Characterization of antigens on human lymphocytes coded by messenger RNA translated in vitro

0161.5890.82.091127.12’$03.00;0 0 1982 Pergamon Press Ltd.

Mol~c~rlov Immunoloyy Vol. 19. No. 9. pp. 1127-l 138. 1982. Printed in Great Britain.

CHARACTERIZATION OF ANTIGENS ON HUMAN LYMPHOCYTES CODED BY MESSENGER RNA TRANSLATED IN VITRO ROBERT *American

TDepartments

W. ALLEN,*

Red Cross

SOLDANO

FERRONEI:

and JAMES A. HOCH

Blood Services,

Missouri&Illinois Region, 4050 Lindell Boulevard. MO 63108, U.S.A. of Cellular Biology and Molecular Immunology, Scripps Clinic and Research 10666 North Torrey Pines Road. La Jolla. CA 92037. U.S.A. (First rrcricrd

6 Junuur~ 1982: uccepred

itr wvised

firm

22 March

St. Louis. Foundation,

1982)

Abstract- Translation and immunoprecipitation were used to identify messenger RNAs (mRNAs) coding for surface antigens expressed on human lymphoblastoid cells. The mRNAs were extracted from several human lymphoid cell lines as well as from hbroblastoid lines. These mRNAs were translated in t’itro, and the translation products were reacted with xenoantisera raised against the antigens on human lymphoid cells. Products immunoprecipitated by these antisera were analysed by electrophoresis and fluorography. Four antisera immunoprecipitated a polypeptide with a mol. wt (MW) of approxtmately 32,000 (~32) from translations programmed with mRNA extracted from all the cell lines. Two antisera immunoprecipitated, in addition to ~32, another polypeptide with a MW of approximately 25.000 (~25) only from translations programmed with RNA from lymphoid cell lines. ~25 mRNA in the different lymphoid cell lines fell into three basic abundance classes as determined by in rim translation and immunoprecipitation. Cells from two Burkitt’s lymphomas (Raji and Daudi) did not express detectable p25 mRNA. Two T-lymphoblastoid lines (Molt-4 and 1301) contained five- to IO-fold less p25 mRNA than the B-lymphoid cell lines (Victor, RPMI-8866. RPMI-6410. RPMI-8226 and RPMI-1788). Both p32 and p25 were expressed on the cell surface inasmuch as lymphoblastoid cells adsorbed antibodies to both polypeptides. Human fibroblast. Raji or Daudi cells adsorbed anti-p32 antibodies from the antiserum but not anti-p25. Quantitative absorptions of the antiserum wtth T- or B-lymphoblastoid cells was used to determine the relative amounts of p32 and p25 expressed on the cell surface. B-Lymphoblastoid cells were found to express two- to five-fold more p25 on the cell surface than T-lymphoblastoid cells. p25 does not represent an immunoglobulin light-chain precursor inasmuch as a IOOO-fold excess of unlabeled human Ig did not compete wtth p25 translated irl rirro for binding by Its respective antibody,

INTRODUCTION

The diverse surface antigens expressed by mammalian cells fall into two general categories: those present on virtually all cells and those restricted to a single or limited number of cell types. Surface markers expressed on subpopulations of hemopoietic cells reflect their distinct developmental commitments that ultimately result from programmed patterns of gene expression (Till & McCulloch, 1980; Katz, 1977). The developmental expression of surface markers during lymphopoiesis has been analysed extensively (McKenzie & Potter, 1979; Boyse & Old, 1978). Although the temporal expression of markers on T-cells (Crouse et ml., 1980) and B-cells (Hammerling et a/., 1975, 1976) has been defined, little is known of the t Present address: College of Physicians & Surgeons of Columbia University. 630 W. 168 St., New York. 10032, U.S.A.

molecular mechanisms that modulate marker synthesis. The expression of surface Ig during B-lymphocyte development has been analysed genetically using recombinant DNA techniques that make available cloned DNAs for use in the identification and/or quantitation of gene sequences or messenger RNA (mRNA). We now know that control over surface Ig expression is exerted at both the DNA and mRNA levels (Tonegawa et al., 1977; Sakano et al., 1979; Perry rt al., 1980~~ h; Rogers rt al., 1980). The mRNAs that code for antigens on human lymphocytes reflect the developmental program of gene expression in lymphoid cells and are a source of nucleic acid probes to study hemopoiesis. Through the use of in vitro translation and immune precipitation with xenoantisera raised against human lymphoid cell surface antigens. we have identified and partially characterized mRNAs that code for two such antigens. One of these mRNAs is 1127

1128

ROBERT

W. ALLEN.

SOLDANO

present in virtually all human cell lines thus far examined, and the other is restricted in its distribution to cells of lymphoid origin.

MATERIALS AND METHODS Cell

CLlltLlre

All cell lines were maintained in 25 or 75-cm2 flasks (Corning) in RPMI-1640 medium supplemented with 10% heat-inactivated horse serum (Irvine Scientific). Cultures were incubated at 37’C in a humidified atmosphere of 5% CO,:95% air. Cell lines were subcultured by dilution every 334 days; in the case of anchorage-dependent lines, subculturing was done with the aid of 0.25% trypsin (Difco Laboratories, Detroit, Michigan-l :250). The lymphoblastoid cell lines used in this study were obtained from Dr Jun Minowada of Institute, the Park Memorial Roswell American Type Culture Collection, or stocks maintained at the Research Institute of the Scripps Clinic. Mass cultures of lymphoblastoid cells were propagated in 2-l. Erlenmeyer flasks with screw caps and shaken gently. Monoclonal hutsan cells

