Published January 1, 1991

Differentiation of T Cell Lymphokine Gene

Expression : The In Vitro Acquisition of T Cell Memory
By Stefan Ehlers and Kendall A. Smith

From the Department of Medicine, Dartmouth Medical School, Hanover,

New Hampshire 03756

Summary
A simple in vitro experimental system was devised to reflect the in vivo generation of a T cell
anamnestic response so that T cell differentiation could be examined at the level of lymphokine
gene expression . Comparison of neonatal and adult T cells revealed that both populations expressed
the genes for interleukin 2 (IL2) and its receptor, but only adult T cells were capable of transcribing
mRNAs for IL-3, IL-4, 11,5, 11,6, interferon y, and granulocyte/macrophage colony-stimulating
factor. However, neonatal T cells could be induced to undergo functional differentiation in vitro,
thereby acquiring the capacity to express the lymphokine gene repertoire characteristic for adult

Downloaded from jem.rupress.org on June 15, 2011

T cells . These data suggest that the T cells generated from neonatal blood by a primary stimulation
in vitro are functionally indistinguishable from the T cells in adult blood that presumably have
undergone primary stimulation in vivo. Therefore, we propose that the term "memory cell"
be applied to those T cells that can be identified by their differentiated state of inducible effector-
lymphokine gene expression.

M emory is the hallmark of an intact immune system,

required for immunity against invading microorganisms
By comparison, T cell memory has thus far eluded a simi-
larly distinctive molecular genetic analysis due to two fun-
and for the efficacy of vaccination in preventing infections. damental experimental problems. First, it has been impos-
Immunologic memory is recognized by the capacity of a pre- sible to unequivocally discriminate between the cells
viously immunized host to respond more rapidly and with responsible for the characteristic features of the naive and
greater intensity to a secondary antigenic challenge (1) . The memory immune responses. Second, the molecules involved
cellular basis for memory lies in the selective clonal expan- in T cell effector functions had not been identified until re-
sion of antigen-specific lymphocytes, a concept originally in- cently. Consequently, putative naive and memory T cell popu-
troduced by Burnet (2) . Therefore, when compared with a lations could not be prepared in sufficient quantity and pu-
naive host, a primed host responds to a secondary antigenic rity to allow the application of biochemical and molecular
exposure more effectively simply because more antigen-reactive genetic approaches that might delineate qualitative differences
cells are present (3) . between them.
In addition to stimulating the expansion of antigen-reactive More recently, several investigators have made use of pheno-
clones, the primary exposure to antigen has also been thought typic markers to define the cellular compartment in adult
to initiate a permanent qualitative change in the cells. For peripheral blood that contains memory T cells . For instance,
example, B cells respond to antigen priming by undergoing the T cell subsets identified by CD45RO, CD29, or CD58
molecular genetic changes that result in their differentiation have been shown to have adhesion and activation characteris-
to secretory cells, recognized morphologically as a change tics that are consistent with the concept of an accelerated sec-
to a plasma cell phenotype. In addition, due to genetic recom- ondary recognition event during a memory response (7-13).
bination and deletion events, Ig H chain class switching occurs, However, there is no way of knowing whether the subset
resulting in a permanently differentiated genotype (4) . Also, defined by the absence of these putative memory markers is
as a result of somatic hypermutation of V region genes, affinity truly antigenically naive, particularly since adults have been
maturation of the antigen-binding region takes place, ulti- exposed repeatedly and over a long period of time to a myriad
mately resulting in the preferential selection and persistence of antigens in the environment . Functional comparisons be-
of those cells secreting antibodies with the highest affinities tween T cell subpopulations derived from the blood of im-
(5, 6). As a consequence of these findings, the concept of munologically experienced individuals and separated on the
B cell memory no longer includes only the quantitative ex- basis of their differential expression of these phenotypic markers
pansion of specific B cell clones, but has progressed to a mo- may therefore not necessarily measure naivete vs. memory.
lecular genetic definition of a truly differentiated memory cell. Indeed, an alternative interpretation of the dissimilarities found

25 J . Exp. Med . © The Rockefeller University Press - 0022-1007/91/01/0025/12 $2 .00

Volume 173 January 1991 25-36
Published January 1, 1991

in these populations has been proposed which holds that the New York); 1HT4, reactive with CD25 (from Dr. Ellis Reinherz,
CD45 isoform markers more appropriately distinguish be- Boston, MA) . Rabbit anti-mouse FITC conjugate was from Zymed
tween functionally diverse T cell subsets that evolve and per- Laboratories, San Francisco, CA, and was used at 1:50 final dilu-
sist as independent lineages (14-16) . tion. Human gamma globulins used to block nonspecific FcR-
To circumvent the uncertainties inherent in trying to identify mediated staining were purchased from Organon Teknika.
and separate naive and memory T cells from adult blood,
we have made use of a biological distinction between naivete Cell Culture Reagents
and memory and drew on the naturally occurring sources Complete medium consisted of RPMI 1640 (Gibco Laborato-
of the beginning and endpoints of the development of im- ries, Grand Island, NY) supplemented with 50 U/ml penicillin,
munity. In this regard, neonatal T cells derived from umbil- 50 14g/ml gentamycin, 200 Eag/ml L-glutamine, and 10% heat-
ical cord blood are naive with respect to antigenic stimula- inactivated calf serum (HyClone Laboratories, Logan, UT) . Ho-
tion and represent a population of truly unprimed cells. By mogeneous rIL2 was obtained from Takeda Chemical Industries,
Osaka, Japan, and supplied as a 1-mg/ml solution in 5 mM ammo-
comparison, T cells derived from adult peripheral blood have nium acetate buffer (pH 5). In some experiments, 10 /AM aphidicolin
had ample antigenic exposure and were used as a source of (Sigma Chemical Co., St. Louis, MO) was used to completely in-
in vivo primed cells that have acquired "memory". hibit [3H]TdR incorporation .
With regard to a molecular characterization of T cell im-
mune responsiveness, it is now established that T cell-de- Cell Purification
rived lymphokines actually execute the cellular and molec- Umbilical cord blood was collected in preheparinized syringes
ular reactions that ultimately are recognized as cellular from healthy full-term neonates immediately after vaginal delivery.
immunity: T cell proliferation, T cell help for B cell prolifer- Adult blood was drawn from healthy volunteers. Mononuclear cells

