Eur. J. Immunol. 1991. 21: 2565-2573
Karen E. Chaffin and Roger M. Perlmutter Howard Hughes Medical Institute and the Departments of Biochemistry, Immunology, and Medicine (Medical Genetics), University of Washington, Seattle
Pertussis toxin blocks thymocyte emigration
2565
A pertussis toxin-sensitive process controls thymocyte emigration Although it is well known that essentially all peripheral Tcells are derived from bone marrow progenitors that mature in the thymus, the mechanism whereby thymocytes gain access to peripheral compartments is obscure.We have learned that this process is sensitive to pertussis toxin (PT). Transgenic Ick-PT mice were generated which express the catalytic subunit of PT in all thymocytes. In a previous study we observed that Tcell receptor signaling is unimpaired in these cells despite the virtual elimination of their Gi protein signal transduction elements through endogenous PT activity. Here we demonstrate that mature T lineage cells accumulate in Zck-PT thymuses and fail to populate peripheral lymphoid organs. The accumulating cells closely resemble normal peripheral T lymphocytes with respect to cell surface phenotype and responses to allogeneic spleen cells, yet perform poorly in in wivo homing assays. This migratory defect does not result from deficient expression of common homing receptors or alterations in intracellular CAMP concentrations. Based on these results, we propose that a novel PT-sensitive signaling pathway, almost certainly involving a Gi protein, is required for thymocyte emigration.
1 Introduction
iments could not discriminate between nonspecific effects of pertussis holotoxin, which includes elements that perMany cell surface receptors require specific guanine nucleo- meabilize cells to permit entry of the S1 subunit, and tide-binding G proteins to communicate with intracellular specific effects resulting from ADP-ribosylation of Gi a effector enzymes involved in second-messengerproduction subunits [15-171. [l]. Members of the G protein family of signal transduction molecules are composed of three subunits, a, andy.The a To evaluate directly the importance of Gi proteins in T cell subunit of the G protein determines its receptor and signaling,we generated transgenicanimals in which just the effector specificity and participates directly in signal trans- S1 subunit of PT was selectively expressed in thymocytes duction [1-3].The less heterogenous part of the G protein, under the control of the proximal promoter of the lck gene. the fircomplex, may also regulate effector enzyme activity Thymocytes of Zck-PT transgenic animals express active S1 [4-61. The G proteins consitute a diverse family of signal and exhibit essentially complete depletion of functional Gi transduction molecules that may exist in 20 or more protein signal transduction elements in vivo [18]. Such different forms [7]. Explicit functions have been defined for depletion should severely limit Gi-dependent signal transonly a few of these. For example, the cl, subunits of G, duction in transgenic thymocytes, permitting investigation proteins couple catecholamine receptor signals to adenylyl of Gi-dependent signaling pathways in otherwise normal T cyclase, while the at subunits of Gt (transducin) couple lineage cells. Despite the virtual absence of functional Gi signals from retinal photoreceptors to cGMP phosphodies- proteins in these cells, thymocytes from Zck-PT animals responded to TcR-CD3 stimulation with appropriate intraterase [l]. cellular Ca2+ accumulation, IL 2 secretion and proliferaLymphocytes and thymocytes contain appreciable levels of tion [MI. We concluded that Gi proteins probably do not Gg and Gi3 proteins (our unpublished results, [8, 91). In participate in TcR-CD3 signal transduction. many other cell types Gi proteins mediate ligand-dependent alterations in adenylyl cyclase [l] and/or phospholi- These results notwithstanding, the Zck-PT mice manifest an pase C activity [lo, 111, and in cardiac atrial cells aj subunits intriguing lymphoid defect which suggests an alternate role directly regulate K+ channel permeability[2,12]. Although for Gi proteins in T lineage cells. Adult transgenic animals the function of Gi proteins in lymphoid cells is unknown, possess thymuses containing abnormally high numbers of previous studies suggested that signalingfrom theTcR-03 apparently mature thymocytes but have only very few T complex might in part depend upon the action of these or lymphocytes at peripheral sites [18]. Here, we describe other G proteins [13]. This conclusion emerged from experiments examining the development, cell surface pheexperiments in which addition of pertussis toxin (PT) to notype and functional behavior of Zck-F'T thymocytes lymphocyte cultures blocked TcR-mediatedactivation [141. which demonstrate that a PT-sensitive mechanism reguThe catalytic subunit of PT, S1, is a mono-ADP-ribosyl- lates thymocyte emigration. transferase that specifically modifies and inactivates the aj subunits of Gi proteins [l].Unfortunately, previous exper2 Materials and methods [I 96341 Correspondence: Roger M. Perlmutter, Howard Hughes Medical Institute SL15, University of Washington, Seattle, WA 98195, USA 0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1991
2.1 Mice
DBA/2, SJL and C57BL/6J mice were obtained from The Jackson Laboratories (Bar Harbor, ME). BLG.PLThy-1.1
+
Oo14-2980/9lI1010-2565$3.50 .25/0
2566
K. E. Chaffin and R. M. Perlmutter
mice were generously provided by Dr. Philip Greenberg (University of Washington).We have previously described the lck-PT transgenic mice. Briefly, these were generated by injecting (C57BL/6J X DBA/2J) F2 mouse zygotes with a DNA construct consisting of the thymocyte-specific, proximal lck promotor driving expression of the PT S1 ADPribosyltranaferase structural gene in the context of the human growth hormone gene [18]. For experiments on thymic ontogeny, transgenic fetuses or pups derived from timed matings were identified by probing dot blots of tail DNA with a fragment of the transgene. The present experimentsutilized animals derived from third and fourth generation backcrosses of three independent transgenic lines to C57BL/6 mice. Mice from all three lines exhibited very similar phenotypes. 2.2 FCM
Eur. J. Immunol. 1991.21: 2565-2573
separated by 15% SDS-PAGE,transferred to nitrocellulose and immunoblotted with antiserum to pertussis toxin S1 subunits (a gift from Dr.Witold Cieplak, NIH) as described [181* 2.4 Assays for CTL and proliferation responses to allogeneic stimulation
LD assays for cytotoxic precursor cells were modified from Ryser and MacDonald [25]. Briefly, MLR microcultures (micro-MLR) were generated by plating variable numbers (0.3 x lo4-40 x lo4) of freshly isolated thymocyte or splenocyte responder cells (containing > 95% live lymphoid cells as assayed by trypan blue exclusion) into round-bottom wells of microtiter tissue culture plates with 5 x 105 irradiated (2000 rad) DBA/2 (H-2d)or self (H-2b) spleen stimulator cells in 0.2 ml of medium A supplemented with 10% FCS, 1 mM nonessential amino acids (Gibco/BRL), 50 mh4 a-MM (Sigma # M-6882, St. Louis, MO) and 11% v/v rat spleen Con A SN. Cultured cells were resuspended 7-8 days later and one-half volume (100 pl) was removed and combined with an equal volume of medium containing 5000 Wr-labeled P815 (H-2d) target tumor cells in round-bottom microtiter wells. Plates were centrifuged at 200 x g for 3 min and then incubated at 37°C for 4 h, after which 100 pl SN from each well was assayed for released W r using a gamma counter. For all experiments, spontaneous W r release from P815 targets was < 15% of detergent lysis. Each responder cell population was derived from a single animal, therefore microcultures positive for allospecific responses were defined as those producing W r cpm 2 SD greater than the means of %r cpm released by an equal number of cells stimulated with irradiated splenocytes from the same individual. Responses of five to ten micro-MLR cultures stimulated with either allogeneic or self splenocytes were assayed for each of seven twofold responder cell dilutions for each responder population. CTL precursor frequencies of transgenic and littermate control responder cell populations assayed concurrently were then calculated using Poisson statistics.
