MHC class I antigens as surface markers of adult erythrocytes during the metamorphosis of Xenopus

DEVELOPMENTAL

BIOLOGY

128,198206 (1988)

MHC Class I Antigens as Surface Markers of Adult Erythrocytes during the Metamorphosis of Xenopus MARTIN

F. FLAJNIK~ AND LOUIS Du PASQUIER

Base1 Institute for Immwwlogg~ Base& Switzerlnnd Accepted March 15, 1988 An alloantiserum produced against Xenopus MHC class I antigens has been used to distinguish different erythrocyte populations at metamorphosis. By analysis using a fluorescence-activated cell sorter (FACS) analyzer, tadpole (stage 55) and adult erythrocytes have distinct volume differences and tadpole cells have no MHC antigens on the cell surface. Both tadpole and adult erythrocytes express a “mature erythrocyte” antigen marker, recognized by its monoclonal antibody (FlF6). During metamorphosis, immature erythrocytes, at various stages of differentiation, which express adult levels of cell-surface MHC antigens by 12 days after tail resorption, are found in the bloodstream. These immature cells are biosynthetically active, produce adult hemoglobin, and mature by 60 days after the completion of metamorphosis. Percoll gradient-density fractionation has shown that all of the cells in the new erythrocyte series express adult levels of MHC antigens but there is only a gradual increase in the amount of “mature erythrocyte” antigen. Tadpole erythrocytes, which are biosynthetically active during larval stages, produce small amounts of surface MHC antigens before the metamorphic climax and then become metabolically inactive. They are completely cleared from the circulation by 60 days after metamorphosis. Erythrocytes from tadpoles arrested in their development for long periods of time expressintermediate levels of MHC antigens, suggesting a “leaky” expression of these molecules in the tadpole cells. The most abundant erythrocyte cell-surface proteins from tadpoles and adults, as judged by two-dimensional gel electrophoresis, are very different. Q 1988 Academic PITSS. h.

tal regulation of a specialized cell product like hemoglobin also occurs for a ubiquitous molecule. One set of Morphological and biochemical events that occur ubiquitous surface antigens, the polymorphic class I during amphibian metamorphosis have been carefully major histocompatibility complex (MHC)3 molecules, chronicled for many years. It has long been known that are developmentally regulated at metamorphosis (Du there are distinct larval and adult forms of certain Pasquier et al, 1979; Flajnik et al, 1986). Unlike class II serum and intracellular proteins (Jurd and MacLean, antigens which can be expressed on the cell surface of 1910; Reeves, 1975; Just et aA, 1977,198O; Knowland and some tadpole tissues (Flajnik et al, 1987a,b), class I Westly, 1979; Broyles, 1981; Widmer et a& 1981, 1983; antigens are absent in larval tissue and rapidly appear Hosbach et al, 1983; Bisbee et CAL, 1976). The cellular and on all tissues within 1 week after the completion of histological basis for restructuring the different organ metamorphosis. These molecules, the classical strong systems at metamorphosis has also been well docu- transplantation antigens, guide recognition events of mented (reviewed in Fox, 1981). In the hematopoietic certain T-lymphocyte subsets (reviewed in Ploegh et al, system, previous studies have shown that immature 1981). The rapid appearance of class I antigens at the erythrocytes enter the bloodstream at metamorphosis metamorphic climax has provided a marker to trace the and produce adult-type hemoglobin, while the remaintransitions in various cell lineages at this time (Flajnik ing larval erythrocytes appear to be rapidly cleared et uL, 1987a). Further, the mode of expression of MHC from the circulation at or shortly after metamorphosis molecules has been examined to determine whether (Just et al, 1977, 1980; Widmer et cd, 1981, 1983; Hos- they are expressed by larval cells meant for destruction bath et aL, 1983; Dorn and Broyles, 1982). at metamorphosis or only by immature adult cells that The cell-surface protein expression associated with restructure the organism. In the present work these the changes that occur at metamorphosis is largely uncharted. It is also not known whether the developmenINTRODUCTION

1To whom reprint requests should be sent at his present address: University of Miami, Department of Immunology and Microbiology, P.O. Box 016906, Miami, Florida 33101. 2The Base1 Institute for Immunology was founded and is supported by F. Hoffmann-La Roche & Co., Ltd., Basel, Switzerland. 0012-1666/88 $3.00 Copyright All rights

Q 1998 by Academic Press, Inc. of reproduction in any form reserved.

