Differential Expression of Integrins on Human Thymocyte Subpopulations
By Christopher F. Mojcik, Daniel R. Salomon, Andrew C. Chang, and Ethan M. Shevach
lntegrins represent a candidate group of cell surface receptors that maycontrol the homing and population of the thymus by T-cell precursors and the subsequent migration of developing thymocytes through the thymic architecture. We have used multiparameter flow cytometric methods to characterize the expression of several members of the integrin family (a3pl. a4pl. a5pl. a6p1, and aLp2) on thymocyte subpopulations and have correlated integrin expression with other well-defined thymocyte differentiation markers. a4p1 was expressed by all thymocytes, but expression was highest on CD4-CD8- double-negative (DN) cells, high on CD'CD8+ double-positive (DP) cells, and lowest on mature single-positive (SP) cells. a3p1, a5pl. and a6p1 were present on 13%, 63%. and 26% of thymocytes, respectively, with maximal levels of expression on DN and SP cells, and low levels of expression on DP cells. Simultaneous analysis of a4p1, a5p1,and CD3 expression suggested a pathway of Tcell differentiation in the thymus in which the majority of the DN cells were a4plh'a5plh', the DP cells a4plh'a5p1'"~-, and the most mature SP cells were a4pl'oa5pl'nt. The stagespecific expression of integrins strongly implies their functional involvement during T-cell maturation in the thymus. This is a US government work. There are no restrictions on its use.
HE ADVENT OF multiparameter flow cytometry has facilitated the dissection of the maturational process of T lymphocytes in the thymus. A variety of phenotypic markers has been described whose differential expression permits the identification of discrete stages during murine or human thymocyte development.'" Thus, in humans, the most immature T-cell precursors enter the thymus as CD4-CD8-CD3- (triple-negative [TN]) cells,',4initially express CD4,4-6then CD8, and as they begin to rearrange their T-cell receptor (TCR) genes and upregulate CD3Y become CD4'CD8+CD3'"'"' (double-positive [DP]) thymocytes. After positive and negative selection, the surviving cells continue to increase their levels of CD3 expression, and downregulate either CD8 or CD4 to become CD4+CDS-CD3*'or CD4-CD8CCD3h'(single-positive [SP]) thymocytes, respectively,1.4.7-1 I These mature cells are then ready to leave the

thy mu^.'.^

A variety of secondary markers, including CDI, CDS, CD7, and more recently, CD28, CD34, and CD69, have enhanced our understanding of this developmental scheme. CD73,'2-'4 and CD3412,14-16are markers of hematopoietic precursors in fetal liver and bone marrow (BM), and the most immature (TN, DN) thymocytes. InBMCD3'CD7+ cells include T, B, natural killer (NK), andmyeloid potential cells." Whereas thymocyte CD34 expression is virtually lost by the DP stage,I4.l6 CD7 expression remains stable throughout thymocyte development, and increases slightly on CD3hi cell^.'^,'^ CD1 is primarily a marker of cortical DP cells,""'* CDS3,'4,'8 CD2819.20 present on the majority of thymoand are
From the Cellular Immunology Section, Laboratory of Immunology, kboratory of Molecular Structure, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD: and the Department of Molecular andExperimental Medicine, Scripps Clinic and Research Foundation, La Jolla, CA. Submitted April 26, 1995; accepted July 19, 1995. Address reprint requests to Ethan M. Shevach, MD, Laboratory of Immunology. NIAID, NIH, Bldg IO, Room llN315, 10 Center DrMSC 1892, Bethesda, MD 20892-1892. The publication costsof this article were defrayed part by page in charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. section 1734 solely to indicate rhis fact. This is a US govenunent work. There are no restrictions on its use. 0006-4971/95/8611-0036$0.00/0

