1858 G. C. Coutinho, S. Delassus, F! Kourilsky et al. Eur. J. Irnrnunol. 1994. 24: 1858-1862

Graqa Cunha Coutinhomn, Sylvie Delassus., Philippe Kourilsky., Ant6nio BandeiraA and Antonio Coutinhoa

Unit6 d’lmmunobiologie, CNRS URA 359A and Unit6 de Biologie Mol6culaire du G b e , INSERM U 2 V , lnstitut Pasteur, Pans

Developmental shift in the patterns of interleukin production in early post-natal life

In mice, immunological tolerance to self is established in the perinatal period, when tolerance susceptibility to allogenic tissues is higher than in adults. We have now investigated whether this could result from developmental regulation of effector functions of T cells exposed to specific antigens, by studying the “natural” or T cell receptor-induced expression of several interleukin genes. We used qualitative and quantitative polymerase chain reaction methods to study interleukin (1L)-2, IL-4, IL-10 and interferon-y mRNA expression by splenic cells at different ages.The results show that newborn peripheral cells (up to day 7), in contrast to the T lymphocytes of adult mice, express high levels of IL-4 and interferon-y, and very low levels of IL-2 messenger spontaneously and upon specificT cell activation. This characteristic phenotype depends on intrinsicT cell properties, as it is not due to the newborn environment.

1 Introduction

Classical observations have shown that the “immature” immune system of newborns is highly susceptible to tolerance induction by allogenic antigens [ l].These findings are compatible with the notion that the normal immune system “learns” the antigenic composition of self in the embryonic and perinatal period [2]. In addition to the deletional mechanism occurring in the thymus throughout life, peripheral unresponsiveness to antigens that are not expressed in the thymic environment must also be predom- inant during this period, which covers the peripheralization of t h e first T cells produced in the animal.Thus, in the first 3 weeks of post-natal life, the absolute numbers of lympho- cytes recovered from the spleen of normal mice increase exponentially (31. This increase in peripheral lymphocyte numbers is largely independent of exogenous antigenic stimulation [4], and it is essentially due to thymic emigra- tion [5, 61. Whether post-thymic tolerance results from dominant or recessive mechanisms is still a matter of debate [7], but the question is essential for our understanding of self-tolerance and our ability to manipulate immune re- sponsiveness for medical purposes. One important ques- tion is whether, upon exposure to “self” antigens, newly formed T cells emigrating from the thymus are simply inactivated or respond along “tolerant” pathways that ensure lack of self aggression, possibly as a consequence of a developmental genetic program.The analysis of peripher- al T cells in the perinatal period could be informative in this respect.

Two major functional phenotypes of CD4+ T cells have been described based on distinct patterns of interleukin (IL) production, namely the inflammatory (Thl) and helper (Th2) types [8]. Because inflammatory responses are related to the destruction and immune rejection of target tissues [9 ] , if a “tolerant”c1ass of T cell responses to antigen were to be considered, it would be expected to display characteristics of the helper effector type. We evaluated the “spontaneous” or anti-TCR-mediated gene expression for several IL by splenic cells during post-natal development, using qualitative and quantitative polymerase chain reac- tion (PCR). While the spontaneous production of ILs by newly formed Tcells arriving in the spleen should give indication of their interactions with “self antigens”, their response to TCR-dependent activation would provide information on their genetic/developmental determina- tion.

2 Materials and methods

2.1 Animals

BALB/c and C57BL/6 mice bred under conventional conditions were obtained in the facilities of the Pasteur Institute and used at either 1 (newborn), 7 or 15 days or 8 weeks of age.

2.2 Cell culture

Spleen cells (3 x lo6) were promptly used for RNA extraction or set in culture (3 x lo6 cells/ml) in 24-well Costar plates in complete medium in the presence of anti-CD3 [lo] antibody (Ab) (5 pg/ml). For Ab-mediated lysis, cells were suspended at 5 x 107/ml with anti-Thy-1 (J.l.J) [ l l ] at 4°C for 45 min, spun down, incubated with guinea pig (‘) at 37”c for 45 min3 and then extensively washed. Thy-1-depleted populations were rou- tinely analyzed by flow cytometry, and always contained less than 1 % CD3+ cells.

[I 126421

Supported by the grant BD/1024/90-ID from J N I n (Portu- gal).

