T h e aorta can act as a site of naive C D 4 + T-cell priming

strate that naïve T cells can be primed directly in the vessel wall with both kinetics and frequency of T-cell activa­ tion found to be similar to splenic and lymphoid T cells. Aortic homing of naïve T cells is regulated at least in part by the P-selectin glycosylated ligand-1 receptor. In experimental atherosclerosis the aorta supports CD4+ T-cell activation selectively driving Th1 polarization. By contrast, secondary lymphoid organs display Treg expansion. Conclusion Our results demonstrate that the aorta can support T-cell priming and that naïve T cells traffic between the circula­ tion and vessel wall. These data underpin the paradigm that local priming of T cells specific for plaque antigens contributes to atherosclerosis progression. and by what mechanism are they recruited?; (ii) Can T-cell priming occur directly in the aorta and where do T-cell/DC contacts take place?; and (iii) How does the aortic CD4+ T-cell phenotype compare with T cells in SLOs and are there divergent effects on these distinct populations af­ ter induction of pathology? Our results demonstrate that the aorta can act as a site of naïve CD4+ T-cell priming and selectively induce a localized Th1 immune re­ sponse in early experimental atherosclerosis.


Introduction
Aortic adaptive immune responses play a role in atherosclerosis,1 with several immune cell subsets identified in human2,3 and murine4,5 vessels; however, to date the precise mechanisms leading to T-cell activation in the arterial wall remain poorly understood.
Several studies have investigated whether vascular resident antigen presenting cells (APCs) retain the ability to present antigen locally to T cells using reductionist approaches such as adoptive transfer of model antigen loaded dendritic cells (D C s),6 in vitro co-culture systems,7 and explanted aortas.8Building on these approaches, antigen presentation has been demonstrated in vivo by constitutive aortic plasmacytoid D Cs (pD Cs).9,10More recently, it was discovered that in the advanced stages o f atherosclerosis in apoE-/-mice, the vessel wall orchestrates the formation of artery tertiary lymphoid organs that control vascular T-cell responses with concomitant reduction in adjacent plaque size without involvement of secondary lymphoid organs (SLO s).4 Therefore, current evidence suggests that aortic A PC s possess the capacity to present antigen and that the aorta may act as a site of T-cell priming in atherosclerosis.
In earlier stage pathology, it is assumed that adaptive immune responses are co-ordinated in SLOs and/or the vessel wall, but naïve T cells have yet to be identified to reside constitutively in vascular tissue and it remains controversial as to whether T-cell priming can occur within the aorta.Therefore, in this investigation, w e set out to address

A nim als
B6.129P2-Apoe(tm 1Unc)/J (apoE-/-) mice, C57BL/6 (wild type; W T ), O T-II, and T E a mice w ere used in this study.A ll animals w ere euthanized by carbon dioxide.No anaesthetic agent was used.For full details see Supplementary material online.A ll the procedures w ere performed in accordance with local ethical and U K Home Office regulations and con form to the guidelines from Directive 2010/63/EU of the European Parliament on the protection ofanimals used for scientific purposes.

Flow cytom etry
Cell suspensions from aortas w ere prepared by enzyme digestion as pre viously described.4,9 For experiments that involved removal of aortic ad ventitia, a modified digestion protocol was performed to allow removal of adventitia.11Spleens and renal lymph nodes (rLN s; abdominal aortic draining) w ere digested in collagenase D (Sigma-Aldrich, Irvine, UK).
Intracellular staining was performed using a cytofix/cytoperm kit (BD Biosciences, Oxford, U K) o r a True-NuclearTM Transcription factor buffer set (Biolegend, London, U K) according to the manufacturer's instructions.For BrdU staining, a BD Pharmigen FITC BrdU flow kit (BD Biosciences) was used according to the manufacturer's instructions.
Because T cells represent a major population (~20% ) o f blood leuco cytes coupled with the fact that aortic naïve T cells were likely to be rare, we included an additional control for all T-cell experiments to en sure that T cells analysed were 100% aortic resident: 3 min prior to the experiment endpoint, animals w ere injected i.v. with a CD45.1 antibody (OT-II mice) o r a CD 45.2 antibody (W T , TEa, and apoE-/-mice) to label circulating blood leucocytes for subsequent exclusion from analysis (Supplementary material online, Figure S1B).Experiments w ere analysed on a LSR II, LSRFortessa (BD Biosciences) or a MACSQuant Analyzer (Miltenyi Biotec, Bisley, U K) using FlowJo (FlowJo L LC , Ashland, OR, USA).For full details and antibodies used see Supplementary material online.