and polyclonul

xrnountihodirs

to

The monoclonal antibodies Q2/70, Q2/80, Q5/6 and Q5/13 are directed to distinct antigenit determinants expressed on subpopulations of human-Ia-like antigens and different from those defining the serological polymorphism of this system. The method of preparation and the serological and immunochemical specificities of these antibodies have been described elsewhere (Quaranta et al., 1980). Antisera 3634 and 3636 are from rabbits immunized with antigens solubilized from the lymphoid cell lines Wil-2 and RPMI-1788, respectively, by 3 M KC1 and partially purified by ultracentrifugal flotation in KBr (Ferrone rt al., 1977). Both antisera, as shown serologically and immunochemically, contain antibodies to the heavy chain of HLA-A, -B and -C antigens and to Ia-like antigens (Ferrone rt cd., 19780, h). In particular, xenoantiserum 3634 has been used to purify human Ia-like antigens whose primary structure bears a high degree of homology with murine I-E coded subregion antigens (Allison et ul., 1978). Antiserum 6521 is from a rabbit hyperimmunized with the cultured human T-lymphoid cell line Molt-4.

FERRONE

and JAMES

A. HOCH

Antiserum 8823 is from a rabbit immunized with Ia-like antigens isolated from an NP-40 extract of cultured human B-lymphoid cells by binding to an immunoabsorbant composed of Staphqdococcus uureu~ Cowan I strain (SAC) bacteria and polyclonal xenoantibodies to antiIa-like antigens (Wilson et al., 1979). According to serological and immunochemical assays, this antiserum contains antibodies to human la-like antigens. Absorption

of antisrra

One hundred microliters of a 1: 10 dilution of antiserum was absorbed for 2 hr at O’C with 1 x 10h cells of all the lines tested. The cells were removed by centrifugation after which 20~1 of the supernatant was reacted with products from translations programmed with Victor RNA. For absorption with erythrocytes, equal volumes of undiluted antiserum and packed, washed erythrocytes were mixed and allowed to stand at 0, C for 2 hr. The erythrocytes were obtained from heparinized blood collected from healthy volunteers. Two microliters of absorbed antiserum was reacted with products from translations programmed with Victor RNA in these experiments. RNA extraction RNA was extracted from the various cell lines with the basic method tha Cox (1968) described in which cells were suspended in 8 M guanidine-HCl (Eastman-Kodak) in 25 mM potassium acetate, pH 5.0, at lo:,;, (w/w). The suspension was transferred to a Dounce homogenizer fitted with a tight pestle and given approximately 50 strokes. One half-volume of 9.5% ethanol was added and the RNA was precipitated during a 2-12-hr period at -20’C. The nucleic acid pellet was resuspended in 25 mA4 EDTA, pH 7.0, and excess guanidineHCl was removed with an equal volume of chloroform:n-butanol (4: 1). RNA was precipitated from the aqueous phase with two volumes of 4.5 M potassium acetate, pH 6.0, overnight at -2O’C. RNA was resuspended at 2 mg/ml in sterile distilled HZ0 and was stored in liquid nitrogen. RNA ,fractionution

011 Sepharow

CL-4B

A 2.5 x 90cm column of autoclaved Sepharose CL-4B (Pharmacia, Piscataway, New Jersey) was poured and equilibrated at room temperature in 0.1 M sodium acetate, pH 5.0,

Lymphocyte Antigen Messenger RNA

containing 1 mM EDTA. Total RNA (10 mg), dissolved in 4 ml of buffer, was heated at 70-90°C for 2 min and quickly chilled before loading on the column. Two milliliter fractions were collected and their ODZeO measured. Fractions containing RNA were brought to 0.2 M NaCl and were precipitated with two volumes of cold 95% ethanol during overnight incubation at -20°C. The RNA pellets were washed once with cold 70% ethanol, dried under N,, resuspended in sterile distilled water at 1 mg/ml and translated in the standard manner. Agarose gel electrophoresis ence of methylmercury

of RNA

in the pres-

Agarose gel electrophoresis of RNA was carried out in a fume hood in 1.2% gels containing 10 mM methylmercury (Alfa Ventron, Danvers, Massachusetts). Gels were equilibrated in 50 mM borate buffer, pH 8.2, containing 10 mM sodium sulfate. Five to ten micrograms of size-fractionated or total RNA was loaded per sample well and electrophoresed overnight at a constant voltage of 40 V with continuous buffer recirculation. Gels were stained for 30 min in 0.5 M ammonium acetate containing 1 lg/ml ethidium bromide (Bio-Rad, Richmond, California) and destained under running water for l-2 hr. Stained RNA was visualized under short-wave U.V. light and was photographed with Polaroid Type 51 film. In vitro protein

1129

with total RNA at 300/1g/ml bated at 23%25°C for 2-3 hr.

and

were incu-

Immunoprecipitution

Completed translations were made to 50 mM potassium phosphate buffer. pH 6.8, 0.2 M NaCl, l:/, NP-40, 1 mM leucine and 1 mg/ml bovine serum albumin (BSA). The translation products were pretreated with 100 ~1 of packed SAC per milliliter of translation for 15 min at room temperature to remove translation products that reacted nonspecifically with the absorbant. The SAC were prepared as described by Kessler (1975). Preabsorbed translations were reacted with either 2 ~1 of normal rabbit serum or 2 ~1 of antiserum for 2 hr at room temperature. Antigenantibody complexes were removed from solution with 20~~1 of loo/;, SAC in phosphate buffer containing 1’:; NP-40. O.ll,, sodium dodecyl sulfate (SDS), 1 mM leucine or methionine, and 1 mg/ml BSA. Immune complexes were eluted from the SAC into 10 mM sodium phosphate, pH 7.0, 1”: SDS and lo:, mercaptoethanol during a 5-min incubation in boiling water. The polypeptides were analysed on 12.55; polyacrylamide gels containing O.lo/, SDS by Laemmli’s (1970) method. Gels were stained with Coomassie blue, destained, impregnated with Enhance@ (New England Nuclear Co.. Boston, Massachusetts), dried, and exposed to X-ray film (Kodak. XRP) at -70°C for 5 days.