Downloaded from jem.rupress.org on June 15, 2011

ation and antibody secretion, as well as the recruitment and were isolated by means of Ficoll-Hypaque discontinuous gradient
activation of macrophages and leukocytes have all been as- centrifugation . In some instances, heavy red cell contamination of
signed to distinct lymphokines (17). Moreover, sensitive assays umbilical cord mononuclear cells necessitated further purification
are now readily available to detect all of these molecules. by centrifugation over a 65% Percoll (Pharmacia LKB Biotech-
Thus far, most studies directed toward uncovering func- nology, Piscataway, NJ) gradient . Interface cells were washed three
tional disparities between naive and memory T cells have fo- times in normal saline and incubated at a final concentration of
cused on potential differences in activation parameters (18-20). 5 x 106 cells/ml in 5 MM L-leucine-methyl-ester hydrochloride
(Sigma Chemical Co.), in serum-free medium at room tempera-
However, by analogy to the genetic differentiation process ture for 45 min. This procedure has been shown to effectively elim-
of B cell memory, we hypothesized that a fundamental func- inate monocytes and a large proportion of NK cells (24, 25) . The
tional distinction between naive and memory T cells might cell mixture was centrifuged at 450 g for 20 min over a 43.5%
arise by differences in the expression of distinct genes, partic- Percoll gradient to separate dead cells and residual low density NK
ularly those encoding the mediators of cellular immunity. cells and macrophages from the high density lymphocytes, as de-
Accordingly, we examined gene expression for a variety of scribed (26). The cell pellet was washed twice in complete medium,
lymphokines by taking advantage of the PCR as a highly and to eliminate residual Fc-Receptor positive cells, N5 x 10' cells
sensitive method to detect mRNA transcripts (21-23) . Our were incubated in a final volume of 2 .5 ml for 20 min at 37*C
results, reported here, indicate that the differential capacity in 100-mm bacteriological petri dishes (Fisher Scientific Co ., Pitts-
to express lymphokine genes underlies and determines the burgh, PA) that had been coated with 10 mg/ml human gamma
globulins in PBS immediately before use; nonadherent cells were
marked functional difference between unprimed and primed then collected and the procedure repeated once. The remainder of
T cells. Moreover, during the course of these studies, we have the cells were incubated for 45 min at 37 °C at a maximum concen-
developed the cellular and molecular methods to permit one tration of 2.5 x 107 cells/ml in Lympho-Kwik TH (One Lambda,
to investigate the mechanisms responsible for this acquisi- Los Angeles, CA), to which 3138 (ceCD56 mAb) was added to give
tion of an extended lymphokine repertoire. a final dilution of 1 :100 of ascites. Treatment with this cocktail
of mAbs and complement selectively lyses HLA-DR* cells, B
cells, residual granulocytes, NK cells, and CD8 * cells. A final den-
Materials and Methods sity centrifugation was performed to separate the dead cells, and
Antibodies the cell pellet was washed twice in complete medium .
T cell activation was performed using 64.1, an IgG2a mAb Assessment of Cell Purity
directed against CD3 (from Dr. Ellen Vitetta, Dallas, TX) . Affinity-
purified goat anti-mouse IgG (Fc specific) was purchased from Or- Cells at different stages of purification were incubated with
ganon Teknika, West Chester, PA. For indirect fluorescent staining, saturating concentrations of mAbs specific for various surface
the following mAbs were used at saturating concentrations : aT11 markers for 60 min on ice. Indirect fluorescent staining was per-
(clone 3PT2H9), reactive with CD2 (from Dr. Ellis Reinherz, formed with 1:50 dilution of rabbit anti-mouse Ig FITC for 60
Boston, MA) ; OKT3, OKT4, and OKT8, reactive with CD3, min on ice. Cells were then fixed in 2% paraformaldehyde-PBS
CD4, and CD8, respectively (Ortho Pharmaceuticals, Raritan, NJ); until evaluation by FACScan (FRCS is a registered trademark of
N901 and 3118, IgG and IgM, respectively, directed against CD56 Becton Dickinson & Co.).
(from Dr. Jerome Ritz, Boston, MA) ; Bl, reactive with CD20 Umbilical cord blood mononuclear cells from density centrifu-
(Coulter Immunology, Boston, MA) ; 3C10, reactive with CD14, gation were 58-70% CD2 * , 44-65% CD3* , 25-39% CD4* ,
and 9.3C9, reactive with HLADR and DQ (from Dr. Gilla Kaplan, 9-21% CD8*, 6-15% CD56*, 7-10% CD20', 10-20% CD14*,

26 Differentiation of T Cell Lymphokine Gene Expression

Published January 1, 1991

and 18-20% HLA-DR* . Umbilical cord blood cells purified by PCR-assisted mRNA Amplification
L-leucine-methyl-ester, FcR panning, and modified Lympho-Kwik
TH were 97-99% CD2* , 96-99% CD3 * , 95-98% CD4* , and Preparation of cDNA . Total RNA preparation was performed
0-1% CD8* . Staining for B cells, NK cells, monocytes, and HLA- essentially as described (28) . In brief, after indicated time intervals,
DR* cells was undetectable above background control levels . Less supernatants were removed from the microtiter wells, and ice-cold
than 1/1,000 cells stained positive for nonspecific esterase (a-naph- PBS was added. Cells were removed by steady pipetting and wells
thyl-acetate esterase ; Sigma Chemical Co .) . These highly purified were inspected for complete cell recovery. Cells were counted
umbilical cord blood-derived Th cells are referred to as neonatal (Coulter Electronics Inc., Hialeah, FL) and aliquoted to
T cells in the text . 106/Eppendorf tube (1 .5 ml), pelleted for 10 s in a table-top Ep-
Adult peripheral blood cells purified by L-leucine-ester, FcR pan- pendorfmicrocentrifuge (Brinkmann Instruments Co ., Westbury,
ning, and modified Lympho-Kwik TH were 95-98% CD2* , NY), and lysed in 500 itl lysis solution, consisting of 4 M guanidine-
95-98% CD3* , 95-98% CD4* , 0-1% CD8* , and did not ex- thiocyanate (Boehringer Mannheim Biochemicals, Indianapolis, IN),
press detectable CD14, CD20, CD56, or HLA-DR. Less than 25 mM Na-citrate (pH 7), 0.5% N-lauroylsarcosine (Sigma Chem-
1/1,000 cells stained positive for nonspecific esterase . These cells ical Co .), and 100 mM 2-ME (Sigma Chemical Co.) . Lysates were
are referred to in the text as adult T cells. prepared at least in duplicate, vortexed, and stored at -70°C until
further processing . After thawing, 50 Al of 2 M Na-acetate (pH
Cell Stimulation and Culture 4.0), 500 lil of water-saturated acid phenol (Bethesda Research
Laboratories, Bethesda, MD), and 100 tcl of chloroform-iso-amyl
96-well, flat-bottomed tissue culture plates (Corning Glass alcohol (49:1) were added to the lysates with thorough vortexing
Works, Corning, NY) were coated with 50,ul of a 20- Wg/ml solu- after each addition . The mixture was then chilled on ice for 15
tion of goat anti-mouse IgG (Fc specific) in Tris-HCI (pH 9.5) min and spun at 10,000 g for 15 min at 4° C. The aqueous phase
for at least 3 h at room temperature. Immediately before use, wells was recovered and RNA was precipitated in an equal volume of

Downloaded from jem.rupress.org on June 15, 2011

were washed three times with normal saline. Unless indicated other- 2-propanol at -20°C for at least 90 min. Precipitates were pelleted
wise, cells were added to the wells at a final concentration of 2 at 4°C, washed once with 75% ethanol in diethylpyrocarbonate-
x 10 5/well in the presence of 1 pg/ml 64 .1 ot-CD3 mAb in com- treated double-distilled water (DEPC-ddH20) and repelleted at
plete medium . This procedure of crosslinking the CD3 structure 4°C at 10,000 g for 15 min. Vacuum-dried pellets were resuspended
is referred to throughout the text as stimulation via a-CD3 . For in 10 p.l DEPC-ddH 20 containing 2 ug oligo-dT (12-18 mer;
further expansion, cells were harvested after 48 h by vigorous pipet- United States Biochemical Corp ., Cleveland, OH) and incubated
ting and incubated at densities not exceeding 106 /ml in complete at 65*C for 10 min. After cooling on ice, the mixture was incubated
medium containing saturating amounts of rIlr2 (1 nM). On the with 10 ttl of 2 x RTbuffer (100 mM Tris-CI, pH 8.3, 150 mM
day of restimulation, cells were washed twice, incubated at 37°C KCL, 6 mM MgCl2, 20 mM DTT) and final concentrations of
for 45 min, and washed once more to remove residual rIL2. Sec- 200 U of MMLV reverse transcriptase (Bethesda Research Labora-
ondary stimulation was performed identically to primary stimula- tories), 1 mM dNTPs (United States Biochemical Corp.), 100 Ftg/ml
tion . In some experiments, the following mitogens were used to acetylated BSA (Sigma Chemical Co.), and 25 U of RNAsin
activate the cells: 1 hg/ml PHA (Wellcome, Research Triangle Park, (Promega Biotec, Madison, WI) for 60 min at 37°C. Tubes were
NC) or 11AM Ionomycin (Calbiochem-Behring Corp., San Diego, then heated to 95°C for 5-10 min, and 80 P.1 of dH20 was added
CA) with 5 ng/ml PMA (Sigma Chemical Co.) . to the 20-P1 reaction mixture. Samples were stored at 4*C until
further use. RT reactions were performed simultaneously for ly-
Thymidine Incorporation sates derived from adult and neonatal blood.
After 24-72 h in culture, DNA synthesis was assessed by adding PCR Conditions. Lymphokine-specific primer pairs IL-2, -3, -4,
0.5 p,Ci/well ['H-methyl]thymidine (10 Ci/mmol; ICN Biochem- -5, -6, IFN-y, and 0-Actin(Y) were purchased (Clontech, Palo Alto,
icals, Irvine, CA) for the final 2 h of culture. Triplicate cultures CA) or synthesized on a Cyclone DNA synthesizer (Biosearch,
were harvested onto glass fiber filters, and radioactivity was counted San Rafael, CA) and purified using NENSORB PREP columns
by liquid scintillation . Under the specified conditions peak prolifer- (DuPont Co., Wilmington, DE): 0-actin (5'), ILl0, GM-CSF, and
ation occurred at 48 h. ['H]TdR incorporation was calculated as IL-2R p55. Sequences are specific as ascertained by computer as-
cpm/10" cells/h. sisted search of updated versions of GenBank. 5' and 3' primers
were complementary to sequences in the first and last exons, respec-
Measurement of 11,2, IFN-y and Granulocyte/Macrophage tively, or spanned exon-exon junctions and are therefore mRNA
CSF (GM-CSFY specific. 5' primers were designed to recognize sequences no far-
ther than 1,250 by upstream of the polyT tract of the cDNA . The
Supernatants of stimulated cells were harvested after indicated control 0-actin 5' primer recognizes a sequence exactly 1,252-1,222
time intervals, centrifuged to remove cells and debris, and stored by upstream . Primer sequences are shown in Fig. 1. In preliminary
at -70*C until protein determinations were performed. IL2 con- experiments, 10-fold serial dilutions of irrelevant DNA (A-phage)
centrations were determined by the CTLL2 bioassay as described were performed into cDNA samples from both adult and neonatal
(27) . INF-y and GM-CSF concentrations were determined by T cells and subjected to 25 cycles of PCR using primers amplifying
specific ELISA (Endogen, Boston, MA). ELISAs were performed a 500-bp fragment (GeneAMP kit; Perkin-Elmer Cetus, Emeryville,
exactly as specified by the manufacturers, and protein concentra- CA). The limit of detection was determined to be in the order
tions are listed in picomoles. The limits of detection were 2.5 pM of 10^ molecules whose amplification product could reproducibly
for INF-y and 1 pM for GM-CSF. be visualized as a faint band after gel electrophoresis . Subsequently,
5 pl of cDNA (representing RNA derived from 5 x 10° cells to
ensure a sensitivity of detection of at least one molecule per cell,
1 Abbreviation used in this paper: GM-CSF, granulocyte/macrophage colony- allowing for losses during RNA preparation and reverse transcrip-
stimulating factor. tion) was amplified in 0.5 ml GeneAmp reaction tubes (Cetus Corp.,

