Regulation of global CD8+ T-cell positioning by the actomyosin cytoskeleton

18  Download (0)

Full text

(1)

ƐՊ|Պ !$Ɛ ĺ (  ( Ѷ

+

$ 

 $ ($     *   "     ( ( "

 !    "  ҃) $!      $ $ !  "

The adaptive immune system faces the formidable challenge to re-spond to a highly diverse variety of pathogenic organisms that have breached epithelial barriers or pose a potential danger. Signature molecular patterns derived from pathogens bind to cognate pathogen- associated molecular pattern (PAMP) receptors expressed

by innate immune cells to inform them on the presence and nature of the threat. The characteristics of the PAMP signaling and other cues induce adaptive immune reactions, which take into account pathogen size, and therefore its susceptibility to phagocytosis, the intra- vs extracellular location as well as the anatomical site of patho-gen entry (eg, gut vs skin). Based on this information, the ensuing adaptive immune responses can be classified into four types.1 Type

II immune responses are triggered by extracellular, multicellular

or-ganisms such as helminths, which are too large to be phagocytosed.

Type III immune responses are triggered by extracellular bacteria and

fungi, once they have breached epithelial barriers such as the epi-dermal layer of skin. Type IV immune responses are elicited against

!;]†Ѵ-|bomo=]Ѵo0-Ѵ Ѷ

+

$Ŋ1;ѴѴrovb|bombm]0‹|_;-1|ol‹ovbm

1‹|ovh;Ѵ;|om

;mv(ĺ"|;bm

Պ|ou-!†;=

Department of Oncology, Microbiology -m7ll†moѴo]‹ķ&mbˆ;uvb|‹o= ub0o†u]ķ ub0o†u]ķ"‰b|Œ;uѴ-m7

ouu;vrom7;m1;

Jens V. Stein, Department of Oncology, Microbiology and Immunology, University of ub0o†u]ķ ub0o†u]ķ"‰b|Œ;uѴ-m7ĺ

Email: jens.stein@unifr.ch

†m7bm]bm=oul-|bom

"‰bvv-|bom-Ѵ o†m7-|bomķu-m|ņ ‰-u7†l0;uĹ!"ƔōƐƕƏƖѵƖ-m7 31003A_172994; Swiss Cancer League, u-m|ņ‰-u7†l0;uĹ "ŊƒƔƑƓŊƏѶŊƑƏƐƓ

"†ll-u‹

CD8+ T cells have evolved as one of the most motile mammalian cell types, designed to continuously scan peptide–major histocompatibility complexes class I on the sur-faces of other cells. Chemoattractants and adhesion molecules direct CD8+ T- cell homing to and migration within secondary lymphoid organs, where these cells colo-calize with antigen- presenting dendritic cells in confined tissue volumes. CD8+ T- cell activation induces a switch to infiltration of non- lymphoid tissue (NLT), which differ in their topology and biophysical properties from lymphoid tissue. Here, we provide a short overview on regulation of organism- wide trafficking patterns during naive T- cell recirculation and their switch to non- lymphoid tissue homing during activation. The migratory lifestyle of CD8+ T cells is regulated by their actomyosin cytoskeleton, which translates chemical signals from surface receptors into mechanical work. We explore how properties of the actomyosin cytoskeleton and its regulators affect CD8+ T cell function in lymphoid and non- lymphoid tissue, combining recent findings bm|_;=b;Ѵ7o=1;ѴѴlb]u-|bom-m7-1|bmm;|‰ouhu;]†Ѵ-|bom‰b|_|bvv†;-m-|ol‹ĺ bm-ѴѴ‹ķ we hypothesize that under certain conditions, intrinsic regulation of actomyosin dy-namics may render NLT CD8+ T- cell populations less dependent on input from extrin-sic signals during tissue scanning.

 + )  ! "

lymphoid and non-lymphoid tissues, extracellular matrix barriers, naive and effector/memory CD8+ T-cell trafficking, regulation of actomyosin cytoskeleton

http://doc.rero.ch

Published in "Immunological Reviews 289(1): 232–249, 2019"

which should be cited to refer to this work.

(2)

extracellular pathogens before these have breached body barriers, such as against gut microbiota. Here, we focus on Type I immune re-sponses, which are specifically directed against intracellular patho-gens. These include some bacteria and parasites, such as Listeria monocytogenes and plasmodium, respectively, which have evolved to thrive inside host cells. Yet, the most common trigger for Type I immune responses are viral infections, which may affect any nucle-ated cell anywhere in the body including hematopoietic and stro-mal cell types. In recent years, it has emerged that cancerogeneous cells are also eliminated in this kind of immune response, which is exploited in checkpoint inhibitor therapies. Type I immunity involves  γ- secreting Th1 CD4+ T cells, inflammatory “M1” macrophages,

]Ŋruo7†1bm]  1;ѴѴvķ  1;ѴѴvķ -m7 1‹|o|oŠb1  Ѷ+ T cells. CD8+

T cells recognize cognate class I peptide–major histocompatibility complexes (pMHC) presented on the surface of antigen- presenting cells (APC) during priming and on infected or cancerogeneous cells in target organs. This is because class I pMHC reflect the intracellular proteome including pathogen- derived non- self peptides or peptides from neoantigen proteins produced by mutations of the genome. Once engaged by the recognition of cognate class I pMHC, effec-tor CD8+ T cells (T

) induce apoptosis of target cells. Thus, Type I

immune responses uniquely involve the directed killing of host cells to reduce or eliminate viral reservoirs. To accomplish this feat, naive CD8+ T cells have to identify rare APC, usually dendritic cells (DCs),

to become activated and generate large numbers of cytotoxic CD8+

T during clonal expansion. Once released into the circulation, CD8+ T

perform a “search and destroy” mission directed against

their own host cells. After contraction, memory CD8+ T cells

con-tinuously patrol lymphoid and non- lymphoid tissues (NLT) for rapid recall responses.

The process of target cell elimination has been extensively studied in simplified in vitro systems. In sum, effector CD8+ T cells

directly bind to the target cell expressing cognate pMHC to direc-tionally deliver the content of their cytotoxic granules including per-forin and granzyme B. Pores in the target cell membrane created by perforin and other molecules allow granzyme- triggered caspase activation, which in turn sparks a cell- autonomous apoptotic pro-gram.2 The close juxtaposition of target cell and CD8+ T cell ensures

high- precision, selective killing through apoptosis and subsequent elimination of cellular and viral particles by macrophages. This pro-cess also avoids expro-cessive inflammation associated with pyroptosis or necrosis. Recent work has identified actomyosin cytoskeleton- mediated force as an important mechanism to potentiate cytotoxic $Ŋ1;ѴѴ -1|bˆb|‹ĺ $_†vķ om1; - |-u];| 1;ѴѴ bv b7;m|b=b;7ķ u;|uo]u-7; Ŋ actin flow at the immunological synapse (IS) enhances the pore- forming activity of perforin.3

 †u|_;ulou;ķ 1ou|b1-Ѵ -1|bm m;|‰ouh reorganization at the IS facilitates centrosome repositioning and granule polarization required for targeted delivery of the cytotoxic cargo to the target cell.4 Taken together, the “destroy” part mediated

by CD8+ T

cells is well characterized on a cellular and molecular

level, including biophysical force generation.5-7

What about the “search” part? Numerous studies have uncov-ered the remarkable adaptability of CD8+ T cells to disseminate into

non- lymphoid organs during an immune response, leading immu-nologists to take their tissue infiltration capacity as a given. Yet, it is now recognized that CD8+ T cells may be excluded from tissues.

Truncated presence of CD8+ T cells and other leukocytes in solid

tumors strongly correlates with a poor prognosis for successful checkpoint inhibitor therapy.8-10 These observations have rekindled

interest in how cytotoxic T cells find their target cell inside complex microenvironments in vivo. pMHC molecules are transmembrane proteins anchored on the cell surface and therefore require direct close juxtaposition of membranes to trigger a signal in responsive T cells. As a search strategy, CD8+ T cells have developed a migratory

lifestyle based on tunable control of their actomyosin cytoskeleton dynamics.11,12 Thus, in contrast to, for example, stationary antibody-

secreting plasma cells, CD8+ T cells continuously undergo vigorous

cell shape changes and cell body displacement to probe their en-vironment and interrogate the surfaces of other cells for cognate pMHC. Chemoattractant receptors and integrins are essential factors to recruit blood- borne CD8+ T- cell populations into target

tissues. In addition, they play important roles in positioning CD8+

T cells close to APCs or into specific microenvironments. The bio-chemical signals triggered by chemoattractant receptors, integrins, and the TCR itself are transformed within cells into mechanical work through the activity of the actomyosin cytoskeleton. This enables CD8+ T cells to exert forces on neighboring cells and extracellular

matrix (ECM) for cytotoxicity and migration. Since any nucleated cell can become infected, CD8+ T cells have acquired the ability to

infil-trate and inspect highly diverse tissue environments, which differ in topography, adhesion molecules, rigidity, and confinement.13,14 Yet,

in contrast to the “destroy” part, the regulation of actomyosin cyto-skeleton for successful “search” is less comprehensively understood. In this review, we will present a short overview over CD8+ T- cell

trafficking patterns before and after encountering cognate pMHC, highlighting their scanning behavior in lymphoid organs and their switch to NLT trafficking in response to external cues from chemo-attractants and adhesion molecules. With this setting as the physio-logical background, we will outline some basic principles of how the dynamic actomyosin cytoskeleton network influences the efficacy of CD8+ T- cell scanning, combining recent findings in cell biology

‰b|_=;-|†u;vo=|bvv†;-m-|ol‹ĺ bm-ѴѴ‹ķ‰;‰bѴѴ;ŠrѴou;|_;_‹ro|_-esis that some CD8+ T cells may perform environmental scanning

without external cues from chemoattractants and adhesion mole-cules, pending on the balance of actin regulators and its network architecture, for autonomous homeostatic surveillance of NLT.