Cells were teasedfrom lymphoid organs in cool (10 "-15 "C) medium A (RPMI 1640 (GibcolBRL, Grand Island, NY) containing 50 p~ 2-ME, 3 p~ glycine, 200 U/ml penicillin, 60 pg/ml streptomycin and 2 mM L-glutamine) supplemented with 5% or 10% FCS. Splenocyte and fetal thymocyte cell preparations were depleted of erythrocytes by alkaline lysis. For FCM analysis, cells were stained with saturatingconcentrations of antibodies at 4 "C as described [19]. mAb used were: FITC-conjugated 53-6.71 (CD8), PE-conjugated GK 1.5 (CD4), and biotinylated 30-H12 (Thy-1.2) from Becton Dickinson (Mountain View, CA); biotinylated RM2-5 (CD2) from PharMingen (San Diego, CA); IM781 (Pgp-1) [20], J l l d [21], and FD441.8 (LFA-1) kindly provided by Dr. Andrew Farr (University of Washington); Mel-14 [22], a gift from Dr. W. M. Gallatin (ICOS, Bothel1,WA); and biotinylated 500AA.2 ((333) [23], biotinylated GK1.5 (CD4), and FITC-conjugated 145.2C11 (CD3) [24]. For one- or two-color analysis, PE-conjugated streptavidin (SA) was used to detect binding of biotinylated reagents, and PE-conjugated goat anti-rat IgG absorbed against mouse IgG was used to detect Mel-14, Jlld, LFA-1 and Pgp-1 (both from Caltag, San Francisco, CA). Three-color FCM was performed using a two-step stainingprocedure with PE-CD4, FITC-CD8, and Alloantigen-specific proliferation by responder cell popubiotinylated CD2, CD3, or Thy-1.2 followed by RED613- lations was assessed by addition of 1 pCi [3H]dThdin 25 pl SA (GibcolBFU) to detect CD2 and CD3 or Duochrome- medium to micro-MLR cultures after removal of half the SA (Becton Dickinson) to detect Thy-1.2. In other cases, a cells for the CTL assay. After 16-20 h, micro-MLR cells four-step staining procedure was employed beginning with were assayed for [3H]dThd incorporation by use of a Pgp-1, J l l d , LFA-1, or Mel-14 developed with PE-goat Betaplate cell harvestedcounter (Pharmaciflallace Oy, anti-rat IgG, then biotinylated CD4 and FITC-CD8 plus Turku, Finland). 1: 100 normal rat serum (NRS) developed with RED613SA plus NRS. Multiparameter FCM was performed on a FACScan instrument (Becton Dickinson) and data 2.5 In vivo homing assay (10-50 000 events/sample) were analyzed using Consort 30, Paint-a-Gate and FACScan Research Software (Becton Freshly isolated thymocytes from 6- to 8-week-old male Dickinson). A FACStar PLUS (Becton Dickinson) instru- transgenic and control donor animals were washed twice with HBSS (Gibco/BRL =#310-4025), resuspended at 2.5 x ment was used for cell sorting experiments. 108-3.5 x lo8 cells/ml in HBSS and injected into the tail veins of 10-13-week-old male BL6.PL-Thy-1.1 recipient mice (5 x lo7-7 x lo7thymocytes in 0.2 ml per mouse) that 2.3 Immunoblots had been subjected to sublethal (880 rad) irradiation Whole cell lysates of thymocytes from control and lck-PT 18-24 h before cell transfer. Mock-injected recipients mice were prepared by hypotonic lysis as described [18] or received only HBSS. Recipient animals were killed 24 h by extraction in 50 mM Tris, pH 7.4, 150 mh4 NaCl, 1% after injection and thymus, spleen, mesenteric and inguinal Triton X-100,2 m~ EDTA, 1m~ PMSF, 1pg/ml aprotinin LN, kidneys, and a sample of peripheral blood were and 1 pg/ml leupeptin. Solubilized proteins were then assayed for cells of donor origin by FCM after staining with
Eur. J. Immunol. 1991. 21: 2565-2573
Pertussis toxin blocks thymocyte emigration
2567
mAb to CD4, CD8, and Thy-1.2 (see above). Approximately identical numbers of Thy-l.2+ cells were recovered (5 x 104-20 x 104 in different experiments) from animals injected with control and transgenic thymocytes.
As illustrated in Fig. 1, thymuses from adult Ick-PT mice contain abnormally high proportions of phenotypically mature CD4+CD8- and CD4-CD8+ cells, typically threeto fivefold more than control thymuses. These singlepositive (SP) cells from Ick-PT thymuses also express high levels of functional TcR-CD3receptors ([18] and below). In addition to thymic abnormalities, all mice from the three 2.6 CAMPassay Ick-PT lines have only very low numbers of peripheral Freshly isolated thymocytes were washed twice in cold T cells from birth to 10 weeks of age (see below). This HBSS, resuspended in HBSS at 5 x 106-25 x 106 cells/ml, phenotype suggested that PT activity in newly differenand lysed by adding 1/10 volume 1M HCl and vortexing. tiated SPthymocytes might block the ability of these cells to After complete evaporation, CAMP present in reconsti- leave the thymus. tuted samples was determined by RIA (RainedDupont; Wilmington, DE). Data presented are derived from assays To explore this hypothesis, we examined the development of Ick-PT mice. During fetal and neonatal ontogeny, on four transgenic and four control thymocyte samples. differentiation and emigration of T lymphocytes occur in more or less synchronous waves [27, 281. Information regarding the kinetics of these processes may, therefore, be 2.7 Thymocyte homotypic adhesion assay extracted from studying thymuses at appropriate developThymocyte homotypic adhesion was assayed using a mod- mental stages. In addition, since large numbers of Tcells ification of the procedure of Rothlein and Springer [26]. first begin leaving the thymus just before birth, the Briefly, 1, 5 or 10 million freshly prepared control and efficiency of thymocyte emigration might potentially be trangenic thymocytes in 100 pl medium A plus 10% FCS revealed by the phenotypes of young animals. We, therewere plated into flat-bottom microtiter wells. Individual fore, employed FCM to analyze Ick-PT thymocytes from wells were stimulated with 0,0.001,0.001,0.1,1, or 10 mg embryonic day 16 (E16) through the neonatal period and at PMA/ml or with 1:25, 1:50, 1:100, 1:200, or 1:400 1 and 2 weeks of age. Subset proportions obtained from dilutions of 145.2C11 (CD3) ascites fluid. After a 3-min individual transgenic animals were averaged and are plotcentrifugation at 500 X g, cultures were placed at 37 "C and ted in Fig. 2 as the percent of average values from examined microscopically at 19min intervals for evidence littermate control animals for each time point. of cell clumping. Cultures scored as positive for clumping activity were qualitatively defined as those containing From El7 onwards, Ick-PT thymuses contain about 50% fewer total thymocytes than thymuses of littermate control multiple aggregates of five or more cells. mice (Fig. 2). S1 activity may therefore be somewhat toxic to developingTlineage cells. Notwithstandingthis cell loss, the representation of the very immature CD4-CD8-, 3 Results immature CD4+CD8+, and the relatively mature 3.1 Single-positiveT cells accumulate in Zck-PT thymuses CD4+CD8- and CD4-CD8+ cell subsets did not differ significantly among Ick-PTanimals and littermate controls during ontogeny during fetal life. In an attempt to dissect the role of Gi proteins in Tcell signal transduction,we generated mice that express the PT Normal subset representation during this period indicates ADP-ribosyltransferase subunit, S1, in their thymocytes that thymocytes from Ick-PTmice are capable of pursuing a under control of the proximal Ick promoter [18].Three lines of these Zck-PT mice have been propagated for five Cell Number % CD4+CD8- + CD4-CD8+ % CD3hi generations, and all transgene-bearing progeny of the original founder mice display a characteristic phenotype.
t
t
.