synthesis

The translation system of Roberts & Patterson (1973) derived from wheat germ (General Mills, Vallejo, California) was used with only minor modifications. The concentrations of components were as follows: 1.5 mM magnesium acetate 4Hz0, 120 mM potassium acetate, 20mM HEPES (Sigma Chemical Co., St. Louis, Missouri), pH 7.6, 2 mM dithiothreitol (Sigma), 1 mM ATP, 20 PM GTP, 40pg/ml creatine kinase, 8 mM creatine phosphate (the latter four from Boehringer Mannheim, Indianapolis, Indiana), 120 PM amino acids (except leucine), 0.4 mM spermidine (Sigma) and C3H]leucine (250-500 &i/ml) (Amersham, Arlington Heights, Illinois, 136 Ci/mmole). The standard volume of translation was 100 ~1. Before the initiation of translation, wheat germ lysates were made mRNA-dependent by digestion of endogenous mRNA with micrococcal nuclease as described by Pelham & Jackson (1976). Translations were programmed

Oliyo dT cellulose

(hromatograph~

Poly A+ RNA was absorbed at room temperature onto oligo dT cellulose (Collaborative Research) in 10 mM Tris-HCl, pH 7.5, containing 0.5 M NaCl and 1 mM EDTA. The column was washed with this buffer until the ODzho dropped below 0.05. Poly A + RNA was eluted from the resin with sterile distilled water, brought to 0.2 M NaCl, and precipitated with two volumes of cold 95”<, ethanol.

RESULTS In vitro

translation

qf’ mRNA

jkom

Iymphohlas-

toid cell lines

Our initial experiments were designed to detect mRNA coding for lymphocyte-specific surface antigens. The mRNA extracted from Victor B-lymphoblastoid cells was used to program the in vitro wheat germ translation sys-

1130

ROBERT

W. ALLEN,

SOLDANO

tern in the presence of C3H]leucine used at high specific activity to label the products of translation. Putative lymphocyte-specific products programmed by this mRNA, first immunoprecipitated with xenoantisera or monoclonal antibodies, were, subsequently, detected fluorographically in dried SDS-polyacrylamide gels. Six xenoantisera known to contain antibodies to HLA-A, -B, -C and -DR antigens were used to immunoprecipitate such translation mixtures. Of these. four precipitated several products in addition to those precipitated by normal rabbit serum (Fig. 1). One such product was a polypeptide with a nominal mol. wt (MW) of 32.000 (~32). although the anti-p32 titer varied from serum to serum. Two other antisera (3634 and 6521) precipitated a polypeptide of approximately 25.000 MW (~25) (Fig. I. lanes B and C). Faintly detectable polypeptides in the 70,000 MW region of the gel were also precipitated by the four xenoantisera. However. the faintness of these bands made rigorous identification and analysis difficult. Antiserum 3634 precipitated from the surfaces of Victor cells both the X- and P-chains of HLA-DR antigens and three additional polypeptides of nominal MWs 34.000, 27,000 and 26,000 (Quaranta et ol.. 1980). However, four different monoclonal antibodies to human Ialike antigens (Wilson et 01.. 1980) failed to precipitate any translation products coded by mRNA from the Victor cells in our experi-

24,000

FERRONE

and JAMES

A. HOCH

ments. Thus, if p32 and ~25 are related to Ialike antigens they are not recognized by any of the monoclonal antibodies to human la-like antigens tested, or by several xenoantisera to these antigens that also failed to react with p32 and ~25 synthesized irk rittv. Therefore. the presence of antibodies to p32 and ~25 in the xenoantiserum primarily directed toward human Ia-like antigens may bc fortuitous.

An mRNA preparation from Victor B-cells was chromatographed on Sepharose CL-4B, and the fractions were analysed by i/l vitro translation and immunoprecipitation with 3634 antiserum. RNA eluted from the column in two major peaks (Fig. 2A) corresponding to 28 and 1% ribosomal RNA as determined by mcthylmercury agarose gel electrophoresis (Fig. 2C). A small peak that eluted before 28s RNA probably represented contaminating DNA fragments, because these fractions did not program detectable translation products in rite. The mRNA for p32 appeared most abundant in fractions eluting after 1% RNA and before fractions containing mRNA for ~25 (Fig. 28). Size estimates for mRNA in column fractions enriched for the p32 and ~25 messages were determined from electrophoretic migration of RNA in agarose gels containing methylmer-

-p25

amPalwmrrrQo A B C

D

E

Fig. I. Translation products. synthesized from lymphoid cell (Victor) mRNA rextiLe wth xenoantisera to B-lymphoid cells (see Materials and Methods). Mol. wt markers shown at the left include: bovine serum albumin (66.000), ovalbumin (43,000) pcpsinogen (35.000) and trypsinogen (24,000). The products illustrated resulted from precipitation with: lane A~ normal rabbit serum. lane B -antiserum 3634. lane C-antiserum 6521, lane D antiserum 3636, lane E- -antiserum 8823.