ƐĺƐՊ|Պ Ѷ

+

$Ŋ1;ѴѴu;1bu1†Ѵ-|bom-m7rv;-u1_bm

Ѵ‹lr_ob7ou]-mv

Activation of CD8+ T cells takes place in organized secondary

lym-phoid organs (SLO), such as lymph nodes (LN) and spleen. Although this task appears at first daunting given the low frequency of pMHC- specific T cells before clonal expansion (approximately 1 in 10ƔŊѵ

CD8+ T cells15) and the large surface of the host organism, CD8+

T cells are remarkably efficient in their search for rare cognate

(3)

  & !  Ɛ Պ Regulation of T- cell trafficking in lymphoid tissues. Snapshots from intravital 2PM image sequence showing T cells (green) in the parenchyme of a mouse popliteal LN. (i- iii) High magnification snapshots with graphical reconstruction and cell dynamics. (i) Top. -u-1ou|;Š‰b|_=;;7bm]-u|;uboѴ;ĺ$_; !m;|‰ouhbvv_o‰mbm]u-‹ĺo||olĺ1|ol‹ovbm7‹m-lb1vbmlo|bѴ;$1;ѴѴvĺoѴ-ubŒ;7$1;ѴѴv ];m;u-|;urƑņƒŊl;7b-|;70u-m1_;7 Ŋ-1|bmm;|‰ouhv-||_;Ѵ;-7bm];7];ķ‰_b1_Ѵ;-7v|oruo|u†vbomv-m7ķlou;blrou|-m|Ѵ‹ķ];m;u-|;- u;-u‰-u71ou|b1-Ѵ Ŋ-1|bm=Ѵo‰Őbm7b1-|;70‹0Ѵ-1h-uuo‰vőĺ$_uo†]_;m]-];l;m|‰b|_1‹|orѴ-vlb1|-bѴvo=bm|;]ubmv-m7o|_;u|u-mvl;l0u-m; molecules, retrograde flow generates low- adhesive friction with the substrate for forward cell displacement. At the trailing edge, or uropod, Ŋ-1|bm1-0Ѵ;v-u;1om|u-1|;7|_uo†]_-1|bˆb|‹o=momŊl†v1Ѵ;‹ovbmĺ$_bv_;Ѵrv|oruor;Ѵ|_;m†1Ѵ;†v-v|_;0b]];v|ou]-m;ѴѴ;|_uo†]_ small pores. Arp2/3 activity is modulated by chemokine receptor signaling and requires Rac1/2 and Cdc42 activity. The most important Rac  =oulo|bѴb|‹bm$1;ѴѴvbv Ƒĺ1|ol‹ovbm1om|u-1|bѴb|‹u;t†bu;v!_oŊ!Ŋl;7b-|;7r_ovr_ou‹Ѵ-|bomo=|_;u;]†Ѵ-|ou‹Ѵb]_|1_-bm o=‹ovbm-m7bv=-1bѴb|-|;70‹u;Ѵ;-v;o=|_;!_o-1|bˆ-|ouu_ Ɛ=uol7;roѴ‹l;ubŒbm]lb1uo|†0†Ѵ;vĺ!_oķ‹oƖ0ķu;7†1;v !_oŊ$Ѵ;ˆ;Ѵv-||_;Ѵ;-7bm];7];ĺŐbbő$orĺ (‰b|_;lb]u-|bm]$1;ѴѴĺ$_; !m;|‰ouhbvv_o‰mbm]u-‹ĺo||olĺmvb7;Ŋo†|vb]m-Ѵbm] Ő0Ѵ†;-uuo‰vő0‹!ƕ-1|bˆ-|bomѴ;-7v|ou-rb7bm7†1|bomo=-m;Š|;m7;7 ŊƐ1om=oul-|bomom0Ѵoo7Ŋ0oum;$1;ѴѴvĺ"_;-u=ou1;v;Š;u|;7 0‹0Ѵoo7=Ѵo‰bm7†1;-_b]_Ŋ-==bmb|‹ ŊƐ1om=oul-|bomķ‰_b1_bm|†um-1|bˆ-|;vo†|vb7;Ŋbm1‹|ovh;Ѵ;|-Ѵu;-uu-m];l;m|=ou1u-‰Ѵbm]Ő0Ѵ†; -uuo‰vőĺŐbbbő$orĺ;7†ѴѴ-u‹Ѵ‹lr_-|b1ˆ;vv;ѴŐ(őĺo||olĺv_-ѴѴo‰"Ɛ]u-7b;m|bvu;t†bu;7|oruolo|;$Ŋ1;ѴѴ;]u;vv|_uo†]_ ƑŊ 7;r;m7;m| Ŋ-1|bmŊ=bѴѴ;7ruo|u†vbom=oul-|bomĺ"1-Ѵ;0-uoˆ;uˆb;‰ķƔƏμm and inserts, 20 μm

(4)

pMHC- presenting cells. In fact, there is a broad participation of vir-tually all available clones during adaptive immune responses.ƐƔķƐѵ

bu1†Ѵ-|bm]m-bˆ; ѵƑ+ CCR7+ CD44low T cells are selectively

re-cruited from the blood circulation into LN via high endothelial ven-†Ѵ;vŐ (őĺ$_uo†]_|_;v;t†;m|b-Ѵ-1|bˆb|b;vo= ѵƑŐŊv;Ѵ;1|bmőķ !ƕ -m7  ŊƐķ ‰_b1_ 0bm7 |o |_;bu ;m7o|_;Ѵb-Ѵ 1o†m|;uѴb]-m7v peripheral node addressin (PNAd), the homeostatic chemokines CCL19/CCL21, and ICAM- 1/- 2, respectively, T cells engage in a mul-tistep sequence of selectin- mediated rolling and integrin- dependent arrest.17

m0ub;=ķ1omv|b|†|bˆ; ѵƑ-1|bˆb|‹r;ulb|v$Ŋ1;ѴѴ-||-1_-ment to sulfated and sialylated sugar residues O- linked to the protein backbone of PNAd components. The resulting fast on- rate binding with short half- lives results in a rolling movement along the HEV surface, which presents heparan sulfate- bound CCL19 and CCL2118

Ő b]†u;Ɛőĺ ƐƖ -m7 ƑƐ 0bm7 |o !ƕ ;Šru;vv;7 om uoѴѴbm] $ 1;ѴѴvķ‰_b1_Ѵ;-7v|o ŊƐ-1|bˆ-|bom=ou=bul-7_;vbomĺ$_;-0u†r| arrest through CCR7 “inside- out” signaling involves instantaneous |u-mv=oul-|bom o= |_; 0;m|  ŊƐ 1om=oul-|bom =o†m7 om 0Ѵoo7Ŋ borne cells into an extended conformation for binding to ICAM mol-ecules copresented on the HEV surface.19 This process is mediated

0‹ |_; vl-ѴѴ $-v; !-rƐ -m7 bv v†rrou|;7 0‹ l;1_-mb1-Ѵ v_;-u forces exerted by continuous blood flow.20 Active, ligand- engaged

 ŊƐ mo| omѴ‹ =bulѴ‹ -uu;v|v uoѴѴbm] 1;ѴѴv 0†| -Ѵvo |ub]];uv Ŋ-1|bm polymerization through “outside- in” signaling. This leads to T- cell flattening and crawling along the high endothelial cell surface. T- cell crawling requires ICAM- 1 on the HEV surface, whereas ICAM- 2 or VCAM- 1 play no discernible role in this process.21  ŊƐŊŊƐŊ

dependent T- cell crawling is reminiscent of the mesenchymal mode o= 1;ѴѴ lo|bѴb|‹ķ ‰_;u; bm|;]ubmŊ|ub]];u;7 -7_;vbom -m7 Ŋ-1|bm =ou-mation permits cell body translocation along a 2D surface.22 During

crawling, T cells actively probe the endothelial cytoskeleton for per-missive sites for transendothelial migration (TEM).23 As a final step,

T cells transmigrate through or in between HEV, which is facilitated by nuclear lobe protrusions.24

The observation that T cells are able to move away from an excess of luminal CCL21 is puzzling and the underlying molecular mecha-nism is not completely understood to date. It has been proposed that apical chemokine- mediated TEM is mediated by shear flow, in a pro-cess called chemorheotaxis.ƑƔķƑѵ Indeed, chemorheotaxis is readily

recapitulated for T cells in in vitro flow chambers. When CCL21 is added on top of an endothelial cell layer, naive T cells rapidly trans-migrate below the endothelial cell layer under flow.27 In this setting,

lack of shear flow impairs transmigration. Mechanistically, one con-ceivable scenario is that shear stress- extended integrins assemble bm|u-1;ѴѴ†Ѵ-u-1|bˆ-|ouvo=|_;vl-ѴѴ$-v;71ƓƑĺ$_bvbm|†uml-‹ lead to filopodia and podosome formation at the interface between crawling lymphocytes and the underlying endothelial surface to ini-tiate TEM.

Alternatively, T cells crawling on endothelium have been shown to form protrusions that tap into intracellular pools of prestored chemokine inside vesicles within endothelial cells to promote trans-migration.28

 †u|_;ulou;ķ _b]_ ;m7o|_;Ѵb-Ѵ 1;ѴѴv |_;lv;Ѵˆ;v l-‹ contribute to T- cell transmigration by acting as a shuttle to deliver

their cargo pending on the “crowdedness” of the surrounding inter-stitium.29 According to this model, endothelial cells sense the density

of the LN interstitium and act as gatekeepers to maintain constant lymphocyte numbers inside the lymphoid tissue volume. One caveat to this proposal is that the LN volume itself is likely to change rapidly and frequently upon expansion or contraction of the surrounding |bvv†;ĺ bm-ѴѴ‹ķ-mbllo0bѴbŒ;71_;lohbm;]u-7b;m|-uo†m7|_; ( may contribute to the directed entry of lymphocytes across the HEV barrier.30 This may further explain how transmigrated T cells are able

to cross the endothelial basement membrane (BM) consisting of col-lagen IV (Col IV), laminins, and other ECM components.