8 Abbreviations used: MHC, major histocompatibility complex; Hb, hemoglobin; B2m, 02,-microglobulin; mAb, monoclonal antibody; FCS, fetal calf serum; BSA, bovine serum albumin; LG, Xenspus Lewis x Xenspus Qiui hybrid; APBS, amphibian phosphate-buffered saline; A/B/A, APBS/O.l% BSA/0.05% sodium azide; TCA, trichloroacetic acid, IEF, isoelectric focusing. 198

FLAJNIKAND DU PASQUIER

199

Erythrocyte Changes at Metamorphosis

noprecipitation and immunofluoreseence (Hsu and Du Pasquier, 1984). Cell staining. Blood cells were collected by heart puncture into amphibian phosphate-buffered saline (APBS) containing 10% fetal calf serum (PCS) and 20 MATERIALS AND METHODS III/ml heparin (Du Pasquier et ah, 1985). Blood was Animals. Tadpoles and adults were reared as pre- usually pooled from 5 to 10 tadpoles of the same stage viously described (Flajnik et al, 1986). A family of iso- determined with the Nieuwkoop and Faber table (1967). geneic Xenopus laevis X Xenopus gilli (LG) hybrids was Adult blood cells were collected from the dorsal tarsus used in all experiments (Kobel and Du Pasquier, 1975). vein of individual frogs (Du Pasquier et al, 1985). Cells LG3 tadpoles were used since the alloantiserum ob- were washed three times in APBS, counted, and resustained against LG3 cells was specific for MHC antigens pended to 5 X 10’ cells/ml in APBS/O.l% bovine serum (Fig. 1). The four segregating MHC haplotypes of each albumin (BSA)/0.05% NaN, (A/B/A). One hundred family member are shown in Fig. 1. microliters of the cell suspension was mixed with an Antibodies. The alloantiserum, LG15 (a/c MHC type) equal volume of FlF6 mAb undiluted supernatant or anti-LC3 (b/d MHC type), was used in fluorescence and heat-inactivated normal LG3 or LG15 anti-LG3 alloanimmunoprecipitation experiments (Flajnik et al., 1934, tiserum diluted 1/5 in A/B/A in 96wei1 round-bottom 1986). The monoclonal antibody mAb FlF6 recognizes plates (Costar), and incubated at 4°C for 1 hr. Cells an epitope expressed on both tadpole and adult erythropreviously incubated with frog sera were washed three cytes and was used as a positive control. FlF6 immunotimes with 250 ~1 of A/B/A and then incubated for 1 hr precipitates a l&kDa molecule from lysates of adult at 4°C with 100 ~1 of protein A-purified llD5 at 50 cell-surface iodinated erythrocytes (Plajnik et al., tg/ml. The cells were washed three times and resus1987b). The llD5, a mouse anti-Xenopus IgY (IgG ana- pended in 50 ~1 of 20 pg/ml goat anti-mouse Ig coupled log) monoclonal antibody, was used for indirect immu- to FITC (Nordic). After 1 hr at 4°C the cells were

questions have been approached in one system that has been developed for many years to study metamorphic events, the changes in erythrocyte populations (Broyles, 1981; Just et al, 19’77,198O;Dorn and Broyles, 1982).

T :: j :., ; :;I -’ . L L/ ‘i

b ‘LG’CLONEe

+ 69

’ LG5 lvc

!

3 5 6 7 14 15 17 46

CLASS I HEAVY a CHAIN

-46

f’

-30

&.i

‘L LGlh

LG46 ale

a/d

FLUORESCENCE

INTENSITY

,

FIG. 1. (a) FACS profiles of adult erythrocytes from MHC-typed members of the LC family stained with control normal X&opus serum or with the LG15 (a/c) anti-LG3 (b/d) ahoantieerum. (b) MHCclass I molecules from radioiodiuated adult erythrocyte cell-surface proteins from members of the LG family immunoprecipitated with the LG15 (a/e) anti-LG3 (b/d) alloantiserum by the protein A method. The molecular weight standards (“C-labeled, Amersham) are noted on the right. Lysates containing approximately lo6 cpm corresponding to 5 x 106 cells were used for each immunoprecipitation with 15 ~1 of Xenop~~ serum and 15 pl of protein A-purified llD5 (1 mg/ml). Proteins were separated on a 7-18% gradient gel.