Correspondence: Graqa C. Coutinho, Unite d’Immunobiologie, lnstitut Pasteur. 25. Rue du Docteur Roux, F-75724 Paris Cedex IS. France (Fax: 33 145688921)

Key words: Ontogeny / Interleukin expression / Interleukin-4

0014-2980/94/0808- 1858$10.00 + .25/0 0 VCH Verlagsgesellschaft rnbH, D-69451 Weinheirn, 1994

Eur. J. Immunol. 1994. 24: 1858-1862

2.3 Polymerase chain reaction

RNA was isolated by a guanidium thiocyanate/phenol/ chloroform procedure [12], and corresponding cDNA was produced. PCR amplification was performed as previously described [13]. IL-specific primer pairs for: IL-2 [14], IL-4 [ 151, IFW-y [ 161, IL-10 [ 171 and fl-actin [ 181 were purchased (Clontech, Palo Alto, CA). Samples were analyzed after amplification by ethidium bromide gel electrophoresis.

IL-4 predominance in early post-natal life 1859

fluorescent primer, are plotted in Fig. 1B as a function of the number of copies in the internal standard, giving the number of copies in the experimental sample. Table 1A show the results of ex vivo determinations in freshly collected spleen cell populations. When three independent measurements were done with different cDNA and differ-

A

2.4 Quantification of mRNA expression

The quantitative titration of the different mRNA was performed as previously described [19, 201 by PCR run to saturation (40 cycles). This technique involves the co- amplification of a standard DNA with the cDNA to be measured, and the detection of the amplification products by an elongation reaction performed with a labeled oligo- nucleotide primer (run-off reaction). As the standard and wild-type DNA differ in length by four base pairs, the run-off products from the two populations can be identified when loaded on an automated gel sequencer. Custom- designed software was used to measure both the length and the intensities of the products. It has been shown that the relative intensities of the two products represent the ratio of the two populations in the tube prior to amplification, even when this amplification reaches the plateau [19]. Thus, by varying the number of standard copies added to the cDNA of interest, a curve can be obtained that allows a precise evaluation (less than 15 YO error) of the number of cDNA copies in the experimental sample.The results are corrected to the total number of cells in the sample, extraction and reverse transcription efficiencies, by quantitating expres- sion of the house-keeping gene HPRT in the same samples [201.

3 Results

3.1 Patterns of IL gene expression in the spleen during post-natal development. Overexpression of IL-4 in the neonatal period

At birth, splenic CD3+ T cells are roughly 30 to 100 times less frequent (0.3 to 1 %) than in adults (30 %), and consist of only some loo00 cells. Thereafter, T cell numbers increase exponentially, over 100-fold in the first week of life. Given that splenicT cell frequencies at birth are below 1 YO, we avoided potential errors due to selective purifica- tion procedures by using PCR to investigate the kinetics of gene expression for IL-2, IFN-y, IL-4 and IL-10, either spontaneously or upon polyclonal stimulation of splenic T lymphocytes. Using quantitative PCR, we determined the number of mRNA copies for IL-2, IL-4, IFN-y and IL-10 (per loo00 copies of HPRT) obtained in spleen cells at different time points during ontogeny.

I I

Figure I. Quantification of HPRT and IL-4 cDNA copies. (A) Quantification of HPRTand IL-4 cDNA copies in newborn spleen. The same amount of cDNA was mixed with 5 X 104, 105 and 5 x lo5 copies of standard HF'RTor IL-4 plasmids (shown in the panels left to right), and the PCR and run-off analyses performed as described [21].The figure shows the raw data obtained from the respective run-off products run on the automated sequencer. The two peaks in each panel correspond to the two cDNA populations (standard and experimental), and the size and intensities of the different bands are given. (B) Automated calculation of the cDNA copy number per sample.The ratio of the intensities of the standard and experimental peaks is reported as a function of the number of standard copies, as described [22].