Once inside lymphoid tissue, T cells rapidly polarize and move along CCL21- and ICAM- 1- expressing fibroblastic reticular cells Ő !vőo=|_;$Ŋ1;ѴѴŒom;ĺ31  !v=oul-vrom];ŊѴbh;Ѵoov;ƒ m;|-work that serves as a robust structural scaffold for T- cell surveillance o= Ѵ‹lr_ob7 |bvv†;ĺ ! ‰u-r |_;lv;Ѵˆ;v -uo†m7 v;1u;|;7  fibers consisting of collagen I (Col I), Col IV, ERTR7, and laminins, ‰_b1_=oul|_; !1om7†b|v‹v|;l=ouvl-ѴѴloѴ;1†Ѵ;|u-mvrou|ĺ32 Intravital twophoton microscopy (2PM) imaging of the LN microen-vironment in live, anesthetized mice has uncovered that migrating T 1;ѴѴv-7-r|-roѴ-ubŒ;7-l;0ob71;ѴѴv_-r;-m7=oѴѴo‰-u-m7ol !Ŋ guided motility pattern with high basal speeds of 12- 15 μm/min, that is, more than one cell diameter per minute33 (Minor differences in

cell speeds between distinct laboratories may be in part due to tech-nical considerations such as sampling frequency34). Thus, within the

technical limitations of 2PM imaging, there is no discernable cluster- dependent motility or directed migration on a population level.35 It

is, therefore, common to observe fast migrating T cells moving in opposite directions within the same field of view, occasionally even crossing paths. In addition to speed and directionality, a useful pa-rameter to quantify T- cell motility is the motility coefficient (MC), expressed in μm2/min. In physical terms, the MC corresponds to

the 3D diffusion coefficient of a gas particle randomly moving in Brownian motion. In LNs, the MC of T cells can exceed 70 μm2/min,

indicating a high ability to scan lymphoid tissue.

b0uo0Ѵ-v|b1u;|b1†Ѵ-u1;ѴѴv=†Ѵ=bѴѴ-|Ѵ;-v|=o†u=†m1|bomv=oum-bˆ; $Ŋ1;ѴѴlb]u-|bomķv†uˆbˆ-Ѵ-m7-1|bˆ-|bomĹ buv|ķ|_;v;1;ѴѴv-1|-v]†b7-ance cue for highly motile T cells by producing the promigratory chemokines CCL19 and CCL21, as well as ICAM- 1. Of note, CCR7 ligands are not the only promigratory factor inside lymphoid tissue for T cells. Autotaxin- generated lysophosphatidic acid (LPA) stimu-lates T- cell motility by activating cell contractility.ƒѵŊƒѶ";1om7ķ !v

secrete IL- 7, which together with CCL19 and CCL21 acts as survival- promoting cytokine.39

$_bu7ķ v†v;|_; !m;|‰ouh-vv|u†1|†u-Ѵ scaffold for DC attachment and scanning by T cells.40 bm-ѴѴ‹ķ !v

control immune responses through the release of prostaglandin E2, nitric oxide, and other mediators.41

The average dwell time of a T- cell population in a given mouse bv-rruoŠbl-|;Ѵ‹ѵŊѶ_o†uv-m7bv7;|;ulbm;70‹|_;0-Ѵ-m1;o= the egress- promoting sphingosine- 1- phosphate receptor 1 (S1P1) and the retention- promoting CCR7.42 Since S1P levels within

lym-phoid tissue are low in contrast to blood and lymph, recirculating T cells gradually increase S1P1 surface levels from very low levels right

(5)

after entry into LN until reaching levels to become responsive to S1P in efferent lymphatic vessels. This returns cells via downstream LN and the thoracic duct into the blood circulation to ensure constant turnover of the TCR repertoire.43 A detailed 2PM analysis of

acti-vated WT and S1P1ƴņƴ effector T cells in reactive LN uncovered that

both populations probed efferent lymphatic vessels but only WT T cells were able to cross into the lymphatic volume for egress.44

This observation suggests that a shallow S1P gradient suffices to promote lymphocyte migration across the lymphatic vessel and its †m7;uѴ‹bm]Ő b]†u;Ɛőĺ$_;v;7-|--ѴvoѴ;m7v†rrou||o|_;mo|bom that despite vivid interstitial migration, T cells require local chemoat-tractants to cross dense tissue barriers such as BM in shear flow- free ;mˆbuoml;m|vķ-r_;mol;mom;ŠrѴ-bm;70‹bm|ubmvb1 Ŋ-1|bmm;|‰ouh properties (see below).

Naive T- cell motility within lymphoid tissue has evolved to bal-ance two opposing aims: On the one hand, T cells need to scan a large number of APC before their S1P1 surface levels start to pro-mote egress via efferent lymphatic vessels.45 On the other hand, T

cells must be able to dwell long enough on individual DCs to inte-grate sufficient signals for activation. Recent work has confirmed that high intrinsic motility allows T cells to find rare DCs within large tissue volumes.ƓѵŊƓѶ Yet, it also allows cells to detach from DCs

pre-senting low amounts of cognate pMHC or low- affinity pMHC during the sampling phase of T- cell activation.49 This is in line with

pioneer-ing work by Dustin and colleagues suggestpioneer-ing competitive roles be-tween T- cell motility vs activation.50 Nonetheless, T cells can arrest

eventually for prolonged periods (h) even on “suboptimal” DCs once they have accumulated enough signals through short contacts.51 To

ensure prolonged LN dwell time under these circumstances, surface †ru;]†Ѵ-|bomo= ѵƖ_-v;ˆoѴˆ;7|o0bm7"ƐƐbmcis and induces its bm|;um-ѴbŒ-|bomĺ ѵƖ†ru;]†Ѵ-|bombvbm7†1;70o|_0‹vr;1b=b1$! signaling on cognate T cells and, transiently and to lower surface Ѵ;ˆ;Ѵvķ 0‹ |‹r;    om -ѴѴ $ 1;ѴѴvĺ52 †u|_;ulou;ķ $Ŋ1;ѴѴ -1|bˆ-|bom leads to a transient loss in general migratory capacity of these cells, which is not well understood on a molecular level. This low motility presumably contributes to keep activated T cells trapped inside lym-phoid tissue for full signal integration.53,54

ƐĺƑՊ|Պ Ѷ

+

$Ŋ1;ѴѴ-1|bˆ-|bom-m7vb]m-Ѵbm|;]u-|bom

‰b|_bmѴ‹lr_ob7ou]-mv

Productive CD8+ T cell- DC interactions leading to expansion and

ef-fector differentiation require pMHC class I (signal 1) and costimula-|ou‹loѴ;1†Ѵ;v ѶƏ-m7 ѶѵŐvb]m-ѴƑőom vķ‰_bѴ;1‹|ohbm;v such as IL- 2 and IL- 12 provide signal 3. CD8+ T cells linearly integrate

the input of these three signals to adapt their expansion according to the stimulatory strength.55 On the other hand, early TCR signaling

;ˆ;m|v v†1_ -v ! r_ovr_ou‹Ѵ-|bom -u; o=|;m 1_-u-1|;ubŒ;7 0‹ - digital, all or nothing response.Ɣѵ How can T cells adapt the level of

input signal for an adequate response when initial signaling events are rather binary? One answer derives from intravital observation of CD8+ T- cell interactions with DCs presenting identical amounts of

signal 2 and signal 3 but loaded with altered peptide ligands (APLs)

to grade signal 1 strength. In this experimental setting, early T DC interaction dynamics are essentially identical, with an immedi-ate arrest of CD8+ T cells irrespective of functional potency of the

APL used for stimulation. Such a behavior correlates well with a digital behavior of early TCR signaling. In contrast, the total dura-tion of these interacdura-tions was determined by the funcdura-tional affin-ity of the peptide. Thus, CD8+ T cells detached from low potency

cognate pMHC- presenting DCs earlier than from medium- or high potency- presenting DCs and constituted an early- wave CD8+ T

population.57 The affinity- dependent kinetics of T- cell detachment

from DCs is therefore emerging as a mechanism to tune the extent of the signal integration in an analog manner while maintaining early binary responses.

T- cell detachment from DCs occurs efficiently in the lymphoid tissue microenvironment.58 How this process unfolds under

physi-ological conditions is currently unclear, also because T- cell detach-ment from APCs is difficult to reproduce in vitro. However, early evidence suggests that chemokines may be involved in this pro-cess.50 A recent study by Bousso and colleagues has uncovered that

once T cells had disengaged from DCs, they become temporarily un-responsive to additional pMHC on DCs in part owing to defective Ca- signaling responses.59 This appears particularly relevant for low-

affinity- primed T cells, whereas high- affinity- primed CD8+ T cells

re-tain a longer window for additional signal integration by DCs. Whole LN reconstructions after light sheet fluorescent microscopy imaging have uncovered that recently primed CD8+ T cells upregulate the

inflammatory chemokine receptor CXCR3 to accumulate at inter-follicular regions of LN, where incoming DCs enter lymphoid tissue. There, these cells engage in productive secondary encounters for prolonged activation before becoming late- wave CD8+ T

popula-tions.57*!ƒŊ7;r;m7;m|u;Ѵo1-ѴbŒ-|bom|o! vbv-Ѵvoo0v;uˆ;7bm

central memory CD8+ T cells (T

CM), both in steady- state and shortly

after viral infections. This process ensures a rapid encounter with incoming DCs and virus- capturing macrophages.ѵƏķѵƐ In sum,

chemo-kines regulate both encounter with DCs at distinct locations within lymphoid tissues and perhaps also their detachment from these cells as prerequisite for egress.

ƐĺƒՊ|Պ Ѷ

+

$

bm=bѴ|u-|bombm|o$v

oѴѴo‰bm] |_;bu -1|bˆ-|bom bm  -m7 vrѴ;;mķ  Ѷ+ T undergo a switch from a SLO- targeted to a non- lymphoid tissue (NLT) traffick-ing pattern. This is important since infected cells are often localized at epithelial barrier tissues with restricted access for naive T cells. $_†vķv|-u|bm]-v;-uѴ‹-vƒѵŊƓѶ_o†uvrov|rublbm]-m7r;-hbm]0;-tween 3 and 7 days postinfection, CD8+ T

enter the blood

circula-tion and invade multiple organs. Accordingly, T- cell activacircula-tion leads to a reprogramming of chemoattractant receptor surface levels, such as de novo expression of inflammatory chemokine receptors CXCR3, CCR4, CCR9, and CCR10, as well as increased levels of the hyaluro-m-mu;1;r|ou ƓƓ-m7|_;bm|;]ubmv ŊƐ-m7(ŊƓĺѵƑ In parallel,

T- cell activation leads to the expression of glycotransferases includ-ing enzymes for core2- O- glycan synthesis and fucosyltransferases

(6)

(-m7(ķ‰_b1_lo7b=‹Ŋv;Ѵ;1|bm]Ѵ‹1oruo|;bmѴb]-m7ŊƐŐ"ŊƐő|o present ligands for endothelial selectins.ѵƒķѵƓ In sum, these changes

promote the recruitment of CD8+ T

into infected or injured

or-gans. The anatomical location of the SLO where naive T cells first see their Ag and become activated determines the imprinting of their homing capacity toward mucosal vs skin- associated tissues. Thus, vitamin A derivatives from nutrient uptake imprint CCR9 and α4β7- mediated gut- homing capacity on T cells activated in gut- draining Peyer's patches and mesenteric LN. In turn, vitamin D derivatives induced by sunlight lead to selectin ligand synthesis and CCR10 expression, which imprint skin- homing capacities in activated T cells.ѵƔķѵѵ