A ssessm ent of Th1 cytokine expression
W T and apoE-/-mice were injected i.p. with 300 mg Brefeldin A (Sigma-Aldrich) to block cytokine release and culled 5 h later.Aortas, spleens, and rLN s w ere harvested and stained for intracellular cytokines via flow cytometry as detailed in Supplementary material online.

Naïve C D 4 + T cells exist constitutively in the adventitia and intima/medial layers of W T aorta
The combination o f perfusion and removal of extra-aortic tissue resulted in a pure vascular preparation (Supplementary material online, Figure S1).(E-G) In a separate experiment, OT-II mice were injected with ovalbu min or PBS and culled 72 h later.Single cell suspensions prepared for flow cytometry analysis of (E) aorta, (F) renal LNs, and (G) spleen were stained with CD62L and CD44 to once again distinguish naïve from activated cells.All plots are gated on CD4+Va2+Vp5.Results are from two independent experi ments, with each group containing five-pooled aortas.Spleen, LNs, and blood were analysed individually.

Prim ing of C D 4 + T cells occurs within the aorta
Evidence exists that T-cell activation occurs within the aorta,6,8,12 but T cells to W T mice (Figure 1E).Following administration of antigen, there was a switch towards a more activated CD44hi phenotype with a con comitant reduction in naïve T cells (Figure 1E).This was also true for rLN s and spleen (Figure 1F and G   In summary, these results clearly demonstrate that OT-II T cells are being primed locally within the vessel wall.

Naïve T-cell recruitm ent to the aorta is suppressed by blockade of PSGL-1
Having  o f cells w ere low, they were 10 times higher than that observed in W T aortas (Figure 5 E).

Early stage atherosclerosis is associated with vascular T-cell expansion, activation and selective Th1 polarization
On examination of rLN s and spleen (Figure 6), the C D 4 + populations were equivalent between W T and apoE-/-animals both for rLN s (Figure 6A) and spleen (Figure 6 E).Both tissues showed a modest increase in T EM cells (~5% ) in apoE-/-mice (Figure 6B

Supplem entary m aterial
the following questions: (i) Do naïve C D 4 + T cells reside in the aorta and by what mechanism are they recruited?;(ii) Can T-cell priming occur directly in the aorta and where do T-cell/D C contacts take place?; and (iii) How does the aortic C D 4 + T-cell phenotype compare with T cells in SLO s and are there divergent effects on these distinct populations af ter induction o f pathology?O u r results demonstrate that the aorta can act as a site of naïve C D 4 + T-cell priming and selectively induce a localized Th1 immune re sponse in early experimental atherosclerosis.