Lymphocyte

Antigen

cury. The results of such an analysis are shown in Fig. 2C. Column fractions enriched for the mRNA coding for p32 migrate with an approximate MW of 6 x lo5 corresponding to approximately 1800 bases. The mRNA coding for ~25 migrates with an approximate MW of 4 x 10’ or approximately 1200 bases. Thus, p32 and ~25 appear to be coded by unique mRNA molecules. In addition, the results of chromatography on oligo dT cellulose indicate that both mRNAs are polyadenylated (data not shown). Distribution c?j’tnRNA coding jbr p32 und ~25 in ,sezvruI cell linrs Although both p32 and p25 were immunoprecipitated by antiserum directed toward lymphocyte cell surface antigens, no evidence existed that their synthesis was lymphocyte-

Messenger

RNA

1131

specific. Therefore, we sought mRNAs coding for p32 and ~25 among cell lines of several phenotypes. The mRNAs extracted from these cells were translated in vitro, and the translation products were reacted with antiserum 3634 with the results shown in Fig. 3. The mRNAs from the Victor, RPMI-8866, RPMI-6410, RPMI-1788 and RPMI-8226 B-lymphoblastoid cell lines coded for both the p32 and p25 translation products, all with equivalent profiles and quantities (Fig. 3, lane B). Comparatively, mRNA from Molt-4 and 1301 T-lymphoblastoid cells coded for an equivalent amount of the p32 product but coded for much less of the ~25 product (Fig. 3, lane C). The nonlymphoid HeLa cell line had mRNA that coded for p32 but little or no ~25 (Fig. 3, lane E). Table 1 shows further data from these

3.Or (A)

1

120

I

130

Fraction

140

150

I

160

No

(B) 66,000

-

43,000

-

24,000

-

A

BC

DE

Fig. 2. (A) and (B). Lrywd

FG OH p. 1132.

A

170

1132

ROBERT

W.

ALLEN,

SC&DAN0

FERRONE

and JAMES

A. HCXH

ABCDEFGHIJKLMN Fig. 2. IC). Fig. 2. (At ~ra~t~onat~o~ of Victor RNA. Total RNA extracted from Victor felts u’as iiaetionated on Sepharose CL-4B as described in Materials and Methods. Two-miiiiiiier fractions were collected and RNA was quantitated by absorbance at 260 nm. [B) ~mmunopre~~p~~t~on 4vith antiserum 3634 of translation products synthesized in &r-o using mRNA fractions eluted from Sepharosc CL-45. Polypepunfractionatcd RMA; lanes tides immunoprecipitated from translations programmed with: lane A B-G mRNA from fractions 133, 138, 144. 150. I56 and 162.respectively. The mol. wt markers are as described in the legend of Fig. 1 except for the lack of pepsinogen (35.000). (C) Auxlysis of RNA from selected column fractions in 1.2”,, agdrosc gels containin g methylmercurq. Five micrograms of RNA from selected fractions of the Sepharose CL-4B cluate were clectrophorcsed in l.?“,, agarose gels containing iOmM methyimer~~lry as described in Materials and Methods. Additions to the legend for Fig. 2C are as foitaws. Sample lanes: lane Am-total RNA: lanes B, Do I(. M and N contain RNA from column fractions I@. 1% 152 I44 I%. 132. 126. IX, 114. $08 and IO2 respectively. Moi. WI markers are in lanes C and L and are also represented by the arrows to the left of the figure. From top to bottom. the mol. wts (x Ii)“) are 2.59, 2.30, 1.94. 1.5% 1.12. O.XY. 0.63, 0.60. 0.48. 0.44 :md 0.36.

cell lines compared to mRNAs from Raji and Daudi B-lymphoblasts originally from Burkitt’s lymphomas; their mRNAs coded for p32 alone. These were the only lines of B-celt origin tested whose mRNA did not code for detectable arnoL~~ts of p25. Add~tio~ally~ normal diploid human fibroblasts (HN-10) had mRNA with a profile identical to that of the HeLa cells (Table 1). The relative abundance of p32 and ~25 mRNAs in B- and T-lymphoid cell Iines was quantitated by sci~ltiilation counting of the radioactivity in p32 and ~25 after they were excised from the dried ~oly~crylamjde gels (Table 2). The radioactivity incarporated into p32 was approximately equivalent in translations programmed with RNA extracted from either Victor or Molt-4 cells. In contrast fiveto IO-fold more ~25 was synthesized in translations programmed with Victor RNA than in transIations programmed with MoIt4 RNA. The relative amounts of ~32 and $5 mRNA in T and B-lymphoid cells were also analysed

densitomctrically. The individual lanes of autoradiographs of dried gels containing side by side immunoprecipitates from translations pro~r~~mmed with Victor or Molt-4 mRNA were scanned with a recording densitometer. The positions of ~32 and ~25 bands on the densitometer tracing were noted and the area under these peaks was calculated. The results of this analysis (not shown) were similar to the counting results discussed earlier; namely the ~32 bands were of equivalent density in aL~t~~rad~ogr~lrns of immunoprecipitates programmed with either Victor or Molt-4 mRNA whereas the density of ~25 bands in immunoprecipitates programmed with Victor RNA was approximately five-fold greater than translations programmed with Molt-4 mRNA.