27 Ehlers and Smith

Published January 1, 1991

Size of
Amplified
mRNA 5' Sense Primer 3' Antisense Primer Fragment
(b p)

-Actin 5'-TGACGGGGTCACCCACACTGTGCCCATCTA3' 5'-CTAGAAGCATTGCGGTGGACGATGGAGGG3' (661bp)

IL-1P 5'-CTTCATCTTTGAAGAAGAACCTATCTTCTT-3' 5'-AATTTTTGGGATCTACACTCTCCAGCTGTA-3' (331bp)

IL-2 5'-ATGTACAGGATGCAACTCCTGTCTT-3' 5'-GTCAGTGTTGAGATGATGCTTTGAC-3' (458 bp)

IL-3 5'.ATGA000GCCTGCCCGTCCTG-3' 5'-GCGAGGCTCAAAGTCGTCTGTTG-3' (449 bp)

IL-4 5'-ATGGGTCTCACCTCCCAACTGCT-3' 5'-CGAACACTTTGAATATTTCTCTCTCAT-3' (456 bp)

IL-S 5'-GCTTCTGCATTTGAGTTTGCTAGCT-3' STGGCCGTCAATGTATTTCTTTATTAAG-3' (291 bp)

IL-6 5'-ATGAACTCCTTCTCCACAAGCGC-3' 5'-GAAGAGCCCTCAGGCTGGACTG-3' (628 bp)

IFNy 5'-ATGAAATATACAAGTTATATCTTGGCTTT-3' 5'-GATGCTCTTCGACCTCGAAACAGCAT-3' (501 bp)

Figure 1 . Primer sequences used

Downloaded from jem.rupress.org on June 15, 2011

GM-CSF 5'-ACACTGCTGAGATGAATGAAACAGTAG-3' 5'-TGGACTGGCTCCCAGCAGTCAAAGGGGATG3' (286 bp)

for PCR-assisted mRNA ampli-

IL-2R p55 5'-TTATCATTTCGTGGTGGGGCAGATGGTTTA-3' 5'-TCTACTCTTCCTCTGTCTCCGCTGCCAGGT3' (391 bp) fication .

Emeryville, CA) in the presence of 200 nM final concentration of thresholds of neonatal vs . adult T cells, so that an optimal
5' and 3' primers, 200 AM dNTPs, 0.5 U of Taq polymerase (Cetus dose of 1 Ag/ml ci-CD3 was chosen for all subsequent ex-
Corp.), and PCR buffer containing 2.5 AM MgC12, 50 mM KCL, periments.
10 mM Tris-CI (pH 8.3), 4nd 0.001% gelatin in a final volume Several approaches were used to evaluate whether activa-
of 25 Al . The reaction mixture was overlaid with a drop of light
tion by ci-CD3 delivered equally potent signals to the two
mineral oil, and PCR was performed in a DNA thermal cycler
(Cetus Corp .) for 25 cycles : 45-s denaturation at 94°C, 45-s an- cell populations via the TCR/CD3 complex. First, IL-2
nealing at 60°C, and 1.5-min extension at 72°C . The reaction production was monitored and found to be comparable (neo-
product was visualized by subjecting to electrophoresis 20 Al of natal T cells: 40 .3 ± 22 PM vs. adult T cells; 43 .3 ± 21
the reaction mix at 80 V for 70 min in 2% agarose in 0.5x TBE pM, n = 8) . Second, IL-2R expression as measured by using
buffer containing 0.5 Ag/ml ethidium bromide. 1 Ag of HaeIII- IL-2R ci-p55 mAbs and flow cytometry showed equivalent
digested Ox174 DNA (Clontech) was run in parallel as molecular and maximal levels after 72 h of culture, in that >97% of
weight markers (providing bands at 1,353, 1,078, 872, 603, 310, the cells stained positive. Third, consistent with these findings,
281, and 234 bp). Specificity of the amplified bands was validated proliferation monitored by [3 H]TdR incorporation was com-
by their predicted size and restriction enzyme digests giving ap-
propriately sized fragments. PCRassisted mRNA amplification was
3000
repeated at least once for each RNA sample, and has been per-
formed on six separate samples from adult T cells and >25 sepa-
rate primary and secondary preparations from neonatal T cells to
date, yielding identical results to the ones described in Results.
as
U d
2000

Results
Activation of Neonatal T Cells vs. Adult T Cells. To meaning-
1000
fully compare functional parameters of neonatal and adult
T cells, it was important to ensure that both cell populations
could be activated maximally and equally. To do so, we capital-
ized on the finding of Geppert and Lipsky (29) that purified 0

T cells can be efficiently stimulated after depletion of acces-

sory cells by crosslinking solely the TCR/CD3 complex, the JaCD3 mAb] (Vg/ml)
cx-CD3 mAb 64 .1 being particularly potent . A comparison
of the concentrations of cx-CD3 required to activate neonatal Figure 2 . Proliferation of neonatal (0) and adult (*) T cells after 48 h
of activation with various doses of tx-CD3 crosslinked via immobilized
and adult purified CD4+ T cells when crosslinked via an goat anti-mouse IgG. [3H]TdR incorporation was assessed during the last
immobilized bridging antibody is shown in Fig. 2. In three 2 h of culture. Data are from an experiment representative of three per-
separate experiments, there were no differences in the response formed . The SD of triplicate cultures were always <10% .