2 . 4
In one set o f experiments, OT-II mice w ere injected i.v. with 200 mg chicken ovalbumin (O V A ) and culled at 2 4 ,4 8 , o r 72 h later.Phosphatebuffered saline (PBS) served as naïve control.T w o hours prior to culling, mice w ere injected with 1 mg BrdU i.p.Single cell suspensions w ere pre pared for aorta, rLN s, and spleen, and C D 4 + O T-II+ T cells stained for proliferation markers: BrdU and Ki-67 as described in Supplementary material online.In a separate series o f experiments, OT-II mice w ere injected with 1 mg/kg FTY720 (Sigma-Aldrich) and 2 mg/kg anti-P-selectin (RMP-1) and anti-E-selectin (RME-1) antibodies (Biolegend) to block T-cell vascu lar recruitment, prior to receiving 200 mg O V A .Treatm ent with FTY720 and anti-P-selectin/anti-E-selectin was repeated on Day 2 (24 h following first dose) to maintain effective plasma concentrations.The control group received 200 mg O V A in addition to respective vehicle (dH2O ) and isotype control antibodies (mouse IgG1, k and mouse IgG2a, k).Mice w ere culled 48 h following O V A administration (Supplementary material online, Figure S2).Single cell suspensions o f aorta, rLN s, and spleen were obtained for flow cytometry analysis o f C D 4 + OT-II prolif eration as assessed by Ki-67.Blockade of aortic T-cell recruitm ent W T mice w ere treated with 100 mg anti-P-selectin glycosylated ligand (PSGL-1; 4RA10) o r equivalent dose of corresponding isotype (Bio X Cell, W e st Lebanon, N H , USA) at 0 and 24 h to block aortic T-cell re cruitment.Mice w ere culled at 48 h after first dose.Aortic single cell sus pensions w ere obtained for flow cytometry quantification of total C D 4 5 + cells, C D 4 + T cells, and naïve C D 4 + T cells.
Results are expressed as mean ± standard error of the mean (SEM) of n animals/groups o f pooled tissues for each experiment.The Student's t-test was used to compare two groups.Analysis o f variance (A N O V A ) was used for comparing three o r more groups with Tukey's multiple comparison post-test being applied as described in Figure legends.GraphPad Prism 6 software (San Diego, C A , USA) was used.A P-value <0.05 was taken to indicate statistical significance.

T
o determine the constitutive repertoire o f T cells that reside in the whole naïve aorta, w e employed high resolution cytometry by time of flight (C y T O F ) which revealed multiple distinct C D 4 + and C D 8 + T-cell populations, two y § populations and unidentified T-cell subtypes (Supplementary material online, Figure S3).Inspection of this data reveals that the aorta harbours a mix o f C D 44low and CD44hi T cells (Supplementary material online, Figure S3).Using flow cytometry, we found ~40% of aortic C D 4 + T cells expressed the classical naïve pheno type: C D 6 2 L+ C D 4 4 lo (Figure 1A).This compares with approximately two-thirds o f C D 4 + T cells being naïve in renal lymph nodes (rLN s), spleen, and blood (Figure 1B-D ).Further characterization demonstrated that a clear C D 4 + T-cell population could be observed both within the adventitia and intima/media (Supplementary material online, Figure S4B).
Analysis o f the T-cell population revealed naïve C D 4 + T cells in both compartments but with a higher proportion of activated T cells (CD44hi) being present within the adventitia (Supplementary material online, Figure S4C).T cells w ere also visualized by confocal microscopy and found to exist in both the intima and adventitia o f the naïve aorta Downloaded from https://academic.

