T cell messenger RNA

Eur. J. Immunol. 1977. 7: 43-48 The alternative possibility is that low levels of antibody act with normal, nonsensitized lymphocytes moving through the thyroid. A process of antibody-dependent cell-mediated tissue damage has been demonstrated in virro in the guinea pig and i n the human [20, 211. If this mechanism is involved, it is possible that the antibody of good responder mice is better able t o cooperate with such “killer” lymphocytes than is antibody from poor responders. As yet we have not been able t o d i a tinguish the antibodies of good and poor responder mice. They are both resistant t o 2-mercaptoethanol degradation and have the sedimentation properties of 7 S globulins.

To distinguish these two models it will be necessary to analyze different classes of antibodies in high and low responders, and, furthermore, t o determine if antibodies of high responder animals differ markedly in their ability t o cooperate with normal lymphocytes or t o activate a n Arthus type of reaction. I n either model, cell-mediated immunity does not play a primary pathogenic role. Received September 15, 1976.

5. References 1 McDevitt, H.O. and Benacerraf, B., Adv. Immunol. 1969.11: 31. 2 McDevitt, H.O., Deak, B.D., Shreffler, D.C., Klein, J., Stimpfling, J.H. and Snell, G.D., J. Exp. Med. 1972.135: 1259. 3 Vladutiu, A.O. and Rose, N.R.,Science 1971.174: 1137.

T.H. Rabbi-’,

A. Fonter’, M. Smith

and S. Gillam’ MRC Laboratory of Molecular Biology, Cambridge’ and Department of Bioche mistry, University of British Columbia, Vancouver+

43

4 Tom&, V., Rose, N.R. and Shreffler, D.C., J. Immunol. 1974. 112: 965. 5 Lennon, V.A. and Byrd, W.J., Eur. J. Immunol. 1973.3: 243. 6 Anderson, L.G., Tarpley. T.M., Talal, N., Cummings, N.A., Wolf, R.O. and Schall, G.L., Clia Exp. Immunol. 1973.13: 355. 7 Spitler, L.E., von Muller, C.M., Fudenberg, H.H. and Eylar, E.H.. J. Exp. Med. 1972.136: 156. 8 McMaster, P.R.B., Lerner, E M and Exum,E.D., J. Exp. Med. 1961.113: 611. 9 Flax, M.H., Jankovd, B.D. and Sell, S., Lob. Invest. 1963.12: 119.

10 Clinton, B.A. and Weigle, W.O., J. Ex$. Med. 1973.136: 1605. 11 Rose, N.R., Twarog, F.J. and Crowle, A.J., J. Immunol. 1971. 106: 698. 12 Nelson, D.S. and Boyden, S.V.,Immunol. 1963. 6: 264. 13 Sonozaki, H. and Cohen, S., Cell. Immunol. 1971.2: 341. 14 Bultmann, B., Bigazzi, P.L., Heymer, B. and Haferkamp, O., Z. Immunitatsforsch. Exp. Klin. Immunol. 1971. 142: 267. 15 Rose, N.R.,Vladutiu, A.O., David, C.S. and Shreffler, D.C., Clin. Exp. Immunol. 1973. 15: 281. 16 Vladutiu, A.O. andRose, N.R.,J. Immunol. 1971.106: 1139. 17 Nakamura, R.M. and Weigle, W.O., J. Exp. Med. 1969.130: 263. 18 Clagett, J.A., Wilson, C.B. and Weigle, W.O., J. Exp. Med. 1974. 140: 1439.8 19 Toma%, V. and Rose,N.R., Clin. Immunol. Immunopathol. 1975.4: 511. 20 Wasserman, J., Packalen. Th., Perlmann, P. and Perlmann, H., Int. Arch. Allergy AppL Immunol. 1971.40: 905. 21 Calder, E., McLeman, D. and Irvine, W.J., Clin.Exp. Immunol. 1973.15: 467.

Immunoglobulin-like messenger RNA in a mouse T cell lymphoma RNA-driven complementary DNA (cDNA) hybridization experiments have been carried out in order t o detect complementary sequences in RNA prepared from a mouse T cell lymphoma line (EL4). In conditions where efficient hybridization of L-chain cDNA with homologous P3 myeloma mRNA was observed, poor hybridization was observed with EL4 mRNA unless a low criterion of hybrid formation was employed (i. e. hydroxyapatite fractionation). The hybrid formed between EL4 mRNA and L-chain cDNA was found t o melt about 5 “C below the homologous hybrid indicating that the sequence detected in EL4 mRNA is similar but not identical with t h e P3 mRNA sequence. However, a similar sequence was detectable in a CA-producing myeloma cell line but in this line t h e concentration of t h e sequence was found t o be an order of magnitude lower than in EL4 cells.