28 Differentiation of T Cell Lymphokine Gene Expression

Published January 1, 1991

However, as the purified T cells did not proliferate in response

to PHA, total mononuclear cells stimulated with PHA had
.ia0 to be compared with purified T cells stimulated with PHA
gam and PMA . Maximal [3 H]TdR incorporation occurred after
48 h and was found to be similar for all experimental groups
u
c u
.d o
(Fig . 3, columns 5 and 6) .
z Production of IFN-,y and GM-CSF during Primary Stimula-
tion. Having ensured equivalent activation results as defined
by IL2 production and responsiveness, the secretion of IFN--y
and GM-CSF was monitored after 24 and 48 h of culture.
As reported previously by Bryson et al . and Wilson et al .
Figure 3 . Proliferation of neonatal and adult T cells and neonatal (30, 31), and shown in Fig . 4 (top), neonatal T cells did not
mononuclear cells using various stimuli . (1) Adult T cells (n = 10) stimu- produce IFN-.y , even though adult T cells secreted readily
lated with 1 Fog/ml soluble a-CD3 ; (2) neonatal T cells (n = 41) stimu- detectable amounts. Neonatal T cell GM-CSF production also
lated with 1 Wg/ml soluble a-CD3 ; (3) adult T cells (n = 10) stimulated was undetectable after 24 h of culture in 24 of 30 samples
via crosslinked a-CD3 ; (4) neonatal T cells (n = 41) stimulated via cross-
linked a-CD3 ; (5) neonatal T cells (n = 6) stimulated with 1 Fog/ml PHA examined, and at the limit of detection (1-1.7 pM) in the
and 5 ng/ml PMA; (6) neonatal mononuclear cells (n = 8) stimulated remaining six samples. (Fig. 4, bottom) . By comparison, adult
with 1 )cg/ml PHA. Background [ 3 H]TdR incorporation ± SD was 5.3 T cells secreted GM-CSF by 24 h and even higher levels by
± 2 cpm/104/h for neonatal T cells, 2.9 ± 1 for adult T cells, and 9.5 48 h. It is noteworthy that in contrast to IFN--y secretion,
± 1 for neonatal mononuclear cells. [3H]TdR incorporation was assessed neonatal T cell supernatants also contained GM-CSF when

Downloaded from jem.rupress.org on June 15, 2011

48 h after activation during the last 2 h of culture . Error bars represent 1 SD.
tested at 48 h. Therefore, neonatal T cells seem capable of
mounting a GM-CSF secretory response, although it is delayed
parable in the two populations when measured at the peak compared with the adult T cell response. In separate experi-
of the response, which occurred after 48 h of culture (Fig. ments (not shown), we have found the induction of GM-
3). As an additional test to ascertain whether neonatal T cells CSF secretion to be entirely IL-2 dependent .
were optimally activated, PHA was used in place of a-CD3 . Lymphokine Gene Expression Detected by PCR-assisted mRNA
Amplification. Since our methods of TCR/CD3 triggering
resulted in equivalent activation of neonatal and adult T cells
as monitored by 11,2 production, yet resulted in deficient
IFN-y and GM-CSF secretion, it appeared as though neo-
natal T cells might be unable to express some lymphokine
genes. Therefore, to examine lymphokine gene expression more
comprehensively, the PCR technique was chosen to provide
the most sensitive assay for the presence or absence of mRNA.
As shown in Fig. 5, before in vitro activation, both neonatal
T cells and adult T cells were devoid of mRNA for any of
the lymphokines tested . In contrast, as shown in Fig . 6, adult
T cells contained readily detectable levels of mRNA for IL-2,

Figure 4 . IFN-.y (top) and GM-CSF (bottom) concentrations in super-

natants of stimulated neonatal and adult T cells . Supernatants were har-
vested 24 and 48 h after activation via a-CD3 . Protein determinations
were performed by specific ELISAs . (1) Neonatal T cells, 24 h (n = 25) ;
(2) neonatal T cells, 48 h (n = 25) ; (3) adult T cells, 24 h (n = 10); Figure 5 . PCR-assisted mRNA amplification of adult (top) and neo-
(4) adult T cells, 48 h (n = 10). Error bars represent 1 SD . natal (bottom) T cells immediately before stimulation .

29 Ehlers and Smith

Published January 1, 1991

ture (24 h) . By comparison, neonatal T cells were incapable

of producing any of the lymphokines, with the notable ex-
ceptions of IL-2 and GM-CSF (Fig. 8). GM-CSF mRNA was
reproducibly detected in neonatal T cells after 24 h of cul-
ture, and in some experiments it appeared as early as 9 h after
stimulation . However, the remainder of the lymphokines re-
mained undetectable up to 48 h of culture.
Lymphokine mRNA Expression in Response to Alternative
Stimuli. Some reports have indicated that naive T cells might
have a higher activation threshold than primed T cells (18-20) .
Therefore, it was important to test alternative pathways of
stimulation to determine whether neonatal T cells were simply
stimulated suboptimally, or whether they were actually in-
capable of expressing most of the lymphokine genes. The
combinations of PHA/PMA and Ionomycin/PMA at con-
Figure 6 . PCR-assisted mRNA amplification of adult (top) and neo- centrations that resulted in optimal proliferation had no effect
natal (bottom) T cells 3 h after primary activation via u-CD3 . on the inducibility of any of the lymphokine mRNAs tested
(Fig. 9). Again, only 1172 and IL-2R p55 mRNAs were de-
tected under these conditions, thereby confirming the results
IL-3, IL-4, IL-5, IFN-y, GM-CSF, and IL-2R p55 as early obtained using a-CD3 for stimulation .

Downloaded from jem.rupress.org on June 15, 2011

as 3 h after stimulation . By comparison, neonatal T cells only Comparison of Neonatal T Cells with Neonatal Mononuclear
expressed detectable mRNAs for IL-2 and the IL-2R p55 chain Cells. Since the T cells were extensively purified before stim-
at this time interval . ulation, it was possible that cells exhibiting a lymphokine
To examine the kinetics of lymphokine mRNA expres- profile typical of adult cells could have been lost during the
sion, cells were harvested after 3, 6, 9, 24, and 48 h of stimu-
lation . Adult T cells showed the same lymphokine expres-
sion profile throughout the study period (Fig. 7), IL-6 mRNA
being only faintly detectable and only after prolonged cul-

Figure 7. PCR-assisted mRNA amplification of adult T cells at various Figure 8 . PCR-assisted mRNA amplification of neonatal T cells at var-
times after stimulation via a-CD3 . From top to bottom : 6, 9, 24, and ious times after stimulation via ct-CD3 . From top to bottom : 6, 9, 24,
48 h after activation . and 48 h after activation .

30 Differentiation of T Cell Lymphokine Gene Expression

Published January 1, 1991

a saturating concentration of rIL2 . At varying intervals, the

cells were restimulated via ct-CD3 using the same protocol
developed for the primary stimulation .
After 7-10 d of culture, the cells cease proliferating, and
express no lymphokine mRNAs (Fig . 11, top) . Then, upon
restimulation via the TCR/CD3 complex, the in vitro primed
neonatal T cells express within 3 h the mRNAs for all of
the lymphokines produced by adult T cells, with the excep-
tion of IIr5 . As shown in Fig. 11, a detailed time course after
restimulation revealed that IL5 mRNA does appear, but only
after a lag period of ti9 h. It then persists along with the
other lymphokine mRNAs for as long as 48 h. Also, when
tertiary stimulation of neonatal T cells was performed, IL5
mRNA was apparent after 3 h. No lymphokine transcripts
were detected in the absence of restimulation. Also, when
IIr2 was replenished without restimulation of the TCR/CD3
complex, only GM-CSF mRNA was apparent . Control neo-
natal T cells placed only in medium and then stimulated after
5 d via ot-CD3 still produced only IIr2 and IL2R mRNA,
Figure 9. PCR-assisted mRNA amplification of neonatal T cells after in a fashion identical to neonatal cells that were stimulated