F
ig u r e 1 Naïve T cells constitutively reside in C57BL/6 wild-type aorta and are reduced in OT-II murine aorta following antigen administration.(A-D) C57BL/6 wild-type mice were culled and single cell suspensions prepared for flow cytometry analysis of (A) aorta, (B) renal lymph nodes, (C) spleen, and (D) blood.C D 4+ T cells were stained with CD62Land C D 44to distinguish naïve from activated cells.All plots are gated on C D 4+ TC R -P+ T cells.Results are from two independent experiments, with each group containing five-pooled aortas.
including T cells bound to the aortic endothelium (Supplementary material online, Figure S7A).T he presence o f naïve T cells within periph eral tissues remains a matter o f controversy.T o further validate our finding that the aortic C D 6 2 L+ C D 4 4 -cells are truly naïve, we employed T Ea mice that consist of a C D 4 + T-cell population expressing a single T-cell receptor (T C R ) specificity and are also RA G deficient, thus T cells remain naïve in the absence o f Ea52-68:1-Ab complex (their cognate antigen).By employing this model, we are able to determine if naïve T cells are capable of trafficking to and maintaining a clear resident population within the aorta, even under non-inflammatory conditions.This was indeed what w e observed: In T Ea mice, the vast majority of C D 4 + T cells in the aorta are C D 6 2 L+ C D 4 4 -, conclusively demon strating that naïve C D 4 + T cells do indeed reside within the aorta (Supplementary material online, Figure S5A).Data from renal lymph nodes are presented for comparison where similarly, most C D 4 + T cells were C D 6 2 L+ C D 4 4 -naïve cells (Supplementary material online, Figure S5B).The aorta, therefore, acts as a reservoir o f naïve C D 4 + T cells that may form interactions with aortic resident antigen presenting myeloid cells.T o fully delineate the repertoire of myeloid cells within the naïve aorta, w e performed C y T O F analysis of C57BL/6 aortas which revealed 12 types of myeloid cells including multiple major histocompatibility complex Class II (M HC-IIhi) populations consisting o f distinct cD C and macrophage subsets (Figure 2).
in vivo proof o f naïve T-cell priming is lacking.Conventional approaches of assessing naïve T-cell homing and antigen-dependent activation via adop tive transfer and subsequent priming in recipient mice is not feasible given that recruitment of donor lymphocytes to the naïve aorta make up only a minor fraction o f the total aortic lymphocyte population.6Therefore, we adopted the method of only investigating endogenous T-cell homing and priming using OT-II mice.The use of OT-II mice, where the majority o f C D 4 + T cells have a T C R specific for chicken ovalbumin (O V A ) 323 339 in the context of I-Ab (alloantigen of H-2b bearing mouse strains, in cluding mice on C57Bl/6 background), allows us to assess whether the aorta can act as a site of priming by injecting OT-II mice with O V A so that the great majority of aortic C D 4 + T-cell-DC contacts will involve T cells with a T C R specific for the O V A peptide presented by aortic cDCs.PBS-treated OT-II aortas contained a similar proportion of naïve C D 4 +

2 F ig u r e 2
) indicating each site harbours an antigen-activated C D 4 + population in parallel with a reduced frequency A Mass cytometry reveals myeloid populations in C57BL/6 aorta.(A) Myeloid cells were gated as Lin-CD11 blo"hl and clustered using viSNE on the expression of cell surface and intracellular markers.Expression levels of selected myeloid markers in the resulting viSNE clustered cell populations is shown for a representative C57BL/6 mouse.(B) Twelve cell populations consisting of monocytes (Ly6C+ and Ly6C-), conventional Types 1 and 2 dendritic cells (cDC1 and cDC2), neutrophils, five macrophage subsets, and two unidentified populations.Doughnut plot shows mean proportion of subsets in aortas of n = 6 mice.(C) Heatmap showing the relative expression level of 20 cell markers within the 12 myeloid cell subsets identified by the viSNE clustering shown in B.

FF ig u r e 3 F ig u r e 4 T
Figure S7C).DC-T-cell contacts were not observed in PBS-treated mice.
identified naïve T cells within the aorta, w e investigated the mechanism o f naïve T-cell homing from the circulation into aortic tissue.PSGL-1 is the major leucocyte ligand for P-selectin, while also showing affinity for E-and L-selectin19 but its role in vascular recruitment o f naïve C D 4 + T cells remains unknown.T o investigate this, we injected C57BL/ 6 W T mice with the anti-PSGL-1 antibody, 4RA10,20 on two consecutive days and studied C D 4 + T-cell populations within the aorta 48 h following the first injection (Supplementary material online, FigureS8).Since PSGL-1 is common to all leucocytes, there was a decrease in the number of total leucocytes (>80%; 2.7-fold) in the vessel of anti-PSGL-1-treated animals (Supplementary material online, FigureS8A).A marked reduction was also identified for total C D 4 + T cells (88%; 4.8-fold; Supplementary material online, FigureS8B).Interestingly, both activated and naïve T cells w ere reduced by anti-PSGL-1 treatment with naïve C D 4 + T cells re duced by 96% which equated to a 5.7-fold reduction in this population (Supplementary material online, FigureS8C).O u r data demonstrate that PSGL-1 has a key role in promoting naïve T-cell recruitment to the aorta.