200

DEVELOPMENTALBIOLOGY

washed four times with A/B/A and finally resuspended in 1 ml A/B/A containing 20 pg/ml propidium iodide. The stained cells were examined on a FACS analyzer (Flajnik et al, 1987b) or sorted on a cell sorter (both Becton-Dickinson). Cell Coulter volume on the FACS analyzer was measured on a linear scale. Fluorescence intensity was measured on a three-decade logarithmic scale. The volume and fluorescence scales are displayed in Fig. 2. Radioactive cell labeling, immunoprecip8ation, and SDS-PAGE. Cell-surface proteins were iodinated by the lactoperoxidase method as described (Cone and Marchalonis, 19’74;Flajnik et al, 1984).For biosynthetic

VOLUME 128,1988

amide gels (Rovera et aL, 1978) as described by Kobel and Wolff (1985). Urea, acetic acid, 2-mercaptoethanol, and pyronine Y (final concentrations: 6 1M,5, 5, and 0.05%, respectively) were added to 50 ~1 of the NP-40 lysate described above (~5 X 10’ cells), mixed, and incubated at room temperature for at least 30 min before loading the gels. Gels were preelectrophoresed for 1 hr at 4”C, loaded with the aid of erythrocyte lysate, then electrophoresed for 18 hr, stained, dried, and fluorographed at -70’ with the aid of dimethyl sulfoxide/2,5diphenyloxide (Flajnik et aL, 1984; Bonner and Laskey, 1974). Percoll gradients. Discontinuous Percoll gradients

labeling, 10’ cells/ml were preincubated for 30 min at 27’C in RPMI-1640 medium (Roswell Park Memorial Institute-1640, GIBCO) lacking methionine. After addition of 200 PCi of [“Slmethionine (Amersham, 1120 Ci/mmole), the cells were incubated for 18 hr at 27°C (Flajnik et al, 1984;Kaufman et a& 1985),washed three times with cold APBS, and lysed at 10’ cells/ml in an NP-40 lysis buffer (16, 21). The lysates were spun for 3

were run as first described by Ratcliffe and Julius (1982) and modified for Xenopus cells by Schwager (1985). Cells at the interphase of each fraction were

min at 15,000g to remove nuclei and cellular debris. Two 5~1 replica samples were either counted directly or after TCA precipitation and treatment with Hz02 to remove any radioactive methionine charged to tRNA. Lysate corresponding to 2 X lo6 cells was incubated with the LG15 anti-LG3 alloantiserum and the immu-

carrier of a cytocentrifuge (Shandon-Southern), and 50 ~1 of the cell suspension was added. The cells were spun at 600 rpm for 5 minutes; the slides were then air-dried

noprecipitate was analyzed for MHC antigens as described (Flajnik et aZ., 1984; Kaufman et al., 1985;

speci$cit@. The alloantiserum LG15 (a/c) anti-LG3 (b/d) specifically recognizes MHC-

Kessler, 1975; Laemmli, 1970). A&d/urea gels. Larval and adult globin chains were analyzed by electrophoresis on acid/urea polyacryl-

mals with a b or d MHC haplotype (LGS:b/d, LG5:b/c,

CONTROL

Fl F6

collected, washed twice with APBS, and used for labeling, cell sorting, or histological analysis. Qtospin preparations. Cell pellets were resuspended to 5 X lo6 cells/ml in L-15 medium containing 5% FCS.

Seventy-five microliters of FCS was pipetted in each

and stained with May Griinwald-Giemsa (Merck). RESULTS

Alloantiserum

linked determinants on adult erythrocytes (Fig. la). Surface staining is observed only with cells from aniLG14:a/d). The antiserum recognizes MHC molecules encoded by the d haplotype best, since LG5 cells are stained less intensely than LG14. No staining is detectable with erythrocytes from any of the animals with an a/c MHC (LG6, 7, 15, 17, 46; all share the MHC but

ANTI-bd

differ at many other loci) and there is no immunoprecipitation

of any radioiodinated

molecules from the

cell-surface of a/c erythrocytes (Fig. lb). Class I antigens, composed of /&-microglobulin (&m) and polymorphic heavy chains (Ploegh et al, 1981; Flajnik et a& 1984), are immunoprecipitated from LG3, LG5, and LG14 cell lysates (Fig. lb). Another molecule at approximately 55 kDa probably represents a class I heavy chain covalently linked to &m (Flajnik, unpublished), which is much more apparent under nonreducing con, VOLUME

FIG. 2. FACS analyzer profiles of LG3 adult and tadpole erythro-

cytes (stage 55) stained with control normal Xenqpus serum, the mAb FlF6, and the LG15 (a/c) anti-LG3 (b/d) alloantiserum. The small clouds of spots approximately twice the volume of the major cloud in the tadpole cells stained with the mAb FlF6 are doublets caused by agglutination.