Complementary to changes in CD8+ T

homing molecules,

tis-sue inflammation promotes increased adhesiveness of postcapillary venules through surface expression of P- and E- selectin, inflamma-tory chemokines such as the CXCR3 ligands CXCL9 and CXCL10, and the integrin ligands ICAM- 1/VCAM- 1, and MAdCAM- 1 in muco-sal tissues.ѵƑ$_bvbvl;7b-|;70‹|_;Ѵo1-Ѵu;Ѵ;-v;o=ŊƐ-m7$ α

by tissue macrophages that act as sentinels through their PAMP re-ceptors. In addition to recruitment to inflamed organs, CD8+ T cells

often accumulate in NLT irrespective of the presence of antigen or obvious inflammation in these organs. Such non- specific CD8+ T

accumulation likely depends on the levels of systemically circulating ŊƐ-m7$ α that increase adhesion receptors and chemokines on all postcapillary EC surfaces.ѵƕ In addition, skin and mucosal tissues,

which are constantly exposed to microbiota or microbiota- derived molecules, display constitutive adhesiveness of their postcapillary venules. As example, dermal postcapillary venules express E- selectin on their surface, which permits baseline adhesion of rolling leuko-cytes in the absence of overt inflammation.ѵѶ In sum, homing of

blood- borne TN and T is critically dependent on adhesion mole-cules and chemoattractant receptors that respond to organ- specific complementary receptors on endothelial cell surfaces, in a manner akin to a combinatorial area code.ѵƖ

Once accumulated in their target organ, the presence of cognate Ag boosts CD8+ T

numbers at sites of infection and contributes

to their long- term retention.70 In skin viral infections, CD8+ T use

CXCR3 to accumulate at viral foci that release the ligands CXCL9 and CXCL10.71,72 In contrast to the directed migration of

neutro-phils to sites of tissue injury or bacterial deposition,73,74 epidermal

CD8+ T cells achieve this by a subtle modulation of their migration

angle distribution toward the source of CXCR3 ligands.72 Similarly,

CXCR3 facilitates dermal CD8+ T

migration to the epidermis. 75

In adjuvant- inflamed skin cells, the αv integrin contributes to tissue scanning of dermal effector CD4+ T cells.ƕѵ Thus, chemokines and

integrins regulate the migration of T in acutely inflamed tissue, at least in some settings.

ƐĺƓՊ|Պ ‹m-lb1|bvv†;v†uˆ;bѴѴ-m1;0‹l;lou‹ Ѷ

+



$1;ѴѴv

After the clearance of a viral infection, most CD8+ T

die through

apoptosis in the contraction phase. The remaining memory cells are

classified into several subsets. Central memory CD8+ T cells (T CM)

continue to patrol secondary lymphoid organs similar to naive T cells. As outlined above, these cells preferentially localize or become rapidly recruited after infection to interfollicular regions, where Ag is first transported to by incoming migratory DCs. A recent study has uncovered that TCM efficiently infiltrate NLT without the need to become first activated in LNs or spleen.77 This function was mostly

attributed to another memory CD8+ T- cell subset, the effector

mem-ory T cells (TEM). Originally, TEM had been defined in human blood samples as CD45RO+

 l;lou‹ $ 1;ѴѴv ‰b|_ Ѵo‰ !ƕ -m7  ѵƑ levels.78 ѵƑƴ CCR7ƴ CD44high CD8+ T

EM are also readily

identifi-able in mice. Yet, their precise function during recall responses is incompletely understood, and CD8+ T

EM may actually comprise a

variety of distinct subpopulations. In this context, recent work has identified three types of effector/memory CD8+ T cells, which are

characterized by varying degrees of CX3CR1 expression. CX3CR1+

CD8+

$1;ѴѴv-rr;-u-uo†m77-‹ѵrov|ˆbu-Ѵbm=;1|bom-m71-m0;7b-vided into CX3CR1int and CX3CR1high subsets. Only the CX3CR1int

population was effectual in peripheral tissue surveillance.79 The

pre-cise function of CX3CR1 and its ligand CX3CL1 is not clear, since the absence of this receptor does not prevent the formation of the three effector/memory T- cell subsets.79

A recently described memory CD8+ T- cell subset derives from

|bvv†;Ŋbm=bѴ|u-|bm]!ŊƐlow CD8+ T . Instructed by tissue- specific 1†;vķ v†1_ -v $ Ŋβ and IL- 15 in skin, these cells become resident in the target tissue after pathogen clearance and hence are called tissue- resident memory T cells (TRM). Many TRM †ru;]†Ѵ-|;  ѵƖ and CD103 in epithelial- rich tissue.80-83 In contrast to other memory

T cell subsets, TRM do not enter blood or lymph in large numbers, but remain inside their tissue of residence as a self- maintaining popula-tion for prolonged periods of time. While TRM were first described in skin following local viral infections, they have now been uncovered in virtually all organs, including gut, salivary glands, liver, lung, geni-tourinary tract, central nervous system, and SLO.84,85 Their function

bv0;v|†m7;uv|oo7bmvhbm-m7|_;];mb|o†ubm-u‹|u-1|ĺ oѴѴo‰bm]- re- exposure to the original pathogen- derived pMHC, TRM rapidly se-1u;|; γ and other cytokines. This in turn activates tissue macro-phages and other cells of the innate immune system and provokes a tissue- wide alert status.81,83 Thus, the presence of T

RM in barrier

tis-sue locally reverses the paradigm that innate immunity precedes the onset of adaptive immune reactions. A second consequence is the rapid recruitment of circulating memory T cells to contain expanding microbes.82 †u|_;ulou;ķ Ѷ+ T cells with a T

RM phenotype have

been identified in tumor tissue and their presence strongly cor-relates with a good prognosis.Ѷѵ

bˆ;m|_;bubm=bѴ|u-|bombm|o-‰b7;u-m];o=7bv|bm1||bvv†;vķ Ѷ+ TRM constitute an ideal tool to address how cells from the same starting pool of naive T cells adapt to specific microenvironments. Intravital imaging of distinct TRM populations in epidermis, liver, and genitourinary tracts has provided evidence of a continuous scanning behavior, even in the absence of replicating pathogens.87-90 This

re-flects their dependence on recognizing membrane- bound pMHC in order to fulfill their function of host cell surveillance. However, TRM

(7)

speeds and MC as a proxy for scanning efficacy differ vastly within different tissues, which may reflect physical constraints of their mi-croenvironments. In the epidermis consisting of tightly packed layers of keratinocytes, TRM speeds become as low as 1- 2 μm/min, which restrains their dissemination away from the site of initial recruit-ment.87,88 In the connective tissue of dermis, T

RM speeds increase

|o ѵŊѶμm/min.91 How do motile T

RM achieve their long residence

time in tissue containing blood and lymphatic vessels without being flushed away or migrate out of the tissue? A key step in establish-ing tissue residency is the downregulation of the transcription factor  Ƒķ‰_b1_ruolo|;v;Šru;vvbomo="ƐƐ-m7!ƕĺ92 Accordingly, forced expression of S1P1 counteracts the establishment of long- term resident CD8+ T- cell populations. Therefore, unresponsiveness

to S1P and potentially CCL21 secreted by lymphatic vessels helps to maintain TRM within tissue. In skin epidermis, TRM typically localize close to the dermal- epidermal junction both in mouse and human tis-sue sections.93 Their retention in this part of the basal keratinocyte

layer may be imposed by local chemoattractants or specific adhesive interaction to prevent them from exiting the epidermis by the con-stant renewal of upper layers from proliferating basal keratinocytes. In support of this, CD103 (αE integrin) binds E- Cadherin on epithelial cells and contributes to local TRM retention but does not influence mi-gration speeds in epidermis.75Ѵom]|_;v-l;Ѵbm;ķbm1u;-v;7 ŊƐ

levels are required for long- term retention of sinusoid- patrolling TRM in liver.90 Taken together, regulation of adhesiveness to the

sur-rounding epithelium and low surface levels of receptors for egress- promoting chemoattractants contribute to TRM tissue residency.

ƑՊ|Պ !$Ƒ ĺ!  &  $    Ѷ

+



$҃   $ $ ++$  $  +"  

 + $"   $ 

Integrins and chemoattractant receptors trigger intracellular signal-ing pathways that lead to changes in shape and dynamic behavior. Eukaryotic cells typically express four types of protein families that determine their shape and biophysical properties: the microtubu-lar network, intermediate filaments, septins and the actomyosin network. Each of these scaffolds plays important roles for CD8+

T- cell function. Thus, microtubules in effector CD8+ T cells deliver

cytotoxic granules to target cells.2 †u|_;ulou;ķ 7;rѴ;|bom o= |_;

lb1uo|†0†Ѵ-u m;|‰ouh bm1u;-v;v -1|bˆ; !_oŊ$ Ѵ;ˆ;Ѵvķ ‰_b1_ bm turn induces contraction and protrusion retraction.94 The role for

intermediate filaments such as vimentin in CD8+ T cells is not well

studied, although these filaments contribute to the rigidity of cir-culating lymphocytes.95 Septins consist of 13 members, which form

hetero- oligomeric complexes and higher order structures such as rings.Ɩѵ In CD8+ T cells, septins are required for persistent motility

and proliferation.97,98 Here, we focus on the actomyosin

cytoskel-eton, arguably the most critical network to convert biochemical sig-nals into mechanical work for CD8+ T- cell shape and displacement.

Ŋ-1|bm bv 1olrov;7 o= roѴ‹l;ubŒ;7 Ŋ-1|bm lomol;uvĺ )_bѴ; |_; -77b|bom o= Ŋ-1|bm lomol;uv |o ru;Ŋ;Šbv|bm] 0-u0;7 Ŋ-1|bm ;m7v

o11†uv u-rb7Ѵ‹ķ |_; bmb|b-Ѵ roѴ‹l;ubŒ-|bom o= bm7bˆb7†-Ѵ Ŋ-1|bm 0;-yond a dimer/trimer requires nucleation factors.99 Prominent

nu-cleation factors of CD8+ T cells are the heptameric Arp2/3 complex

and the formins mDia1 and mDia2. The Arp2/3 complex becomes -1|bˆ-|;7 0‹ 71ƓƑŊ$ -m7 !-1Ɛņ!-1ƑŊ$ ˆb- |_; m†1Ѵ;-|bomŊ promoting factors WASP and SCAR/WAVE, respectively, to cre--|; m;‰ Ŋ-1|bm =bѴ-l;m|v om ru;=oul;7 Ŋ-1|bm =bѴ-l;m|v bm - ƕƏŦ -m]Ѵ;ĺ$_†vķurƑņƒŊl;7b-|;7m†1Ѵ;-|bom1u;-|;v-0u-m1_;7 Ŋ-1|bm network and is preferentially generated at the leading edge of mi-grating CD8+

 $ 1;ѴѴv |o =oul rv;†7oro7v Ő b]†u;Ɛőĺ v;†7oro7v are evolutionary conserved actin- filled protrusions characteristic of ameboid cell motility and require both WASP and SCAR/WAVE.100

In high- resolution microscopy, pseudopods actually appear as inter-leaved microlammelipodia.101 The precise actin network

architec-|†u;bv7;|;ulbm;70‹ Ŋ-1|bmŊ0bm7bm]ruo|;bmvĺ ņ("ruo|;bmv protect barbed ends from capping protein and thus contribute to filament elongation. Similarly, the actin nucleation factor mDia pro-tects growing actin filaments, whereas cofilin destabilizes existing filaments to promote actin turnover.102 The Arp2/3 inhibitor Arpin

-m7 1ouombm -Ѵvo Ѵblb| urƑņƒŊl;7b-|;7 Ŋ-1|bm m;|‰ouh ;Šr-m-sion.103-105$_†vķ0u-m1_;7 Ŋ-1|bmm;|‰ouhv-u;_b]_Ѵ‹7‹m-lb1‰b|_

constant filament turnover.