Control Thymus
1
Transgenic Thymus
P
1%
"
CD4+CD8+
'p
t 9 %
200
n
4
0
m'p 100
101
118
103
-
1 0 4 100
CD8
E16E17EIBEi0 E
2
6
I4
E16E17E18E10 E
2
6
14
Age (days) 101
102
109
id
Figure 1 . CD4 and CD8 expression by Ick-PT thymocytes. Thymocytes of adult transgenic and littermate control mice were incubated with mAb to CD4 and CD8 and analyzed by FCM. The percentage of cells in each quadrant is indicated.
Figure 2. Ontogeny of Zck-PT thymuses.Thymocytesof transgenic and control mice of various ages [specifiedas days of embryoniclife (E) or after birth] were counted and cell surface antigens assayed by FCM. Data for transgenic thymuses are shown as percent of control at each time; each bar represents mean comparisons of between 4 and 17 thymuses.
2568
Eur. J. Immunol. 1991. 21: 2565-2573
K. E. Chaffin and R. M. Perlmutter
normal maturation sequence to the SP stage [27, 281. In contrast to earlier time points, the representation of the more mature CD4+CD8- and CD4-CD8+ thymocytes in Ick-PT thymuses increased dramatically during the first week of life, such that Ick-PT thymuses contain a threefold excess of this mature, CD3hisubset 6 days after birth. After 1week of age, mature subset representation in Ick-PT as compared to control thymuses remained consistently elevated. The increase in cells with mature cell surface phenotype during the 1st week of life is the most strikingly abnormal characterstic of Ick-PT thymocyte development. Importantly, this does not result from development regulation of transgene expression. Fig. 3 A demonstrates that Ick-PT thymocytescontain comparablelevels of S1 protein from at least El7 to 6 weeks of age. There is also no significant development regulation of the Gi target of PT in fetal thymocytes: both the ai2 chain of GQand the clo chain of GO proteins are present at equivalent levels in thymocytes examined at any of these ages (data not shown). Moreover, both the immature CD4+CD8+ and the more mature SP thymocyte subsets from Ick-PT mice contain S1 protein, although the more mature populations contain three- to fivefold less transgene product (Fig. 3 B). These results agree with previous reports concerning the regulation of the Zck-proximal promoter [29-321. Hence, the timing of the more severe developmental disturbance in Ick-PT thymocyte maturation does not result from late expression of the transgene product or its cellular targets. The accumulation of SP thymocytes documented in Fig.2 is reflected in the splenocytes of Ick-PT mice. As shown in Fig. 4, FCM studies performed at 1 week of age revealed that < 1/1OOOIck-PT splenic lymphocytes displayed the T cell markers CD4 or CD8. In contrast, appreciablenumbers of such cells were readily detected among control splenocytes. Similar deficits in Thy-1.2+ and CD3+ cells were observed in the spleens and in the peripheral blood of Zck-PT mice at 1 , 2 and 6 weeks of age ([18] and data not shown). Although some mature T cells do eventually
(B) s1+
Figure 3. S1 expression by lck-PT thymocytes. Whole cell lysates were separated by SDS-PAGE on a 15% gel, transferred to nitrocellulose, and immunoblotted with a polyclonal rabbit antiserum to S1. (A) Lysates of total thymocytes from kk-PT mice of indicated ages and adult littermate control, 100 pg proteidlane. (B) Lysates of sorted populations of lck-PT thymocytes, 2 X lo6 cellsflane.
% CD4+CD8*O
t
% CD4-CDEt
t
Age (weeks)
Figure 4. Ontogeny of Ick-PT spleens. Erythrocyte-depleted splenocytes from transgenic and control mice of various ages were incubated with mAb to CD4 and CD8 and analyzed by FCM. Each bar represents mean results of between two and four spleens. (0) control; (m) transgenic.
appear in the spleens of Zck-PT mice, the number rarely exceeds 15% of the normal value in mice up to 10 weeks of age. Hence, increased numbers of phenotypically mature thymocytes are found in Ick-PT thymuses during the period when such cells would normally begin to appear in the peripheral lymphoid circulation, and peripheral T lymphocyte colonization is severely delayed or disrupted in Ick-PT animals. These results support the idea that thymocyte emigration is severely blocked in Zck-PT mice. 3.2 Transgenic SP thymocytes are Mel-14s, Jlld-, CDP and Pgp-1-
In contrast to peripheral T lymphocytepopulations, normal SP thymocytes frequently manifest immature characteristics with respect to expression of other cell surface markers [27,33-361. If matureT lineage cells accumulate in Ick-PT thymuses instead of emigrating to the periphery, then Ick-PT SP thymocyte populations should contain more mature-appearing cells than normal SP thymocyte populations. We, therefore, utilized three-color FCM to compare Mel-14, Jlld, CD3 and Pgp-1 expression by Ick-PT and control SP thymocytes and peripheral T cells. More than 50% of SP Ick-PTthymocytes expressed the Mel-14antigen at levels comparable to or greater than those found among SP splenic T cells, while only 20%-30% of control SP thymocytes expressed appreciablelevels of Mel-14 (Fig. 5). Similarly, transgenic SP thymocytes contained greater proportions of Jlld- cells than did normal SP thymocytes, resembling splenocytes in this respect since nearly all normal splenic T lymphocytes are Jlld- (Fig. 5 and 127, 211). In addition, when compared to control CD4-CD8+ thymocytes, a consistently greater proportion of transgenic CD4-CD8+ thymocytes expressed high levels of CD3 (Fig. 5 and data not shown). Both transgenic SP thymocytes and normal splenic T cells also expressed CD3 at slightly lower cell-surface densities than did control SP t hymocytes. SP thymocytes from Ick-PT animals also contained greatly increased proportions of Pgp-1- cells compared with normal SPlthymocytes which express low but detectable levels of Pgp-1 (Fig. 5). Recent thymic emigrantsreported-
Eur. J. Immunol. 1991. 21: 2565-2573
Pertussis toxin blocks thymocyte emigration
".
2569
8000-
C
0 ._ 4d
6000-
P u
Mel-14
1
4
4000--
P)
1
1
-' C .--u .-
E,
2000-
L +
I
M
04 1000
1€4
1 E5
Cells/well
J1 ld-
Figure 6. Proliferation by Zck-PT thymocytes after allogeneic stimulation. Non-limiting numbers of splenocytesfrom one control (Con) animal and thymocytes from two control and one transgenic (Tg) animal were cultured with irradiated DBA/2 stimulator cells, then analyzed 8 days later for proliferation by assaying r3H]dThd incorporation. Responder cell preparations originally contained 23.8% (+,Tg thymus), 7.3 and 8.9% @A, Con thymuses) or 33.5% (0,Con spleen) SP cells. Each point represents the mean response of ten individual microcultures.
