] - [10] .

To determine whether CD4⁺ Fyn^−/−CD45RB^hi T cells were impaired inducers of disease, groups of SCID recipients were injected with either WT CD4⁺ CD45RB^hi T cells, or the same phenotypic subset derived from Fyn^−/− animals. As illustrated in Figure 5(a), SCID recipients of WT naïve inducer T cells developed colitis starting at 6 weeks post injection. By 13 weeks, 100% of the recipients of WT inducer T cells had succumbed to enteropathy. In striking contrast, recipients of Fyn deficient inducer T cells remained symptom free for the duration of the 16-week observation period (Figure 5(a)).

Towards formally demonstrating that Fyn deficiency was at the root of impaired disease induction by CAR⁺CD4⁺ Fyn^−/−CD45RB^hi T cells, the latter were transduced with either Ad/GFP or Ad/WT-Fyn and their colitogenic capacity assessed. As illustrated in Figure 5, Fyn transduction restored the disease inducing capacity of Fyn^−/− cells.

The symptoms correlated with profound expansion of inducer T cells rescued in the Lamina Propria of the large bowel of SCID recipients. As illustrated in Figure 5(b), colitogenic WT CD4⁺ CD45RB^hi T cells expanded ~100-fold over the course of the assay, in contrast to the ~2-fold expansion of Fyn^−/− CD4⁺ CD45RB^hi T cells. Notwithstanding the lack of expansion of Fyn^−/− T cells, their upregulated expression of CD69 was comparable to that observed in WT inducer T cell populations, evidence that they were indeed engaging environmental antigens in the gut (Figure 5(c)). Figure 5(d) illustrates the characteristic histology and gut pathology induced by the WT colitigenic T cells, with flattening of the villi and T cell infiltration, not observed in recipients of their Fyn^−/− counterparts. Importantly, these characteristic indicators were rescued upon ectopic expression of Fyn (Figure 5(b) and Figure 5(d)), formally demonstrating that Fyn predicated the disease inducing capacity of CD4⁺CD45RB^hi T cells (Figure 5(a)).

The obligate role of Fyn in supporting colitogenic T cells in vivo appears directly related to its role in supporting IL-2 production. As illustrated in Figure 6(a) and Figure 6(b), of the 22 cytokines assessed subsequent to TcR/CD4 co-aggregation of CD4⁺ primary T cells derived from either WT or Fyn^−/− animals, only IL-2 production was significantly affected. Further, as illustrated in Figure 6(c), IL-17 production by Fyn^−/− T cells was within 2-fold of levels derived from WT T cells when cytokine induction was assessed in circumstances known to skew CD4⁺ responders towards Th17 cells [21] [22] .

IL-2 mediated expansion of CAR⁺ CD4⁺ CD45RB^hi T cells was directly assessed by treating SCID recipients with 6 daily injections of PBS, mAb specific for IL-2 (S4B6), or its isotype control. As illustrated in Figure 7, ~10% of donor WT CD4⁺ CD45RB^hi T cells rescued in the spleen and lymph nodes of these recipients had divided less than 6 times at day 7 post injection in SCID recipients treated with either PBS or isotype control. In contrast, in recipients treated with S4B6 the proportion of cells that had divided less than 6 times increased ~3-4 fold (Figure 7).

Analysis of Fyn^−/− CAR⁺ CD4⁺ CD45RB^hi T cells in this assay revealed that ~35% of cells divided less than 6 times, consistent with their impaired expansion in comparison to WT cells (Figure 5(b) and Figure 7). Further, anti-IL2 treatment increased the proportion of cells dividing less than 6 times to >90% at day 7 (Figure 6). These results support the conclusion that autocrine IL-2 mediates the expansion of both WT and Fyn^−/− CD4⁺T cells in vivo and provides the mechanistic basis for the non-colitogenic capacity of Fyn^−/− CD4⁺ CD45RB^hi T cells.

Fyn deficiency has not been previously associated with compromised immunity. As Fyn-dependent IL-2 production and T cell expansion was profoundly impaired in Fyn deficient T cells, we next sought to examine the impact of Fyn deficiency on immunocompetency of Fyn deficient mice in vivo, noting that IL-2, likely produced by CD4⁺ T cells, is the key cytokine involved in the early programming of naïve CD8⁺ T cell effector and memory differentiation [23] - [25] in circumstances of viral infection.

A flu-infection model [11] [12] was used to directly assess the immune-competence of Fyn-deficient animals. As illustrated in Figure 8(a), >90% of Fyn-deficient hosts succumbed 7 - 8 days post infection with either dose of A/PR-8 (H1N1) used, exhibiting up to 30% loss of body weight within the first week. In contrast, 60% of wild-type (WT) hosts survived at the higher dose and 80% at a 10-fold lower dose of flu virus (Figure 8(a)).

Figure 8(b) illustrates lung pathology of uninfected WT; and WT and Fyn-deficient hosts at day 6 after PR-8 infection. Both WT and Fyn-deficient hosts exhibited chronic inflammatory peribronchiolitis. The WT hosts presented with minimal alveolar involvement, characterized by vascular congestion and focal inflammatory cell infiltration. In contrast, the Fyn-deficient hosts exhibited extensive and severe alveolar congestion with infiltrating blood and polymorphonuclear leukocytes; and succumbed to respiratory failure due to pulmonary consolidation (Figure 8(b)).

The primary immune response to PR-8 infection is known to involve both T-cell lineages as well as B-cells, all of which localize to the lung toward resolving infection [26] . Day 6 is the inflection point in the primary response reported herein. As illustrated in Figure 8(c), the number of CD4⁺ and CD8⁺ T cells localized to the lungs of infected WT and Fyn-deficient hosts is comparable at day 6 post infection; as is the frequency of contained PR8-specific CD8⁺ T cells (Figure 8(c)). It merits comment that flu-specific T cells of both lineages were assessed and while enumeration of CD8⁺ flu-specific T cells binding the dominant MHC Class I binding PR8-derived peptide was reproducible, results assessing binding of one of many MHC Class II PR8-derived peptides were not (data not shown). Notwithstanding, these results support the conclusion that neither the T cell repertoire [1] - [3] nor T cell homing capacity is compromised in the absence of Fyn.

As illustrated in Figure 8(c), the survival of WT hosts correlated with a 20-fold and 9-fold increase in the number of lung infiltrating CD4⁺ and CD8⁺ T cells, respectively, at day 9 post infection, and a concomitant 20-fold increase in the number of contained flu-specific CD8⁺ T cells (Figure 8(c)). Viral titers assessed over the course of PR8 infection in both WT and Fyn-deficient hosts mirrored the physiological outcome of infection. Specifically, while there was a significant difference in viral titers observed at day 3 post-infection, possibly due to Fyn-dependent differences in an innate component of the immune response [4] , the viral titers at day 6 post- infection were comparable in WT and Fyn-deficient hosts (Figure 8(d)). The Fyn-deficient animals were moribund at this time point and succumbed shortly thereafter, while the primary adaptive immune response in the WT animals [26] , as expected, increased over the following 3 days and resolved the viral burden (Figure 8(c) and Figure 8(d)).

These results provide the first physiological evidence that Fyn deficient animals are profoundly immune- compromised. As formal proof of Fyn-dependent IL-2 production underpinning in vivo CD4⁺ T cell expansion was demonstrated in the colitis model system reported herein, a rational hypothesis underlying the observation that Fyn-deficient animals succumb to PR8 infection is that Fyn-deficient PR-8 specific T cells fail to sufficiently expand at the critical inflection point of the adaptive immune response; that in turn would suggest a central role for Fyn in T cell homeostasis.