1. Introduction

B lymphocytes are known t o have antibody molecules o n their cell surface, and these antibodies act as receptors for antigen (for review see [ 1 I). T lymphocytes have a helper [I 15181

Correspondence: T.H. Rabbitts, MRC Laboratory of Molecular Biology, Hills Road, Cambridge CB2 2QH, England Abbreviations: cDNA Complementary DNA SDS Sodium dodecyl sulfate pT: Polythymidylicacid at:Product of RNA concentration (moles nucleotide liter-1) and time in seconds Tm: Melting temperature

function in antibody-mediated immune responses and can themselves bind antigen in a specific fashion [2]. The nature of the T cell receptor, however, remains unclear. It has been variously postulated t o be immunoglobulin (Ig) [3] or products of t h e Ir locus [4, 51. If t h e T cell receptor is indeed a n Ig molecule, it might be expected that the molecules should be detected by techniques which are capable of detecting B cell receptors. Conflicting reports have been published [3, 61 so that the nature of t h e T cell receptor remains unsettled.

If

cells do produce Ig the cytoplasm of such cells would contain the messenger RNA (mRNA) f o r t h e Ig chains. In

44

Eur. J. Immunol. 1977. 7: 43-48

T.H.Fbbbitts, A. Forster, M. Smith and S. Gilkm

addition, if T cells do express Ig genes this presumably means that T cells as well as B cells undergo the differentiation events involving V and C gene integration. With this background in mind we have carried out RNA-driven complementary DNA (cDNA) hybridization experiments designed to detect Ig L-chain mRNA (CK class) in the cytoplasmic RNA of a @-positive mouse T cell lymphoma (EL4). The results indicate that such cells express a gene which is similar t o but not identical with the CK gene. This gene product, however, may not be confined to T cells since it was also detected in a mouse myeloma (P8) cell line which produces a Ch L-chain.

2. Materials and methods 2.1. Cell lines and media

P3-X63 is a clone of the P3 mouse myeloma cell line originally derived from MOPC 21 [7] producing IgGl(K). EL4 is a mouse T cell lymphoma (obtained through Dr. A.J. Munro from Dr. Bolyston) and P8 is a subline of MOPC 104E producing only X-chains. All cells were grown in Dulbecco's modified Eagle's medium plus 10 % heated horse serum (Gibco, Grand Island, N.Y., USA).

2.2. mRNA and cDNA preparation Total mRNA was prepared from postnuclear supernatant by a modification of a procedure previously published [8]. Either fresh cells weve used for RNA preparation or cells which had been frozen (- 70 "C) as a pellet in a small quantity of medium. In the latter case, the pellet was thawed rapidly at 37 "C. In the procedure described below the volumes are for approximately 1 O9 cells. The cells were suspended in 20 ml ice-cold 10 mM Tris-HCl pH 7.5, 10 mM MgC12, 10 m~ KCl and 2 ml 10 % Nonidet-P40 (BDH Chemicals, Poole, Dorset, England) were added. Lysis was completed by gentle hand homogenization and the nuclear fraction was removed by centrifugation for 5 min at 1000 x g in a Sorvall bench-top centrifuge at 4 "C. One ml 10 % sodium dodecyl sulfate (SDS) was added t o the supernatant and the mixture heated for 1.5 min at 100 "C. After chilling, proteinase K (BDH) was added to give 1 mg/ml and the suspension incubated for 1 h on ice. A further 1 ml 10 % SDS was added, the mixture warmed to room temperature and 24 ml redistilled water-saturated phenol-chloroform (50: 50) added. This mixture was shaken for 10 min (on an orbital shaker) and the aqueous phase recovered by centrifugation for 10 min at 10 000 x g (at 25 "C). The aqueous phase was re-extracted with an equal volume of fresh phenolchloroform, and finally the RNA was precipitated by the addition of one tenth volume 2 M sodium acetate buffer pH 5.0 plus 2.5 volume ethanol (overnight, -20 "C). The RNA precipitate was recovered by centrifugation, dried in YUCUO and dissolved in 1 ml binding buffer (500 mM NaCl, 1 0 mM TrisHCI pH 7.4, sterilized by the addition of 1 pl of diethylpyrocarbonate/lOO ml). The RNA solution was passed (at room temperature) through a column of oligo(dT)-cellulose, washed with 10 ml binding buffer and the fraction bound t o the column (poly(A)-containing fraction) was eluted with 1 ml 10 mM Tris-HCI pH 7.4 (treated with diethylpyrocarbonate as above).