CD3 +
1-
Pgp-l+ Figure 5. FCM analysis of SP Zck-PT thymocytes. Thymocytes from adult transgenic and thymocytes and splenocytes from adult littermate control mice were incubatedwith mAb to CD4, CD8 and Mel-14, Jlld, Pgp-1 or CD3 and analyzed by three-color FCM. (-) Tg thymus, (..--)Con thymus, (----)Con spleen.
ly express little or no Pgp-1, while the low to high levels of Pgp-1 on splenicT cells are thought to result from exposure to antigen [34, 351. In contrast to SP populations, Fig. 5 shows that both control and transgenic CD4+CD8+thymocytes expressed equivalent levels of Mel-14, Jlld, and CD3, and similar levels of Pgp-1. Thus, results in Fig. 5 clearly support the view that Zck-PT SP thymocyte populations contain relatively mature T lymphocytes. In fact, fully 15%-20% of total Ick-PT thymocytes manifest the cellsurface phenotypes of recent thymic emigrants found in the peripheral lymphoid tissues of normal mice. 3.3 Increased frequency of CD4-CDW CTL precursors among Zck-PT thymocytes
To assess the functional maturity of Zck-PT thymocytes, we examined the ability of these cells to respond to an allogeneic challenge. First, transgenic and control thymocytes and control splenocytes were incubated in bulk cultures with irradiated DBAD (H-2d) stimulator spleen cells. Proliferative responses were assayed 8 days later by As shown in Fig. 6, approximately threeontrol thymocytes in response to allogeneicstimulation, correspondingwell with the threefold higher representation of SP cells in the transgenic preparations. Indeed, H-2d stimulator cells
induced proliferation of Zck-PT thymocytes and control splenocytes to an approximately equal extent, a result consistent with the higher content of SP cells in these populations. Thus, the accumulating SP cells in Zck-PT thymuses are capable of normal, mature poliferative responses to stimulation by APC. Stimulationof normal CD4-CD8+ thymocytes yields fewer CTL than the same number of splenic CD4-CD8+ T cells because thymic CD4-CD8+ populations contain immature cells incapable of mediating cytolyses [27,36-381. To assess further the functional maturity of Zck-PT thymocytes, we therefore, enumerated H-2d-specific CTL precursors in transgenic and control CD4-CD8+ cell populations.Transgenic and control thymocytes and control splenocytes were stimulated in LD cultures with irradiated DBA/2(H-2d)or self (H-2b)spleencells, then assayed 7-8 days later for CTL activity against H-2dtumor cell targets. Fig. 7 A shows data from a representa experiment demonstrating that, as expected from their greater proportion of CD4-CDV cells, total Zck-PTtransgenic thymocytes generated significantly larger numbers of H-2d-specificCTL than did total control thymocytes. Fig.7B compares data on H-2dspecificC l l precursors in transgenic and control lymphoid populations after all values were normalized to equivalent numbers of CD4-CD8+ cells. Transgenic CD4-CD8+ thymocytes produced two to three times as many H-2d-specific CTL effectors as the contra1 CD4-CD8+ thymocytes. In fact, the values obtained for CD4-CD8+ Zck-PT thymocytes are indistinguishable from those obtained for normal CD4-CD8+ splenocytes. Hence, by this sensitive assay of functional maturity, CD4-CD8+ Ick-PTthymocytes resemble normal peripheral, rather than relatively immature thymic, CD4-CD8+ populations. 3.4 Mature Zck-PT thymocytes exhibit defective homing in vivo
Previous studies have established that treatment of mature lymphocytes with pertussis holotoxin blocks the ability of
K. E. Chaffin and R. M. Perlmutter
2570
(A)
Eur. J. Immunol. 1991.22: 2565-2573
Responder Cells/Well 0.0
2.OE5
1 .OE5
3.OE5
4.OE5
- 400 -
microcultures;allo-specificresponder wells are defined in Sect. 2.4. The number of responder cells giving37% non-respondingcultures corresponds to reciprocal CTL precursor frequency for a given cell population: 1/40000 (Tg) or 1/280000 (Con) for the experiment shown.Thymocytepreparations originallycontained 10.2% (Tg) or 4.1% (Con) CD4-CD8+ cells. (B) Summary of data from three separate LD CTL precursor assays. CTL precursor frequencies in Tg and Con cell populations were normalized to proportions of CD4-CDSf cells originally present, the percent of control responses for Tg populations was calculated, and the results of different experiments were averaged. ( ) Tg thymus / Con thymus, ( 0 )Tg thymus / Con spleen. Errors bars are 1 SD.
0
22
$3 05
300-
U+
w o
L=ae
6+-
:2 0
200-
0
2 ' a
______ __-________
100
I
20 U Q k
*
n\\\\\\\\\\\y
0-7
these cells to exit the bloodstream and enter lymphoid organs [39,40]. Since phenotypically matureT lymphocytes accumulate in the thymuses of Zck-PT mice, it seemed plausible that the mechanisms regulating thymocyte emigration and homing of peripheral lymphocytes might be related or identical. We addressed this hypothesis by examining the in vivo homing capabilities of Ick-PT thymo-
cytes. Equal numbers of control or transgenic thymocytes (all genetically Thy-1.2) were injected into the tail veins of irradiated BL6.PL-Thy-1.1 congenic recipients. After 24 h, organs of recipient mice were removed and assayed for cells of donor origin by FCM after staining with antibodies to Thy-1.2. For one experiment, Thy-1.2+ cells were also analyzed for CD4 and CD8 expression by three-color FCM.
Table 1. I n vivo homing of lck-PT thymocytesa)
Esp. 1 I-
Donor
Mock Id-PI' Control Mock Ilk-PT Control Mock Ick-PT
%,
Yo
organ
Thy- 1.2 ' ccllsh'
Thy-].2'
I hynius
I .
0 0.1
0.I Splccn
0 1.5.0 10. I
Mcscnteric LN
0 0.4 4.4
1nguin;il LN
0 0.4 6.6
BlOOd
kk-PT Con t rol
Mock k.k - PT Control
r-
Rccipicnt
Control
Mock Id-PT Control Mock
,
0
33. I 6.8
h:idnc y ntl
cells 0 0. I 0. I
0.2 36.5 10.5 0.4 2.0 6.4
-
Exp. 2 1
% Thy-1.2' CD4+CD8 cells
K, Thy-1.2' CDJ-CDS+ cclls
nd
nd
23.9 6.4
8.9 2.9
I .2 4.4
0.4 1.6
0.9
0.5
7.4
5.5
0.3 1.7
45.7 0.4
22.4 0.2
I .9
0.7 0
0
0.2 72.9 I .2 0.2 3.7 0.9
0.I
a) Thy-1.2+ thymocytes from control or lck-PT animals were injected into the tail veins of irradiated Thy-1.1 congenic recipients. After 24 h, recipient organs were assayed for donor cells by FCM (see Sect. 2.2 for details). b) The percentage of lymphoid cells in the recipient organ bearing lymphoid markers.