MATERIALS AND METHODS

Cell prepararion. Thymus fragments were obtained during elective cardiac surgery for valvular malformations on children aged 3 days to 7 years. Thymi were minced and pressed through a 60-mesh stainless steel screen. Thymocytes were then separated by Ficolldiatrizoate gradient (Organon Teknika, Durham, NC). Antibodies. The following antibodies were used Leu-5b antiCD2 phycoerythrin (PE; Becton Dickinson, San Jose, CA), OKT3 anti-CD3 (American Type Culture Collection [ATCC], Rockville, MD), HIT 3a anti-CD3 CyChrome (Pharmingen, San Diego, CA), 13B8.2 anti-CD4 PE (Amac, Inc, Westbrook, ME), S3.5 anti-CD4 biotin (Caltag Laboratories, South San Francisco, CA), BLla antiCD5 fluorescein isothiocyanate (FITC; Amac, Inc), 8H8 anti-CD7 FITC (Amac, Inc), B9.11 anti-CD8 FITC or PE (Amac, Inc), Leu 28 anti-CD28 PE (Becton Dickinson), HPCA-2 anti-CD34 PE (Becton Dickinson), Leu 23 anti-CD69 unconjugated or FITC (Becton Dickinson), P1B5 anti-a3/31 (Life Technologies/GIBCO BRL, Grand Island, NY), HP2/1 anti-a4pl (Amac, Inc), L25.3 anti-a4pl PE (Becton Dickinson), SAM1 anti-a5pl (Amac, Inc), GoH3 anti-a6pl (Amac, Inc), 25.3 anti-aLp2 (Amac, Inc), WT31 anti-TCRap (Becton Dickinson), goat F(ab'), antimouse IgG and light chains (human Ig adsorbed) FITC or PE (TAGOfBiosource,Camarillo, CA), preimmune goat F(ab')z FITC or PE (TAGO/Biosource), mouse IgGl and IgG2 isotype controls (Amac, Inc), SA-Tricolor (Caltag Laboratories), and purified mouse IgG (Sigma Chemical CO, Louis, MO). St Act-l MoAb was a kind gift of Dr Steven Shaw (National Cancer Institute). Cytojuoromerric analysis. Freshly prepared thymocytes ( IO6 cells) in Hanks' Balanced Salt Solution (HBSS)/3% fetal calf serum (FCS)/O.1% NaN3 (sorter buffer) were incubated with 2 pg purified MoAb per sample for 20 minutes at 4C. After washing, cells were incubated with 50 pL per tube of a 1:150 dilution of PE- or FITCconjugated goat F(ab')* antimouse IgG for 20 minutes at 4C. Cells were again washed, then unbound secondary antibody sites were blocked with 20 pg per tube of purified mouse IgG for 20 minutes. Without washing, 2 kg FITC- or PE- (or biotin-) conjugated MoAbs were added for 20 minutes at 4C. When biotin-conjugated MoAbs were used, cells were washed, then incubated for 20 minutes at 4C with SA-Tricolor as a final step. Cells were then washed and resuspended in sorter buffer for analysis. At least two isotypematched control antibodies were used. Flow cytometry wasperformed on a fluorescence-activated cell sorter (FACScan; Becton Dickinson) and analyzed with Lysis I1 software (Hewlett Packard, Palo Alto, CA). Analysis involved FlTC and PE, with Tricolor (excited at 488 nm) used as a third color. Forty thousand cells were collected per sample for analysis. Compensation and quadrant parameters were set with FITC/PE/Tricolor positive and negative controls; nonviable cells were excluded by light scatter and propidium iodide (using an additional gate at the extreme range of FL3).

RESULTS

Expression of a401 on thymocyte subpopulations. As previously 98.3% -+ 1.6% of human thymo-
cytes expressed a 4 p l with a relatively homogeneous level of expression as a single peak on fluorescence histogram (Fig 1, top panel). Human postnatal thymocytes can be easily divided into four (DN3.3% 2 1.0%, DP 78.7% C 4.9%, CD4SP 11.8% 5 2.9%, CD8SP 6.0% k 1.9% [n = 221) major subpopulations and further analysis of the expression of a 4 p l on these subpopulations using multiparameter staining uncovered an interesting heterogeneity (Fig 1, bottom panels). The a 4 p l - fraction contained predominantly DN cells, and of the total DN cells, 58% were a4p1- (Fig 1, Rl). However, further analysis of these a4p1- cells showed that the majority were also CD7- and CD3- (Fig 2 and data not shown). Thus, they are not T-lineage cells and will be excluded from further analysis. Additionally, approximately 12.5% of the DNs were y6 TCR+. The a4p1'" population (Fig 1, R2; 45% to 50% of a4pl' cells) was depleted of DN cells (14.1% of total) and contained roughly half (47.0%) of the total DP cells; surprisingly, this population contained virtually all the CD8+SP cells (86.4%) and 60% of the CD4+SP thymocytes. These CD4'SP cells were almost exclusively CD4h'. In contrast, the a 4 p l h ' subset (Fig 1, R3) was enriched for (immature) CD4'" SP cells; CD8+SP and CD4hiSP (mature) thymocytes were virtually absent. Indeed, the brightest 5% of a 4 p l + cells (R4 in Fig 1) were 81% DP and contained 15.7% CD4'" SP cells. Two-color immunofluorescence analysis using other known thymocyte differentiation markers was usedto further characterize the a 4 p l h i and a4pl'" subsets (Fig 2). The a 4 P l - fraction is negative for all T-lymphoid markers. The cy4Pl'O subset contained the majority of CXBTCR~ (89.1% of the &TCRhi cells were a4Pll0), CD3h' (92.7%), CDShi (74.9%), CD28h' (80.0%), CD69hi(93.3%) thymocytes (Fig 2, a, b, e-g). The a4pll" cells were also CD34- (only 1.9% of a4Pl'" thymocytes were CD34+), and CD1ai""Lo(Fig 2, c and d). On the other hand, the a4/31h' subset contained phenotypically less mature thymocytes, specifically a& TCR1O-lnt 0.i"' , CD5'", CD28lo, and CD69". This subset CD3' also contained 85.6% of the CD34+ cells, and the majority of CDlahi cells. Indeed, as levels of TCR, CD3, CD5, CD28, and CD69 increased, the level of a 4 p l tended to decrease. Cells bearing the y6 TCR, CD2 and CD7 (data not shown), were equally divided between a4,81'" and a 4 p l h i subsets. Preliminary data indicate that of the a 4 p l + fraction, 10% are y6 TCR+,which represents approximately one third of all the y6 TCR+ cells in the thymus (data not shown). Taken together, these studies demonstrate that the level of expression of a4p1 varies markedly during different stages of thymocyte development with the highest levels of expression seen on the CD4" SP cells and on the least mature DP populations. Expression o a5PI on thymocyte subpopulations. a 5 p l f was present on 63.2% 2 14.7% of thymocytes with a relatively heterogeneous level of expression as indicated by a broad peak on the histogram (Fig 3, top panel). When the a 5 p l + population was split (at the channel of peak fluorescence) into a5@lhi a5/3li0 subpopulations, we also oband served a correlation between the levels of expression of a 5 p l and CD4/CD8 (Fig 3, bottom panels). The a5Pl'"