In contrast to Arp2/3, the nucleation factors mDia1/2, also hmo‰m-v=oulbmvķl;7b-|;-1|bmm†1Ѵ;-|bom|o1u;-|;Ѵbm;-u Ŋ-1|bm filaments. Upon phosphorylation of the regulatory myosin light chain (MLC), non- muscle Myosin II assembles into bipolar filaments |_-|;m-0Ѵ;Ѵbm;-u Ŋ-1|bm=bѴ-l;m|v|oloˆ;bm-m-m|br-u-ѴѴ;Ѵl-m-ner, leading to contraction.ƐƏѵ Myosin- mediated contractility of the

1ou|b1-Ѵ Ŋ-1|bmm;|‰ouhbmu;v|bm]1;ѴѴvu;v†Ѵ|vbm|_;vr_;ub1-Ѵ1;ѴѴ shape observed in the absence of extracellular polarity- inducing agents. Upon induction of polarity (eg, by addition of chemokines), pMLC accumulates at the trailing edge of migrating cells, where it l;7b-|;v†uoro71om|u-1|bѴb|‹Ő b]†u;Ɛőĺom|u-1|bѴb|‹v;uˆ;vv;ˆ;u-Ѵ critical functions. It is required to detach the trailing edge containing  ŊƐ-m7o|_;u-7_;vbomloѴ;1†Ѵ;vv†1_-v ƓƓ-‰-‹=uolvb|;v of adhesion.107 In addition, uropod forces are required to push the

nucleus through narrow pores in vitro and in vivo. Thus, inhibition of uropod contractility using the Myosin II inhibitor blebbistatin or |_;!bm_b0b|ou+ƑƕѵƒƑѴ;-7v|o-r_;mo|‹r;‰_;u;|_;Ѵ;-7bm] edge still migrates toward a chemotactic source, whereas the nu-cleus becomes physically stuck.108 Uropod contractility also leads

to an anterograde flow of cytoplasm. Depending on the anchorage of the leading edge cortical actomyosin cytoskeleton to the plasma membrane, the resulting hydrostatic pressure may lead the detach-ment of the plasma membrane from the underlying cytoskeleton.109

This leads to the formation of membrane blebs, which are rapidly re-=bѴѴ;7‰b|_ Ŋ-1|bmĺ110 Bleb formation has been reported for a variety of cell types in vivo, such as during zebrafish gastrulation. There, bleb formation contributes to successful steering of cell migration.111,112

Blebs do not appear to play a role during trafficking of naive T cells bmvb7;Ѵ‹lr_ob7|bvv†;ķ‰_;u; Ŋ-1|bmŊ=bѴѴ;7rv;†7oro7ruo|u†vbomv are predominant.113 Bleb formation may in turn constitute one of the

migration strategies acquired by CD8+ T cells in NLT. Yet, it remains

(8)

unclear whether CD8+ T cells in NLT are able to develop sufficient

contractility required for bleb formation. This is experimentally chal-lenging to address, since blebs constitute transient structures diffi-cult to resolve by intravital imaging.

$_; 0u-m1_;7 Ŋ-1|bm m;|‰ouh bv 1omv|-m|Ѵ‹ u;lo7;Ѵ;7 0‹ =-1|ouv v†1_ -v ]Ѵb-Ѵ l-|†u-|bom =-1|ou Ő őĺ   7bv-vv;l0Ѵ;v Arp2/3- containing branched junctions without severing actin fil-aments, which may render these filaments amenable to Myosin II- l;7b-|;71om|u-1|bom-||_;†uoro7ĺ11ou7bm]Ѵ‹ķ bvo=oulv-u; highly expressed in CD8+ T cells throughout all stages of immune

u;vromv;vՉ‰‰ĺbll];mĺou]őĺ bm-ѴѴ‹ķ‹ovbm-1|bˆb|‹1om|ub0†|;v

to disassembly of actin filaments for cell- scale actin treadmilling.114

uol|_;v;ruor;u|b;vķ†m;Šr;1|;7=;-|†u;v1-m;l;u];v†1_-v|_; universal coupling of cell speed and the straightness of movement: the faster cells migrate, the higher is their directionality persistence. High migratory speed stabilizes cell directionality by transporting intracellular polarity cues, such as Myosin II, to the trailing edge |_uo†]_=-v|u;-u‰-u7 Ŋ-1|bm=Ѵo‰ĺ115

m bm|ubmvb1 h;‹ =;-|†u; o= Ŋ-1|bm roѴ‹l;ubŒ-|bom ‰-v u;1;m|Ѵ‹ †m1oˆ;u;7 0‹ _b]_Ŋu;voѴ†|bom bl-]bm] o= 0u-m1_;7 Ŋ-1|bm -m]Ѵ; distribution at the leading edge.ƐƐѵ With increasing membrane

load, which may correlate with encountering a dense barrier during

  & !  Ƒ Պ Role of chemokines in barrier crossing and perception of resistance by T cells in non- lymphoid tissue. A, As an example, CD8+ T migration in skin is pictured. In order for dermal CD8+ T

to accumulate in the epidermis, they have to cross the basement membrane

Őőv;r-u-|bm]7;ulbv=uol;rb7;ulbvĺm|_;-0v;m1;o=-||u-1|bom0‹1_;lohbm;vu;Ѵ;-v;7=uolh;u-|bmo1‹|;vķ|_;Ѵ;-7bm]Ŋ;7]; Ŋ-1|bm dynamics are wired to form leading edge protrusions away from dense barriers such as the ECM of a BM. This is in part because fast- growing Ŋ-1|bm=bѴ-l;m|v-11†l†Ѵ-|;ru;=;u;m|b-ѴѴ‹‰_;u;|_;rѴ-vl-l;l0u-m;;Šr;ub;m1;vѴ;vvu;vbv|-m1;ĺ)_;mv|bl†Ѵ-|;70‹1_;lohbm;vķ chemokine receptors on CD8+ T

Ѵ;-7|oѴo1-Ѵrbh;vo=ruo|u†vbˆ; Ŋ-1|bm-1|bˆb|‹|obmv;u|rv;†7oro7v|_uo†]_r;ulbvvbˆ;]-rvo=|_;

BM ECM mesh. The nucleus likely acts as a ruler to avoid nuclear rupture, which would occur when cells try to pass through non- permissive ]-rvĺķuorov;7r;u1;r|bomo=u;vbv|-m1;|o];m;u-|; Ŋ-1|bmruo|u†vbomv-vbm7b1-|;70‹-uuo‰v=uol-lb]u-|ou‹$Ŋ1;ѴѴr;uvr;1|bˆ;

(9)

lb]u-|bomķ Ѵ;-7bm] ;7]; Ŋ-1|bm 0u-m1_bm] -m]Ѵ;v 0;1-l; 0uo-7;u with shorter fragments. In contrast, where membrane load was decreased, expanding actin filaments growing perpendicular to the plasma membrane outran filaments growing at steeper angles. When reaching the plasma membrane, fast perpendicular filaments become protected from capping proteins by factors such as VASP/ ENA and formins.ƐƐѵ This leads to faster membrane protrusion

speeds and, by force coupling, efficient cell body translocation. This model provides an explanation for the remarkable mechanosensi-tivity displayed by leukocytes during their migration through 3D collagen matrices, where these cells choose the path of least resis-tance117

Ő b]†u;Ƒőĺ

Pseudopod protrusion speeds alone do not determine the path of tissue- infiltrating T cells. Rather, the nucleus as the biggest or-ganelle is likely to play a central role for T- cell decision taking on which path to follow in geometrically complex environments. Thus, it has emerged in recent years that nuclear deformability deter-mines the physical limitations of cell migration. The stiffness of the nuclear envelope is determined by the expression of the structural proteins lamin A and lamin C.118 In leukocytes, their expression is

low, indicating a relatively soft nuclear envelope in line with their motile lifestyle.119 When cancer cells are forced through narrow

pores in microchannel systems, their nuclear membrane may rupture, leading to the recruitment of the ESCRT repair mecha-nism.120,121 It is conceivable that similar mechanisms maintain T-

cell nuclear integrity intact during prolonged tissue surveillance, in particular in tight epithelial environments. Maintaining nuclear integrity also imposes a physical limit to T- cell migration in the ab-v;m1;o=ruo|;oѴ‹|b17;]u-7-|bomo=|_; ĺ ou$Ŋ1;ѴѴ0Ѵ-v|vķ|_; minimum pore size is around 4 μm2, which correlates to a circle

with a diameter of approximately 2.5 μm.122 Taken together,

lead-bm] ;7]; Ŋ-1|bm 7‹m-lb1v -m7 m†1Ѵ;-u u;v|ub1|bom -u; blrou|-m| parameters for path finding in complex environment. How is cell

0o7‹ 7bvrѴ-1;l;m| -11olrѴbv_;7 0‹ 7‹m-lb1 Ŋ-1|bm u;lo7;Ѵbm] and contraction?