The mRNA was recovered from the eluted fraction by addition of one tenth volume 2 M sodium acetate and 2.5 volume ethanol in the presence of 50 pg E. coli tRNA carrier. The MOPC 21 K-chain 13 S mRNA fraction was prepared from P3 cells as described [9]. The amount of mRNA was estimated by hybridization with [3H]poly(U) as described [lo]. cDNA was prepared from the respective mRNA fractions using reverse transcriptase. The K-chain cDNA (designated L-chain cDNA) was prepared using 1 3 S L-chain mRNA template plus the oligonucleotide T2G3T as primer. T2G3 was chemically synthesized by the conventional diester method of Khorana and coworkers [ 11 ] and was a generous gift from Drs. C.C. Cheng and G.G. Brownlee. The T2G3 was converted t o T&T by the enzymatic addition (using polynucleotide phosphorylase) of thymidine diphosphate [ 121. Purity and chain length of the product was assessed by high pressure anion exchange (RPC-5) chromatography [ 131. cDNA was prepared with this oligonucleotide in a reaction containing 20 pg/ml 13 S L-chain mRNA, 10 pg/ml T2G3T, 50 pg/ml actinomycin D, 50 mM Tris-HC1 pH 8.3, 20 mM dithiothreitol, 6 0 mM KCl, 6 mM MgC12, 0.5 mM dTTP, dATP and dGTP plus 50 pCi t3H]dCTP (Radiochemical Center, Amersham, Bucks, England, 25 Ci/mmol). Incubation was for 2 h at 37 "C, cDNA made from total mRNA was prepared using 80 pg/ml mRNA and 20 pg/ml pTlo as primer (P.L. Biochemicals, Milwaukee, Wisc., USA). After incubation all cDNA preparations were purified in the same way. NaOH was added t o give 0.5 N, the mixture was boiled for 5 min, neutralized with an equal volume 1 M NaH2P04 and passed through a 15 ml column of Sephadex SP-50 equilibrated with 0.3 M NaCl, 0.01 M sodium acetate buffer pH 5.0. The breakthrough peak contained the [3H]cDNA. Mouse globin cDNA was made as above but with 20 pg/ml globin mRNA (a kind gift from Dr. P. Harrison) and 20 pg/ml pTlo primer.

2.3. Hybridization procedures RNA-cDNA hybridizations were carried out at 70 "C in 0.24 M or 0.1 M sodium phosphate buffer pH 6.8,'containing 0.05 % SDS. The mixtures of RNA, cDNA and buffers were prepared at appropriate concentrations of RNA and aliquots taken into siliconized capillaries. The capillaries were sealed and incubated at 70 "C after boiling for 1 min. At the appropriate Crt values (Cr = moles RNA nucleotide/liter and t = time in seconds) the samples were diluted to 0.012 M sodium phosphate buffer (for single strand specific S1 nuclease digestion) or 0.1 2 M buffer (for hydroxyapatite fractionation). The percentage hybridization was determined for each time point by S1 nuclease digestion at 45 "C or hydroxyapatite procedures at 65 "C as described previously [ 141.

For the melting curves, 4 pg EL4 mRNA, P3 mRNA or P8 mRNA were incubated t o Crt = 7 with 3 ng L-chain cDNA. The mixtures were diluted 50 times with 0.12 M sodium phos- an aliquot removed. The phate, placed at 65 "C for 5 mir temperature of the mixtures was increased by 4 "C increments, equilibrated at each temperature for 5 min, after which an aliquot was removed. All aliquots were then assayed for percentage of cDNA as hybrid by hydroxyapatite fractionation at 65 "C. The curves are drawn by normalizing the initial percentage value t o 100 %.