Finally, we
sought to assess differences in the phenotype of C D 4 + T cells in the aorta, rLN s, and spleen of W T vs. apoE-/-mice.A s already demonstrated by others,6 apoE-/-aortas contained significantly more (~400% ) C D 4 + T cells compared with W T (Figure 5A).In apoE-/-aortas, w e observed a lower frequency of naïve C D 4 + T cells and a significant Downloaded from https://academic.oup.com/cardiovascres/article/116/2/306/5454687 by guest on 21 December 2020 F ig u r e 5 CD 4+ T cells show a more activated phenotype in atherosclerotic aortas.Wild-type and apoE-/-mice were maintained on a high-fat diet for 10-12weeks.Mice were culled and aortic single cell suspensions assessed via flow cytometry for (A) CD 4+ T-cell number per vessel, (B) naïve (CD62L+CD44-) vs. effector memory (CD62L-CD44+) T-cell phenotype, (C) activation status, and (D) proliferation.(E) Additionally, aortas were stained for detection o fT RM cells (S1PR1-CD103+CD44+CD69+) cells.(A) Data shown represent n = 6-7 groups with each group comprising 3-8 pooled aortas from 6 to 7 independent experiments.(B-E) Data shown represent n = 3-4 groups with each group comprising 6-8 pooled aortas from 3 to 4 independent experiments.Individual data points represent average value per group; horizontal bars denote mean.Results are presented as mean ± SEM.The Student's unpaired t-test; *P < 0.05, **P < 0.01.switch from na ive (C D 6 2 L+ C D 4 4 -) to an effector memory (T EM) population (C D 6 2 L-C D 4 4+ ) was noted (Figure 5 B) concomitant with increased surface expression of the T-cell activation marker CD69 (Figure 5C).N o significant difference was noted for T-cell proliferation (Figure 5D).C D 4 + tissue-resident memory cells (T RM) have been identified in 21 22 23 several tissues including lung2 , and skin but have never been investigated in the vasculature.Using a well-defined marker profile for T R M , namely Sphingosine-1-phosphate receptor 1 (S1PR1)-C D 1 0 3 + C D 4 4 + C D 6 9 + , w e identified a small population fitting the T RM profile in the aortas of apoE-/-mice and while the absolute number

EF
and F, respectively) which is in contrast to the ~20% increase observed in apoE-/-aorta.Also, in contrast to the aorta, no increase in C D 4 4 + C D 6 9 + T cells was ob served in the rLN s (Figure6C) o r spleen (Figure6G).The cells o f rLN s in apoE-/-mice did; however, show a higher rate of proliferation compared with W T , a phenotype not observed in the spleen (Figure6D and H, respectively).The major pro-atherosclerotic T-cell response in atherosclerosis is driven by Th1 cells.24W e , therefore, employed direct in vivo intracellular cytokine stainingto quantify T cells actively secretingTh1 cytokines with out exogenous stimulation.25Following 12w eek's high-fat diet (H FD ), mice w ere injected with Brefeldin A 5 h before culling.C D 4 + T cells w ere assessed for the expression of the signature Th1 cytokine: IFN-y in addition to T N F-a (Figure 7A-C ).There w ere ~10 times more IFN-Y+ C D 4 + T cells in apoE-/-aortas compared with W T .Aortic Th1 cells w ere not actively producing T N F-a (Figure 7A).No significant differences in the Th1 phenotype w ere observed in either rLN s o r spleen (Figure 7B and C, respectively).W e considered that the increased frequency o f C D 4 4 hi C D 4 + T cells in the apoE-/-spleen and rLN s may be associated with Treg polarization and expansion which is known to play a protective role in the pathol ogy.24C D 2 5 + Fo xP 3 + Tregs were detected in both W T and apoE-/aortas and although a trend towards less Tregs was observed in apoE-/aortas, this did not reach significance (Figure 7D).This was in contrast to SLO s where T regs formed a larger proportion of the total C D 4 + popu lation with this value increasing in apoE-/-rLN s (Figure 7E) and spleen (Figure 7F).In summary, early stage atherosclerosis is associated with vascular T-cell expansion, activation and selective Th1 polarization, whereas SLO s are skewed towards enhanced Treg expansion as ob served previously.26ig u r e 6 C D 4+ T-cell phenotype in renal lymph nodes and spleen of apoE-/-mice vs. wild-type.Wild-type and apoE-/-mice were maintained on a high fat diet for 10-12 weeks.Mice were culled and single cell suspensions from (A-D) rLNs to (E-H) spleen were assessed via flow cytometry for (A, E) C D 4+ T-cell frequency, (B, F) naïve vs. effector memory T-cell phenotype, (C, G) activation status, and (D, H) proliferation.Data shown represent n = 20-21 mice from 3 to 4 independent experiments.Individual data points represent average value per mouse; horizontal bars denote mean.Results are presented as mean ± SEM.The Student's unpaired t-test; *P < 0.05, **P < 0.01, ***P < 0.001.