ƑĺƐՊ|Պ Ŋ-1|bmm;|‰ouh7‹m-lb1v-m7$Ŋ1;ѴѴlo|bѴb|‹

Mesenchymal cell migration (eg, fibroblast migration) is achieved by Ŋ-1|bmruo|u†vbomv-||_;Ѵ-l;ѴѴbro7b†l|or†v_|_;l;l0u-m;=ou-ward, followed by new integrin- mediated focal adhesions and trailing edge contraction.123,124 In this migration mode, elevated substrate

adhesiveness actually stalls migration speeds owing to the necessity to detach focal adhesions. Similarly, T- cell blasts placed on a 2D sur-face (which has been the experimental standard over the last dec--7;vőu;t†bu;bm|;]ubmvķ|‹rb1-ѴѴ‹ ŊƐķ-m7-mbllo0bѴbŒ;7bm|;]ubm ligand, such as ICAM- 1, on the surface for cell displacement.125 This

is because integrins are required for cell attachment on the surface in 2D systems, as well as for outside- in signaling to the actomyosin 1‹|ovh;Ѵ;|om-m7=ou=ou1;|u-mvlbvvbom=uolu;|uo]u-7; Ŋ-1|bm=Ѵo‰ Ő b]†u;Ɛőĺ"†1_-lo|bѴb|‹lo7;l-‹r-u|b-ѴѴ‹u;=Ѵ;1|r_‹vboѴo]b1-Ѵ$Ŋ cell migration on endothelial surfaces under flow, although in the lat-ter case shear forces contribute to integrin activation.20

In contrast, ameboid motility in 3D environments does not in-volve strong substrate anchoring. In general, there are four modes of focal adhesion- free migration: (a) swimming migration in blebbing cells, (b) cell substrate intercalation by insertion of protrusions into preformed or deformable gaps of the environment, (c) chimneying by exerting force to ECM or adjacent cells perpendicular to the di-rection of migration in order to keep the cell body in place, and (d), flow- friction- driven force transmission by retrograde cortical actin flow.ƐƑѵ Recent work has shown that the latter mode is used by

naive T cells for migration under in vitro and in vivo confinement. When these T cells are placed under confinement, for example, by squeezing these cells between a coated plate and an agarose layer, on ICAM- 1 coated plates in the presence of CCL19 or CCL21, these

  & !  ƒ Պ Models of extrinsic vs intrinsic regulation of cell motility. A, Naive CD8+ T cells (TN) under confinement are immotile unless -1_;lohbm;bvruoˆb7;7|obm7†1;roѴ-ubŒ-|bom-m7u;-u‰-u7 Ŋ-1|bm=Ѵo‰ĺm|;]ubmѴb]-m7vruolo|;1;ѴѴ0o7‹|u-mvѴo1-|bom|_uo†]_‰;-h interactions, without inducing adhesion. B, CD8+ T

ķ and TMEM1-mbmrubm1brѴ;|†m;|_;bubm|ubmvb1 Ŋ-1|bm|u;-7lbѴѴbm]-m7ņou1om|u-1|bѴb|‹

independent of extrinsic cues for spontaneous motility. The precise contribution of intrinsic regulation during CD8+- mediated immune

surveillance remains unknown

(10)

cells show remarkable high migration speeds reminiscent of values o0v;uˆ;7bmѴ‹lr_ob7|bvv†;Ő b]†u;vƐ-m7ƒőĺ&m7;u|_;v;1om7b-|bomvķ ŊƐ0;1ol;v7bvr;mv-0Ѵ;=ou1;ѴѴ-||-1_l;m|vbm1;1;ѴѴv-u; physically forced into contact with the environment. High- resolution bl-]bm] o= Ŋ-1|bm 7‹m-lb1v bm 1om=bm;7 $ 1;ѴѴv ;Šrov;7 |o !ƕ Ѵb]-m7vv_o‰1omv|-m| Ŋ-1|bmroѴ‹l;ubŒ-|bom-||_;buѴ;-7bm];7];ĺ -m‹m;‰Ѵ‹=oul;7 Ŋ-1|bm=bѴ-l;m|v-u;u-rb7Ѵ‹r†v_;7|o‰-u7|_; back of the cell, presumably owing to deformation resistance of the Ѵ;-7bm] ;7]; rѴ-vl- l;l0u-m;ĺ $_bv u;v†Ѵ|v bm =-v| Ŋ-1|bm u;|uo-grade flow tuned by the strength of chemokine receptor signaling. $_; vr;;7 o= |_; u;-u‰-u7 Ŋ-1|bm =Ѵo‰ 7;|;ulbm;v |_; lour_oѴ-ogy of the cell, with high flow speeds corresponding to elongated shapes.113 m |_; u;7†1|bombv| †m7;u -]-uov; v;||bm]ķ  ŊƐ

|u-mv-mits weak frictional interactions to the substrate but does not it-v;Ѵ= 1om|ub0†|; |o o†|vb7;Ŋbm Ŋ-1|bm roѴ‹l;ubŒ-|bomĺ $_†vķ  ŊƐ 0;1ol;v;m]-];70‹u;|uo]u-7; Ŋ-1|bm=Ѵo‰|o‰;-hѴ‹bm|;u-1|‰b|_ its ligand ICAM- 1.127$_uo†]_|_bv1Ѵ†|1_=†m1|bomķu;-u‰-u7 Ŋ-1|bm

flow slows down and becomes transformed into forward movement, which is tuned by ICAM- 1 density. In contrast to the mesenchymal mode of motility, high substrate adhesiveness does not induce lym-phocyte arrest because no focal adhesions are formed.

In vivo morphometric analysis of CCR7ƴņƴķ ŊƐƴņƴ and double-

7;=b1b;m| $ 1;ѴѴv lb]u-|bm] om |_; ! m;|‰ouh o= Ѵ‹lr_ob7 |bvv†; have confirmed key aspects of the proposed model, such as cell ;Ѵom]-|bom bm |_; -0v;m1; o=  ŊƐ -m7 1;ѴѴ v_ou|;mbm] bm |_; -0-sence of CCR7. In addition, while ab-0-sence of either molecule re-sulted in a speed decrease, double- deficient T cells showed a more pronounced loss of cell body translocation. These data clearly demonstrate a complementary input by both modules for cell trans-location in interstitium, in contrast to their sequential engagement during extravasation.113 In sum, the mode of motility in naive T cells

within lymphoid tissue is consistent with a cortical actin flow model proposed by Bray and White128 and permits a continuous sliding

mode of T cells while limiting their adhesive interactions with the surrounding environment. One of the implications of this model is |_-| v;Ѵ;1|bˆ;  ŊƐ ;m]-];l;m| l-‹ 0; u;v|ub1|;7 |o ;m1o†m|;uv with cognate pMHC- presenting DCs, as has been suggested by in vitro studies of the IS and the reduced interaction time of CD8+ T

cells with ICAM- 1- deficient DCs in vivo.129,130 Yet, a recent study has

1_-ѴѴ;m];7|_;1om1;r||_-| ŊƐŊŊƐbm|;u-1|bomv-u;u;t†bu;7 for T- cell arrest on pMHC- presenting DCs during the first hours of their encounters.131

$_†vķ|_;ru;1bv;uoѴ;o= ŊƐ7†ubm]$Ŋ1;ѴѴ-1-tivation in vivo and IS formation is still not well understood.

ƑĺƑՊ|Պ!;]†Ѵ-|ouvo=|_;-1|ol‹ovbm1‹|ovh;Ѵ;|ombm

 Ѷ

+

$1;ѴѴv

Biochemical information triggered by chemokines and other me-diators is transformed by leukocytes into mechanical work through the activity of the actomyosin cytoskeleton. Which factors in turn orchestrate the actomyosin network assembly and disassembly? As bm -ѴѴ ;†h-u‹o|b1 1;ѴѴvķ l;l0;uv o= |_; vl-ѴѴ $-v; =-lbѴ‹ķ !_oķ Rac and Cdc42, play central roles in T cells. These proteins cycle

0;|‰;;m Ŋ0o†m7bm-1|bˆ;v|-|;v-m7$ŊѴo-7;7-1|bˆ;v|-|;vķ in which these molecules interact with downstream effector mole-cules. Rac1 and Rac2, both of which are expressed in leukocytes, ac-tivate Arp2/3 complexes via SCAR/WAVE complexes. Lack of either Rac1 or Rac2 only partially reduces speeds of naive T cells within lymphoid tissue, suggesting largely overlapping functions of these bvo=oulv =ou or|bl-Ѵ 1_;lohbm;Ŋ|ub]];u;7 Ŋ-1|bm ];m;u-|bomĺ132 Accordingly, in lymphoid tissue of plt/plt mice lacking the promigra-tory factors CCL19 and CCL21, T- cell speeds decrease to a similar level between WT T cells and T cells lacking either Rac1 or Rac2. This suggests that under suboptimal migration conditions, one Rac isoform suffices to transmit signals for cell motility. In contrast, lack of both Rac1 and Rac2 precipitates a strong loss in T- cell motility.132

As outlined above, Cdc42 uses WASP complexes for Arp2/3 ac-|bˆ-|bomĺ ;|-bѴ;7 ! $Ŋ0-v;7u;rou|;u-1|bˆb|‹l;-v†u;l;m|v_-ˆ; uncovered that while both Rac and Cdc42 activity localizes to the leading edge of migrating leukocytes, local Cdc42 signals precede Rac activation before cell turning.133

 bm-ѴѴ‹ķbm1om|u-v||o=b0uo0Ѵ-v|v and other mesenchymal cells, Rho activity is restricted to the trail-bm];7];-m7l;7b-|;v†uoro71om|u-1|bѴb|‹ˆb-!Ŋl;7b-|;7 r_ovr_ou‹Ѵ-|bomŐ b]†u;Ɛőĺmu;v|bm]1;ѴѴvķv†1_-vbmѴ‹lr_o1‹|;v =u;v_Ѵ‹bvoѴ-|;7=uol0Ѵoo7ķѴ;ˆ;Ѵvo=-1|bˆ;ķ$ŊѴo-7;7!-1ƐņƑ-m7 71ƓƑ-u;Ѵo‰ĺ!_oŊ!Ŋ‹ovbm-1|bˆb|‹bv-ѴvoѴo‰0†|Ѵbh;Ѵ‹ to be present, indicated by the spherical shape of cells under these conditions (most cells adapt a spherical shape in suspension owing to their intrinsic cortical actomyosin cytoskeleton contractility).