MOJCIK ET AL

LW Fhoresceace
Fig 1. Flow cytometric analysis of a4gl expression on human thymocytes. Three-color immunofluorescence staining was performed as I detailed in Materialsand Methods. Forty thousand cells were collected per sample on a and analyzed FACScen, with Lysis I software. Nonviable cells were excluded by light scatter and propidium iodide gating. Compensation and quadrant parameters were set with FITC/PE/Tricolor positive and negative controls. (Top panel) Single-color analysisa4g1 with of negative control overlaid. Gates(Rl-R4)set as indicated. Positive staining chosen based on negativecontrol. Peak channel of immunofluorescence wasarbitrarily chosen to demarcate the hi and lo staining fractions. (Bottom panels) Analysis of CD4 and CD8 staining of m4g1 stained, gated populations. Within each quadrant, upper set of numbers the represent the percentage of gated population within the quadrant, whereas the lower set of numbers (parentheses) represent percentage of totel thymocytes found in that quadrant. These percentages were derivedfrom the following calculation: (% of Thymocytesin Quadrant x % of Total Cells in Gated Population)/(%of Total Thymocytes in Quadrant.)
population (Fig 3, R2) was predominantly DP and contained few DN (6.1% of total DN), CD4+SP (6.9% of total CD4+SP), or CDX'SP (14.4% of total CD8+SP) thymocytes. The a5/31h' subset (Fig 3, R3; 50% of total cells) was enriched for CD4+SP (86.4% of total) and CDX'SP (58.5% of total), and contained 39.6% of the DN cells; however, the brightest 5% of a 5 p l + thymocytes were markedly enriched for CD4"SP cells and DN cells (Fig 3, R4). We next compared the level of expression of a5pl with the other markers that define thymocyte differentiation. The relative fluorescence intensity of a 5 p l staining tended to be highest in immature (aPTCR'O)and mature (apTCRh') compartments, and lowest in the intermediate (aPTCR"') subset (Fig 4a). The a5p l hi brightest subpopulation was enriched for cells with immature phenotypes (Figs 4, a, b, df): a/?TCR'" (75.3% of the brightest a5plh' were a/?TCR'"), CD3'" (75.2%), CD28'" (77.9%), CD69" (48.3%), and CD34hi(58.0%). In fact, as cells lose CD34, there is a concomitant decrease in a5@1(compare with CD34 on a 4 p l h ' thymocytes, in which levels remain constant [Fig 2dl). A similar relationship was observed for CD7 as the brightest a 5 p l + cells were also CD7h' (Fig 4c). Cells bearing the yGTCR were virtually all a5/?lh',with about one third in the a 5 p l h ' brightest 5% (not shown). When we analyzed thymocytes for the simultaneous expression of a 4 p l and a501 (Fig 5A), three distinct subpopu-
lations could be identified: a4pl'" a5pl'", a4plh' a5Pli0/~, and a4plh' a5pl When the level of expression of CD3 was correlated with levels of expression of a 4 p I l d p l (Fig 5B), the relationships between the relative levels of expression of these integrins, the CD3 complex, and the program of T-cell differentiation in the thymus weremore clearly defined. The a4pl'" a5pl'" cells (Fig 5A, R1) contained predominantly CD3'"' (36%) and CD3h' (55%) thymocytes (Fig 5B, bottom panel), a4plh' a5fll'"- cells (Fig 5A, R2) contained equivalent numbers of CD3" (45%) and CD3"'' (49%) thymocytes (Fig 5B, bottom panel), while the a 4 p 1h' a5plh' cells (Fig 5A, R3) were predominantly CD3- cells (84%, Fig 5B, bottom panel). The CD3 staining profile of whole thymus is shown in Fig 5B (top panel) for comparison purposes. These data suggest a pathway of T-cell differentiation inthe thymus in whichthe majority of the precursor cells are a4plhi a5plh' (R3), the DP population a4plh' a5pl'"'(R2), and the most mature SP cells a4pl'" a5P1'"' (RI). a3pl and a601 expression. Both a 3 p l and a 6 p l were expressed on low percentages of human thymocytes (12.8% ? 5.7% [Fig 6A, top panel] and 26.0% ? 11.0% [Fig 6B, top panel], respectively). Both the a 3 p l - and the a601 subsets (Fig 6, A and B, bottom panels, R1) were enriched in CD4SP cells (63.5% and 64.8%of total, respectively) and DN cells (70.4% and 71.1% of total, respectively). When the levels of expression of a 3 p l and a 6 p l were correlated

INTEGRINS ON HUMAN THYMOCYTES
Fig 2. Two-color immunofluorescence analysisO f human thymocytes stained for a4/31 and counterstained with apTCR, CD3, CDla, CD34, CD5. CD28. or CD69, as indicated. Regions demarcating hilintllo e staining were s t based on single-color profile for each individual marker. Mean f SD for each marker is as follows: Cola 75.1 +- 9.6 (n = 9); CD3 hi 21.3 +6.5, int 40.4 & 7.3 (n = 20); CD5 99.0 f 0.5 (n = 11); CD798.0.7 (n = 11); CD28 23.9 f 4.1 In = 11); CD34 5.2.5 (n = 23); CD69 26.5 f 4.6 In = 20); a/3TCR hi 21.1 f 3.3, int 39.0 +- 3.8 (n = 15); 9TCR 1.5 f 0.7 (n = 12).