W
e have conclusively identified naïve T cells within the aorta o f several mouse strains.W e also illustrated how the aorta supports local T-cell priming in a model system.Indeed, in OT-II mice, aortic C D 4 + T cells can be primed as efficiently as lymphoid T cells.Naïve T cells also de pend, at least in part, on PSGL-1 for aortic homing.W e have also per formed a comparative assessment o f the C D 4 + T-cell phenotype between the atherosclerotic aorta, an aortic draining lymph node and the spleen, highlighting the importance o f the local aortic immune re sponse.A significant Th1 response evolved in the aorta, whilst T cells in SLO s were biased more towards a regulatory phenotype.These data support the hypothesis that atherosclerosis induces local vascular Th1 cell responses and that this local response is quite distinct from the more tolerogenic responses observed in SLOs.Naïve T cells have been identified in peripheral tissues including brain, pancreas, lung, skin, and testes.27Here, w e demonstrate that naïve C D 4 + T cells constitutively reside in the aorta of W T mice.W e found similar proportions o f T cells amongst total leucocytes in both the ad ventitia and intima/media with naïve T cells being detected in both com partments.W e confirmed by microscopy that T cells can be found bound to non-inflamed aortic endothelium and resided in close proxim ity to D C s indicating T cells may be able to enter the aortic wall from the lumen.Indeed, microvessels are generally absent in the healthy aortic wall,28 which strongly suggests that the major point of entry in naïve vessels is via the arterial lumen.The presence of naïve T cells within the intima/media is intriguing, given that the intima is the principal location of A PC s in the naïve aorta7 where they accumulate in the subendothelial space.Here, w e have fully delineated the repertoire of myeloid cells within the naïve aorta by C y T O F analysis revealing 12 types o f myeloid cells including multiple MHC-IIhi A P C populations.This offers the poten tial for naïve T cells to encounter A PC s in the environment where lipid accumulation/oxidation and atherogenesis begins.Much o f the evidence regarding local T-cell activation and clonal expansion directly in the ves sel is circumstantial such as co-localization immunohistochemistry of ath erosclerotic plaques3 or the use o f artificial systems ex vivo,8 hence it is unknown if T-cell priming can occur in the aorta in vivo.W e w ere able to demonstrate activation o f aortic OT-II T cells following antigen with ki netics paralleling those observed for splenic OT-II cells with peak prolif eration observed at 72 h.A t 72 h, we also observed BrdU uptake, indicating cells w ere undergoing mitosis.This was following a 2-h pulse of BrdU, which would strongly indicate the BrdU + cells were o f a local nature (i.e. in the target tissue of interest).T o confirm this hypothesis, w e pre-treated OT-II mice with FTY720 and blocking antibodies to P-and E-selectin prior to administration of antigen, thus ensuring any clonally expanded OT-II T cells within the aorta could only derive from endogenous naïve T cells encountering an aortic A P C presenting ovalbu min peptide on MHC-II.The result from this experiment confirmed that priming of naïve C D 4 + cells occurred in the aortic wall and is direct evi dence that A PC s within the vessel can uptake and present antigen locally Downloaded from https://academic.oup.eom/cardiovascres/article/116/2/306/5454687 by guest on 21 December 2020A