ƑĺƒՊ|Պ v-m7v-u;1ub|b1-Ѵu;]†Ѵ-|ouvo=

-1|ol‹ovbm7‹m-lb1v

After stimulation with chemoattractants, CD8+ T cells acquire a

po-larized shape with a leading edge and a trailing edge, and separated -1|bˆb|b;vo=vl-ѴѴ$-v;vĺ"l-ѴѴ$-v;v|_;lv;Ѵˆ;v-u;-1|bˆ-|;7 0‹;mŒ‹l;vķ‰_b1__;Ѵr|oѴo-7$bmrѴ-1;o= bm|o|_;0bm7-bm] ro1h;| |o -1|bˆ-|; vl-ѴѴ $-v;vķ vo 1-ѴѴ;7 ]†-mbm; ;Š1_-m]; =-1|ouvŐ vőĺ$_;-rruoŠbl-|;Ѵ‹ѶƏ vb7;m|b=b;7bmlo†v;-m7 human genome do so by stabilizing the nucleotide- free form of small $-v;vķ ‰_b1_ |_;m -ѴѴo‰v |_; u;rѴ-1;l;m| o=   0‹ |_; lou; -0†m7-m|$ĺ$_;u;-u;v;ˆ;u-Ѵ=-lbѴb;vo= v‰b|_7bv|bm1|-1-tive domains.134

$_;Ѵ-u];v|v†0]uo†ro= v-u;|_; 0Ѵ_oloѴo]‹ (DH)- containing proteins including Vav, which is well characterized for its role during TCR signaling. Yet, the involvement of Vav proteins in primary CD8+ T- cell motility is minor.135 Another member, Tiam1,

participates in CXCL12- induced T- cell migration through Rac activa-tion but its role in CCR7- mediated migraactiva-tion and in vivo trafficking has not been investigated in depth.Ɛƒѵ

A second family is constituted by dedicator of cytokinesis Ő őruo|;bmvĺ =-lbѴ‹l;l0;uv;Šru;vv|_; _oloѴ-o]‹u;]bomŐ !őƑ-v1-|-Ѵ‹|b1-ѴѴ‹-1|bˆ;7ol-bmĺ)_bѴ; Ɛbv ;Šru;vv;7bmlov|momŊ_;l-|orob;|b11;ѴѴ|‹r;vķ Ƒbv-_;l--|orob;|b1-ѴѴ‹;Šru;vv;7!-1Ɛņ!-1Ƒ ĺ;m;|b17;Ѵ;|bomo= Ƒ u;v†Ѵ|vbmv|uom]Ѵ‹7blbmbv_;7 Ŋ-1|bmroѴ‹l;ubŒ-|bom-=|;u1_;loh-ine stimulation and leads to a phenocopy of Rac1 and Rac2- deficient

(11)

T cells, that is, virtually abolished mobility of T cells in lymphoid tissue.132,137,138

 Ƒ bv |_;u;=ou; |_; l-bm !-1Ɛņ!-1Ƒ   bl-rѴb1-|;7bm1_;lohbm;Ŋbm7†1;7urƑņƒ-1|bˆ-|bombm$1;ѴѴvĺ Ƒ also leads to Rac activation downstream the S1P1, which delays ;]u;vvo= ƑŊ7;=b1b;m|$1;ѴѴv=uolѴ‹lr_ob7|bvv†;ĺ138

m|u-ˆb|-Ѵbl-]bm]o= ƑŊ7;=b1b;m|$1;ѴѴvbmѴ‹lr_ob7|bvv†; _-v b7;m|b=b;7 |‰o uoѴ;v -11olrѴbv_;7 0‹ ƑŊ!-1Ŋ-Šbv 7†ubm] bll†m;u;vromv;vĺ buv|ķ-vru;7b1|;70‹lo7;Ѵbm]ķѴ-1ho=lo|bѴb|‹ v|uom]Ѵ‹u;7†1;v|_;Ѵbh;Ѵb_oo7o= Ƒƴņƴ T cell encounters with

u-u; v7bv|ub0†|;7|_uo†]_o†|Ѵ‹lr_ob7|bvv†;ĺ";1om7ķ ƑŊ driven motility is required to allow T cells to find and cluster around rare “optimal” DC, that is, DCs with high levels of cognate pMHC. Ƒ7o;vvo0‹ruolo|bm]$Ŋ1;ѴѴ7;|-1_l;m|=uol v‰b|_Ѵo‰ levels of cognate pMHC.49 This supports the notion that high

inter-stitial motility of T cells has not only evolved to search for antigen but serves also to limit T- cell activation: only signals, which induce a strong stop signal, or which are repeatedly encountered over the course of several hours, are licensed to give rise to T- cell activation. Thus, chemokines regulate T- cell responses not only by guiding to APC but additionally by maintaining high basal motility as quality control for the strength of the activatory signal.

Ƒ -Ѵvo u;]†Ѵ-|;v Ŋ-1|bm 7‹m-lb1v -| |_; " o= -1|bˆ-|;7 T cells.139 $_;u;ķ |_; Ѵbrb7Ŋ0bm7bm] !Ɛ 7ol-bm o= Ƒ 0bm7v

phosphoinositide- 3,4,5- phosphate (PIP3) of the peripheral supramo-Ѵ;1†Ѵ-u-1|bˆ-|bom1Ѵ†v|;u|o7ubˆ;1;m|ubr;|-Ѵ Ŋ-1|bmroѴ‹l;ubŒ-|bomĺ This leads to the transportation of TCR to the central supramolecular activation cluster and exerts forces that may help the TCR differen-tiate between low- and high- affinity ligands.ѵķƐƓƏ As a consequence,

urƑņƒŊ7;r;m7;m| Ŋ-1|bm roѴ‹l;ubŒ-|bom -| |_; " bv v|uom]Ѵ‹ u;-7†1;7bm ƑŊ7;=b1b;m|$1;ѴѴvĺ$_bv1ouu;Ѵ-|;v‰b|_7;=;1|bˆ;$Ŋ1;ѴѴ activation, despite the fact that most downstream signaling events except Rac activation remain intact.141

 bm-ѴѴ‹ķ Ƒ=†m1|bomvbm T- cell activation beyond facilitating DC encounter and direct TCR sig-naling. The homeostatic chemokine CCL21 acts as a costimulatory molecule during naive T- cell activation, which is in most part medi--|;70‹!-1Ŋ7ubˆ;m !-1|bˆ-|bomĺ142$_;blrou|-m1;o= Ƒ=ou the functioning of the immune system is reflected in the early- onset v;ˆ;u;bll†mo7;=b1b;m1‹bm_†l-mvѴ-1hbm] Ƒ;Šru;vvbomĺ143

Ѷbv-71ƓƑ -m7-ѴѴo‰v v|om-ˆb]-|;];ol;|ub1-ѴѴ‹ complex environments.144 CD8+

$1;ѴѴvѴ-1hbm] Ѷv_o‰1ol-parable expansion during an adaptive immune response, but their memory populations contract rapidly.ƐƓƔķƐƓѵ Human CD8+ T cells with

7;1u;-v;7ou-0v;m| Ѷ;Šru;vvbomo‰bm]|o];m;|b1l†|-|bomv undergo cell death in collagen matrix owing to disintegration of cell shape, a process termed cytothripsis.147 In line with disturbed T- cell

|u-==b1hbm]ķbm_b0b|bomo= ѶŊ71ƓƑŊl;7b-|;7lb]u-|bom-l;Ѵbo-rated the disease course in a mouse model of multiple sclerosis.148

u_ Ɛ Ő-Ѵvo hmo‰m -v rƐƐƔ  ou Ѵv1ő bv - !_o   _b]_Ѵ‹ ;Šru;vv;7 bm m-bˆ; -m7 -1|bˆ-|;7 $ 1;ѴѴvĺ m b|v -0v;m1;ķ !_oŊ$ loading and migration are severely impaired in response to chemoat-tractants.149mroѴ-ubŒ;7$1;ѴѴѴbm;vķu_ Ɛbvv;t†;v|;u;70‹|_;

microtubule network at the trailing edge. Microtubule disassembly results in the release of this factor, leading to activation of Rho. This

bm|†um|ub]];uv!Ŋ7;r;m7;m|r_ovr_ou‹Ѵ-|bom-m7†uoro7 contractility94

Ő b]†u;Ɛőĺm-77b|bom|o|_;;Š-lrѴ;v]bˆ;m-0oˆ;ķ$ 1;ѴѴv;Šru;vvlou;|_-mƑƏ vķ-m7|_;bu;Šru;vvbomr-||;umvo=|;m change during activation (www.immgen.org). Yet, the roles of most l;l0;uvo=|_; -m7 0Ѵ=-lbѴb;v=ou|_;bmˆbˆo=†m1|bomo= CD8+ T cells remain largely unknown to date.

1|bˆ-|;7 vl-ѴѴ $-v;v 1Ѵ;-ˆ; $ |o   |o u;|†um |o |_; u;v|bm] v|-|;ķ 0†| 7o vo ‰b|_ u;Ѵ-|bˆ;Ѵ‹ vѴo‰ hbm;|b1vĺ $-v;Ŋ -1|bˆ-|bm] ruo|;bmv Ővő 0bm7 |o |_; -1|bˆ;ķ $Ŋ0o†m7 =oulv o= !_oķ !-1ķ -m7 71ƓƑĺ  0bm7bm] 1-|-Ѵ‹Œ;v |_; bm|ubmvb1 $-v; -1|bˆb|‹|o1omˆ;u|vl-ѴѴ$-v;v|o|_; Ŋ0o†m7ķbm-1|bˆ;=oulĺ "blbѴ-u|o vķ|_;u;-u;lou;|_-mѵƏv=o†m7bmlo†v;-m7 _†l-m ];mol;ķ |_†v o†|m†l0;ubm] |_; m†l0;u o= vl-ѴѴ $-v;vĺ $_; 1-|-Ѵ‹|b1 7ol-bm bv 1omv;uˆ;7 0;|‰;;m vķ robm|bm] |o - 1ollom;ˆoѴ†|bom-u‹oub]bmĺ"blbѴ-u|o vķ Ѷ+ T cells express lou;|_-mƑƏb7;m|b=b;7l;l0;uvՉ‰‰ĺbll];mĺou]őķ-m7|_; physiological function of most of these factors is unknown to date. Of note, in an unbiased genetic screen to identify factors facilitating intratumoral CD8+ T- cell accumulation in a mouse model of

mela-mol-ķ|_;-†|_ouvb7;m|b=b;7|_;!_ou_Ɣ-v-mbm_b0b|ou of CD8+ T

infiltration.