Downloaded from jem.rupress.org on June 15, 2011

primary stimulation with PHA and PMA (top) or Ionomycin and PM when freshly isolated . Moreover, the capacity to express an
(bottom). extended lymphokine mRNA profile was also reflected in
the secretion of the gene products : all supernatants harvested
purification procedure. Alternatively, costimuli originating 24 h after secondary stimulation contained readily detectable
from accessory cells might be necessary for neonatal T cells levels of IFN -'y (19.2 ± 4.1 pM, n = 25) and GM-CSF (71.4
to express lymphokine genes other than 11,2 . Therefore, total ± 7.8 pM, n = 25).
mononuclear cells were stimulated with 1 Fag/ml PHA and
tested for cytokine mRNA expression. As shown in Fig. 10,
the mRNAs for IIr3, IL4, IIr5, and IFN-y still remained
undetectable under these conditions, while IL-10, IL-6, and
GM-CSF mRNAs were easily amplified, as expected from
the presence of monocytes and NK cells.
The Differentiation ofNeonatal T Cells In Vitra The fore-
going experiments served to illustrate a distinct difference
between neonatal and adult T cells in their capacities to ex-
press lymphokine genes. However, this inability to express
lymphokine genes upon a primary stimulation in vitro might
be explained simply by an immaturity on the part of the neo-
natal cells. Consequently, it was crucial to ascertain whether
the primary in vitro stimulation could actually mimic the
in vivo situation, promoting a switch, so that upon secondary
stimulation, an adult-type lymphokine repertoire could be
expressed. Therefore, neonatal T cells were stimulated as be-
fore for 48 h, then washed out of a-CD3 and cultured with

Figure 11 . PCR-assisted mRNA amplification of secondary neonatal

T cells. Neonatal T cells were primarily stimulated with a-CD3 and ex-
Figure 10. PCItassisted mRNA amplification of neonatal mononuclear panded in rlIL2 for 7 d. From top to bottom: immediately before restimu-
cells after 3 h of primary stimulation with PHA. Mononuclear cells were lation, 3 h after restimulation via a-CD3, 9 h after restimulation, and
prepared by a 65% Percoll-gradient centrifugation . 48 h after restimulation .

31 Ehlers and Smith

Published January 1, 1991

The switch in the capacity to express the lymphokine genes was deficient in both the mRNA expression and the secre-
first becomes detectable after 3 d of culture and is stable for tion of these lymphokines relative to the putative memory
as long as 14 d, after which the cells begin to die if not re- cell subset . Separate studies by Lewis et al. (38) using Northern
stimulated. These results suggested that the switch could not blot analysis to examine INF-y mRNA expression by neo-
be explained by the selective outgrowth of a minor subset natal T cells supported the impression that antigenically naive
of already differentiated cells present at the initiation of cul- T cells were somehow defective in their capacity to respond
ture. However, to confirm this interpretation, neonatal T cells to stimulation by the production ofat least some lymphokines.
were stimulated in the presence of aphidicolin to prevent DNA However, the mechanism underlying such a functional de-
replication . As aphidicolin specifically blocks DNA polymerase ficiency has remained controversial and ill defined, and thus
I and II without affecting RNA synthesis or blastogenesis, far, it has not been approached with a molecular genetic hy-
it is an ideal agent to prevent proliferation without overall pothesis in mind. Rather, most investigators have focused on
metabolic inhibition (32). After 5 d of stimulation in the uncovering differences in the activation parameters between
presence of aphidicolin, viable cells were recovered and re- naive and memory T cells, and have not directly posed the
stimulated via ot-CD3 . Again, a lymphokine profile typical question as to whether primed T cells might differ from un-
of differentiated T cells was detected. primed T cells qualitatively by their ability or inability to
express multiple effector lymphokine genes .
Discussion To devise an experimental approach that would allow an
Primary antigenic exposure in vivo initiates changes in the unambiguous interpretation of the results in terms of gene
immune system that can ultimately lead to immunity. One expression, several conceptual and technical problems had to
crucial feature of immunity is the capacity of the host to be taken into consideration .

Downloaded from jem.rupress.org on June 15, 2011

"remember" the priming antigen, and thereby respond to (a) It was crucial to begin the analysis with -a homoge-
it more efficiently; this property of the immune system is neous population of antigenically virgin T cells, since any
conventionally termed immunologic memory (1, 33) . The contamination of the naive T cells with already primed cells
precise nature of the events governing the acquisition of would have precluded a meaningful analysis . However, the
memory in vivo has escaped direct experimental scrutiny, partly exact cellular compartments responsible for memory and na-
because of the inaccessibility and complexity of the cellular ivete have remained uncertain, so that it has been difficult
components involved. As a first step toward an adequate in to identify representative cell subsets in adult blood and to
vitro representation of the processes determining T cell separate them from one another. In this regard, several pheno-
memory, we devised a simple experimental model system that typic markers have recently been proposed to distinguish be-
makes use of the biologically defined cellular sources of na- tween naive and memory T cells . For instance, it has been
ivete and memory, yet applies a reductionist approach with shown that the T cell subsets defined by CD26, CD29,
respect to T cell activation and the measurement of T cell CD45RO, or CD58 are capable of responding to recall an-
responsiveness. This in vitro system demonstrates a clear tigens and therefore, by definition, contain memory T cells
qualitative difference between unprimed and primed T cells, (7-9) . A number of investigators have made use of these
in that antigenically virgin cells can respond to activation markers to separate cells and have reported differences in acti-
by proliferating, yet cannot secrete effector lymphokines other vation parameters and functional and phenotypic character-
than IL2. By comparison, antigenically experienced cells are istics of the subsets obtained (7, 10, 18-20, 36, 37, 39-43) .
shown to be capable of expressing many, if not all, of the A distinct disadvantage of this strategy lies in the inevitable
lymphokines involved in an immune/inflammatory response crosscontamination with the reciprocal subset, which essen-
immediately upon stimulation. Most important, naive T cells tially precludes the application of sensitive molecular genetic
can and do respond to a primary in vitro activation by a approaches that might detect the products of those few con-
differentiation process, which then enables them to express taminating cells . In addition, the marker CD45RO, for ex-
the entire repertoire of effector lymphokines immediately upon ample, does not per se qualitatively delineate "memory", since
subsequent restimulation . Therefore, this in vitro system of many other cell types of the hematopoietic lineage express
T cell differentiation mimics the accelerated nature and func- it, including immature thymocytes (44, 45) . Accordingly, to
tional superiority ofthe in vivo anamnestic immune response. circumvent the ambiguities inherent in identifying and separ-
It should be emphasized that the early studies by Tedder ating putative naive and memory T cells in adult blood, we
et al., (34, 35) comparing putative naive and memory T cell capitalized on the only known natural source of purely naive
subsets separated from adult blood found that memory cells cells, those found in newborns. By definition, neonatal T cells
could provide help for the generation of antibody-forming derived from umbilical cordblood represent a population com-
cells, while naive T cells were deficient in this regard . How- prised primarily of virgin T cells, in that antigenic exposure
ever, at this time, T cell help had not yet been ascribed to in utero is rare. This state of functional immune naivet6 is
distinct lymphokines, so that it was impossible to proceed reflected in the fact that neonates are less efficient than adults
to a molecular dissection ofthe phenomenon . Subsequently, at mounting an effective immune response to a variety of
Sanders et al . (7, 18), Salmon et al . (36), and Lewis et al. infectious agents (46). By comparison, a considerable propor-
(37) did examine the production of lymphokines, notably tion of T cells derived from adult blood have memory, in
IFN-y and IL4, and found that the putative naive cell subset that normal individuals have had extensive exposure to en-

32 Differentiation of T Cell Lymphokine Gene Expression

Published January 1, 1991

vironmental microbes and have developed long-lasting im- itor mRNA expression (21-23). The particular reaction condi-
munity to most infectious agents. tions chosen were designed to detect as few as one mRNA
(b) To interpret any differences in the lymphokine gene molecule/cell, given that all the cells in the population under
expression between neonatal and adult T cells as attributable study were responding to activation. Within these limits of
to the T cells themselves, it was critical to rigorously purify detection, neonatal T cells expressed easily detectable levels
them from contaminating granulocytes, B cells, monocytes, of IL-2 and IL2R p55 mRNA transcripts, yet did not ex-
and NK cells, which are also known to produce some cytokines press any of the other lymphokines tested for. Therefore, at
(47-49) . this sensitivity level, it can only be concluded that the vast
(c) In view of several reports that have suggested that neo- majority of neonatal CD4+ T cells cannot express the genes
natal monocytes are relatively immature or might even have encoding most of the T cell lymphokines upon only a pri-
suppressive effects, it was important to use a system depleted mary stimulation .
of accessory cells (50, 51) . Moreover, Inaba and Steinman (52) Finally, having uncovered a fundamental difference between
have provided evidence that unprimed and primed T cells neonatal and adult T cell lymphokine gene expression, it was
have distinct requirements for APC . As well, phenotypic obligatory to demonstrate that a primary stimulation of neo-
studies have suggested that primed cells have a higher den- natal T cells would lead to their differentiation in vitro, al-
sity ofa number of activation molecules on their surface (10, lowing them to express an adult-type lymphokine pattern
11) . Most of these molecules facilitate cell-cell interaction, upon restimulation. Otherwise, the lack of the capacity to
while some have also been implicated in auxiliary signaling. express the majority of lymphokine genes could have been
Since most of these structures find their counterparts on mono- interpreted as merely a sign of immaturity on the part of
cytes, it has been suggested that a complex of triggering mol- the neonatal T cells . As secondary stimulation of neonatal