Log Fluorescence

with the levels of expression of the a@TCR,CD3, and CD34, the highest levels of expression were present on CD3'" and CD3h' thymocytes, with lower levels present on the CD3'"' subset (Fig 7, B and E). The majority of the CD34h' cells were a6Plh',but a3Pl'" (Fig 7C and F). &LP2 integrinexpression in human thymocytesubsets. As previously reported,- almost all human thymocytes expressed aLP2 (99.0% +- 0.6%) with a relatively homogenous level of expression (Fig 8A, top panel). aLP2- (Fig SA, R1) cells were virtually (92.9%) all DN, but almost half (45.9%) of the DN population was a L p 2 ~ 8A, R3). DP thymo(Fig cytes were slightly enriched in the ( Y L P ~ ~ " population (Fig 8A, R2), CD4'SP enriched in the C K L P subset (Fig 8A, ~~'
R3), and CD8+SP cells were equally divided between the two. Expression of aLP2 was moderate (ie, in the lower half of the ~yLp2~' subset) on CD3Cbearing cells (Fig 8B, panel c, R2), and the level of expression tended to increase with thymocyte maturation. Thus, @TCRh',CD3h',CDSh' (not shown), CD7h'(not shown), CD2Sh',and CD69h'thymocytes tended to be ~yLp2~' 8B, panels a, b, d, e). (Fig

DISCUSSION

We have analyzed the expression of a number ofmembers (a3p1, a4p1,aSPl, a601, and aLP2) of the integrin superfamily on human thymocytes and have correlated their expression with the expression of a number of other membrane

Log Plnoreseence

>and SP stages. As DP T cells with active a4Pl expressed almost the same level of 0401 as DP T cells with inactive a4P1, integrin function appears to switch off before downregulation of integrin expression. On peripheral T cells, the converse appears to be true as Shimizu et a136have shown that integrin activation occurs before upregulation of expression. The a4 chain can alternatively pair with the 07 chain.34 Although the HP211 MoAb does not discriminate between a4Pl and a407,4 the Act-l MoAb is specific for cy4P7. Previous reports4 indicate that unstimulated human thymus is negative for a4P7, and in our hands only 3.3% 5 1. l % (n = 3) of thymocytes stained with this antibody. Therefore, we believe that HP2/1 almost exclusively recognizes cy401 in the thymus. Although 05Pl is also expressed by the majority of thymocytes (60% to 70%), its pattern of expression is quite distinct from that of ( ~ 1Although both of these integrins. were expressed at high levels on DN T cells, the level of a501 decreased on DP cells, yet was again highon SP cells. Almost all of the constitutive capacity of thymocytes to adhere to FN appeared to be mediated by a4pl because anti-05p1 MoAbs had no effect on adhesi0n.2~In contrast, a5P1, but not a4/3l, functioned as a costimulatory molecule for FN and anti-CD3 induced T-cell proliferation of SP and DP/transitional thymocytes. Because cu5pl acquired the capacity to deliver a costimulatory signal at the putative time of positive studies are currently underway to determine whether this property of a5pl is a prerequisite to or a consequence of thymocyte selection. The patterns of expression of 0301 and a6Pl integrins on thymocyte subsets were quite similar to a5Pl even

Fig 3. Flow cytometric analysis of a5p1 expression on human thymocytes. (Top panel) Single-color analysis &@l.Gates (Rl-R41 set as of indicated. Note that region R4 represents a subset of region R3. (Bottom panels) Analysis of C M and CD8 staining of &@l stained, gated populations. Percentagesof cells within each quadrant were calculated as described in the legend to Fig 1.
antigens that have previously been used to define pathways of T-cell maturation within the thymus in humans and experimental animals. Because integrins have been shown both to control the homing and migration of cells as well as to function as costimulatory molecules for the activation of peripheral T lymphocytes and thymocytes, the regulation of their expression on various thymocyte subsets should provide important insights into their possible functional role at different stages of thymocyte maturation. The control of integrin function may exist at several different levels including selective expression of an integrin on a thymocyte subset(s) as well as quantitative differences in the level of expression of integrins that are widely expressed on all thymocytes. In addition, integrins may exist in active and inactive states in terms of their ability to bind their target ligands and we have previously shownz4that a major subset of thymocytes expresses cy401 in a constitutively active form. By the use of multiparameter flow cytometric techniques, we have obtained clear evidence for differential expression of most of the integrins studied (Table 1). a4Pl is expressed on almost all thymocytes (98%), yet its pattern of thymocyte expression appears to be unique. Its level of expression was highest on DN thymocytes, decreased slightly on DP T cells, and was 10-fold lower on SP cells. The high level of expression of a401 on DN and DP cells appears to correlate with its functional activity as Leavesley et a145have demonstrated that a40 1, but not aL/% or platelet endothelial cell adhesion molecule-l (PECA l), was constitutively active on CD34+ progenitor BM cells and mediated adhesion to VCAh4-1 and we have previously shownz5that a4Pl is constitutively activated on thymocytes at least by the DP stage, but is inactive at the transitional
Fig 4. Two-color immunofluorescence analysis of human thymocytes stained for cr5/31 andcounterstained with apTCR, CD3. CD7. CD34 CD5, CD28. or CDW, as indicated.
though they were only expressed on a minority of thymocytes (10% to 30%). Their level of expression was highest on the least mature (DN) thymocytes, decreased on DP cells, and increased (though not to level seen on DN cells) on SP cells. Wadsworth et a146have shown that murine a6 paired with either the p1 or p4 chains, but in the thymus preferentially associated with p4, and that the expression of a6p4 was restricted to murine DN thymocytes. Because the MoAb used in our studies was a6-specific. we have not yet formally studied the &chain pairing of a 6 on human thymocytes; however, preliminary data with the a604 specific MoAb S3/ 41 shows no staining of thymocytes. We have previously