T cell messenger RNA

Eur. J. ImmunoL 1977. 7: 43-48

3. Rerultr 3.1. Use of a specific hexanucleotide to produce pure reverse

transcripts of K-chain mRNA Mouse L-chain mRNA is difficult t o prepare in pure form and therefore cDNA made from impure fractions will in part reflect the impurity of the mRNA [ 9 ] . Recently we have been investigating the use of defined oligonucleotides t o specifically initiate synthesis of mouse L-chain cDNA transcripts even when this mRNA is among other contaminating species. One particular oligonucleotide T&3T has been found t o act as a site-specific primer for the equivalent of amino acid residue 204 within the C, region [IS]. These sequencing studies indicated the lack of priming on mRNA species other than the L-chain, and this observation has been quantitated by measuring the back-hybridization of cDNA (made with the hexanucleotide) t o its mRNA template. The measurement of the rate of hybridization between cDNA and an excess of the mRNA template is a means by which purity of mRNA can be quantitatively assessed [ 161. In our experiments we were concerned not only with the rate of hybridization but also with the proportion of cDNA which became hybridized in the kinetic transition which characterizes the L-chain mRNA component of the curve. Figure 1 shows hybridization of hexanucleotide-primed cDNA with the Lchain mRNA and the hybridization of rabbit globin cDNA t o globin mRNA. The latter mRNA consists of essentially only 01 and Pglobin so that the half point of the mRNAcDNA hybridization reaction (Crt lp) represents that expected for an overall sequence of about 4.32 x lo5 daltons. The Crtl,2 value for globin cDNA is 6.17 x 1O-4. When L-chain was hybridized t o its template, a single hybridizing transition occurred with Crt112 = 4 x and consisting of practically all the cDNA. This indicates that the mRNA used in this experiment is about six times less pure than globin mRNA. Since the

45

major mRNA species of the L-chain mRNA fraction is the Lehain mRNA itself (as shown by fingerprints of 32P-labeled mRNA) and the cDNA hybridizes to the mRNA with a single transition involving 90-95 7% of the cDNA, the cDNA must consist of at least 90-95 % L-chain cDNA transcripts.

3.2. Detection of CK-like sequences in EL4 and P8 cells When the hexanucleotide-primed Lehain cDNA was hybridized with total cytoplasmic poly (A) containing RNA (total mRNA) prepared from P3 myeloma cells a single hybridization transition occurred (involving around 90 % of the cDNA) with Crt,12 = 3.1 x lO-*(Fig. 2). This transition corresponds to the fastest hybridizing component observed when P3 total cDNA ( i e . cDNA prepared from total cytoplasmic mRNA) was hybridized with P3 total mRNA* and occurs at a Crt value corresponding t o the fastest hybridizing material of EL4 total cDNA annealed with EL4 total mRNA (Fig. 2). This latter hybridization curve shows a wide range of hybridizing components similar to those observed with, for example, mouse L-cell mRNA [ 171, Hela cell mRNA [ 161 and mouse myeloma mRNA*. EL4 total mRNA shows hybridizing components up t o values as high as Crt = 30. When L-chain cDNA was axpealed with the EL4 total mRNA a single hybridization transition was observed but the hybridization differed in two major respects from that observed when the Lchain cDNA was annealed with the homologous P3 mRNA (Fig. 2). Firstly, the final hybridization value observed was only about 35 % compared to 90 % with P3 mRNA and, secondly, the Crtl12 = 9 x 10-l compared with Crtl12 = 3.1 x for P3 mRNA.

"t 80

P

Figure 2. Hybridization of L-chain cDNA and total EL4 cDNA with total mRNA. 5 pg RNA (20 pg/ml) and 3 ng [sH]cDNA were annealed at 70 OC in 0.1 M sodium phosphate buffer, containing 0.05 % SDS. At the Crt values shown, samples were removed and the percentage -5

rn

-4

-3 -2 Log Crt

-1

0

Figure 1. Hybridization of L-chain cDNA and rabbit globin cDNA to the respective mRNA templates. 1 pg of rabbit globin mRNA and 3 ng [3H]L-chain cDNA were separately annealed at 70 O C in 1 ml 0.24 M sodium phosphate buffer pH 6.8, containing 0.05 ISDS.

At the Crt values shown, samples were removed and tiie percentage of cDNA present as hybrid was determined by S1 nuclease digestion. (0-0) globin cDNA vs. globin mRNA; (0-0) L-chain cDNA vs. L-chain mRNA.

of cDNA present as hybrid was determined by S1 nuclease digestion. (A-A) L-chain cDNA us. P3 mRNA; (0-0) L-chain cDNA vs. EL4 mRNA; (0-0) Total EL4 cDNA Js. EL4 mRNA. The curves shown in Fig. 2 were all prepared using single strand specific Si nuclease to measure hybrid formation. This procedure discriminates against both single-stranded tails and mismatched hybrids, so allowing only well-matched

* B. Mechler and T.H. Rabbitts, in preparation.