Downloaded from jem.rupress.org on June 15, 2011

ecules involving the TCR and these accessory structures is T cells did result in the expression of a lymphokine reper-
more easily formed with primed than unprimed cells (11, 53) . toire typical of T cells exposed to antigenic stimulation in
Thus, only an accessory cell-depleted system avoids the ex- vivo, our data support the interpretation that neonatal T cells
perimental ambiguity and difficulties in the interpretation of really are mature : they can respond to TCR/CD3 activation
results that arise when stimulation is performed in the pres- in vitro by proliferating, and also by differentiating to fully
ence of monocytes and other non-T cells. competent, lymphokine-secreting effector cells. It is impor-
(d) Some investigators have emphasized that primed T cells tant to note that our data are derived from whole cell popu-
express a higher density of ancillary molecules, such as CD2, lations and therefore contribute no information as to func-
CD18/CD11a, CD28, CD29, CD44, CD54, CD58, etc., con- tions of single cells. However, it is worthy of emphasis that
tributing to their lower threshold for activation (8, 11, 53, all attempts to demonstrate clonal segregation or restriction
54). Therefore, to compare neonatal and adult T cells meaning- of lymphokine gene expression by human T cell clones similar
fully, it was mandatory to devise an efficient and equal system to that reported with murine T cell clones have failed to date
for activating the cells, one that functioned solely via a trig- (36, 60-62) .
gering structure present in equal densities on both popula- When considering the nature of a primary immune re-
tions, and one that would be indifferent to any dissimilarities sponse compared with an anamnestic response, our results
in T cell accessory molecules . This was accomplished by cross- are consistent with the characteristics of the phenomenon
linking exclusively the CD3/TCR complex via immobilized recognized in vivo as immunologic memory. Thus, if our
ci-CD3, which has been shown to be an effective stimulus in vitro findings are reflective of what actually occurs upon
for T cells even in the absence of accessory cells (29). introduction of antigen in vivo, the first several days are taken
(e) It was necessary to demonstrate that the signals trans- up by the IL-2-promoted clonal expansion of lymphocytes.
duced through the TCR/CD3 complex were equivalent for Assuming that antigen persists during this period, the ex-
both neonatal and adult T cells, particularly since several reports panded clones of naive T cells that have undergone the differen-
have implied that neonatal T cells are somehow immature tiative switch would then be capable ofreceiving a secondary
(55-59) . When monitored by IL2/IL2R-mediated prolifer- stimulation from persisting antigen . Ultimately, the secre-
ation, crosslinking with a-CD3 yielded similar results with tion of the rest of the lymphokines would direct the recruit-
identical ot-CD3 dose-response relationships for both neo- ment and activation ofthe remainder of cells that participate
natal and adult T cells (Figs. 2 and 3) . Moreover, when trig- in a fully developed immune response. By comparison, a
gering via the TCR/CD3 complex was bypassed entirely using previously primed host that already has an expanded antigen-
phorbol ester and calcium ionophore, identical lymphokine experienced population is capable of secreting all of the
patterns were observed as with a-CD3-mediated activation lymphokines immediately upon antigenic stimulation . Con-
(Fig. 9). Therefore, in this experimental activating system, sequently, the tempo of the anamnestic response is more rapid
it is unlikely that differential lymphokine mRNA expression compared with a primary response because of two fundamental
by neonatal and adult T cells is attributable to differing acti- differences in the T cell populations : a primed host has an
vation thresholds or due to differences in known intracel- expanded number of antigen-reactive cells, and these cells are
lular signal transduction pathways. already differentiated so that they can release all ofthe immune-
(f) To implicate a switch in gene expression, it was im- inflammatory lymphokines immediately after antigen trig-
portant to use a sensitive method such as the PCR to mon- gering.

33 Ehlers and Smith

Published January 1, 1991

With respect to additional mechanisms thought to be in- compared with and complemented by other criteria that have
volved in the phenomenon ofimmunologic memory, several been advanced, especially for the phenotypic delineation of
studies have indicated that memory cells also differ from naive memory T cells. In this regard, most data derived from
cells in their tissue distribution and recirculation patterns (12, CD45RO- or CD58-based separations of adult T cells are
13, 63) . These characteristics may help explain the observed congruent with this hypothesis, although a similarly com-
rapid accumulation of primed cells at inflammatory foci and prehensive analysis oflymphokine gene expression in the com-
may also contribute to the known longevity ofcell-mediated plete absence of accessory cells has not yet been attempted
immunity (11-13). Furthermore, the response to antigen (36, 39). Further proof for the validity of this definition is
during a secondary exposure may begin more rapidly because to be expected from in vivo experimental systems that afford
primed T cells may be more sensitive to lower antigen con- a detailed examination of successive stages in the physiologic
centrations (10, 11, 18-20) . However, in addition to these development of cell-mediated immunity. Clinical data from
characteristics of memory, the data presented here imply that pathophysiologic conditions may prove even more useful, in
much of the functional superiority of the anamnestic immune that a possible cause for immunodeficiencies could be the failure
response lies in the differentiated state of a primed T cell that to acquire, or alternatively, the loss of, memory T cell func-
enables it to immediately secrete precisely those effector lym- tion. Similarly, the concept of T cell tolerance may benefit
phokines that are required for a full-fledged, coordinate im- from the hypothesis advanced here: it is entirely conceivable
mune defense. that an anergic T cell genetically resembles a naive cell, but
From the way our in vitro experiments were planned and one that cannot undergo the switch to a functional memory
performed, differences in activation thresholds, known sig- phenotype. It is also attractive to speculate that some pathogens
nalling pathways, and accessory cell contributions could be may have evolved strategies that selectively interfere with this

Downloaded from jem.rupress.org on June 15, 2011

excluded as responsible for the two distinct functional T cell differentiation process of the T cell as a means to evade cel-
phenotypes. Instead, all of our data are consistent with the lular immunity.
interpretation that by analogy to B cells, T cells also undergo At this stage, it is premature to speculate in detail as to
an activation-dependent differentiation event that permits their the molecular mechanisms that might account for the ap-
secretion of multiple effector lymphokines upon restimula- parent coordinate change in lymphokine gene expression that
tion. Accordingly, we propose that the term "memory T cell" underlies the switch from naivete to memory. However, since
be reserved for those T cells identified by their differentiated the model system described in this report provides all of the
state of lymphokine gene inducibility. T cell memory thus components necessary for the differentiation ofT cells in vitro,
entails not only the clonal expansion of specific T cells, but thereby recapitulating the developmental history ofthe anam-
also encompasses a differentiation event that allows additional nestic T cell response in vivo, a molecular dissection of the
lymphokine gene expression . mechanisms that determine and govern the development of
This tentative definition provides for a testable working T cell memory is feasible.
hypothesis of T cell memory. In particular, it can now be

We gratefully acknowledge the friendly cooperation provided by the doctors, nurses, and midwives col-
lecting suitable cord blood samples at the Alice Peck Day Memorial Hospital birthing center, Lebanon,
NH. We also thank Gary Ward for expert technical assistance in performing thecytofluorographic analyses
and Dr. Gilla Kaplan for helpful suggestions with the manuscript.
This work was supported by grants from the National Institutes of Health (RO1 CA-17643), The Council
for Tobacco Research, Inc., USA (1715), and a postdoctoral training grant to S. Ehlers from the Deutsche
Forschungsgemeinschaft (EM01/1-1) .
Address correspondence to Kendall A. Smith, Immunology Research Laboratory, Dartmouth Medical
School, Hanover, NH 03756 .