shown that 10% to 15% of thymocytes adhered to laminin in the absence of stimulation, whereas 39% to41% of thymocytes bound laminin after activation with phorbol myristate acetate (PMA) or MnZ+, have not yet determined if the We cells that express constitutive laminin binding activity are composed of a unique subset of cells that express a 1 and/ or a 1. Both laminin or merosin will function as costimulatory molecules for thymocyte proliferation in the presence of plate-bound anti-CD3, and this costimulatory function can be inhibited by either anti-a3@1 or anti-a6@1 9! The pattern of expression of a.501,a3p1, and a601, with high levels on the least and most mature populations and

Log Fluorcsecnce

mt LI "I
Fig 5. Three-color immunofluorescence analysis of human thymocytes stained with a4@1 (PE), a 5 p l (FITC), and CD3 (CyChrome). (A) Analysis of a4p1 versusa5pl. Gateparame(B) ters set as indicated. Analysis of CD3 staining ungated (total of thymus; top panel) or gated, a4plla5Bl stained thymocytas (bottom panels).Numbersrepresent percentage of total thymocytes (top panel) or gated population (bottom panels) within each region.
intermediate levels on the DP cells, is similar to only one of the classical thymocyte differentiation markers, CD7, a member of the Ig gene superfamily whose structure is similar to murine Thy-l. CD7 is expressed at high levels on CD34+ DN thymocytes:' intermediate on DP cells, and high on SP cells (data not shown). Anti-CD7 MoAbs can costimulate proliferation of peripheral T cells in the presence of antiCD3 or PMA''*'* or can also be directly mitogenic for T ~ e l l sor NK, cells5' after cross-linking. Interestingly, one ~ ~ ~ ~ consequence of anti-CD7-induced activation is integrin-mediated adhesion to FN, VCAM-l, or I C A " 1 without upregdation of integrin expre~sion.~~~'' Anti-CD7-mediated integrin activation appears to be mediated by a mechanism distinct from that produced by anti-CD3 or PMA stimulat i ~ n. Because CD7 may function as an adhesion molecule,'* '~ it is quite possible that interaction of CD7 with its physio-
logic ligand may play a role in the activation of integrin function during T-cell development in the thymus. aLP2 was expressed at low-intermediate levels onDN and DP cells and at higher levels on SP cells. This pattern of expression is one common to several of the classical thymocyte differentiation markers including CD5, CD28, and CD69. It is likely that aLP2 interactions with ICAM-1 (CD54) may play critical roles at multiple steps during the process of thymocyte differentiation, as Singer et a157-59 have shown that both the binding of thymocytes to thymic epithelium (TE) as well as TE-dependent, phytohemagglutinininduced, SP-thymocyte activation were inhibited by MoAbs to aLP2 or I C A " 1. Similarly, addition of MoAbs to aLP2 or ICAM-1 to mouse thymic organ cultures resulted in arrest of T-cell development at the DN stage, which suggests that aLp2 has a developmentally restricted role at an early stage

~ o g Fluorescence

' F 3 )

/a6pl-

77.83 (85.66)
~ (35.17). ,~ ~ p ~ + j26 56.06' (14.34
Fig 6. Multiparameter analysis of human thymocytes stained with a3pl or d p l. (A) Top panel, sinset as indigle-color analysis of a381. Gates (R") f cated. Numbers represent percentage o total thymocytes within gated region. Bottom panel, analysis of C M and CD8 staining of a3p1 stained, gated populations. Percentages of cells were determined as described in the legend t o Fig 1. (B) l o p panel, singlecolor analysis of a681. Gates IRl-RZ) set as indicated. Bottom panel, analysis of CD4 and CD8 staining of stained, gated populations.
of development. Murphy etalM have raised the possibility that in patients with Down syndrome overexpression of aLP2 or ICA"1 may lead to inappropriate thymocyte developmental interactions with TE, enhanced deletionof cortical thymocytes, and decreased numbers of peripheral T cells. For the most part, our results agree with earlier, less comprehensive studies of integrin expression by human thymoCytes.30.35.42.43 Some of the minor differences between these studies may relate to the specific MoAbs used, to staining techniques, or to our use of fresh preparations of thymocytes in contrast to cells that had been fixedafter staining or stored at 4C for 18 hours before staining. Our results on the expression of a 4 P l on human thymocytes appear to be quite distinct from studies of the expression of a401 on murine thym o c y t e ~ in ~. ~ ~ the expression of n4Pl was highest ~ which on DN thymocytes, low on DP, and very low on SP. However, we have recently used two other MoAbs to reexamine the expression of a401 on murine thymocyte subpopulations and found that its pattern of expression was virtually identical to that seen on human thymocytes (Mojcik C, Shevach EM, in preparation). In addition, as in humans, the major subpopulation that expresses a4P1 in a constitutively active form was enriched for immature DP cells rather than the DN cells as previously Studies of the differentiation of human thymocytes in vivo after injection of progenitor cells (CD34+, CD3-, CD4-, CD8-) into allogeneic thymus xenografts in severe combined immunodeficient (SCID-hu) micei5 or in vitro after injection of progenitors into fetal thymic organ cultures (FTOC) have shown the sequential appearance of CD3-CD4+CD8-, CD3-CD4'CD8+, and CD3'CD4+CD8' cells, with the later appearance of mature CD3+CD4'CD8and CD3'CD4-CD8+ populations.'7 The expression and function of several of the integrins also appears to vary