Eur. J. Immunol. 1991. 21: 2565-2573
Pertussis toxin blocks thymocyte emigration
2571
Results shown in Table 1 indicate that a small number of receptors might explain some features of the Zck-PT control thymocytes migrate from the peripheral blood into phenotype. Results of FCM analyses revealed, however, mesenteric and inguinal LN. Typically, 4%-8% of inguinal that SP lck-PT thymocytes expressLFA-1, CD2 and CD4 at or mesenteric LN lymphocytes from animals injected with levels indistinguishable from those of normal SP thymocontrol thymocytes expressed the donor Thy-1.2 marker cytes and normal splenicT lymphocytes (data not shown). after 24 h, representing approximately 0.1% of injected Furthermore, 24 h after stimulation with immobilized mAb cells. The fact that lymphoid cells from the kidneys or to CD3, both control and transgenic SP thymocytes exthymuses of these animals contained
Karen E. Chaffin and Roger M. Perlmutter Howard Hughes Medical Institute and the Departments of Biochemistry, Immunology, and Medicine (Medical Genetics), University of Washington, Seattle
Pertussis toxin blocks thymocyte emigration
2565
A pertussis toxin-sensitive process controls thymocyte emigration Although it is well known that essentially all peripheral Tcells are derived from bone marrow progenitors that mature in the thymus, the mechanism whereby thymocytes gain access to peripheral compartments is obscure.We have learned that this process is sensitive to pertussis toxin (PT). Transgenic Ick-PT mice were generated which express the catalytic subunit of PT in all thymocytes. In a previous study we observed that Tcell receptor signaling is unimpaired in these cells despite the virtual elimination of their Gi protein signal transduction elements through endogenous PT activity. Here we demonstrate that mature T lineage cells accumulate in Zck-PT thymuses and fail to populate peripheral lymphoid organs. The accumulating cells closely resemble normal peripheral T lymphocytes with respect to cell surface phenotype and responses to allogeneic spleen cells, yet perform poorly in in wivo homing assays. This migratory defect does not result from deficient expression of common homing receptors or alterations in intracellular CAMP concentrations. Based on these results, we propose that a novel PT-sensitive signaling pathway, almost certainly involving a Gi protein, is required for thymocyte emigration.
1 Introduction
iments could not discriminate between nonspecific effects of pertussis holotoxin, which includes elements that perMany cell surface receptors require specific guanine nucleo- meabilize cells to permit entry of the S1 subunit, and tide-binding G proteins to communicate with intracellular specific effects resulting from ADP-ribosylation of Gi a effector enzymes involved in second-messengerproduction subunits [15-171. [l]. Members of the G protein family of signal transduction molecules are composed of three subunits, a, andy.The a To evaluate directly the importance of Gi proteins in T cell subunit of the G protein determines its receptor and signaling,we generated transgenicanimals in which just the effector specificity and participates directly in signal trans- S1 subunit of PT was selectively expressed in thymocytes duction [1-3].The less heterogenous part of the G protein, under the control of the proximal promoter of the lck gene. the fircomplex, may also regulate effector enzyme activity Thymocytes of Zck-PT transgenic animals express active S1 [4-61. The G proteins consitute a diverse family of signal and exhibit essentially complete depletion of functional Gi transduction molecules that may exist in 20 or more protein signal transduction elements in vivo [18]. Such different forms [7]. Explicit functions have been defined for depletion should severely limit Gi-dependent signal transonly a few of these. For example, the cl, subunits of G, duction in transgenic thymocytes, permitting investigation proteins couple catecholamine receptor signals to adenylyl of Gi-dependent signaling pathways in otherwise normal T cyclase, while the at subunits of Gt (transducin) couple lineage cells. Despite the virtual absence of functional Gi signals from retinal photoreceptors to cGMP phosphodies- proteins in these cells, thymocytes from Zck-PT animals responded to TcR-CD3 stimulation with appropriate intraterase [l]. cellular Ca2+ accumulation, IL 2 secretion and proliferaLymphocytes and thymocytes contain appreciable levels of tion [MI. We concluded that Gi proteins probably do not Gg and Gi3 proteins (our unpublished results, [8, 91). In participate in TcR-CD3 signal transduction. many other cell types Gi proteins mediate ligand-dependent alterations in adenylyl cyclase [l] and/or phospholi- These results notwithstanding, the Zck-PT mice manifest an pase C activity [lo, 111, and in cardiac atrial cells aj subunits intriguing lymphoid defect which suggests an alternate role directly regulate K+ channel permeability[2,12]. Although for Gi proteins in T lineage cells. Adult transgenic animals the function of Gi proteins in lymphoid cells is unknown, possess thymuses containing abnormally high numbers of previous studies suggested that signalingfrom theTcR-03 apparently mature thymocytes but have only very few T complex might in part depend upon the action of these or lymphocytes at peripheral sites [18]. Here, we describe other G proteins [13]. This conclusion emerged from experiments examining the development, cell surface pheexperiments in which addition of pertussis toxin (PT) to notype and functional behavior of Zck-F'T thymocytes lymphocyte cultures blocked TcR-mediatedactivation [141. which demonstrate that a PT-sensitive mechanism reguThe catalytic subunit of PT, S1, is a mono-ADP-ribosyl- lates thymocyte emigration. transferase that specifically modifies and inactivates the aj subunits of Gi proteins [l].Unfortunately, previous exper2 Materials and methods [I 96341 Correspondence: Roger M. Perlmutter, Howard Hughes Medical Institute SL15, University of Washington, Seattle, WA 98195, USA 0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1991
2.1 Mice
DBA/2, SJL and C57BL/6J mice were obtained from The Jackson Laboratories (Bar Harbor, ME). BLG.PLThy-1.1
+
Oo14-2980/9lI1010-2565$3.50 .25/0
2566
K. E. Chaffin and R. M. Perlmutter
mice were generously provided by Dr. Philip Greenberg (University of Washington).We have previously described the lck-PT transgenic mice. Briefly, these were generated by injecting (C57BL/6J X DBA/2J) F2 mouse zygotes with a DNA construct consisting of the thymocyte-specific, proximal lck promotor driving expression of the PT S1 ADPribosyltranaferase structural gene in the context of the human growth hormone gene [18]. For experiments on thymic ontogeny, transgenic fetuses or pups derived from timed matings were identified by probing dot blots of tail DNA with a fragment of the transgene. The present experimentsutilized animals derived from third and fourth generation backcrosses of three independent transgenic lines to C57BL/6 mice. Mice from all three lines exhibited very similar phenotypes. 2.2 FCM
Eur. J. Immunol. 1991.21: 2565-2573
separated by 15% SDS-PAGE,transferred to nitrocellulose and immunoblotted with antiserum to pertussis toxin S1 subunits (a gift from Dr.Witold Cieplak, NIH) as described [181* 2.4 Assays for CTL and proliferation responses to allogeneic stimulation
LD assays for cytotoxic precursor cells were modified from Ryser and MacDonald [25]. Briefly, MLR microcultures (micro-MLR) were generated by plating variable numbers (0.3 x lo4-40 x lo4) of freshly isolated thymocyte or splenocyte responder cells (containing > 95% live lymphoid cells as assayed by trypan blue exclusion) into round-bottom wells of microtiter tissue culture plates with 5 x 105 irradiated (2000 rad) DBA/2 (H-2d)or self (H-2b) spleen stimulator cells in 0.2 ml of medium A supplemented with 10% FCS, 1 mM nonessential amino acids (Gibco/BRL), 50 mh4 a-MM (Sigma # M-6882, St. Louis, MO) and 11% v/v rat spleen Con A SN. Cultured cells were resuspended 7-8 days later and one-half volume (100 pl) was removed and combined with an equal volume of medium containing 5000 Wr-labeled P815 (H-2d) target tumor cells in round-bottom microtiter wells. Plates were centrifuged at 200 x g for 3 min and then incubated at 37°C for 4 h, after which 100 pl SN from each well was assayed for released W r using a gamma counter. For all experiments, spontaneous W r release from P815 targets was < 15% of detergent lysis. Each responder cell population was derived from a single animal, therefore microcultures positive for allospecific responses were defined as those producing W r cpm 2 SD greater than the means of %r cpm released by an equal number of cells stimulated with irradiated splenocytes from the same individual. Responses of five to ten micro-MLR cultures stimulated with either allogeneic or self splenocytes were assayed for each of seven twofold responder cell dilutions for each responder population. CTL precursor frequencies of transgenic and littermate control responder cell populations assayed concurrently were then calculated using Poisson statistics.