Received for publication 27 August 1990 and in revised form 25 September 1990.

References
Bach, J.F. 1978 . Definition of (immunologic) memory. In Im- 3 . Cerottini, J.-C ., H.D. Engers, H.R. MacDonald, and KT
munology. J.F. Bach, editor. John Wiley & Sons, Inc ., New Brunner. 1974. Generation ofcytotoxic T Lymphocytes in vitro.
York. 848-851. I. Response ofnormal and immune mouse spleen cells in mixed
2. Burnet, F.M. 1959 . The clonal selection theory of acquired leukocyte cultures. J. Exp Med. 140:703 .
immunity. In The Abraham Flexner Lectures : 1958 . Cambridge 4. Lewin, B. 1990 . Generation of immune diversity involves re-
University Press, Cambridge. 49-80. organization of the genome . In Genes IV. B. Lewin, editor.

34 Differentiation of T Cell Lymphokine Gene Expression

Published January 1, 1991

Oxford University Press, NY 705-729. 22 . Rappolee, D.A ., D. Mark, M.J . Banda, and Z. Werb . 1988 .
5. Berek, C., and D. Milstein . 1987. Mutation drift and reper- Wound macrophages express TGF-a and other growth factors
toire shift in the maturation of the immune response. Immunol. in viva : analysis by mRNA phenotyping . Science (Wash. DC).
Rev. 96 :23 . 241:708 .
6. Manser, T, L.J . Wysocki, M.N . Margolies, and M.L . Gefter. 23 . Brenner, C.A ., A.W. Tam, P.A . Nelson, E.G. Engleman, N.
1987. Evolutio n of antibody variable region structure during Suzuki, K.E . Fry and J.W. Larrick. 1989 . Message amplification
the immune response . Immunol. Rev. 96 :155 . phenotyping (MAPPing): a technique to simultaneously mea
7. Sanders, M.E ., M.W. Makgoba, S.O. Sharrow, D. Stephany, sure multiple mRNAs from small numbers of cells. Biotech-
TA . Springer, H.A . Young, and S. Shaw. 1988 . Human niques. 7:1096.
memory T lymphocytes express increased levels of three cell 24 . Thiele, D.L ., M. Kurasaka, and P.E . Lipsky. 1983 . Phenotype
adhesion molecules (LFA-3, CD2, and LFA-1) and three other of the accessory cell necessary for mitogen-stimulated T and
molecules (UCHL1, CD29, and Pgp-1) and have enhanced B cell responses in human peripheral blood: delineation by its
IFN-y production . J. Immunol. 140:1401. sensitivity to the lysosomotropic agent, Lleucine methyl ester.
8. Hailer, D.A ., D.A . Fox, D. Benjamin, and H.L . Weiner. 1986 . J. Immunol. 131:2282.
Antigen reactive memory T cells are defined by Tai. J Im- 25 . Thiele, D.L ., and P.E . Lipsky. 1985 . Modulation of human
munol. 137:414 . natural killer cell function by Lleucine methyl ester: monocyte-
9. Smith, S.H ., M.H. Brown, D. Rowe, R.E . Callard, and P.C .L . dependent depletion from human peripheral blood mononuclear
Beverley. 1986 . Functional subsets of human helper-inducer cells. J. Immunol. 134:786 .
cells defined by a new monoclonal antibody, UCHL1. Immu- 26 . Damle, N.K ., L .V. Doyle, J.R. Bender, and E.C. Bradley. 1987 .
nology. 58 :63. Interleuki n 2-activated human lymphocytes exhibit enhanced
10 . Sanders, M.E ., M.W. Makgoba, and S. Shaw. 1988 . Human adhesion to normal vascular endothelial cells and cause their
naive and memory T cells: reinterpretation of helper-inducer lysis. J. Immunol. 138:1779.

Downloaded from jem.rupress.org on June 15, 2011

and suppressor-inducer subsets. Immunol. Today. 9:195 . 27 . Gillis, S., M.M . Ferm, W Ou, and K.A. Smith. 1978 . T cell
11 . Beverley, P.C .L . 1990 . Is T cell memory maintained by cross- growth factor : parameters of production and a quantitative
reactive stimulation? Immunol. Today. 11 :203 . microassay for activity. J. Immunol. 120 :2027 .
12 . Damie, N.K., and L.V. Doyle. 1990. Ability of human T lym- 28 . Chomczynski, P., and N. Sacchi . 1987 . Single-step method
phocytes to adhere to vascular endothelial cells and to aug- of RNA isolation by acid guanidinium thiocyanate-phenol-
ment endothelial permeability to macromolecules is linked to chloroform extraction. Anal. Biochem. 162:156 .
their state of post-thymic maturation . J. Immunol. 144:1233. 29 . Geppert, TD., and P.E . Lipsky. 1987 . Accessory cell indepen-
13 . Tedder, TF., T Matsuyama, D. Rothstein, S.F. Schlossman, dent proliferation of human T4 cells stimulated by immobi-
and C. Morimoto. 1990. Human antigen-specific memory T lized monoclonal antibodies to CD3. J Immunol. 138 :1660.
cells express the homing receptor (LAM-1) necessary for lym- 30 . Bryson, YJ ., H.S. Winter, S.E . Gard, T.J . Fischer, and E.R .
phocyte recirculation . Eur. J Immunol. 20 :1351. Stiehm . 1980. Deficiency of immune interferon production by
14 . Morimoto, C., N.L . Letvin, A .W. Boyd, M. Hagan, H.M . leukocytes of normal newborns . Cell. Immunol. 15 :191 .
Brown, M.M . Kornacki, and S.F. Schlossman . 1985 . Th e iso- 31 . Wilson, C.B., J. Westall, L. Johnston, D.B. Lewis, S.K . Dower,
lation and characterization of the human helper inducer T cell and A.R . Alpert . 1986 . Decreased production of interferon-
subset . J Immunol. 134:3762. gamma by human neonatal cells: Intrinsic and regulatory
15 . Morimoto, C., N.L . Letvin, J.A . Distaso, WR. Aldrich, and deficiencies . J Clin. Invest. 77 :860 .
S.F. Schlossman . 1985 . Th e isolation and characterization of 32 . Hoy, C.A ., G.C . Rice, M. Kovacs, and R.T. Schimke. 1987 .
the human suppressor inducer T cell subset . J. Immunol. Over-replication o£ DNA in S phase chinese hamster ovary cells
134:1508. after DNA synthesis inhibition . J. Biol. Chem . 262:11927 .
16 . Rothstein, D.M., S. Sohen, J.F. Daley, S.F. Schlossman, and 33 . Cerottini, J.-C., and H.R. MacDonald . 1989 . The cellular basis
C. Morimoto . 1990 . CD4'CD45RA' and CD4'CD45RA - of T cell memory. Annu . Rev. Immunol. 7:77.
T cell subsets in man maintain distinct function and CD45RA 34 . Tedder, TF., LT Clement, and M.D . Cooper. 1985 . Human
expression persists on a subpopulation of CD45RA' cells lymphocyte differentiation antigens HB-10 and HB-11 : I. On-
after activation with Con A. Cell. Immunol. 129:449 . togeny of antigen expression . J Immunol. 134:2983.
17 . Smith, K.A . 1984 . Lymphokine regulation of T cell and B 35 . Tedder, TF., L.T. Clement, and M.D. Cooper. 1985 . Human
cell function . In Fundamental Immunology. WE . Paul, editor. lymphocyte differentiation antigens HB-10 and HB-11. II .
Raven Press, New York. 559-576. Differential production of B cell growth and differentiation
18 . Sanders, M.E ., M.W. Makgoba, C.H . June, H.A . Young, and factors by distinct helper T cell subpopulations . J. Immunol.
S. Shaw. 1989 . Enhanced responsiveness of human memory 134:2983.
T cells to CD2 and CD3 receptor-mediated activation . Eur. 36 . Salmon, M., G.D. Kitas, and P.A . Bacon . 1989 . Production
J. Immunol. 19 :803 . of lymphokine mRNA by CD45R' and CD45R- helper T
19 . Horgan, K.J ., G.A. VanSeventer, Y Shimizu, and S. Shaw. cells from human peripheral blood and by human CD4' T
1990. Hyporesponsiveness of "naive" (CD45RA') human T cell clones. J. Immunol. 143:907 .
cells to multiple receptor-mediated stimuli but augmentation 37 . Lewis, D.B., K.S. Prickett, A. Larsen, K. Grabstein, M. Weaver,
of responses by co-stimuli . Eur. J. Immunol. 20 :1111. and C.B. Wilson . 1988 . Restricted production of interleukin
20 . Byrne, J.A ., J.L . Butler, and M.D . Cooper. 1988 . Differentia l 4 by activated human T cells. Proc. Nad. Acad. Sci. USA.
activation requirements for virgin and memory T cells. J. Im- 85 :9743.
munol. 141:3249. 38 . Lewis, D.B ., A. Larsen, and C.B. Wilson . 1986. Reduced
21 . Mullis, K.B., and F.A . Faloona. 1987 . Specific synthesis of DNA interferon-gamma mRNA levels in human neonates : Evidence
in vitro via a polymerase-catalyzed chain reaction . Methods En- for an intrinsic T cell deficiency independent of other genes
zymol. 155:335 . involved in T cell activation . J. Exp. Med. 163:1018.