Fig 7. Two-color irnmunofluorescence analysis of human thymocytes stained for a 3 p l or a6pl and counterstained with (AI apTCR, (B) CD3, or (Cl CD34.
during this development process and allows us to propose possible functional involvement of these molecules during the different stages of thymocyte development. If one confines the analysis to a comparison of the absolute levels of expression of the various integrins, it superficially appears that the majority of the earliest precursors (DN population, Figs 1 and 3, R1, bottom panels) do not express either a4P 1 or (~501; however, the majority of these DN cells are neither T cells nor precursors of the T-cell lineage. On the other hand, greater than 95% of the CD34h' population expressed high levels of a 4 P l and a5Pl (Figs 2 and 4, and data not shown) and almost all of these cells were CD3-CD4-CD8-. Thus, one can conclude that the earliest T-cell precursors in the thymus probably expressed these integrins upon entry to the thymus and that these integrins may even play a role in the homing of these cells from the BM to the thymus. The expression of high levels of a4Pl (Fig 1, bottom, R3) or a501 (Fig 3, bottom, R4) is maintained on the CD3-CD4I0CD8- population, but while the level of a4Pl is maintained, the level of a5Pl is markedly downregulated on the DP population (Fig 3, bottom, Rl). In contrast, the expression of a 4 P l is downregulated whereas the expression of d B 1 is upregulated upon maturation of the DP cells to CD3'CD4'CD8- or CD3+CD4-CD8+ cells. Thus, a4Pl interactions with FN or VCAM-1 may modulate the transition from CD34+CD3-CD4-CD8- cells to CD3-CD4TD8- cells as well as the transition of the CD3-CD4+CD8- cells to CD3'CD4+CD8+ cells. It is quite possible that a 5 P l may also be involved in some of the early steps of this process or that the downregulation of a 5 p l expression is required for maturation to DP stage. Similarly, as the constitutive activity of a 4 P l is downregulated, while the expression of a5Pl is upregulated, at some point in the transition from DP to SP, integrin interactions

INTEGRINS

ON HUMAN THYMOCYTES

Log Pborescence

Fig 8. Multiparameter analyhuman of sis thymocytes stained with al432. (A) Top panel, singlecolor analysis of -2. Gates (Rl-R3) set as indicated.Numbersrepresentpercentage of total thymocytes within gated region. Bottom panel, analysis of CD4 and CD8 staining of aw2 stained, gated populations. Percentages cells of wara determined as described in the legendto Fig 1. (B) Two-color immunofluorescence analysis of human thymocytes stained for aLf32 and wuntarstained with or aSTCR, CD34. CD3, CD28. CD69 as indicated.

Table 1. Levels of Differentiation on Human Thymocytes
Antigen Formative DN Formative CD4"SP

Transitional Cells

CD34 CD7 CD5 CDla CD3hflCR CD28 CD69 a301 a4p1 a5pl aff LP2

++ ++++ +++

+ +++ +++

+++ +++ ++

+++ ++

++++ ++

with their target ligands may also play a role at later stages of thymocyte maturation. The ability to monitor the development of murine3' or human16thymocytes from precursor cells in vitro in organ cultures should facilitate the evaluation of the function of these integrins at each development step by blocking their interaction with their ligands by the addition of inhibitory MoAbs or peptides. We are currently undertaking R O C and SCID-hu studies to address these issues.
ACKNOWLEDGMENT We thank Drs B.F. Akl and E.A. Lefrak, nurses L. Smith and C. Hammond of Virginia Heart Surgery Associates (Fairfax), and the OR teams of Fairfax Hospital (Fairfax, VA) for their assistance in obtaining the pediatric thymic tissues. REFERENCES 1. Rothenberg EV: The development of functionally responsive T cells. Adv Immunol 51:85, 1992