Cells were teasedfrom lymphoid organs in cool (10 "-15 "C) medium A (RPMI 1640 (GibcolBRL, Grand Island, NY) containing 50 p~ 2-ME, 3 p~ glycine, 200 U/ml penicillin, 60 pg/ml streptomycin and 2 mM L-glutamine) supplemented with 5% or 10% FCS. Splenocyte and fetal thymocyte cell preparations were depleted of erythrocytes by alkaline lysis. For FCM analysis, cells were stained with saturatingconcentrations of antibodies at 4 "C as described [19]. mAb used were: FITC-conjugated 53-6.71 (CD8), PE-conjugated GK 1.5 (CD4), and biotinylated 30-H12 (Thy-1.2) from Becton Dickinson (Mountain View, CA); biotinylated RM2-5 (CD2) from PharMingen (San Diego, CA); IM781 (Pgp-1) [20], J l l d [21], and FD441.8 (LFA-1) kindly provided by Dr. Andrew Farr (University of Washington); Mel-14 [22], a gift from Dr. W. M. Gallatin (ICOS, Bothel1,WA); and biotinylated 500AA.2 ((333) [23], biotinylated GK1.5 (CD4), and FITC-conjugated 145.2C11 (CD3) [24]. For one- or two-color analysis, PE-conjugated streptavidin (SA) was used to detect binding of biotinylated reagents, and PE-conjugated goat anti-rat IgG absorbed against mouse IgG was used to detect Mel-14, Jlld, LFA-1 and Pgp-1 (both from Caltag, San Francisco, CA). Three-color FCM was performed using a two-step stainingprocedure with PE-CD4, FITC-CD8, and Alloantigen-specific proliferation by responder cell popubiotinylated CD2, CD3, or Thy-1.2 followed by RED613- lations was assessed by addition of 1 pCi [3H]dThdin 25 pl SA (GibcolBFU) to detect CD2 and CD3 or Duochrome- medium to micro-MLR cultures after removal of half the SA (Becton Dickinson) to detect Thy-1.2. In other cases, a cells for the CTL assay. After 16-20 h, micro-MLR cells four-step staining procedure was employed beginning with were assayed for [3H]dThd incorporation by use of a Pgp-1, J l l d , LFA-1, or Mel-14 developed with PE-goat Betaplate cell harvestedcounter (Pharmaciflallace Oy, anti-rat IgG, then biotinylated CD4 and FITC-CD8 plus Turku, Finland). 1: 100 normal rat serum (NRS) developed with RED613SA plus NRS. Multiparameter FCM was performed on a FACScan instrument (Becton Dickinson) and data 2.5 In vivo homing assay (10-50 000 events/sample) were analyzed using Consort 30, Paint-a-Gate and FACScan Research Software (Becton Freshly isolated thymocytes from 6- to 8-week-old male Dickinson). A FACStar PLUS (Becton Dickinson) instru- transgenic and control donor animals were washed twice with HBSS (Gibco/BRL =#310-4025), resuspended at 2.5 x ment was used for cell sorting experiments. 108-3.5 x lo8 cells/ml in HBSS and injected into the tail veins of 10-13-week-old male BL6.PL-Thy-1.1 recipient mice (5 x lo7-7 x lo7thymocytes in 0.2 ml per mouse) that 2.3 Immunoblots had been subjected to sublethal (880 rad) irradiation Whole cell lysates of thymocytes from control and lck-PT 18-24 h before cell transfer. Mock-injected recipients mice were prepared by hypotonic lysis as described [18] or received only HBSS. Recipient animals were killed 24 h by extraction in 50 mM Tris, pH 7.4, 150 mh4 NaCl, 1% after injection and thymus, spleen, mesenteric and inguinal Triton X-100,2 m~ EDTA, 1m~ PMSF, 1pg/ml aprotinin LN, kidneys, and a sample of peripheral blood were and 1 pg/ml leupeptin. Solubilized proteins were then assayed for cells of donor origin by FCM after staining with
Eur. J. Immunol. 1991. 21: 2565-2573
Pertussis toxin blocks thymocyte emigration
2567
mAb to CD4, CD8, and Thy-1.2 (see above). Approximately identical numbers of Thy-l.2+ cells were recovered (5 x 104-20 x 104 in different experiments) from animals injected with control and transgenic thymocytes.
As illustrated in Fig. 1, thymuses from adult Ick-PT mice contain abnormally high proportions of phenotypically mature CD4+CD8- and CD4-CD8+ cells, typically threeto fivefold more than control thymuses. These singlepositive (SP) cells from Ick-PT thymuses also express high levels of functional TcR-CD3receptors ([18] and below). In addition to thymic abnormalities, all mice from the three 2.6 CAMPassay Ick-PT lines have only very low numbers of peripheral Freshly isolated thymocytes were washed twice in cold T cells from birth to 10 weeks of age (see below). This HBSS, resuspended in HBSS at 5 x 106-25 x 106 cells/ml, phenotype suggested that PT activity in newly differenand lysed by adding 1/10 volume 1M HCl and vortexing. tiated SPthymocytes might block the ability of these cells to After complete evaporation, CAMP present in reconsti- leave the thymus. tuted samples was determined by RIA (RainedDupont; Wilmington, DE). Data presented are derived from assays To explore this hypothesis, we examined the development of Ick-PT mice. During fetal and neonatal ontogeny, on four transgenic and four control thymocyte samples. differentiation and emigration of T lymphocytes occur in more or less synchronous waves [27, 281. Information regarding the kinetics of these processes may, therefore, be 2.7 Thymocyte homotypic adhesion assay extracted from studying thymuses at appropriate developThymocyte homotypic adhesion was assayed using a mod- mental stages. In addition, since large numbers of Tcells ification of the procedure of Rothlein and Springer [26]. first begin leaving the thymus just before birth, the Briefly, 1, 5 or 10 million freshly prepared control and efficiency of thymocyte emigration might potentially be trangenic thymocytes in 100 pl medium A plus 10% FCS revealed by the phenotypes of young animals. We, therewere plated into flat-bottom microtiter wells. Individual fore, employed FCM to analyze Ick-PT thymocytes from wells were stimulated with 0,0.001,0.001,0.1,1, or 10 mg embryonic day 16 (E16) through the neonatal period and at PMA/ml or with 1:25, 1:50, 1:100, 1:200, or 1:400 1 and 2 weeks of age. Subset proportions obtained from dilutions of 145.2C11 (CD3) ascites fluid. After a 3-min individual transgenic animals were averaged and are plotcentrifugation at 500 X g, cultures were placed at 37 "C and ted in Fig. 2 as the percent of average values from examined microscopically at 19min intervals for evidence littermate control animals for each time point. of cell clumping. Cultures scored as positive for clumping activity were qualitatively defined as those containing From El7 onwards, Ick-PT thymuses contain about 50% fewer total thymocytes than thymuses of littermate control multiple aggregates of five or more cells. mice (Fig. 2). S1 activity may therefore be somewhat toxic to developingTlineage cells. Notwithstandingthis cell loss, the representation of the very immature CD4-CD8-, 3 Results immature CD4+CD8+, and the relatively mature 3.1 Single-positiveT cells accumulate in Zck-PT thymuses CD4+CD8- and CD4-CD8+ cell subsets did not differ significantly among Ick-PTanimals and littermate controls during ontogeny during fetal life. In an attempt to dissect the role of Gi proteins in Tcell signal transduction,we generated mice that express the PT Normal subset representation during this period indicates ADP-ribosyltransferase subunit, S1, in their thymocytes that thymocytes from Ick-PTmice are capable of pursuing a under control of the proximal Ick promoter [18].Three lines of these Zck-PT mice have been propagated for five Cell Number % CD4+CD8- + CD4-CD8+ % CD3hi generations, and all transgene-bearing progeny of the original founder mice display a characteristic phenotype.
t
t
.