46

Eur. J. Immunol. 1977. 7: 43-48

T.H.Rabbitts, A. Forster, M. Smith and S. Gillam

hybrid regions t o be scored as hybrid. In view of the low final hybridization value observed when L-chain cDNA was hybridized t o EL4 the hybridization was repeated using less stringent annealing conditions (0.24 M phosphate buffer instead of 0.1 M buffer), and the hybridization was measured using hydroxyapatite fractionation (in which there is n o digestion of single-strand tails). Under these conditions the Lchain cDNA hybridizes efficiently t o both P3 mRNA (Crtl12 = l o b 2 ) and EL4 mRNA (Crtllz = 8 x (Fig. 3). The Crtl,,2 values obtained indicate that this is between 10 and 3 0 times less of the cross-reacting mRNA species in EL4 than P3 mRNA (Figs. 2 and 3). The difference in hybridization efficiency of cDNA with EL4 mRNA, when S I nuclease and hydroxyapatite were used, is similar to a previous observation with mismatched hybrids [ 181. Since the hydroxyapatite conditions are known to tolerate the cross-reaction of sequences at least up to 8- 15 % difference [ 181 the results indicate that EL4 mRNA contains a sequence related to but not identical with the CK sequence. In view of the somewhat unexpected observation of a CK-like sequence in EL4 cells, we examined (a) the possibility that this CK sequence might also be present in other cells of lymphoid origin and (b) the possibility that another unrelated mRNA sequence (globin) might also be present in EL4 cells. The presence of the C K like sequence in a Ch-producing mouse myeloma line (P8) was investigated by incubating P8 mRNA t o high Crt values with L-chain cDNA as shown in Fig. 3. As with EL4 mRNA, a sequence capable of hybridizing with the L-chain cDNA was observed in P8 mRNA with Crtl12 approximately 5 x lo-'. This CrtllJ value indicates that the concentration of the sequence in P8 cells is about 10 times lower than the corresponding sequence in EL4 cells and, therefore, about lO(1 fold lower than in the homologous CK-producing P3 cells.

The common occurrence of this CK-like sequence in two unrelated cell lines indicated the possibility that a general nonspecific transcription might be occurring in these cells. It has been reported that globin mRNA can be detected in nonerythroid tissues [ 191 so we have looked for globin sequences in EL4 mRNA. Fig. 3 shows that no hybridization occurs between globin cDNA and EL4 mRNA at values as high as Crt25. At this point the hybridization of even P8 mRNA and L-chain cDNA is virtually terminated.

3.2. Melting characteristics of mRNA: cDNA hybrids The disparity between the hybridization results obtained using S1 nuclease and hydroxyapatite indicated that we are observing mismatched sequences rather than perfectly complementary hybrids in EL4 and P8 mRNA. The amount of this mismatching was determined by studying the melting temperature of the hybrids. Fig. 4 compares the melting profiles of the hybrids formed by annealing L-chain cDNA with EL4, P3 or P8 mRNA. The homologous P3 mRNA: cDNA hybrid possesses the highest melting temperature displaying a T, (half-point of melting) in these conditions of around 8 3 OC. The EL4 mRNA:cDNA and P8 mRNA:cDNA hybrids melt with profiles similar to each other and with a T, around 77 OC. These latter hybrids, therefore, show about 5 OC lower melting temperatures than the homologous hybrid which probably reflects about 5 % difference from the CK sequence (see Section 4.).

100

70

P i

/ mm

-2

-1

0

1

2

Log Crt

Figure 3. Hybridization of cDNA with mRNA assayed by hydroxyapatite fraction. 3 ng of L-chain cDNA were annealed with 0.38 pg P3 mRNA or EL4 mRNA (3.8 pg/ml) or 2.8 pg P8 mRNA (114 pg/ml) and 3 ng of [JHIglobin cDNA were annealed with 2.8 pg EL4 mRNA (114 pg/ml). The annealing conditions were 70 O C and 0.24 M sodium phosphate buffer, containing 0.05 % SDS. Samples were taken at time intervals and the percentage hybridization assayed by hydroxyapatite fractionation at 65 OC. ( 0 - 0 ) Globin cDNA YS. EL4 mRNA; (A-A) L-chain cDNA YS. P3 mRNA; (A-A) L-chain cDNA YS. P8 mRNA; (0-0) L-chain cDNA YS. EL4 mRNA.