Table 1. Quantification of interleukin cDNAa)

A. cDNA copy numbers per loo00 copies of HPRT

IL-2 IFN-y IL-4 IL-10

Newborn nd 3900 14600 544 Day 7 70 20849 98571 4434

Adult 75 560 400 1886 Day 15 222 9 625 5 382 357

B. cDNA copy numbers per loo00 copies of HPRT

IL-2 IFNy IL-4 IL-10

Newborn 471 1457 58400 2012 Day 7 997 24358 161714 5307 Day 15 2023 24722 11 395 3 105 Adult 14152 41679 8148 21553

1860 G. C. Coutinho, S. Delassus, I? Kourilsky et al. Eur. J. Immunol. 1994. 24: 1858-1862

A NATURAL EXPRESSION OF IL-4 AND IFNr

B

Y c 3

5

; - a 8 * z L

E

AGE IN DAYS

tions at different ages. The values obtained in one experi- ment are shown in Table 1B. Higher standard deviations were observed with quantification in stimulated cells, which can be explained by the fact that small differences in incubation times may greatly influence the content of IL messenger in cultured cells. As can be seen in Table 1B and Fig. 2B, the expression of interleukin genes upon stimula- tion increases with age for IL-2, IFN-y, and IL-10. In contrast, expression of IL-4 shows a maximum at 7 days of age and decreases thereafter, in spite of the increasing proportion of T cells in these cultures. These quantitative results indicate that newborn peripheral cells predominant- ly express IL-4 over IL-2, a reversed pattern of what is found in adult cell populations. By 7 and 15 days of post-natal life, intermediate patterns were found. Results obtained in two mouse strains (BALBk and C57BL/6) using qualitative PCR were concordant with these (see Fig. 3).

CD3-DEPENDENT INTERLEUKIN EXPRESSION

ant i -CD3 stimulation 12h Spontaneous

jmj A I

AGE IN DAYS

Figure 2. (A) Age dependence of natural expression of IL-4 and IFN-y in the spleen.The number of copies of HPRT, IL-4 and IFN-y was quantitated in spleen mRNA from donors at the indicated ages. IL gene copy numbers were normalized using HPRT expression and corrected for the total number of spleen cells. (B) Age dependence of IL-2, IL-4. IL-10 and IFN-y expression in anti-CD3-stimulated cultures. Spleen cells (3 X loh) from donors at the indicated ages were cultured with anti-CD3 for 12 h and then analyzed for expression of IL genes as above.

ent standard plasmid dilutions, the SD of the mean results was less than 15 % for all determinations. As can be seen, IL-2 mRNA is undetectable in newborn spleen cells, even by a method with enough sensitivity to detect as few as ten molecules. In contrast, we could detect a very high number of IL-4 mRNA copies in newborn spleen, in spite of the very low number of Tcells in the organ. In fact, by calculating IL mRNA copy numbers per organ, thus accounting for the increase in the total number of cells, it can be seen in Fig. 2A that spontaneous IL-4 expression is maximal at 7days of age, and then decreases sharply. Spontaneous expression of IFN-y also increases markedly at 7 days of age, but is kept a high levels throughout the first 2 weeks of life.

Quantification of IL messenger was also performed after

A A 2 4 1 0 r A r 4 l l ) .

Newborn spleen cells

anti-Tlly-l treated Newborn spleen

B

Adult spleen cells

Newborn T-cell depleted t Adult spleen cells

Figure 3. IL gene expression in newborn and adult spleen. Spon- taneous gene expression of fi-actin (A), IL-2 (2), IL-4 (4). IL-10 (lo), IFN-y (y) in newborn and adult spleens. (A) PCR amplifica- tion of newborn spleen cDNA after T cell depletion. Newborn spleen cells were either treated with anti-Thy-I and complement or left untreated, set in culture with anti-CD3 for 12 h, and qualitative PCR was performed. (B) PCR amplification of adult spleen cells diluted in newborn T cell-depleted spleen. Adult spleen cells were diluted in anti-Thy-1 and complement treated newborn spleen cells to the neonatal representation ofT cells.The cell mixture was set in culture with anti-CD3 for 12 h, and qualitative PCR was per- formed.

3.2 Cytokine expression in newborn spleen is T cell dependent

To ascertain the T cell dependence of IL gene expression in neonates, day 1 newborn spleen cell suspensions were depleted of T lymphocytes by anti-Thy-1 and complement, set in culture with anti-CD3, harvested after 12 h, and the same procedure followed as above. The results show the maintenance of mRNA for 6-actin and IL-10, and the loss of detectable products for IL-2, IL-4 and IFN-y (Fig. 3A). Thus expression of IL-2, IL-4 and IFN-y in newborn spleen

12 h of anti-CD3 in v i m stimulation of spleen cell popula- cell populations is T cell dependent.