Studies of mRNA abundance suggested that ~32 is ubiquitous, whereas ~25 is expressed

Lymphocyte

Antigen

Messenger

RNA

1133

~32 24,000

~25

-

B

A

C

DE

Fig. 3. Immune precipitation of translation products programmed with RNA extracted cultured cell lines. RNA was extracted from cell lines and translated in aim as described and Methods. Mol. wt standards are as in Fig. I. Lane A-representative profile of products with normal rabbit serum from translations programmed with mRNA extracted from T-lymphoid cell lines, lane B-antiserum 3634/mRNA from Victor B-cells, lane 3634/mRNA from Molt-4 T-cells, lane D-normal rabbit serum/mRNA from HeLa cells, serum 3634/mRNA from HeLa cells.

predominantly on cells of B-lymphoid origin. Since the antisera we used to recognize p32 and ~25 were raised against cell surface components it was important to determine if p32 and ~25 were actually located on the cell surface. If so, adsorption of the 3634 antiserum with cells expressing p32 and ~25 should remove the antiserum’s reactivity with p32 and/or ~25 synthesized in tlitro. As anticipated, adsorption of the antiserum with Victor cells removed the antibodies for both p32 and ~25 programmed by Victor mRNA (Fig. 4, lane D),

Table

I. Distribution

of mRNA

Cell line used for InRNA isolation

coding

from various in Materials precipitated either B- or C-antiserum lane E-anti-

indicating that both of these proteins are components of the cell surface and exposed to antibody. Adsorption with 18 other B-cell lines, including 11 from Burkitt’s lymphomas, three cell lines originating from Epstein-Barr virus transformed peripheral blood B-cells, and the four B-lymphoblastoid lines shown in Table 2 gave identical results, i.e. removal of antibodies to both p32 and ~25. Adsorption of antiserum with the T-lymphoblastoid cell line Molt-4 (Fig. 4, lane C) or 1301 (data not shown) also removed antibodies to both p32 and ~25 indi-

for p32 and ~25 among Origin

lymphoid

and nonlymphoid

cell lines

Relative mRNA coding capacity* D32 P25

Victor, RPMI-8866, RPMI-6410, RPMI-8226 RPMI-1788

B-lymphoblastoid

High

High

Molt-4, 1301

T-lymphoblastoid

Low

High

Raji, Daudi

Burkitts lymphoma

Not detectable

High

HN-10

Normal diploid skin fibroblast

Not detectable

High

HeLa

Cervical carcinoma

Not detectable

High

*The relative coding capacity was determined empirically from the intensities of the p32 and ~25 bands in autoradiographs after 5 days of exposure. The total amount of acid-precipitable radioactivity was comparable in each translation and the designation of coding capacity for a particular cell line was straightforward (see Fig. 3 for example).

1134

ROBERT

W. ALLEN,

SOLDANO

FERRONE

and JAMES

2. Radioactivity incorporated into ~32 and p25 extracted from B- or T-lymphoblastoid

Table

translated cells

Counts Source Victor

Molt-4

of mRNA

Experiment

Number*

(T-lymphoid)

from

per minute

mRNA

of 3H-

~32+

~25+

2876

2017

2

4260

3828

1

2904

478

2

4431

394

1

(B-lymphoid)

A. HOCH

*In both experiments equivalent amounts of Victor or Molt-4 mRNA we!-c used to program translations which resulted in approximately 4 x lOi cpm of total TCA-precipitable radioactivity per IOO-btl assay. tThe radioactive p32 and ~25 bands were located in the dl-ied SDS gel with the aid of the exposed autoradiogram. Once located. the bands were excised from the gel. rchydratcd. solubilized with Protosol at 37 C. and counted in a liquid scintillation counter.

eating that the ~25 is present on the surface of T-cell lines even though the mRNA for p25 appears to be less abundant. Quantitative absorption experiments were performed using antiserum 3634 and Victor or Molt-4 cells in an attempt to determine the relative amounts of p32 and ~25 expressed on the cell surface. Aliquots of antiserum, absorbed with different numbers of Victor or Molt-4 cells, were tested for residual antibody to p32 and ~25 translated in @it~o from Victor mRNA. The results of these experiments consistently showed that B- and T-lymphoid cells

24,000

express roughly equal amounts of p32 but differing amounts of ~25; with B-lymphoid cells expressing more ~25 than T-lymphoid cells. The relative increase in ~25 expression on Victor cells varied from two- to five-fold greater than ~25 expression on Molt-4 cells. The results of an experiment in which Victor cells expressed about five-fold more ~25 than Molt-4 cells are shown in Fig. 5. It is seen that the end point for detection of anti-p32 antibodies occurs when 1 x IO’ cells of either cell type are used to absorb 10 ,~tlof serum. The end point of detectable anti-p25 antibodies is five-

*

A

I3

C

D

E

Fig. 4. Presence of p32 and p25 on the surfaces of lymphoid and nonlymphoid cell lines. Antiserum absorbed with the different cell lines was reacted with translation products whose synthesis was programmed with Victor mRNA. Mol. wt standards are as in Fig. I. Products precipitated with: lane A- normal rabbit serum, lane B-unabsorbed antiserum, lane C antiserum absorbed with Molt-4 T-cells. lane D---antiserum absorbed with Victor B-cells, lane E antiserum absorbed uith HeLa cells.