C D 25 F
ig u r e 7 Atherosclerosis induces divergent effects on aortic vs. lymphoid tissue T-cell polarization.Wild-type and apoE-/" mice were maintained on a high-fat diet for 10-12 weeks.Mice were then separated into two experimental groups.Group 1 received 300 mg of Brefeldin A 5 h prior to culling and single cell suspensions of (A) aorta, (B) rLNs, and (C) spleen were utilized for intracellular staining of C D 4+ T cells for the Th1 cytokines: IFN-y and TNF-a.Representative plots and graphs show IFN-y+ Th1 T cells.Plots are gated on C D 4+ TC R-P and results represent n = 3 per group with group containing three pooled aortas, three pooled rLNs, and n = 8 spleens from three independent experiments.In Group 2, mice were culled and single cell suspensions of (D) aorta, (E) rLNs, and (F) spleen were assessed via flow cytometry for the presence of C D 4+ Tregs (CD25+FoxP3+).Data shown for aorta represent n = 3-4 groups with each group containing 6-8 pooled aortas.Data shown for rLNs and spleen represent n = 20-21 individual organs.All data derived from 3 to 4 independent experiments.Individual data points represent average value per group/organ; horizontal bars denote mean.Results are presented as mean ± SEM.The Student's unpaired t-test; *P < 0.05, **P < 0.01, ***P < 0.001.to C D 4 + T cells and induce clonal expansion.Using OT-II mice, we were able to visualize T cells and D C s forming cell contacts following an tigen challenge.Such contacts were observable in the adventitia, media, and intima.O f interest, D C s and T cells w ere found to reside in areas of the media where muscle fibres w ere less dense.The observation that D C s and T cells can interact in all the aortic layers indicates that the ar chitecture o f the aortic wall can support local T-cell activation, even in the absence o f a local inflammatory stimulus.W h ilst priming o f T cells in healthy W T aorta is very unlikely given the small polyclonal population that exists with respect o f SLO s, the situation in athero sclerotic plaques may offer an environm ent m ore conducive to T-cell activation.In the context o f atherosclerosis-in the presence of chronic pro-inflammatory stimuli-local antigen presentation could, in theory, take place within the developing lesion, w here T cells and A P C s are m ore highly concentrated with T cells co-localizing with ac tivated D C s .2,3Evidence fo r aortic T-cell priming in atherosclerosis was previously suggested in apoE-/-mice by the fact that the T-cell repertoire within atherosclerotic aortas became m ore restricted o ver the course o f pathology w hile no changes in T-cell clonality could be detected in S LO s.12The mechanism by which naïve T cells enter the aorta is unknown.W e identified PSGL-1, the major leucocyte ligand for endothelial selectins, as a receptor o f interest in regulating homing o f naïve T cells into the aorta; since w e demonstrated that naïve T cells are depleted in the aorta following anti-PSGL-1 treatment.PSGL-1 blockade o r deficiency reduced atherosclerosis formation, adhesive interactions between endo thelial cells and leucocytes, and neointimaformation in apoE-/-mice.29-31Naïve T cells lack a fully glycosylated PSGL-1 so binding to selectins is lower than for activated T cells yet some binding affinity still remains.32Adoptively, transferred T cells have been previously shown to enter ath erosclerotic aortas in a partially L-selectin dependent manner despite Lselectin receptors being absent from aortic tissue.6This apparent dis crepancy can be reconciled by the fact that L-selectin expressed on an endothelial bound T-cell binds to PSGL-1 on an unbound T-cell (a pro cess termed secondary capture).33The initial interaction (primary cap ture) involves a T-cell binding to P-or E-selectin on the vessel wall.In fact, it has been demonstrated that the absence of L-selectin has no ef fect on aortic leucocyte primary capture and rolling.Moreover, following 12weeks HFD, neither T-cell number nor plaque area was altered be tween apoE-/-and apoE-/-/L-selectin-/-mice.34 W e also noticed that T EM cells w ere less affected by PSGL-1 blockade (2.7-fold reduction) compared with naïve cells, making up