150 Yet, the mechanism of action has thus far

not been described, and may conceivably involve changes in T- cell activation as well as altered ability to respond to chemoattractants.

o1-Ѵ v†rru;vvbom o= vl-ѴѴ $-v; -1|bˆb|‹ 0‹ v v;uˆ;v - similar function as local activation, namely to compartmentalize pools of active Rac/Cdc42 and Rho to the leading and trailing edge of polarized cells, respectively. As an example, Myosin IXB (Myo9b) bv-m Ŋ-1|bmŊ0bm7bm]1‹|ovh;Ѵ;|-Ѵlo|ouruo|;bm‰b|_-!_o-1-tivity as “cargo,”151 which accumulates at the leading edge of

polar-ized macrophages and DCs.152 In these cell types, Myo9b deficiency

Ѵ;-7v|obm1u;-v;7v|;-7‹Ŋv|-|;!_oŊ$Ѵ;ˆ;Ѵv-m7-1om|u-1|;71;ѴѴ phenotype that resulted in impaired migration in vitro and in vivo. Absence of Myo9b in T cells also resulted in increased steady- state !_oŊ$ Ѵ;ˆ;Ѵvķ u;7†1;7 bm ˆb|uo lb]u-|bom |o‰-u7 _ol;ov|-|b1 chemokines and lower LN homing in vivo.91 Despite these defects,

Myo9bƴņƴ CD8+ T cells showed similar clonal expansion and

effec-tor differentiation in spleen and LN as their WT counterparts during DC- or virus- triggered inflammation. In contrast, Myo9bƴņƴ effector

CD8+ T cells failed to efficiently seed NLTs and to protect hosts

from skin infection. This phenotype correlated with a strongly re-duced ability to cross dense ECM barriers in vitro.91 Taken together,

Myo9b- dependent repression of Rho activity at the leading edge has evolved to enable effector T cells to negotiate tissue barriers, in particular those formed perpendicular to their migration path. Specifically, T- cell seeding of non- lymphoid epithelial tissues that are separated by a dense BM from underlying connective tissue is critically dependent on carefully balanced Rho activity and for the establishment of protective CD8+ T

RM populations.91 These findings

also highlight the impact of tissue architecture and properties on T- 1;ѴѴbm=bѴ|u-|bom-m7v†uˆ;bѴѴ-m1;Ő b]†u;Ƒőĺ

$_; !_o u;]†Ѵ-|ou -lѵƔ0 bv _b]_Ѵ‹ ;Šru;vv;7 bm m-bˆ; $ 1;ѴѴvķ ‰_;u;b|1om|ub0†|;v|o|_;l-bm|;m-m1;o=Ѵo‰v|;-7‹Ŋv|-|;!_oŊ$ Ѵ;ˆ;Ѵv -hbm |o ‹oƖ0ĺ +;|ķ bm 1om|u-v| |o ‹oƖ0ķ -lѵƔ0 7o;v mo|

(12)

specifically accumulate at the leading edge of polarized T cells but is uniformly distributed along the plasma membrane.153

$_;u;ķ -lѵƔ0 v;t†;v|;uv !_o -m7 ru;ˆ;m|v b|v $ŊѴo-7bm]ĺ &rom 1_;lohbm;Ŋ |ub]];u;7 ";uņ$_u r_ovr_ou‹Ѵ-|bomķ -lѵƔ0 |u-mvb;m|Ѵ‹ 7;|-1_;v =uol|_;rѴ-vl-l;l0u-m;|or;ulb|!_oŊ$=oul-|bomĺ153 A sa-lient feature of CD8+ T

RM cells is their decrease in expression levels of

-lѵƔ0ĺ154$_bvv†]];v|v|_-|Ѵo‰ -lѵƔ0Ѵ;ˆ;Ѵvbm$RM may lead to r;ul-m;m|Ѵ‹bm1u;-v;7!_oŊ$Ѵ;ˆ;Ѵvķ-m7|_†vbm1u;-v;71om|u-1-tility of their cortical actomyosin cytoskeleton. This may reflect an adaptation to more dense microenvironments of NLT as compared to the loose fibroblastic scaffold found in SLO. Despite these few exam-rѴ;vķ|_;uoѴ;=oulov|o=|_;v7†ubm]]Ѵo0-Ѵrovb|bombm]|_uo†]_ integrin regulation and migration control remain unaddressed to date.

ƑĺƓՊ|Պ;‹om7 -m7vĹ-77b|bom-Ѵu;]†Ѵ-|ouvo=

$Ŋ1;ѴѴlo|bѴb|‹

Ѵom] ‰b|_ u;1;m| 7bv1oˆ;ub;v om   -m7  =†m1|bom =ou $Ŋ cell immune surveillance, a number of unexpected regulators of chemokine- triggered T- cell motility have emerged in recent years. While it is beyond the scope of this review to provide a compre-hensive overview, we present a few notable examples. Studies on |_;-1|bmŊ0bm7bm]ruo|;bml‹ovbmƐŐ‹oƐ]ő_-ˆ;†m1oˆ;u;7-Ѵbmh between T- cell motility patterns and activation phenotype. Myo1g accumulates at sites adjacent to sites where the T- cell plasma mem-brane encounters obstacles, and contributes to the formation of a m;‰ Ŋ-1|bmruo|u†vbomm;Š||ov†1_o0v|-1Ѵ;vĺ155 This confers to T cells the ability to meander and efficiently scan their surrounding en-vironment. In turn, Myo1g- deficient T cells display a more directional motility inside lymphoid tissue as compared to WT T cells. 2PM im-aging showed that the decreased meandering ability shortened the average duration of individual T cell- DC contacts. This decrease in contact time was compensated when DCs were abundant within lymphoid tissue, since CD8+ T cells remained able to integrate

suffi-cient information for full activation. In contrast, under conditions of low DC frequency, this aberrant behavior resulted in impaired T- cell activation.155 This observation provides further evidence that the

migratory behavior of T cells is finely tuned to balance efficient tis-sue scanning with sufficient signal integration for activation.

$_;";uņ$_uhbm-v;)Ɛbv0;v|hmo‰m=oub|vuoѴ;bmu;]†Ѵ-|bm] v-Ѵ|_ol;ov|-vbvbmhb7m;‹ĺ&m;Šr;1|;7Ѵ‹ķbm$1;ѴѴvķ)Ɛ-1|v-v- m;]-|bˆ;u;]†Ѵ-|ouo= ŊƐ-1|bˆb|‹‰_bѴ;rovb|bˆ;Ѵ‹u;]†Ѵ-|bm]1_;-lo|-Šbvˆb-|_;hbm-v;v*"!Ɛ-m7"$ƒƖ-m7|_;bom|u-mvrou|;u "ƐƑƑĺ$_†vķѴ-1ho=)Ɛbmrubl-u‹$1;ѴѴvu;v†Ѵ|;7bm-m_‹-peradhesive, hypomotile phenotype with a strongly decreased abil-ity to inspect large tissue volumes.ƐƔѵ

!;]†Ѵ-|ouv o= |_; ľѴ-u];Ŀ $-v;vķ |_-| bvķ |_; αi subunits of !vķ bm r-u|b1†Ѵ-u !"Ɛķ u;]†Ѵ-|; $Ŋ1;ѴѴ lb]u-|bom bm ˆbˆo.157 $_;v; ruo|;bmv 7o vo 0‹ -11;Ѵ;u-|bm] |_; bm|ubmvb1 $-v; -1|bˆb|‹ o=$Ŋ0o†m7αi to fine- tune responsiveness to chemotactic sig-m-Ѵvĺm1om|u-v||o -lѵƔ0ķ!"Ɛbvvr;1b=b1-ѴѴ‹†ru;]†Ѵ-|;7bm$RM. )_bѴ; b|v bm ˆbˆo =†m1|bom u;l-bmv bm1olrѴ;|;Ѵ‹ †m7;uv|oo7ķ !"Ɛ may help to suppress signaling by tissue egress- promoting factors,

bmr-u|b1†Ѵ-u"Ɛoubm1-v;o=;rb7;ulbvķƑƐĺ$_†vķѴ-1ho=!"Ɛ may contribute to facilitate long- term tissue residency in these cell populations by regulating responsiveness to chemoattractants, in addition to decreased S1P1 and CCR7 levels.

ƑĺƔՊ|Պ Ѷ

+

$Ŋ1;ѴѴl;l0u-m;ruor;u|b;v-m7

;Š;u|bomo==ou1;v

In recent years, it has become increasingly clear that regulation of |_; Ŋ-1|bmm;|‰ouh-==;1|v|_;0bor_‹vb1-Ѵruor;u|b;vo=$Ŋ1;ѴѴl;l-branes. One important readout is the plasma membrane tension, which is related to the force needed to deform a membrane. Owing to the task of CD8+ T cells to interact with DCs and scan the surfaces

of other cells, control of membrane tension is critical. Membrane tension is influenced by three components: first, “in- plane” tension describes the tension between lipids of the membrane bilayer and is influenced in part by osmotic pressure. It is generally assumed that local rises of in- plane tension become distributed within mil-liseconds throughout the plasma membrane.158,159 However, this

has been put into question by a recent study.ƐѵƏ Second, membrane-

to- cortex attachment (MCA), or membrane- cytoskeleton adhesion, describes the anchorage of the cortical actomyosin cytoskeleton through adaptors for transmembrane proteins such as phosphoryl-ated Ezrin/radixin/moesin (pERM) or protein binding lipids such as PIP2 of the inner leaflet of the plasma membrane. MCA, therefore, increases membrane tension. During T- cell activation via TCR signal-ing, pERM levels rapidly decrease, allowing for a relaxation of the MCA in T cells. This in turn facilitates attachment to DCs and may r-u|b-ѴѴ‹1olr;mv-|;=ou ŊƐŊŊƐ-7_;vbom7†ubm];-uѴ‹bm|;u-actions.ƐѵƐ In support of this, regulation of stiffness is one of the

factors that controls T cell engagement with APC.ƐѵƑ Third, cortical

tension describes the tension of the actin cortex below the plasma membrane, which depends on myosin activity, the length of actin filaments and their nanoarchitecture.Ɛѵƒ Together, these factors

regulate the deformability of cells and the force needed to induce cell shape change. Compared to most stromal cells, leukocytes are soft and deformable cells. The Young's modulus quantifies the re-sistance of an object to being deformed when a force is applied to it and ranges from more than 10 kPa in fibroblasts to approximately 0.1 kPa in CD8+ T cells.ƐѵƓ Regulation of cell deformability is tightly

regulated and is likely a key feature for CD8+ T- cell infiltration and

scanning of diverse tissues. Nonetheless, there is to date scarce in-=oul-|bom |o 7-|; _o‰ |_; u;]†Ѵ-|bom o= |_; Ŋ-1|bm 1‹|ovh;Ѵ;|om influences the biophysical properties and surveillance ability of dis-tinct T- cell subsets in lymphoid and non- lymphoid tissues.

ƒՊ|Պ !$ƒ ĺ * $!   "     $!   "  

!  &  $    Ѷ

+

$҃  

 " $       $

A major challenge is to translate findings from reductionist in vitro systems to physiological settings in vivo. What are the

Figure

Updating...

References

Related subjects :