Lymphocyte

66,000

-

43,000 35,000’

-

24,000

-

Antigen

Messenger

1135

RNA

~32 P25

18,000I5,000ABCDEFGHIJ Fig. 5. Quantitative absorption of antiserum with Victor and Molt-4 cells. One hundred microtiter aliquots of I : IO diluted antiserum were absorbed on ice with different numbers of Victor or Molt-4 cells as described in Materials and Methods, Twenty microliters of the absorbed sera were then used to immunoprecipitate translation products synthesized from Victor mRNA. Products precipitated with: lane A-normal rabbit serum: lane B-unabsorbed antiserum; lanes C-F-antiserum absorbed with I x 10’. 5 x IO”, I x IO6 and 5 x IO’ Victor cells; lanes G~J~antiserum absorbed with I x 105, 5 x 105. I x 10” and 5 x IO’ Molt-4 cells.

fold higher for Molt-4 cells indicating five-fold less p25 expressed per cell in this experiment. As was mentioned earlier. however, the results of the quantitative absorption experiments for p25 expression varied. The reason for such variation is not clear at this time but may result from differences in the growth status of the absorbing cells. i.e. stationary vs logarithmically growing cells. Antibodies to both p32 and p25 were also adsorbed with populations of B- and T-lymphocytes isolated from peripheral blood (data not shown); thus, both proteins apparently are constituents of normal lymphocytes and not trivial consequences of growth in culture. In similar experiments with HeLa cells, adsorption removed antibody to p32 but not p25 (Fig. 4, lane E). Undoubtedly, then, HeLa cells do not synthesize ~25. The only B-lymphoblastoid cell lines lacking detectable mRNA for p25 were Raji and Daudi, both of which removed the antibody to p32 but not to p25 (Fig. 6). All of the B-cell lines lacking p25 by mRNA analyses and/or adsorption tests originated from Burkitt’s lymphomas. Neither p32 nor p25 was present on erythrocytes, since pooled peripheral red blood cells did not adsorb antibodies for either protein from antiserum 3634 (data not shown).

Competition bet\\‘een p25 and human IgG for anti-p25 binding The similar migration in SDS-PAGE of p25 and the immunoglobulin light chain raised the possibility that p25 might be an immunoglobulin light chain precursor reacting with antiimmunoglobulin antibodies present in the antiserum. To test this possibility an aliquot of translation products programmed with Victor RNA was mixed with excess non-immune human IgG before adding antiserum. Antiserum was then added and the immune precipitates were analysed by SDSPAGE and fluorography in the standard manner. As shown in Fig. 7, the excess human Ig light chains do not compete with either p25 or p32 for binding to their respective antibodies.

DISCUSSION

Through the combined use of in citro translation and immunoprecipitation, the mRNAs coding for two human cell surface antigens, p32 and ~25, were identified and partially characterized. The mRNAs coding for p32 and p25 were polyadenylated and separable with Sepharose CL-4B. We estimate the MW of p32 mRNA to be 6 x 105, corresponding to about 1800 bases, and that of p25 mRNA to be

1136

ROBERT

W. ALLEN,

SOLDANO

FERRONE

and JAMES

A. HOCH

-p32

24,000

4~25

-

A

B

C

D

Fig. 6. Absorption of antiserum 3634 with Raji cells. After absorption with Raji cells or Victor cells (Materials and Methods) the antiserum was reacted with translation products synthesized in tin-o using Victor RNA. Products precipitated with: lane A-normal rabbit serum, lane B-unabsorbed antiserum, lane C-antiserum absorbed with Victor cells, lane D-antiserum absorbed with Raji cells.

4 x 105, corresponding to about 1200 bases. These size estimates were determined from the electrophoretic migration of mRNA from enriched fractions relative to single stranded DNA markers in agarose gels containing methylmercury. Both mRNAs are significantly larger than necessary to code for their respective polypeptides even considering a probable signal sequence (Blobel & Dobberstein, 1975) and a 3’ poly A tail. Nontranslated regions have been reported for other mammalian mRNA molecules (Majzoub et al., 1979; Maurer, 1980). The mRNA coding for p32 was present in approximately equal amounts in all cell lines examined. In contrast to the ubiquitous distribution of p32 mRNA, the mRNA coding for ~25 was restricted to cell lines that originated from lymphoid precursors. The lack of p25 on several cell lines of nonlymphoid origin suggests that p25 may be a marker for lymphoid cells. The B-lymphoid cells had five- to IO-fold more p25 mRNA than T-lymphoid cells when measured by scintillation counting of radioactive p25 bands excised from dried polyacrylamide gels or by scanning autoradiograms with a densitometer. In both analyses, the total amount of protein synthesized in vitro and the relative synthesis of p32 were comparable. The reduced amount of p25 mRNA in T-lymphoid cells detected by in vitro translation may be the result of either a reduced rate

of transcription of ~25 mRNA from its structural gene or a mechanism which modifies ~25 mRNA in such a way that it is rendered unavailable for translation. In the latter case equivalent amounts of p25 RNA would be present in B- and T-lymphoid cells, but the majority of p25 RNA in T-lymphoid cells would be modified such that it is unavailable for translation either in the cells themselves or in translation systems derived from wheat germ or rabbit reticulocytes. Once a cloned DNA probe complementary to ~25 mRNA is available, it should be feasible to address more directly these two possible regulatory mechanisms. The relative abundance of p32 and p25 mRNAs in T- and B-lymphoid cells extended roughly to the amounts of each antigen expressed on the cell surface. Equivalent amounts of p32 were expressed on T- and B-lymphoid cells whereas B-lymphoid cells consistently expressed more ~25 than did T-lymphoid cells. The variation in ~25 expression as determined by quantitative absorption is not totally clear at this time but may reflect either differences in the amount of target p25 synthesized in vitro in individual experiments or may reflect modulation of p25 expression by the absorbing cells as a function of growth status, i.e. stationary vs logarithmic. Two B-lymphoid lines, Raji and Daudi, did not contain detectable p25 mRNA and were

Lymphocyte

24,000

15,000

Antigen

-

-

A

B

C

Fig. 7. Competition between ~32 and ~25 and excess human immunoglobulin for immunoprcapitation by antiserum 3634. Translation products synthewed from Victor mRNA were divided into three equal aliquots. Aliquots were reacted with normal rabbit serum (lane A). antiserum 3634 (lane B). or 5 /~g of non-immune human IgG followed by antiserum 3634 (lane C) and processed for analysis in the standard manner.