Control Thymus
1
Transgenic Thymus
P
1%
"
CD4+CD8+
'p
t 9 %
200
n
4
0
m'p 100
101
118
103
-
1 0 4 100
CD8
E16E17EIBEi0 E
2
6
I4
E16E17E18E10 E
2
6
14
Age (days) 101
102
109
id
Figure 1 . CD4 and CD8 expression by Ick-PT thymocytes. Thymocytes of adult transgenic and littermate control mice were incubated with mAb to CD4 and CD8 and analyzed by FCM. The percentage of cells in each quadrant is indicated.
Figure 2. Ontogeny of Zck-PT thymuses.Thymocytesof transgenic and control mice of various ages [specifiedas days of embryoniclife (E) or after birth] were counted and cell surface antigens assayed by FCM. Data for transgenic thymuses are shown as percent of control at each time; each bar represents mean comparisons of between 4 and 17 thymuses.
2568
Eur. J. Immunol. 1991. 21: 2565-2573
K. E. Chaffin and R. M. Perlmutter
normal maturation sequence to the SP stage [27, 281. In contrast to earlier time points, the representation of the more mature CD4+CD8- and CD4-CD8+ thymocytes in Ick-PT thymuses increased dramatically during the first week of life, such that Ick-PT thymuses contain a threefold excess of this mature, CD3hisubset 6 days after birth. After 1week of age, mature subset representation in Ick-PT as compared to control thymuses remained consistently elevated. The increase in cells with mature cell surface phenotype during the 1st week of life is the most strikingly abnormal characterstic of Ick-PT thymocyte development. Importantly, this does not result from development regulation of transgene expression. Fig. 3 A demonstrates that Ick-PT thymocytescontain comparablelevels of S1 protein from at least El7 to 6 weeks of age. There is also no significant development regulation of the Gi target of PT in fetal thymocytes: both the ai2 chain of GQand the clo chain of GO proteins are present at equivalent levels in thymocytes examined at any of these ages (data not shown). Moreover, both the immature CD4+CD8+ and the more mature SP thymocyte subsets from Ick-PT mice contain S1 protein, although the more mature populations contain three- to fivefold less transgene product (Fig. 3 B). These results agree with previous reports concerning the regulation of the Zck-proximal promoter [29-321. Hence, the timing of the more severe developmental disturbance in Ick-PT thymocyte maturation does not result from late expression of the transgene product or its cellular targets. The accumulation of SP thymocytes documented in Fig.2 is reflected in the splenocytes of Ick-PT mice. As shown in Fig. 4, FCM studies performed at 1 week of age revealed that < 1/1OOOIck-PT splenic lymphocytes displayed the T cell markers CD4 or CD8. In contrast, appreciablenumbers of such cells were readily detected among control splenocytes. Similar deficits in Thy-1.2+ and CD3+ cells were observed in the spleens and in the peripheral blood of Zck-PT mice at 1 , 2 and 6 weeks of age ([18] and data not shown). Although some mature T cells do eventually
(B) s1+
Figure 3. S1 expression by lck-PT thymocytes. Whole cell lysates were separated by SDS-PAGE on a 15% gel, transferred to nitrocellulose, and immunoblotted with a polyclonal rabbit antiserum to S1. (A) Lysates of total thymocytes from kk-PT mice of indicated ages and adult littermate control, 100 pg proteidlane. (B) Lysates of sorted populations of lck-PT thymocytes, 2 X lo6 cellsflane.
% CD4+CD8*O
t
% CD4-CDEt
t
Age (weeks)
Figure 4. Ontogeny of Ick-PT spleens. Erythrocyte-depleted splenocytes from transgenic and control mice of various ages were incubated with mAb to CD4 and CD8 and analyzed by FCM. Each bar represents mean results of between two and four spleens. (0) control; (m) transgenic.
appear in the spleens of Zck-PT mice, the number rarely exceeds 15% of the normal value in mice up to 10 weeks of age. Hence, increased numbers of phenotypically mature thymocytes are found in Ick-PT thymuses during the period when such cells would normally begin to appear in the peripheral lymphoid circulation, and peripheral T lymphocyte colonization is severely delayed or disrupted in Ick-PT animals. These results support the idea that thymocyte emigration is severely blocked in Zck-PT mice. 3.2 Transgenic SP thymocytes are Mel-14s, Jlld-, CDP and Pgp-1-
In contrast to peripheral T lymphocytepopulations, normal SP thymocytes frequently manifest immature characteristics with respect to expression of other cell surface markers [27,33-361. If matureT lineage cells accumulate in Ick-PT thymuses instead of emigrating to the periphery, then Ick-PT SP thymocyte populations should contain more mature-appearing cells than normal SP thymocyte populations. We, therefore, utilized three-color FCM to compare Mel-14, Jlld, CD3 and Pgp-1 expression by Ick-PT and control SP thymocytes and peripheral T cells. More than 50% of SP Ick-PTthymocytes expressed the Mel-14antigen at levels comparable to or greater than those found among SP splenic T cells, while only 20%-30% of control SP thymocytes expressed appreciablelevels of Mel-14 (Fig. 5). Similarly, transgenic SP thymocytes contained greater proportions of Jlld- cells than did normal SP thymocytes, resembling splenocytes in this respect since nearly all normal splenic T lymphocytes are Jlld- (Fig. 5 and 127, 211). In addition, when compared to control CD4-CD8+ thymocytes, a consistently greater proportion of transgenic CD4-CD8+ thymocytes expressed high levels of CD3 (Fig. 5 and data not shown). Both transgenic SP thymocytes and normal splenic T cells also expressed CD3 at slightly lower cell-surface densities than did control SP t hymocytes. SP thymocytes from Ick-PT animals also contained greatly increased proportions of Pgp-1- cells compared with normal SPlthymocytes which express low but detectable levels of Pgp-1 (Fig. 5). Recent thymic emigrantsreported-
Eur. J. Immunol. 1991. 21: 2565-2573
Pertussis toxin blocks thymocyte emigration
".
2569
8000-
C
0 ._ 4d
6000-
P u
Mel-14
1
4
4000--
P)
1
1
-' C .--u .-
E,
2000-
L +
I
M
04 1000
1€4
1 E5
Cells/well
J1 ld-
Figure 6. Proliferation by Zck-PT thymocytes after allogeneic stimulation. Non-limiting numbers of splenocytesfrom one control (Con) animal and thymocytes from two control and one transgenic (Tg) animal were cultured with irradiated DBA/2 stimulator cells, then analyzed 8 days later for proliferation by assaying r3H]dThd incorporation. Responder cell preparations originally contained 23.8% (+,Tg thymus), 7.3 and 8.9% @A, Con thymuses) or 33.5% (0,Con spleen) SP cells. Each point represents the mean response of ten individual microcultures.