Temperature

"C

Figure 4. Melting profiles of hybrids formed between mRNA and L-chain cDNA. EL4 mRNA, P3 mRNA or P8 mRNA (4 pg) were separately annealed with L-chain cDNA (3 ng) to Crt = 7. The annealing mixtures were diluted to 0.1 2 M sodium phosphate and heated in approximately 4 OC increments from 65 OC to 100 OC. After 5 min equilibration at each temperature, an aliquot was removed and diluted with cold 0.12 M phosphate buffer. The percentage of cDNA remaining as hybrid was determined in these diluted aliquots by hydroxyapatite fractionation at 65 OC. The data has been normalized as a percentage of the hybrid content at 65 OC. (0-0) P3 mRNA:L-chain cDNA hybrid; ( 0 - 0 ) EL4 mRNA:L-chain cDNA hybrid; (A-A) P8 mRNA: L-chain cDNA hybrid.

4. Discussion A sensitive procedure for the detection of Ig RNA in a mixed population of RNA species is t o measure the rate of hybridization of radioactive Ig cDNA with the RNA. If the RNA population contains Ig mRNA sequences, these will hybridize to

T cell messenger RNA

Eur. J. Immunol. 1977.7: 43-48 the cDNA. In addition, if the mRNA is in excess over the radioactive cDNA probe, the rate of hybridization will depend on the concentration of the mRNA [ 161. Since RNA excess hybridization analysis has been employed for the present study it is critical t o know the purity of the cDNA used since contaminating non-Ig cDNA species will hybridize with the corresponding RNA in the population. The L-chain cDNA used in the present study was prepared using a specific hexanucleotide primer and has been estimated t o be in excess of 90 % specific L-chain transcripts. In addition, the hexanucleotide primes at a site towards the 3'-end of the mRNA C region, therefore the cDNA does not contain any 3' non-coding sequences [ I 51. Furthermore the average size of the cDNA population was around 300 bases (unpublished data) so that a high proportion of the cDNA molecules are complementary t o CK sequences only. L-chain cDNA hybridized t o both homologous P3 total mRNA and heterologous EL4 total mRNA with a single transition. In the case of P3 mRNA the hybridization involved a high frequency RNA class (ie. an RNA present in high proportion relative t o other mRNA species in the population). The hybridizing sequence present in EL4 mRNA appears to be present at a concentration between 10 and 3 0 times lower than the concentration of L-chain mRNA in P3 cells. Furthermore two pieces of evidence suggest that the EL4 sequence involved is not exactly homologous with the CK sequence. EL4 mRNA will efficiently hybridize L-chain cDNA when hybrid formation is assayed by hydroxyapatite fractionation but rather inefficiently when assayed by S1 nuclease digestion. This suggests that hybrids are being formed which contain regions of mismatched bases. This possibility was confirmed by the finding that EL4 mRNA:L-chain cDNA hybrids were indeed less stable (i.e. possessed a lower T, value) than P3 mRNA:L-chain cDNA hybrids (Fig. 4). The sequence observed in EL4 mRNA, however, does not appear to be exclusive to T cells since we also detected a sequence capable of hybridizing with L-chain cDNA in P8 myeloma cells (Fig. 3). Since we know that sequences with dissimilarities as great as CK and CA are unlikely to crosshybridize in the conditions used [ 181, the sequence detected in the P8 cells is not the CA sequence. In addition it is unlikely t o be 3' non-coding sequences since our L-chain cDNA does not include such sequences. In fact, the melting temperature of the L-chain cDNA hybrid with P8 mRNA and EL4 mRNA is about the same (Fig. 4) indicating that both cell lines transcribe the same gene sequence. A major point of difference, however, is seen in the concentration of this mRNA species in EL4 and P8 mRNA. In the former case, the mRNA seems to be a relatively high abundance species whereas in P8 cells it appears t o be a low abundance species. The significance of this difference in cellular abundance of the mRNA is obscure. Recently it was reported that low levels of globin mRNA could be detected (using techniques similar t o those employed here) in non-erythroid tissues and tissue culture lines [ 191. This report, whether biologically significant or not, emphasizes the need for caution in interpreting data from this type of experiment. In our own experiments, however, we were unable t o detect hybridization of mouse globin cDNA with EL4 mRNA (Fig. 3). Thus, in this respect, our observations do not