ditions. Fluorescent staining of tadpole and adult cells. Fluorescence and volume analysis was carried out on a

FACS analyzer to determine whether tadpole erythrocytes are stained with the anti-MHC reagent. Tadpole and adult cells are readily distinguished by their v01ume (Fig. 2). Both tadpole and adult erythrocytes stain

FLAJNIK AND Du PASQUIER

Eq&hroq&

with the same intensity with the mAb FlF6 used as positive controls in these experiments (Flajnik et aL, 1987b). In contrast, only the adult cells are stained with the anti-b/d alloantiserum (Fig. 2). Tadpole erythrocytes are not stained from stage 52 (3 weeks postfertilization) until stage 5’7(6 weeks postfertilization). At the metamorphic climax (stages 58/59 to 64/65), tadpole cells become slightly positive with the alloantiserum (Fig. 3). At 6 days postmetamorphosis, a new population of cells, which is FIFG-negative, appears. Concomitantly, there is a slight increase in staining intensity with the alloantiserum. At 12 days postmetamorphosis, there is a bimodal staining distribution both with FIF6 and the alloantiserum (Fig. 3). When equal numbers of alloantiserum-stained cells from an animal 12 days postmetamorphosis and from an adult are mixed at a 1:l ratio, erythrocytes with the smallest volume, from the metamorphosing froglet, are stained just as brightly as the cells with the largest volume which are from the adult (Fig. 4). A third population of erythrocytes, derived from the froglet and with an intermediate volume, show little fluorescence (Fig. 4). Cell fractionation and correlation with hemoglobin and class I expression. In order to separate the brightly and weakly stained cells from young froglets, Percoll gradients were used (Schwager, 1985). On discontinuous gradients the erythrocytes from tadpoles up to stage 58159 remain at the top of the gradient and adult cells sediment at the bottom (data not shown). Cells from

FlF6

CONTROL

ANTI-&i

--lr----7 STAGE SSISS, TADPOLE

STAGE64165, METAMORPH

6 DAYS FUSTMETAMORPWSIS-

1PDAYS POSTMETAMORPHOSIS-

VOLUME

t

FIG. 3. FACS analyzer profiles of blood cells from LG3 animals stained at different stages of metamorphosis with control normal Xempus serum,the mAb FlF6, and the LG15 (a/c) anti-LG3 (b/d) alloantiserum.

201

Changes at Metamorphosis ANTI-bti

CONTROL

VOLUME

P

FIG. 4. FACS analyzer profiles of a 1:l mixture of blood cells from LG3 animals 12 days after metamorphosis and during adult life. Cells were stained with control normal Xaopus serum and the LG15 (a/c) anti-LG3 (b/d) alloantiserum. The brightly stained population with the smallest volume and the weakly staining population with the intermediate volume are derived from the metamorphosing froglet; the brightly staining cells with the largest volume are from the ad?llt frog. This was determined by first examining cells separately before mixing.

animals 12 days postmetamorphosis are found scattered at each of the different densities (Fig. 5a). Cytological analyses show that cells from the different fractions contain both mature tadpole erythrocytes that, at this time, enter the gradient and immature adult erythrocytes that are of increasing maturity (as judged by FlF6 staining) in higher density fractions (Fig. 5a). There is an inverse relationship between the number of cells staining brightly with the mAb FlF6 and the alloantiserum (Fig. 5a). Proteins from cells from the different Percoll fractions were labeled biosynthetically with [35S]methionine. Acid/urea gels showed that the predominant globin chains stained with Coomassie blue from fractions 1 and 2 were larval, which correlates well with the number of mature larval erythrocytes detected in cytospin preparations, and also with the number of brightly staining FlFG-positive cells in these fractions. The biosynthesized globin chains in these fractions, however, are of the adult type (Fig. 5b); these are presumably made by the immature adult cells in these fractions. We cannot be sure that there are no tadpolelike globin chains produced by cells present in these fractions, but it is clear that most of the label is incorporated into adult globin. Erythrocytes from tadpoles through stage 57/58 are biosynthetically active although it appears that no globin chains are produced (Table 1 and Fig. 5b). Mature adult erythrocytes do not incorporate r5S]methionine into protein (Table 1 and Fig. 5b). Detergent lysates of cells from these fractions, standardized for cell number, were incubated with alloantisera and class I molecules were immunoprecipitated (Fig. 5b). Increasing amounts of class I molecules are precipitated from lysates of cells with higher and higher densities. Even though tadpole cells are biosyn-

202

DEVELOPMENTALBIOLOGY

.:,’ I/

FRl cFR2

a

FR2 ‘*“,

5

VOLUME 128,198s

FR3 l

mc-

FR4

tp

a

VOLUME

b

GLOBIN A

T12345

PROTEINS

MHC

T12345A

h w

Tl

ADULT GLOBINS

I

66,

46, LARVAL GLOBINS-

CLASS 2345A

--a*#

CC1

I /

30,

!