incapable of absorbing anti-p25 antibodies. Since Raji and Daudi cells originated from Burkitt’s lymphomas, the question arose as to whether all these tumors fail to express ~25. However, screening of 12 other Burkitt’s lymphomas by absorption analysis showed that 11 of the 12 lines did absorb anti-p25 antibodies, indicating that lack of ~25 expression is not characteristic of this particular type of tumor. The fact that Daudi cells do not express ~25 or b,-microglobulin might suggest that, like HLA-A, -B and -C antigens, ~25 is expressed on the cell surface only in association with p,-microglobulin. However, this seems unlikely considering that no mRNA coding for p25 is detectable in Daudi cells and that Raji cells, which express P,-microglobulin, do not express detectable levels of ~25. Quaranta et al. (1980) reported the existence of multiple antigenic forms of human Ia-like antigens. These authors found that when all Ia-like antigens were immunologically depleted from lysates of Victor cells with a murine monoclonal antibody. polypeptides the size of

Messenger RNA

1137

the Ia-like antigens remained that precipitate with antiserum 3634. We have used the procedures of Quaranta et aI. (1980) as an initial approach to identify the matured counterpart of ~25 expressed on the cell surface. We have also found antiserum 3634 to immunoprecipitate a complex pattern of surface antigens in the 25,00&35,000 MW range. The array of polypeptides immunoprecipitated with the antiserum has proven too complex for a definitive identification of the processed counterparts of p32 and ~25. However, our results suggest that the processed counterpart of ~25 may be a polypeptide of MW approximately 27,000 (unpublished results). We are currently analysing immunoprecipitates from labeled cell lysates and in vitro translations with two-dimensional gel electrophoresis which should aid in the identification of the processed forms of both polypeptides. A slight increase in MW of the fully mature form of ~25 resulting from glycosylation might be expected. Similar increases in the MW of processed surface antigens vs their precursors synthesized in vitro have been reported for HLA-A, -B and -C antigens (Ploegh rt al., 1979) the alpha and beta chains of human Ialike antigens (Korman rt ul., 1980) and mouse thymus leukemia antigen (Slomski & Cohen, 1981). Whether or not the p32 and ~25 proteins are related to Ia antigens cannot be determined from this study. The results of the competition experiment with human Ig do, however, indicate that neither p32 nor p25 are related antigenitally to the immunoglobulin light chain or its precursors. Regardless of the true identity of ~25, it is present in cells of lymphoid origin and may be considered a marker of hemopoietic differentiation. Since antiserum 3634 recognized HLA antigens as well as HLA-DR antigens, it was surprising that these polypeptides were not immunoprecipitated from our in vitro translations of lymphoid mRNA. Undoubtedly these polypeptides are synthesized in our translations, but, as several investigators have noted, surface antigens translated irz vitro do not necessarily resemble antigenically their processed counterparts on the cell surface (Ploegh et al., 1979; Korman et al., 1980). Although antiserum 3634 contains high titers of anti-Ia xenoantibodies, these antibodies may react with antigenic determinants dependent upon glycosylation, subunit association. and insertion into the cell

1138

ROBERT

W. ALLEN.

SOLDANO

membrane (Springer et ~1.. 1977; Lancet at nl., 1979) and consequently would not precipitate these polypeptides from the translations. The same would hold true for the inability of the antiserum to precipitate HLA-, -B and -C antigens translated in vitro.