++++ ++ +++

++++ +
Symbols reflect variation in intensity of staining.
2. Reinherz EL, Schlossman SF: The differentiation and function of human T lymphocytes. Cell 19321, 1980 3. Lanier LL, Allison JP, Phillips JH: Correlation of cell surface antigen expression on human thymocytes by multi-color flow cytometric analysis: Implications for differentiation. J Immunol 137:2501, 1986 4. Boyd RL, Hugo P: Towards an integrated viewof thymopoiesis. Immunol Today 12:71, 1991 5. Kraft DL, Weissman IL, Waller EK: Differentiation of CD3-4-8- human fetal thymocytes in vivo: Characterization of a CD3-4+8- intermediate. J Exp Med 178:265, 1993 6. Takeuchi Y, Fujii Y, Okumura M, Inada K, Nakahara K, Matsuda H: Characterization of CD4+ single positive cells that lack CD3 in the human thymus. Cell Immunol 151:481, 1993 7. Shortman K, Vremec D, Egerton M: The kinetics of T cell antigen receptor expression by subgroups of CD4+8+ thymocytes: Delineation of CD4+8+32+ thymocytes as post-selection intermediates leading to mature T cells. J Exp Med 173:323, 1991 8. Benoist C, Mathis D: Positive selection of the T cell repertoire: Where and when does it occur? Cell 58:1027, 1989 9. Guidos CJ, Danska JS, Fathman CG, Weissman IL: T cell receptor-mediated negative selection of autoreactive T lymphocyte precursors occurs after commitment to the CD4 or CD8 lineages. J Exp Med 1722335, 1990 10. Robey EA, Fowlkes BJ, Gordon JW, Kioussis D, von Boehmer H, Ramsdell F, Axel R: Thymic selection in CD8 transgenic mice supports an instructive model for commitment to a CD4 or CD8 lineage. Cell 64:99, 1. Maradon G, Rocha B: Generation of mature T cell populations inthe thymus: CD4 or CD8 down-regulation occurs at different stages of thymocyte differentiation. Eur J Immunol 24:196, 1994 12. Patel DD, Haynes B F Cell adhesion molecules involved in intrathymic T cell development. Semin Immunol 5:283, 1993 13. Campana D, Janossy G, Coustan-Smith E, Amlot PL, Tian W-T, Ip S, Wong L: The expression of T cell receptor-associated proteins during T cell ontogeny in man. J Immunol 142:57, 1989 Huang S, Picker LJ: Flow cytometric 14. Terstappen LW, assessment of human T-cell differentiation in thymus and bone marrow. Blood 79:666, 1992 15. Peault B, Weissman IL, BaumC, McCune JM, Tsukamoto A: Lymphoid reconstitution of the human fetal thymus in SCID mice with CD34+ precursor cells. J Exp Med 174:1283, 1991 16. Galy A, Verma S, Barcena A, Spits H: Precursors of CD3+CD4+CD8+ cells in the human thymus are defined by expression of CD34. Delineation of early events in human thymic development. J Exp Med 178:391, 1993 17. Spits H, Lanier LL, Philips JH: Development of human T and natural killer cells. Blood 85:2654, 1995 18. Merkenschlager M, Fisher AG: Human postnatal thymocytes generate phenotypically immature CD3di, CDSdim, abright CD1 P%eny in organ culture. J Immunol 148:1012, 1992 19. Turka LA, Linsley PS, Paine R 111, Schieven GL, Thompson CG, Ledbetter JA: Signal transduction via CD4, CD8, and CD28 in mature and immature thymocytes: Implications for thymic selection. J Immunol 146:1428, 1991 20. Turka LA, Ledbetter JA, Lee K, June CH, Thompson CB: CD28 is an inducible T cell surface antigen that transduces a proliferative signal in CD3 mature thymocytes. J Immunol 144:1646, 1990 21. Yamashita I, Nagata T, Tada T, Nakayama T: CD69 cell surface expression identifies developing thymocytes which audition for T cell antigen receptor-mediated positive selection. Int Immunol 5:1139, 1993 22. Testi R, Phillips JH, Lanier LL: Constitutive expression of a phosphorylated activation antigen (Leu 23) by CD3bighhuman thymocytes. J Immunol 41:2557, 1988

23. Jung LKL, Haynes BF, Nakamura S, Pahwa S, Fu SM: Expression of early activation antigen (CD69) during human thymic development. Clin Exp Immunol :466, I990 24. Testi R, DAmbrosio D, De Maria R, Santoni A: The CD69 receptor: A multipurpose cell-surface trigger for hematopoietic cells. Immunol Today 15:479, 1994 25. Salomon DR, Mojcik CF, Chang AC, Wadsworth S, Adam DH, Coligan JE, Shevach EM: Constitutive activation of integrin a&, defines a unique stage of human thymocyte development. J Exp Med 179:1573, 1994 26. Utsumi K, Sawada M, Narumiya S, Nagamine J, Sakata T, Iwagami S, Yasumichi K, Teraoka H, Hirano H, Ogata M, Hamaoka T, Fujiwara H: Adhesion of immature thymocytes to thymic stromal cells through fibronectin molecules and its significance for the induction of thymocyte differentiation. Proc Natl Acad Sci USA 88:5685, 1991 27. Sawada M, Nagamine J, Takeda K, Utsumi K, Kosugi A, Tatsumi Y, Hamaoka T, Miyake K, Nadajima K, Watanabe T, Sakakibara S, Fujiwara H: Expression of VLA-4 on thymocytes: Maturation stage-associated transition and its correlation with their capacity to adhere to thymic stromal cells. J Immunol 149:3517, 1992 28. Cardarelli PM, Crispe IN, Pierschbacher MD: Preferential expression of fibronectin receptors on immature thymocytes. J Cell Biol 106:2183, 1988 29. Cardarelli PM, Pierschbacher MD: T-lymphocyte differentiation and the extracellular matrix: Identification of a thymocyte subset that attaches specifically to fibronectin. Proc NatlAcad Sci USA 83:2647, 1986 30. Pilarski LM, Yacyshyn BR, Jensen GS, Pruski E, Pabst HF: PI integrin (CD29) expression on human postnatal T cell subsets defined by selective CD45 isoform expression. J Immunol 147:830, 1. Harvey JE, Jones DB: Distribution of LCA protein subspecies and the cellular adhesion molecules LFA- 1, ICAM-I and p150,95 within human foetal thymus. Immunology 70:203, 1990 32. Fine JS, Kruisbeek AM: The role of LFA-l/ICAM-I interactions during murine T lymphocyte development. J Immunol 147:2852,1991 33. Yong K, Khwaja A: Leucocyte cellular adhesion molecules. Blood Rev 4:211, 1990 34. HynesRO: Integrins: Versatility, modulation and signaling in cell adhesion. Cell 69:11, 1992 35. Hemler ME: VLA proteins in the integrin family: Structures, functions, and their role on leukocytes. Annu Rev Immunol 8:365, 1990 36. Shimizu Y, Van Seventer GA, Horgan KJ, Shaw S: Regulated expression and binding of three VLA (PI) integrin receptors on T cells. Nature 345:250, 1990 37. Miyake K, Weissman IL, Greenberger JS, Kincade PW: Evidence for a role of the integrin VLA-4 in lympho-hemopoiesis. J Exp Med 173:599, 1991 38. Ryan DH, Nuccie BL, Abboud CN, Winslow JM: Vascular cell adhesion molecule-l and the integrin VLA-4 mediate adhesion of human B cell precursors to cultured bone marrow adherent cells. J Clin Invest 88:995, 1991 39. Williams DA, Rios M, Stephwen C, Patel VP: Fibronectin and VLA-4 in haematopoietic stem cell-microenvironment interactions. Nature 352:438, 1991 40. Roldan E, Garcia-Pardo A, Brieva JA: VLA-4-fibronectin interaction is required for the terminal differentiation of human bone marrow cells capable of spontaneous and high rate immunoglobulin secretion. J Exp Med 175:1739, 1. Bellon T, Lopez-Rodriguez C, Rubio MA, Jochems G, Bernabeu C, Corbi AL: Regulated expression of p150,95 (CD1 IcICD18;