47

appear t o be simply the detection of low levels of generalized gene expression in EL4. The difference between the sequence of L-chain cDNA and the heterologous mRNA species (i.e. EL4 and P8) was quantitated by determination of the melting temperature of the cDNA:RNA hybrids (Fig. 4). The T, of both EL4 and P8 mRNA:L-chain cDNA hybrids was found to be about 5 "C lower than the homologous P3 mRNA:cDNA hybrid. The similarity between the EL4 and P8 melting curves indicate that we are observing a similar component in both cell lines. The difference between these melting points and that of the P3 mRNA:cDNA hybrids represents between 3 % and 6 % difference between the L-chain cDNA and the crossreacting RNA sequence of EL4 and P8 [20, 211. Recently, similar results have been reported with RNA prepared from a thymoma cell line and from T cell-rich mouse cell populations [22]. Although there appears t o be some variation in the apparent concentration of RNA from those described in this paper, there is agreement that the sequence detected in the T cells is not identical with the CK sequence. In addition, this paper [22] describes a RNA sequence present in spleen cells which also shows a low melting point when hybridized with L-chain cDNA. The significance of these differences from the CK sequence is not at all obvious. EL4 cells originated from C57 mice while P3 and P8 cells originated from BALB/c mice. Nonetheless, both P8 and EL4 show the same degree of difference from the P3 sequence. Therefore it is not likely that the difference in sequence is due simply t o the genetic background of the mice strains used. A paper by Boylston and Mowbray [23] reports that surface Ig was extractable from EL4 cells but that attempts to identify the material with anti-IgM antisera were inconclusive. Other papers [ 24, 251 indicate that Ig-like molecules are present on T cell surfaces and that these molecules display some but not all the antigenic determinants of B cell Ig. The specific relevance of our hybridization results t o lg production by T cells is rather dubious since, using hydroxyapatite procedures, we were able t o detect hybridizing sequences in a CA-producing myeloma cell line. Rather the point should be made that the observations demand caution in the interpretation of this kind of experiment. We thank Shirley Howe for growing the tissue culture cells and Dr. J. Beard for providing reverse transcriptase enzyme. We also thank Drs. A.J. Munro and B. Mechler for many helpful discussions.

Received August 18, 1976; in revised form October 22, 1976.

5. References 1 Raff, M.C., Nature 1973.242: 19.

2 Rajewsky, K. and Pohlit, H., in Amos, B. (Ed.) Progress in Immunology, Vol. 1, Academic Press, New York and London 1971, p. 337. 3 Marchalonis. J.J. and Cone, R.E., nansplant. Rev. 1973.14: 3. 4 Benacerraf, B. and McDevitt, H.O., Science 1972. 175: 273. 5 Munro, A.J. and Taussig, M.J., Nature 1975. 256: 103. 6 Vitetta, E.S., Bianco, C., Nussenzweig, V. and Uhr. J.W., J. Exp. Med. 1972. 136: 81. 7 Horibata, K. and Harris, A.W., Exp. Cell Res. 1970. 6 0 61 8 Cory, S. and Adams, J.M., J. MoL Biol. 1975. 99: 519.

48

P. Creemers

Eur. J. Immunol. 1977.7: 48-53

9 Rabbitts, T.H., Jarvis, J.M. and Milstein,C., Cell 1975. 6: 5. 10 Bishop, J.O., Rosbash, M. and Evans, D., J. Mol. Biol. 1974.85: 75. 11 Weber, H. and Khorana, H.C., J. Mol. Bwl. 1972. 72: 219. 12 Gillam, S., Waterman, K., Doel, M. and Smith, M., Nucleic Acid Res. 1974. 2: 1649. 13 Egan, B.Z. and Kelmers, A.D., Methods in Enzymology 1974. 29: 469.

17 Ryffel, G.U. and McCarthy, B.J., Biochemistry 1975. 14: 1379. 18 Rabbitts, T.H. and Milstein, C., Duns. Biochem. Soc. 1975. 3: 870. 19 Humphries, S., Windass, J. and Williamson, R., Cell 1976. 7: 267. 20 Bonner, T.I., Brenner, D.J., Neufeld, B.R. and Britten, R.J., J. Mol. Biol. 1973. 81: 123. 21 Ullman, J.S. and McCarthy, B.J., Biochim. Biophys. Acta 1973. 294: 416.