Eur. J. Immunol. 1994. 24: 1858-1862 IL-4 predominance in carly post-natal life 1861

3.3 Newborn pattern of IL expression is not imposed by the cellular environment

At birth, the spleen is a hematopoietic organ, and this or other differential conditions in its cellular environment could selectively interfere with the patterns of IL expres- sion characterized above. In order to investigate this possibility, adult cells were diluted in T cell-depleted, newborn spleen cells, at the frequencies found in newborn spleen. Upon anti-CD3 stimulation, the pattern of IL gene expression was found to be that of the adult type (Fig. 3B). These experiments indicate that differential IL expression patterns of newborn and adult T lymphocytes are intrinsic to the cells, and not imposed by the cellular environment or culture conditions of the producing cells.

4 Discussion

Putative functional differences between the peripheral immune system in neonatal versus adult life were investi- gated by evaluating the patterns of IL gene expression by peripheral T cells during ontogeny, using a combination of qualitative and quantitative PCR methods.We showed that the newborn spleen contains relatively high levels of IL-4 mRNA, while no IL-2 messenger was detected in the same samples. Since IL-4 gene expression was entirely dependent onT cells, and the newborn pattern could not be ascribed to the respective cellular environment, we interpret these results as indicating that splenic T cells that are physiolog- ically activated in the newborn have characteristics of helper T cells. This finding could be due to a developmental control of IL gene expression, to the predominance of a particular T cell type in early post-natal life, or to peculiar conditions under which newly formed T cells are exposed to self antigens. Regardless of the mechanism, this pattern of IL expression would be compatible with a “tolerant” phenotype. The post-natal evolution of peripheral T cell responses to anti-CD3 stimulation gives support to the first of these hypotheses. Thus, under these conditions, which should reveal all the functional potential of splenicT cells at various ages, the same patterns were found.

Age-associated increases in foreign antigen load, resulting from bacterial colonization of mucosae and of an adult-type diet, are maximal at the time of weaning and could be thought to cause this shift in IL expression pattern. On the other hand, we have recently found that adult specific pathogen-free and “antigen-free’’ mice [21] show compar- able IL expression patterns that are undistinguishable from those described here for conventional animals (unpub- lished results) .We cannot discard the possibility that IL-4 or other cytokines are produced by cells other thanT lympho- cytes. However, it as been recently described [22] that purified peripheral T cells from 4-day-old mice produce fourfold more IL-4 than the same number of adult T cells upon anti-CD3 stimulation.

While newborn cells show a pattern that is related to the Th 2 phenotype, we also found that early post-natal T cells produce abundant levels of IFN-y, a lymphokine associated with inflammatory responses. The current classification of CD4 T cells into these two functional phenotypes, there- fore, does not appropriately describe the situation in the neonatal period. On the other hand, while the levels of

expression of IFN-y are kept constant into adult life, IL-4 expression undergoes a very marked quantitative drop from day 7 onwards. Moreover, most T cells in the newborn seem to be “naturally” engaged in IL-4 and IFN-y expres- sion, since in vitro, CD3-dependent activation does not considerably stimulate the levels detected ex vivo. Given these observations with in vitro stimulated cells, we inter- pret the age-dependent decrease in the spontaneous expression of IFNy as representing proportionally fewer naturally activated T cells in the spleen. The three-order- of-magnitude decrease in spontaneous expression of IL-4, on the other hand, is likely to represent the combined effects of lower frequencies of activated T cells, as well as the developmentally regulated limitation in expression revealed in anti-CD3 stimulated cultures. Interestingly, it has been described that single positive CD4+ thymocytes differ from their progeny, naive peripheral CD4+ cells, by producing mainly IL-4, IL-10 and IFN-y, but little IL-2 [23]. In fact, these thymocytes still express such a Th2-like phenotype early after migration to the periphery, where they later switch to producing large amounts of IL-2 and little IL-4 upon stimulation [24]. Our results show that the functional phenotype of recent thymic migrants represents the exclusive T cell type in the periphery during the first week of life, and that it continues to be the dominant T cell population up to 2 weeks of age. These results support the notion that the peripheral Tcell pool grows with the accumulation of recent thymic migrants, rather than by extensive peripheral expansion [6]. Most importantly, the present observations are in line with hypotheses suggesting that self tolerance is not exclusively due to deletion or functional inactivation of self-reactiveT cells, but is at least partly the result of non-inflammatory responses of T cells responding to autologous antigens. Thus, it is possible that the developmental control of IL gene expression described here, imposing a predominant expression of IL-4, relates to the establishment of natural self tolerance. Moreover, the maintenance of this cell developmental program through- out life could ensure the continuous recruitment of CD4 T cells reactive with peripheral tissue-specific antigens into a non-inflammatory (regulatory?) type of effector func- tions, thus contributing to the maintenance of self toler- ance. In experimental systems of tolerance induction to allogeneic tissue grafts during the neonatal period, differ- entiation of specificT cells into aTh2 functional phenotype has also been correlated to the state of tolerance [25].