FERRONE

and JAMES

A. HOCH

Lancet D.. Parham P. & Stromillgcr J. L, (1979) Heavy chain of HLA-A and HLA-B antigens is conformationally labile: a possible role for {~z-microglohulin. Proc. nutn. Acud. Sci. U.S.A. 76, 3844 3848. McKenzie I. F. C. & Potter T. (1979) Murinc lymphocyte surface antigens. .4dr. Inntiuu. 27, 179 3.78. Majzoub J. A., Kronenberg H. M.. Potts J. T.. Rich A. K: Nabener J. F. (1979) Identification and cell-free translation of mRNA codine for a nrecuriror of oarathvrotd secretory protein. J. hi;. Ciirrn: 254, 7449 7455. . Acicfloizird~cm~11rs -The authors express their thanks to Maurer R. A. (1980) Imm~~nochemical isolation of prolacKenton Slaven for excellent technical assistance. R.W.A. is tin messenger RNA. J. hicrl. Cltc~tri.2SS, X54--859. the recipient of a post-doctoral fellowship No. PF1486 Pelham H. R. B. & Jackson R. J. (1976) An clficient mRNA from the American Cancer Society. This research was supdependent translation system from reticulocytc lysates. ported, in part. by grants GM 19516, GM 25891, AI 19189 Eur. J. Bio~hcvz. 67, 247 256. and CA 28555 awarded by the National Institute of GenPerry R. P., Kelley D. E.. Coleclough C., Setdman J. G.. eral Medical Sciences. PHSDHHS, the National Cancer Leder P., Tonegawa S.. Matthyssens G. 6i. Weigcrt M. Institute PHSIDHHS. (1980tr) Transcription of mouse K chain gents: imnlications for allelic exclusion. P,nc. c~trrrt. At’c;;i. Sr,,. f/.S.A. 77, 1937-1941. REFERENCES Perry R. P., Kelley D. E., Coleclough C., Scidman J. G.. Leder P.. Tonegawa S., Matthyssens C. & Weigert M. Allison J. P., Walker. L. E., Russell W. A.. Pellegrino M. A.. Ferrone S., Reisfeld R. A., Frelinger J. A. & Silver J. A. (1980h) Two mRNAs with different 3’ ends encode mcntbrane-bound and secreted forms of immunogiohulin mu (1978) Murine Ia and human DR antigens: homology of amino terminal sequences, Proc. mm. Acud. Sci. U.X.4. chain. Cc/l 20, 3 I3 -3 19. Ploegh H. L., Cannon L. E. & Strommger J. L. (1979) 75, 3959-3956. Blobel G. & Dobberstein B. (1975) Transfer of nroteins Ceil-free translation of the mRNAs for the hcavv and across membranes, J. Cc/l Biol. 6j, 835-851. light chains of HLA-A and HLA-B antigens. Proc~,’/lutn. Boyse E. A. & Old L. J. (1978) The immunogenetics of Acod. Sk U.S.A. 76, 2?73%2777. differentiation in the mouse, Harvey Leer. 71, 23353. Quaranta V.. Walker L. E.. Pellegrino M. Kr Ferrone S. Cox R. A. (1968) The isolation of nucleic acids using gua(1980) Purification of immunologically functional subsets nidinum chloride. %flile~lt. Enz_rtn. 126, 12&I 29. of human la-like antigens on a monoclonal antibody Crouse D. A.. Jordan R. K. & Sham J. G. (1980) Immuno(01513) immunoabsorbailt. J. Itttmnn. 125. 1423-1425. globulin genes B-cell differentiation. In ‘Proc&diqys 8th Roberts B. E. & Patterson B. M. (1973) Efhcient transMid-west Autumn fmmur~oloy~ Con~juwcr (Edited by lation of tobacco mosiac virus RNA and rabbit globin Batisto J. & Knight K. I...). p. 65. ElsevieriNorth-Hol9S RNA in a cell-free system from commercial wheat land, Amsterdam. germ. Proc. ncrtn. Artrd. SC/. C;.S.A. 70, 2330 2334. Ferrone S.. Allison J. P. Ly: Pellegrino M. A. (1978~) Rogers J.. Early P., Carter C., Calame K., Bond M.. Hood Human DR (La-like) antigens: biological and molecular L. & Wall R. (1980) Two mRNAs can bc produced from profile. Contemp. Topics nwlw. fnmmw. 7, 329-28 1. a single immunoglobulin mu gene by alternattv~c RNA Ferrone S., Naeim D., Indiveri F., Walker L. E. & Pelleprocessing pathways. Cell 20, 303 312. grino M. A. (19786) Xen(~antisera to human DR antiSakano H.. Huppi K.. Heinrich G. & Tonegawa S. (1979) gens: serological and imm~~~~ochemical characteri~atioll. Sequences at the somatic recombination sites of tm1ntmunogmrtic.s 7, 349-358. munoglobulin light chain genes. ;vlrnrrgt. torid. 280, Ferrone S., Pellegrino M. A. 8: Reisfeld R. A. (1977) Immu288 294. nogenicity of human B cell antigens solubilized from culSlomski R. & Cohen E. P. (1981) Isolation and cell-free tured human lymphoid cells. J. Imtnmi. 118, 1036- 1041. translation of messenger ribonucleic acids specifying Hammerlmg U., Chin A. F. & Abbot J. (1976) Ontogeny of thymus leukemia antigens. Biocherni.w~ 19, 5659 5664. murine B lymphocytes: sequence of B-cell differentiation Springer T. A.. Kaufman J. F.. Siddoway L. A., Giphart M.. from surface-immunoglobuhn-negative precursors to Mann D. L., Ferhorst C. & Strominger J. L. (1977) Puriplasma cells. Ptw, win. Acad. Sri. Z:.S.A. 73, 2008-2012. fication of HLA-linked B lymphocyte alloantigens in imHamnlerling U.. Chin A. F., Abbot J. & Scheid IM. F. ~nunoI~~gicalIy active form by prcparativc sodium doand studies on their sub(1975) The ontogeny of murine B lymphocytes. J. ~~?z~~~~~~. decyl sulfate gel clectrophoresis IIS, 14251431. unit association. J. hioi. firm. 252, 6201-6207. Katz. D. H. (1977) Lpmphocp D[&mtiation. Reqpzition Till J. E. & McCulloch E. A. (198’0) Hemopoietic stem cell md Rrgulution. Academic Press, New York, differentiation. Rioc~liim. hiophj~.s. AUU 605, 431-459. Kessler S. (1975) Rapid isolation of antigens from cells with Tonegawa S., Brack C.. Hozumi N.. Matthysscns G. & a staphylococcal protein A antibody absorbant: parSchuller R. (1977) Dynamics of immunoglobulin genes. ameters of the interaction of antibody-antigen comImmure. Rrv. 36, 258. plexes with protein A. J. Immun. 115, 1617-1624. Wilson B. S., Indlveri F., Molinaro G. A., Quaranta V. & Korman A. J., Ploegh H. L., Kaufman J. F., Owen M. J. & Ferrone S. (1980) Characterization of DR antigens on cultured melanoma cells by using m[~noclonal antiStrominger J. L. (1980) Cell-free synthesis and processing of the heavy and light chains of HLA-DR antigens. 1. bodies. T~~~~s~~utz~}z Proc. 12, 1255129. rsp. Med. 152, 65S-825. Wilson B. S.. Indiveri F., Pellegrino M. A. & Ferronc S. Laemmli U. K. (1970) Cleavage of structural proteins dur(1979) The absence of DR antigens on human T lyming the assembly of the head of bacteriophage T4. phoid cells: serologic and immunochemical studies with Nature, Lond. 227, 680-685. xenoantisera. Trtlnspl0itn PWC. 11, 712.