35 Ehlers and Smith

Published January 1, 1991

39 . Clement, LT, N. Yamashita, and A.M. Martin . 1988 . The lymphocytes to OKT3-monoclonal antibody-induced mitogen-
functionally distinct subpopulations of human CD4' helper/ esis . Scand. J. Immunol. 23 :91 .
inducer T lymphocytes defined by anti-CD45R antibodies de 52 . Inaba, K., and R.M . Steinman . 1984 . Resting and sensitized
rive sequentially from a differentiation pathway that is regu- T lymphocytes exhibit distinct stimulatory (antigen-presenting
lated by activation-dependent post-thymic differentiation. J. cell) requirements for growth and lymphokine release. J. Exp.
Immunol. 141:1464. Med. 160:1717.
40 . Akbar, AX, L . Terry, A. Timms, P.C.L . Beverley, and G. 53 . Koyasu, S., T Lawton, D. Novick, M.A . Recny, R.F. Siliciano,
Janossy. 1988 . Loss of CD45R . and gain of UCHLI reactivity B.P. Wallner, and E.L . Reinherz . 1990. Role of interaction of
is a feature of primed T cells. J. Immunol. 140:2171. CD2 molecules with lymphocyte function-associated antigen
41 . Schraven, B., M. Roux, B. Hutmacher, and S.C . Meuer. 1989 . 3 in T cell recognition of nominal antigen. Proc. Natl. Acad.
Triggering of the alternative pathway of human T cell activa- Sci. USA. 87 :2603.
tion involves members of the T 200 family of glycoproteins . 54 . Turka, L.A ., J.A . Ledbetter, K. Lee, C.H . June, and C.B .
Eur. J. Immunol. 19 :397 . Thompson . 1990. CD28 is an inducible T cell surface antigen
42 . Byrne, J .A ., J .L . Butler, E.L . Reinherz, and M.D . Cooper. that transduces a proliferative signal in CD3' mature thymo-
1989 . Virgin and memory T cells have different requirements cytes. J. Immunol. 144:1646.
for activation via the CD2 molecule . Int. Immunol. 1:29. 55 . Gerli, R., A. Bertotto, S. Crupi, C. Arcangeli, I. Marinelli,
43 . Matsuyama, T, P. Anderson, J.F. Daley, S. Schlossman, and F. Spinozzi, C. Cernettif, P. Angelella, and P. Rambotti . 1989 .
C. Morimoto. 1988. CD4'CD45R' cells are preferentially Activation of cord T lymphocytes . I. Evidence for a defective
activated through the CD2 pathway. Eur.J. Immunol. 18 :1473. T cell mitogenesis induced through the CD2 molecule. J. Im-
44 . Janossy, G., M. Bofill, D. Rowe, J. Muir, and P.C .L . Beverley. munol. 142:2583.
1989 . The tissue distribution of T lymphocytes expressing 56 . Bertotto, A., R. Gerli, L. Lanfrancone, S. Crupi, C. Arcangeli,
different CD45 polypeptides . Immunology. 66 :517 . C. Cernetti, F. Spinozzi, and P. Rambotti . 1990 . Activation

Downloaded from jem.rupress.org on June 15, 2011

45 . Serra, H.M ., J.F. Krowka, J.A . Ledbetter, and L.M . Pilarski. of cord T lymphocytes. II . Cellular and molecular analysis of
1988 . Loss of CD45R (Lp220) represents a post-thymic T cell the defective response induced by anti-CD3 monoclonal anti-
differentiation event. J. Immunol. 140:1435. body. Cell. Immunol. 127:247 .
46 . Klein, J.O., J.S. Remington, and S.M. Marcy. 1983 . Curren t 57 . Hayward, A.R ., and A.R . Lawton . 1977 . Inductio n of plasma
concepts of infections of the fetus and newborn infant . In In- cell differentiation of human fetal lymphocytes : evidence for
fectious Diseases of the Fetus and Newborn Infant . J.S. functional immaturity of T and B cells. J. Immunol. 119:1213.
Remington, and J.O. Klein, editors. WB. Saunders Company, 58 . Miyagawa, Y, K. Sugita, A. Komiyama, and T Akabane. 1980 .
Philadelphia, PA . 1-26 . Delayed in vitro immunoglobulin production by cord lympho-
47 . Rappolee, D.A., and Z. Werb. 1988 . Secretory products of cytes. Pediatrics. 65 :497 .
phagocytes . Curr. Opinions Immunol. 1:47. 59 . Andersson, U., A.G . Bird, S. Britton, and R. Palacios. 1981 .
48 . Aneg6n, L, M.C. Cuturi, G. Trinchieri, and B. Perussia . 1988 . Humoral and cellular immunity in humans studied at the cell
Interaction of Fc receptor (CD16) ligands induces transcrip- level from birth to two years of age. Immunol. Rev. 57 :5 .
tion of interleukin 2 receptor (CD25) and lymphokine genes 60 . Mosmann, TR ., and R.L . Coffman . 1989 . THI and TH2 cells:
and expression of their products in human natural killer cells. different patterns of lymphokine secretion lead to different func-
J. Exp. Med. 167:452 . tional properties . Annu. Rev. Immunol. 7:145 .
49 . Pistoia, V., S. Zupo, A. Corcione, S. Roncella, L. Matera, R. 61 . Paliard, X., R.D. Malefijt, H. Yssel, D. Blanchard, I. Chre-
Ghio, and M. Ferrarini. 1989 . Production of colony-stimulating tien, J. Abrams, J. DeVries, and H. Spits. 1988 . Simultaneous
activity by human natural killer cells: analysis of the condi production of IL-2, Il:4, and IFN-y by activated human
tions that influence the release and detection of colony- CD4` and CD8' T cell clones . J. Immunol. 141:849 .
stimulating activity. Blood. 74:156 . 62 . Patel, S.S ., A.D. Duby, D.L . Thiele, and P.E . Lipsky. 1988 .
50 . Taylor, S., and Y.J . Bryson . 1985 . Impaired production of 'y- Phenotypic and functional characterization of human T cell
interferon by newborn cells in vitro is due to a functionally im- clones . J. Immunol. 141:3726.
mature macrophage . J. Immunol. 134:1493. 63 . MacKay, C.R., W.L . Marston, and L. Dudler. 1990. Naive
51 . Papadogiannakis, N ., S.-A. Johnsen, and L.B. Olding. 1986 . and memory T cells show distinct pathways of lymphocyte
Monocyte-regulated hyporesponsiveness of human cord blood recirculation . J. Exp. Med. 171:801 .

36 Differentiation of T Cell Lymphokine Gene Expression