4.

TOTAL

BIOSYNTHETIC

w

FIG. 5. (a) Discontinuous Percoll gradient preparations of blood cells from LG3 animals 12 days after metamorphosis. Cells formed bands at densities 1.075 (fraction l), 1.08 (fraction 2), 1.085 (fraction 3), 1.09 (fraction 4), 1.095 (fraction 5). Cytospin preparations from each fraction were stained with May Griinwald-Giemsa. For immunofluorescent staining, aliquots of cells from each fraction were stained with control normal Xenqpus serum, the mAb FlF6, and the LG15 (a/c) anti-LG3 (b/d) alloantiserum and subjected to FACS analyzer analysis. (b) Cells from these fractions were labeled with CssS]methionine as detailed under Materials and Methods. Lysates were separated on acid-urea gels to determine relative amounts of total and biosynthesized tadpole and adult globin chains by Coomassie blue staining and fluorography, respectively. Aliquots of lysates corresponding to equal numbers of cells were incubated with the LG15 (a/c) anti-LG3 (b/d) alloantiserum for MHC class I molecule immunoprecipitation by the protein A method as in Fig. Ib. A, adult; T, tadpole (stages 54-56).

thetically active, they produce no detectable class I molecules. No radiolabeled class I antigens are precipitated from adult erythrocyte lysates since they are not metabolically active. Although the immature adult cells from all the fractions, by FACS analysis, have equivalent numbers of cell-surface MHC antigens, more class I is apparently immunoprecipitated from lysates of cells in the denser fractions. This is in part because those fractions contain fewer old tadpole cells and partly because the most mature adult cells incorporated more [%J]methionine per cell, generating a discrepancy between % gels and fluorescence analysis (Table 1). A polymorphism that has been detected for X la& and

X gilli ,&m may be responsible for the presence of two ,&m bands on these gels (Kaufman and Flajnik, unpublished). The data presented in Fig. 5 suggest that the most immature adult cells in the bloodstream 12 days after tail absorption already express maximum levels of MHC antigens but a low level of FIF6 molecules, while the leftover tadpole cells show the inverse. To examine this further, cells from Percoll fractions 2 and 3 were sorted into FlF6 and alloantiserum bright and dull populations, respectively (Fig. 6). As expected, the cells from the FlFG-bright and alloantiserum-dull populations had a mature morphology (data not shown), were

FLAJNIK AND Du PASQUIER

Erythrocgte Changes at Metamorphosis

TABLE 1 [35S]M~~~~~~~~~ INCORPORATIONINTO PROTEINS FROM BLOOD CELLS SEPARATED ACCORDING TO DENSITY OR SORTED INTO DIFFERENT POPULATIONS(cpm/cell) Cell type

Expt. 1

Expt. 2

Fraction 1 Fraction 2 Fraction 3 Fraction 4 Fraction 5 Adult Tadpole (stages 54-56) FlFG-bright FlFG-dull Anti-b/d-bright Anti-b/d-dull

2.4 34 34 48

3.4 8.8 32 36

54

46

0.1

0.1

22

Expt. 3” 2.8 34 52 68

(2.2) (26) (34) (42) 82 (W 0.12 (0.1)

4.2

2.6

(2.2)

0.8 34 30.4 0.52

Note. Lysates were prepared and counted as under Materials and Methods. Cells in fractions 1 to 5 were from animals 12 days postmetamorphosis as in Fig. 5. The bright and dull populations were sorted and are displayed in Fig. 6. a Counts per minute in parentheses in Experiment 3 are TCA-precipitated counts after treatment with HzOz to remove tRNA charged with [35S]methionine.

biosynthetically inactive (Table l), and contained tadpole globin chains (Fig. 6). The FlFG-dull and alloantiserum-bright populations had an immature morphology (data not shown), were biosynthetically active CONTROL