Received December 20,1993; in final revised form April 29, 1994: accepted May 5, 1994.

5 References

Billingham, R. E., Brent, L. and Medawar, P. B.. Nature 1953. 172: 603. Burnet, M. F., Ausf. J . Sci. 1957. 20: 67. Velardi, A. and Cooper, M. D., J. Immunol. 1984. 133: 672. Forni, L., Heusser, C. and Coutinho, A., Ann. Inst. Pus- teurllmmunol. 1988. 139: 245. Scollay, R. B. and Weissman, E., Eur. J. Immunol. 1980. 10: 210. ModiglianLY., Coutinho, G., Defranoux. O., Coutinho, A. and Bandeira, A., Eur. J. Immunol. 1994. 24: 1223. Coutinho, A . , Coutinho, G. , Grandien, A, , Marcos, M. A. R. and Bandeira, A , , Res. Immunol. 1992. 143: 345.

1862

8 Mosmann, T. R. and Coffman, R. L., Annu. Rev. Immunol. 1989. 7: 145.

9 Mosmann,T. R. and Coffman, R. L., Immunol. Today 1987.8: 223.

10 Leo, 0. M., Foo, M., Sachs, D. H., Samelson, L. E. and Bluestone, J. A., Proc. Natl. Acad. Sci. USA 1987. 84: 1374.

11 Bruce, J., Symington, E W., Mckearn, T. J. and Sprent, J., J. Immunol. 1981. 127: 1496.

12 Chomczynski, P. and Sacchi. N.. Anal. Biochem. 1987. 162: 156.

13 Ehlers, S. and Smith, K. A., J. Exp. Med. 1991. 173: 25. 14 Kashima, N., Nishi-Takaoka, C., Fujita, T., Taki, S.,Yamada,

G., Hamuro, J. and Tanigushi,T., Nature 1985. 313: 402. 15 Noma,Y., Sideras, P., Naito,T., Bergstedt-Linquist, S., Azuma,

C., Severinson, E. ,Tanabe,T., Kinashi,T., Matsuda, F. ,Yaoita, Y. and Honjo, T., Nature 1986. 319: 640.

16 Gray, P. W. and Goeddel, D. V., Proc. Natl. Acad. Sci. USA 1983. 80: 5842.

17 Moore, K. W.,Vieira, P., Fiorentino, D. F.,Trounstine, M. L., Khan,T. A. and Mosmann,T. R., Science 1990. 248: 1230.

G. C. Coutinho, S. Delassus, P. Kourilsky et al. Eur. J. Immunol. 1994. 24: 1858-1862

18 Alonso, S., Minty, A., Bourlet,Y. and Buckingham, M., J. Mol.

19 Pannetier, C., Cochet, M.. Darche, S. and Kourilsky, P., C. R.

20 Pannetier, C., Delassus, S., Darche, S., Saucier, C. and

21 Wostman, B. S., Bruckner-Kardos, E. and Pleasants, J. R.,

22 Adkins, B., Ghanei, A. and Hamilton, K., J. Immunol. 1993.

23 Bendelac, A. and Schwartz, R. H., Nature 1991. 353: 68. 24 Bendelac, A., Matzinger, P., Seder, R., Paul, W. E. and

25 Powell, T. J. J. and Streilein, J. W., J . Immunol. 1990. 144:

26 Delassus, S., Coutinho, G. C., Saucier, C., Darche, S. and

Evol. 1986. 23: 11.

Acad. Sci. Paris 1992. 315: 271.

Kourilsky, P., Nucleic Acids Res. 1993. 21: 577.

J. Nutr. 1982. 112: 552.

151: 6617.

Schwartz, R. H., J. Exp. Med. 1992. 175: 731.

854.

Kourilsky, P. , J. Immunol. 1994. 152: 241 1.