CD3 +
1-
Pgp-l+ Figure 5. FCM analysis of SP Zck-PT thymocytes. Thymocytes from adult transgenic and thymocytes and splenocytes from adult littermate control mice were incubatedwith mAb to CD4, CD8 and Mel-14, Jlld, Pgp-1 or CD3 and analyzed by three-color FCM. (-) Tg thymus, (..--)Con thymus, (----)Con spleen.
ly express little or no Pgp-1, while the low to high levels of Pgp-1 on splenicT cells are thought to result from exposure to antigen [34, 351. In contrast to SP populations, Fig. 5 shows that both control and transgenic CD4+CD8+thymocytes expressed equivalent levels of Mel-14, Jlld, and CD3, and similar levels of Pgp-1. Thus, results in Fig. 5 clearly support the view that Zck-PT SP thymocyte populations contain relatively mature T lymphocytes. In fact, fully 15%-20% of total Ick-PT thymocytes manifest the cellsurface phenotypes of recent thymic emigrants found in the peripheral lymphoid tissues of normal mice. 3.3 Increased frequency of CD4-CDW CTL precursors among Zck-PT thymocytes
To assess the functional maturity of Zck-PT thymocytes, we examined the ability of these cells to respond to an allogeneic challenge. First, transgenic and control thymocytes and control splenocytes were incubated in bulk cultures with irradiated DBAD (H-2d) stimulator spleen cells. Proliferative responses were assayed 8 days later by As shown in Fig. 6, approximately threeontrol thymocytes in response to allogeneicstimulation, correspondingwell with the threefold higher representation of SP cells in the transgenic preparations. Indeed, H-2d stimulator cells
induced proliferation of Zck-PT thymocytes and control splenocytes to an approximately equal extent, a result consistent with the higher content of SP cells in these populations. Thus, the accumulating SP cells in Zck-PT thymuses are capable of normal, mature poliferative responses to stimulation by APC. Stimulationof normal CD4-CD8+ thymocytes yields fewer CTL than the same number of splenic CD4-CD8+ T cells because thymic CD4-CD8+ populations contain immature cells incapable of mediating cytolyses [27,36-381. To assess further the functional maturity of Zck-PT thymocytes, we therefore, enumerated H-2d-specific CTL precursors in transgenic and control CD4-CD8+ cell populations.Transgenic and control thymocytes and control splenocytes were stimulated in LD cultures with irradiated DBA/2(H-2d)or self (H-2b)spleencells, then assayed 7-8 days later for CTL activity against H-2dtumor cell targets. Fig. 7 A shows data from a representa experiment demonstrating that, as expected from their greater proportion of CD4-CDV cells, total Zck-PTtransgenic thymocytes generated significantly larger numbers of H-2d-specificCTL than did total control thymocytes. Fig.7B compares data on H-2dspecificC l l precursors in transgenic and control lymphoid populations after all values were normalized to equivalent numbers of CD4-CD8+ cells. Transgenic CD4-CD8+ thymocytes produced two to three times as many H-2d-specific CTL effectors as the contra1 CD4-CD8+ thymocytes. In fact, the values obtained for CD4-CD8+ Zck-PT thymocytes are indistinguishable from those obtained for normal CD4-CD8+ splenocytes. Hence, by this sensitive assay of functional maturity, CD4-CD8+ Ick-PTthymocytes resemble normal peripheral, rather than relatively immature thymic, CD4-CD8+ populations. 3.4 Mature Zck-PT thymocytes exhibit defective homing in vivo
Previous studies have established that treatment of mature lymphocytes with pertussis holotoxin blocks the ability of
K. E. Chaffin and R. M. Perlmutter
2570
(A)
Eur. J. Immunol. 1991.22: 2565-2573
Responder Cells/Well 0.0
2.OE5
1 .OE5
3.OE5
4.OE5
- 400 -
microcultures;allo-specificresponder wells are defined in Sect. 2.4. The number of responder cells giving37% non-respondingcultures corresponds to reciprocal CTL precursor frequency for a given cell population: 1/40000 (Tg) or 1/280000 (Con) for the experiment shown.Thymocytepreparations originallycontained 10.2% (Tg) or 4.1% (Con) CD4-CD8+ cells. (B) Summary of data from three separate LD CTL precursor assays. CTL precursor frequencies in Tg and Con cell populations were normalized to proportions of CD4-CDSf cells originally present, the percent of control responses for Tg populations was calculated, and the results of different experiments were averaged. ( ) Tg thymus / Con thymus, ( 0 )Tg thymus / Con spleen. Errors bars are 1 SD.
0
22
$3 05
300-
U+
w o
L=ae
6+-
:2 0
200-
0
2 ' a
______ __-________
100
I
20 U Q k
*
n\\\\\\\\\\\y
0-7
these cells to exit the bloodstream and enter lymphoid organs [39,40]. Since phenotypically matureT lymphocytes accumulate in the thymuses of Zck-PT mice, it seemed plausible that the mechanisms regulating thymocyte emigration and homing of peripheral lymphocytes might be related or identical. We addressed this hypothesis by examining the in vivo homing capabilities of Ick-PT thymo-
cytes. Equal numbers of control or transgenic thymocytes (all genetically Thy-1.2) were injected into the tail veins of irradiated BL6.PL-Thy-1.1 congenic recipients. After 24 h, organs of recipient mice were removed and assayed for cells of donor origin by FCM after staining with antibodies to Thy-1.2. For one experiment, Thy-1.2+ cells were also analyzed for CD4 and CD8 expression by three-color FCM.
Table 1. I n vivo homing of lck-PT thymocytesa)
Esp. 1 I-
Donor
Mock Id-PI' Control Mock Ilk-PT Control Mock Ick-PT
%,
Yo
organ
Thy- 1.2 ' ccllsh'
Thy-].2'
I hynius
I .
0 0.1
0.I Splccn
0 1.5.0 10. I
Mcscnteric LN
0 0.4 4.4
1nguin;il LN
0 0.4 6.6
BlOOd
kk-PT Con t rol
Mock k.k - PT Control
r-
Rccipicnt
Control
Mock Id-PT Control Mock
,
0
33. I 6.8
h:idnc y ntl
cells 0 0. I 0. I
0.2 36.5 10.5 0.4 2.0 6.4
-
Exp. 2 1
% Thy-1.2' CD4+CD8 cells
K, Thy-1.2' CDJ-CDS+ cclls
nd
nd
23.9 6.4
8.9 2.9
I .2 4.4
0.4 1.6
0.9
0.5
7.4
5.5
0.3 1.7
45.7 0.4
22.4 0.2
I .9
0.7 0
0
0.2 72.9 I .2 0.2 3.7 0.9
0.I
a) Thy-1.2+ thymocytes from control or lck-PT animals were injected into the tail veins of irradiated Thy-1.1 congenic recipients. After 24 h, recipient organs were assayed for donor cells by FCM (see Sect. 2.2 for details). b) The percentage of lymphoid cells in the recipient organ bearing lymphoid markers.
Eur. J. Immunol. 1991. 21: 2565-2573
Pertussis toxin blocks thymocyte emigration
2571
Results shown in Table 1 indicate that a small number of receptors might explain some features of the Zck-PT control thymocytes migrate from the peripheral blood into phenotype. Results of FCM analyses revealed, however, mesenteric and inguinal LN. Typically, 4%-8% of inguinal that SP lck-PT thymocytes expressLFA-1, CD2 and CD4 at or mesenteric LN lymphocytes from animals injected with levels indistinguishable from those of normal SP thymocontrol thymocytes expressed the donor Thy-1.2 marker cytes and normal splenicT lymphocytes (data not shown). after 24 h, representing approximately 0.1% of injected Furthermore, 24 h after stimulation with immobilized mAb cells. The fact that lymphoid cells from the kidneys or to CD3, both control and transgenic SP thymocytes exthymuses of these animals contained