Fl F6

203

(Table l), and produced adult globin chains and class I ,molecules (Fig. 6). “‘Blocked” tadpoles. LG3 tadpoles were arrested in their development at stage 55 (Nieuwkoop and Faber, 1967) with sodium perchlorate which prevents thyroid function (DiMarzo, 1981). Erythrocytes from an animal blocked for 9 months have an intermediate amount of MHC antigen on their surface as compared to the amount of adult staining (Fig. 7). Cells positive for MHC can be detected 2 months after the initiation of chemical treatment, and the intensity of staining increases over time. Erythrocytes from tadpoles that are blocked at later stages of development have higher levels of MHC antigens than tadpoles blocked at earlier stages (data not shown). Other cell-surface molecules. Erythrocyte cell-surface proteins from LG3 tadpoles, adults, and tadpoles that had been blocked for approximately 6 months were iodinated and analyzed by two-dimensional IEF-SDSpolyacrylamide gel electrophoresis (Fig. 8). Class I antigens were immunoprecipitated from radiolabeled adult lysates and also analyzed in this way (Fig. 8). It is clear that the major erythrocyte surface proteins of isogenic tadpoles and adults are different, although these experiments do not totally rule out post-translational modifications. A black arrow points out the one spot that is unequivocally shared between tadpoles and

ANTI-bd

TOTAL

GLOBIN

MHC CLASS I

m-

J

CELL

SORTER

FIG. 6. Blood cells from LG3 animals 12 days after metamorphosis were run on discontinuous Percoll gradients and fractions 2 and 3 (as in Fig. 5) were collected. An aliquot of 5 X 106cells was removed and stained with control Xenop~.~serum. The rest of the cells were stained with either the mAb FlF6 or the LG15 (a/c) anti-LG3 (b/d) alloantiserum. These cells were sorted on a Becton-Dickinson cell sorter into bright and dull populations. The histograms, derived from the FACS analyzer, show the quality of fluorescent staining and the purity of the separated cells after sorting. Unseparated, bright, and dull populations were labeled with [%]methionine (Table 1). Acid-urea gels were stained for total protein with Coomaasie blue, and immunoprecipitations of class I molecules were carried out as in Fig. 5b. T, tadpole globin chains; A, adult globin chains. The arrowheads pointing toward the MHC class I gel indicate the high-molecular-weight MHC-encoded class I a-chain and the low-molecular-weight Barn.

204

DEVELOPMENTALBIOLOGY

VOLUME 128,198s

m br&Io.,. CONTROL

ADuLT+j---jm

+!

%%FEpj---Jm;

FLUORESCENCE

INTENSITY

*

FIG. 7. FACS analyzer profiles of blood cells from LG3 adult and tadpoles that had been arrested in their development for 9 months with sodium perchlorate. Cells were stained with control normal Xenopus serum and the LG15 (o/c) anti-LG3 (b/d) alloantiserum.

adults. From this experiment and others it is clear that the immunoprecipitated class I molecules have the same size and charge as some of the spots in the area indicated with an open arrowhead in the adult and blocked tadpole gels. Most of the bands are removed by preclearing the labeled lysates with the alloantiserum (data not shown). None of the other proteins are identi-

ta

0

fied, although some may be MHC-encoded but not reactive with the alloantiserum. The major surface proteins of the blocked tadpole are composed of a mixture of most of the adult and tadpole proteins. Figure 8 displays only those cell-surface components of the erythrocytes that are labeled best with the lactoperoxidase technique (Cone and Marchalonis, 1974). Upon longer exposures of these gels, many more surface proteins can be identified, some (approximately 20%) that are definitely shared between tadpoles and adults (data not shown). Such major differences in membrane expression during ontogeny may be limited to the erythrocyte lineage. Most of the surface proteins of tadpole and adult thymic lymphocytes are identical (data not shown). DISCUSSION

The expression and regulation of adult-specific proteins at metamorphosis have been examined in this paper. At least for hematopoietic tissue, the data demonstrate that adult proteins, e.g., hemoglobin or MHC class I, are expressed in a specific adult manner only in the new population of cells entering the circulation at

63

8

4mJ 492 469 446 430

414

ADULT

TADPOLE

MHC CLASS

I

BLOCKED TADPOLE

FIG. 8. Two-dimensional IEF/SDS-polyacrylamide gel electrophoresis of radioiodinated cell-surface proteins from LG3 tadpoles, adults, and blocked tadpoles. After iodination, cells were lysed in IEF sample buffer (Ferreira and Eichinger, 1981). Equal cpms (500,000 cpm) corresponding to approximately 2 X 106cells were loaded onto the first-dimension IEF gels. The gel in the lower right corner displays LG3 class I molecules immunoprecipitated as in Fig. lb. The black arrow points out a protein that is identical on all gels. The open arrowhead indicates the class I molecule in the adult and blocked tadpole gels. The molecular weight standards are noted at the right of each gel. The proteins were separated in the second dimension on 12% gels.