aWp2) and VLA-4 (CD49dCD29; a4/pl) integrins during myeloid cell differentiation. Eur J Immunol 24:41, 1994 42, Zutter MM: Imrnunolocalization of integrin receptors in normal lymphoid tissues. Blood 77:2231, 1991 43. Watt SM, Thomas JA, Edwards AJ, Murdoch SJ, Horton MA: Adhesion receptors are differentially expressed on developing thymocytes and epithelium in human thymus. Exp Hematol20:1101, 1992 44. Murphy M, Insoft RM, Pike-Nobile L, Derbin KS, Epstein LB: Overexpression of LFA-1 and ICAM-l in Down syndrome thymus: Implications for abnormal thymocyte maturation. J Immunol 1505696, 1993 45. Leavesley DI, Oliver JM, Swart BW, Berndt MC, Haylock DN, Simmons PJ: Signals from platelet/endothelial cell adhesion molecule enhance the adhesive activity of the very late antigen-4 integrin of human CD34' hemopoietic progenitor cells. J Immunol 153:4673, 1994 46. Wadsworth S , Halvorson HJ, Coligan JE: Developmentally regulated expression of the p4integrin on immature mouse thymocytes. J Immunol 149:421, 1992 47. Kamata T, h z o n W, Takada Y: Identification of putative ligand-binding sites of the integrin a4@1 (VLA-4, CD49dCD29). Biochem J 305:945, 1995 48. Lazarovits AI, Moscicki RA, Kurnick JT, Camerini D, Bhan AK, Baird LG, Erikson M, Colvin RB: Lymphocyte action antigens. I. A monoclonal antibody, anti-Act-l, defines a new late lymphocyte activation antigen. J Immunol 133:1857, 1984 49. Chang AC, Salomon DR, Wadsworth S, Hong M-J P, Mojcik and a integrins mediate lami& CF, Shevach EM, Coligan JE: nidmerosin binding and function as costimulatory molecules for human thymocyte proliferation. J Immunol 154:500, 1995 50. Schmitt C, Ktorza S , Sarun S, Blanc C, De Jong R, Debre P:
CD34-expressing human thymocyte precursors proliferate inresponse to interleukin-7 but have lost myeloid differentiation potential. Blood 82:3675, 1993 51. Camera AC, Rincon M, Sanchez-Madrid F, Lopez-Botet M, de Landazuri MO: Triggering of co-mitogenic signals in T cell proliferation by anti-LFA-l (CD18, CDlla), LFA-3, and CD7 monoclonal antibodies. J Immunol 141:1919, 1988 52. Ware RE, Haynes B F T cell CD7 mRNA expression is regulated by both transcriptional and post-transcriptional mechanisms. Int Immunol 5:179, 1993 53. Shimizu Y, van Seventer GA, Ennis E, Newman W, Horgan KJ, Shaw S : Crosslinking of the T cell-specific accessory molecules CD7 and CD28 modulates T cell adhesion. J Exp Med 175:577, 1992 54. Lazarovits AI, Osman N. Le Feuvre CE, Ley SC, Cmmpton MJ: CD7 is associated with CD3 and CD45 on human T cells. J Immunol 153:3956, 1994 55. Rabinowich H, Lin W-C, Herberman RB, Whiteside T L Signaling via CD7 molecules on human NK cells: Induction of tyrosine phosphorylation and integrin-mediated adhesion to fibronectin. J Immunol 153:3504, 1994 56. Chan ASH, Reynolds PS, Shimizu Y :Tyrosine kinase activity associated with the CD7 antigen: Correlation with regulation of T cell integrin function. Eur J Immunol 24:2602, 1994 57. Singer KH, Denning SM, Whichard LP, Haynes B F Thymocyte LFA-l and thymic epithelial cell ICA"1 molecules mediate binding of activated human thymocytes to thymic epithelial cells. J Immunol 144:2931, 1990 58. Singer KH, Le P , Denning SM, Whichard LP, Haynes BF: T The role of adhesion molecules in epithelial-T-cell interactions in thymus and skin. J Invest Dermatol 94:85S, 1990 H 59. Haynes BF, Denning SM, Le PT, Singer K : Human intrathymic T cell differentiation. Semin Immunol 2:67, 1990