14 Rabbitts, T.H. and Milstein, C., Eur. J. Biochem 1975. 52: 125.

22 Storb, U., Hager, L., Putnam, D., Buck, L., Farin, F. and Clagett, J., Proc. Nat. Acad. Sci. US 1976. 73: 2461.

15 Milstein, C., Brownlee, C.C., Cheng, C.C., Hamlyn, P.H., Proudfoot, N.J. and Rabbitts, T.H., in 27. Colloquium-Mosbach 1976, SpringerVerlag, Berlin, Heidelberg 1977, p. 75.

24 Hammerling, U., Mack, C. and Pickel, H.C., Immunochemistry 1976.13: 525.

16 Bishop, J.O., Morton, J.C., Rosbash, M. and Richardson, M., Nature 1974.250: 199.

25 Hammerling, U., Pickel, H.C., Mack, C. and Masters, D., Immunochemistry 1976. 13: 533.

Paula Creerners* Radiobiological Institute TNO, R ijswij k

23 Boylston, A.W. and Mowbray, J.F., Immunology 1974. 2 7 855.

The role of leukocyte subpopulations in the indirect leukocyte adherence inhibition assay in the mammary tumor virus system A modification o f the leukocyte adherence inhibition test for the detection of cellular immunologic reactivity of mice t o the mammary tumor virus (MTV) has been described. It involves t h e transfer of t h e leukocyte adherence inhibition factor (LAIF) produced by spleen cells from immunized animals when cultured with antigen t o indicator cells, for which peritoneal exudate cells from normal mice are used. The method proves t o be sensitive a n d highly reproducible. By crude separation of leukocyte subpopulations it became established that for t h e production of LAIF the following sequence is needed: 1. incubation of adherent cells with MTV; 2. transfer of a soluble factor SFI produced by the adherent cells to T cells; 3. transfer of another soluble factor SF2 released by T cells t o the adherent cells.

1. Introduction The leukocyte adherence inhibition (LAI) assay for demonstrating immunologic cellular reactivity t o certain antigens has been developed by Halliday a n d Miller [ 11. They observed that the proportion of cells adhering to glass surface was reduced if antigen was added t o which the animals were sensitized. In this test system, the blocking and enhancing properties of serum could also be established. Later it has been reported that the LA1 assay may be regarded as a depend-

able in v i t r o test for determining immunologic reactivity t o tumor specific antigens in t h e case of melanoma [ 21 and breast cancer [3].

Holt [ 4 ] reported that the LA1 reaction can be inhibited by pretreatment of t h e cells with anti-@ serum and complement and t h a t t h e inhibition of adherence can be transferred t o normal cells by the supernatant of spleen cells incubated with t h e antigen. He therefore concluded that a soluble factor, the T cell-dependent leukocyte adherence inhibition factor (LAIF) is responsible for the inhibition of adherence. His results with the LA1 test did not parallel those of the leuko[I 15241 cyte migration inhibition assay; he therefore suggested that different mechanisms be involved in the two tests. We per* This study was performed under NIH Contract NO1 CP 43328 with formed the LA1 assay with peritoneal cells derived from mice the National Cancer Institute, Division of Cancer Cause and Prevention, Viral Oncology, Department of Health, Education and Welfare, Bethesda, bearing primary mammary tumor virus (MTVtinduced tumors Md., USA, to Dr.Peter Bentvelzen. as indicator cells [ 51 a n d found that the results parallel the MTV-specific proliferation of spleen and lymph node cells Correspondence: Paula Creemers, Radiobiological Institute TNO, which is stimulated by T cells [ 61. 151 Lange Kleiweg, Rijswijk, The Netherlands Abbreviations: LAI: Leukocyte adherence inhibition LAIF: Leukocyte adherence inhibition factor MTV: Mammary tumor virus R L V Rauscher leukemia virus SF,: Soluble factor produced by adherent cells upon contact with MTV SF2: Soluble factor produced by T cells upon contact with SFI PBS: Phosphate buffered saline SE: Standard error

In t h e indirect LA1 assay, which is described below, in vitroinduced reactivity t o MTV is transferred t o normal indicator cells. Evidence is presented that, for LAIF production, an interaction between T cells and adherent cells is necessary, but that direct cell t o cell contact is not required.