FLAJNIK AND Du PASQUIER

Ergthroqte

metamorphosis. Several other results, however, demonstrate that the above affirmation must be qualified. First, tadpole erythrocytes, as suggested by earlier studies (Dorn and Broyles, 1982) and confirmed by these, are metabolically active. They are able, probably in response to hormonal signals, to express a small amount of class I molecules at metamorphosis (Fig. 5a) and at that time undergo their final maturation to become biosynthetically inactive and are later cleared from the circulation (Table 1). Second, the immature adult erythrocytes, regardless of their differentiation state, have adult levels of MHC class I antigens on their surface, but show a more gradual increase in expression of the determinant recognized by the mAb FlF6 (Fig. 5). It may be especially important for the young frog to express high levels of MHC antigens shortly after metamorphosis. These molecules are implicated in the differentiation processes of T-cells (education and tolerance) and high levels are probably necessary during the regeneration of the immune system at metamorphosis (von Boehmer, 1986; Flajnik et aL, 1987a). Third, tadpoles blocked in their development allow a leakage of adult globin expression (Just et al, 1977). We confirmed this observation, although we never found a complete switch to adult globin expression even after 1 year. Cells at different stages in ontogeny can also be distinguished by their Coulter volume. The two distinct volume differences that are always observed are (1) between erythrocytes from tadpoles until stage 5’7 and from adults (Fig. 2), and (2) in the bulk of new adult cells from animals lo-12 days after metamorphosis, especially from Percoll fractions 2 and 3, which are small as compared to either tadpole or adult cells (Fig. 4 and 6). It is dangerous to make more definite statements about the subtle volume difference of the new adult cells as they mature into the final adult stage; there are slight volume differences in the same cell populations depending on whether they were stained with the FlF6 mAb supernatants or with antisera, presumably due to changes in osmolarity (Figs. 3 and 5). In the two special cases noted above, the volume differences are always apparent regardless of the conditions of immunofluorescent staining (Figs. 2,4, and 6). Our data suggest that the immature cells that enter the blood stream after metamorphosis produce only adult globin chains, and not a mixture of larval and adult as was implied by experiments using specific cDNA clones on anemic tadpole cells (Widmer et cd, 1983). On the other hand, since the old tadpole cells produce small amounts of class I antigens before climax, it may not be surprising to also find low amounts of biosynthesized adult globin proteins in these cells. Immunofluorescent staining with adult- and tadpole-

Chaqps

at Metamorphosis

205

specific antibodies after sorting cells for class I expression should solve these problems. Our data suggest that there is also a gradual increase in the expression of class I antigens and other adultspecific cell-surface proteins in cells from tadpoles blocked in their development. Since tadpole erythrocytes before the metamorphic climax are biosynthetically active, we postulate that adult and larval proteins are expressed by the same cells. Consistent with this idea, there is always a unimodal staining distribution for MHC antigens in erythrocytes from tadpoles arrested in their development (Fig. ‘7) which increases in fluorescence intensity the longer the blockage. Further experiments are needed to clarify this problem. The so-called “leakage” of MHC expression by the mature tadpole cells may actually be the normal mode of expression for those tissues not destroyed at metamorphosis. The turnover which occurs within the erythrocyte populations at metamorphosis is probably also manifested in other hematopoietic lineages such as lymphocytes. For example, there is a decrease in cell number in the thymus by as much as two orders of magnitude (Du Pasquier and Weiss, 1975; DiMarzo, 1981). It appears that the thymus is repopulated by extrinsic precursors. The B-cell compartment may change in a similar manner, since the tadpole’s antibody repertoire changes to adult type at metamorphosis (Du Pasquier et al, 1979; Hsu and Du Pasquier, unpublished). The types of macrophages and dendritic cells in the thymus also change (Clothier and Balls, 1985). It is important that some of the tadpole cells remain during metamorphosis in order to protect the organism against pathogens. As with the erythrocyte changes, we propose that the tadpole cells may remain for a while, as the immature adult cells differentiate, to reconstitute the immune and hematopoietic system in general. These turnover events must now be examined in detail for each hematopoietic cell type. Staining procedures with adult-specific antibodies, e.g., directed against MHC antigens, as was done here to discern the erythrocyte shifts at metamorphosis, should resolve these issues. We thank Michaela Manes and Diana Thorpe for technical assistance, Joseph Schwager and Barbara Fagg for critical review of the manuscript, and Judie Hossmann for typing it. We also thank John Just and Hiroko and Johannes Holtfreter for their helpful comments. REFERENCES BISBEE, C. A., BAKER, M. A., and WILSON,A. C. (1976). Albumin phylogeny for clawed frogs (Xengpus). Science 195,785-787. BONNEn, W., and LASKEY, R. (1974). A film detection method for tritium labeled proteins and nucleic acids in polyacrylamide gels. Eur. J. Bioc~

46,33-83.