a greater pro portion o f the C D 4 + cells in the treatment group compared with iso type.T EM cells may use additional receptors, to facilitate binding to endothelial selectins including: T-cell immunoglobulin and mucin domain 1 (TIM-1),35 C D 44,36 and E-selectin ligand-1.37PSGL-1 on naïve T cells can also bind the chemokines, C C L1 9 , and C C L2 1 38 but expression of these chemokines in vascular tissue is low o r absent under non inflammatory conditions.4,39,40Therefore, w e consider these unlikely to contribute to the reduced trafficking observed in W T mice.There may also be chemotactic factors in the vessel wall derived from myeloid cells that contribute to the reduced trafficking o f naïve T cells observed after PSGL-1 blockade, as an indirect effect, due to concurrent reductions in other leucocyte subsets.Finally, we performed a phenotypic analysis of C D 4 + T cells between W T and apoE-/-aortas, under similar H FD conditions.In line with previ ous results,6 here, w e show that aortic C D 4 + T cells display a near four fold increase in numbers in apoE-/-mice compared with W T mice.W e next investigated the relative proportions of naïve vs. activated C D 4 + T cells.Naïve C D 4 + T cells w ere significantly reduced in aortas of apoE-/mice compared to W T , concomitant with an increase in T EM frequency.In contrast, when w e examined rLN s and spleens, w e discovered that the C D 4 + populations w ere equivalent in terms of magnitude with only a minor switch towards an activated phenotype, considerably less than what was observed in the aorta.This is consistent with a lack o f in creased proliferation o f splenic C D 4 + T cells previously observed at 8 weeks HFD in apoE-/-mice.41Th1 cells are pro-atherosclerotic in both humans3,42,43 and animal models.24By employing intracellular cytokine staining to reveal in vivo cy tokine prolifes, w e showed that IFN-Y+ (Th1) T cells w ere 10 times more abundant in apoE-/-aortas.In contrast, we did not detect significant differences in the Th1 population in either the rLN s o r spleen.W e also quantified C D 4 + Tregs, which are known to be atheroprotective in ani mal models.24Tregs w ere a small proportion of total C D 4 + T cells within the aorta with no significant differences between W T and apoE-/mice.In contrast, Tregs grew as a proportion o f C D 4 + T cells both in rLN s and spleen.In support of this data, a study conducted on Foxp3-eG FP/LDLr-/-mice showed a progressive increase in the frequency of splenic C D 4 + Tregs at 4, 8, and 20w eeks HFD , while the total C D 4 + T-cell population remained unchanged.26The immune-mechanisms that drive aortic T-cell activation and Th1 response in apoE-/-mice are likely to be multi-factorial with both anti genic and non-antigenic stimuli contributing to the aortic resident T-cell phenotype.One additional factor worth considering is the potential presence o f B cell follicles, resembling early tertiary lymphoid organs, which can be found in even young apoE-/-mice44 and these structures, if present, may exert immunomodulation on the underlying vascular T-cell response.W e have used a model in vivo system to illustrate the capacity that the naïve aorta has to promote local T-cell priming.O ther approaches that could have further validated these findings such as orthotopic aortic transposition utilizing an aortic graft from a donor with trackable (i.e.fluorescent) cells could also have aided in discriminating local from sys temic immune effects.However, not only are such approaches techni cally challenging, the grafts would yield a very low number o fT cells with respect to an entire intact aorta thus necessitating a large number of sur geries to produce sufficient tissue for quantifiable data.Future studies, however, utilizing such approaches coupled with detailed temporal phenotyping of T C R usage in experimental atherosclerosis would further enhance our knowledge of when and where antigenic stimulation of T cells occurs.In conclusion, the aorta can support T-cell priming and local activation o f C D 4 + T cells is associated with the vascular specific Th1 response we observed in early stage atherosclerosis in the apoE-/-aorta.