BROYLES,R. H. (1981). Changes in the blood during amphibian meta-

206

DEVELOPMENTALBIOLOGY

morphosis. In “Metamorphosis, A Problem in Developmental Biology,” (L. L. Gilbert and E. Frieden, Eds.), pp. 461-490. Plenum, New York/London. CLOTHIER,R. H., and BALLS, M. J. (1985). Structural changes in the thymus glands of Xenqpus luaris during development. In “Metamorphosis” (M. J. Balls and M. Bournes, Eds.), pp. 332-359. Oxford Univ. Press, Oxford. CONE, R. E., and MARCHALONIS, J. J. (1974). Surface proteins of thymus derived and bone marrow derived lymphocytes. Selective isolation of immunoglobulins and antigen by nonionic detergents. Biochem J. 104,435-447. DIMARZO, S. J. (1981). “Ontogeny of Alloreactivity in the Toad X&nopus lads.” Ph.D. Thesis, University of Rochester, Rochester, N.Y. DORN, A. R., and BROYLES,R. H. (1982). Erythrocyte differentiation during the metamorphic hemoglobin switch of Rana cote&ana. Proc Nat1 Acd Sci USA 79,5592-5596. Du PASQUIER,L., BLOMBERG,B., and BERNARD,C. C. A. (1979). Ontogeny of immunity on amphibian: Changes in antibody repertoires and appearance of adult major histocompatability antigens in Xenopus. Eur. J. ImmunoL 9,900-906. Du PASQUIER, L., FLAJNIK, M. F., GUIET, C., and Hsu, E. (1985). Methods used to study the immune system in Xenopus (Amphibia, Amura). In “Immunological Methods,” (I. Lefkovits and B. Pernis, Eds.), Vol. 3, pp. 425-465. Academic Press, New York/London. Du PASQUIER,L., and WEISS,N. (1975). The thymus during the ontogeny of the toad Xeqvus Levis. Eur. J. Immunol. 3,773-777. FERREIRA, A., and EICHINGER, D. (1981). A simplified two-dimensional electrophoresis technique. J. ImmunoL Methods 43,291-299. FLAJNIK, M. F., Hsu, E., KAUFMAN, J. F., and Du PASQUIER,L. (1987a). Changes in the immune system during metamorphosis of Xenopua ImmunoL Today 8,58-64. FLAJNIK, M. F., Hsu, E., KAUFMAN,J. F., and Du PASQUIER,L. (198713). Biochemistry, tissue distribution and ontogeny of surface molecules on Xenom hemopoietic cells. In “Comparative Aspects of Differentiation Antigens in Lymphopoietic Tissues” (M. Miyosaka and Z. Trnka, Eds.), pp. 387-419. Dekker, New York. FLAJNIK, M. F., KAUFMAN, J. F., Hsu, E., MANES, M., PARISOT,R., and Du PASQUIER,L. (1986). Major histocompatibility complex-encoded class I molecules are absent in immunologically competent Xenqpus before metamorphosis. J. Immunol 137,3891-3899. FLAJNIK, M. F., KAUFMAN, J., RIEGERT, P., and Du PASQUIER, L. (1984). Identification of class I major histocompatibility complexencoded molecules in the amphibian Xenopus. Immunogenetics 20, 433-442. Fox, H. (1981). Cytological and morphological changes during amphibian metamorphosis. In “Metamorphosis: A Problem in Developmental Biology” (L. L. Gilbert and E. Frieden, Eds.), pp. 327-362. Plenum, New York/London. HOSBACH,H., WYLER, T., and WEBER, R. (1983). The Xenqpus Levis globin gene family: Chromosomal arrangement and gene structure. Cell 32, 45-53. HSIJ, E., and Du PASQUIER,L. (1984). Studies on Xenopus immunoglobulins using monoclonal antibodies. MoL Immund 21,257-270. JIJRD,R. D., and MACLEAN,N. (1970). An immunofluorescent study of the haemoglobins in metamorphosing Xenopus lo&-~. J. Embryol. Ezp. Morph01 23,299-309.

VOLUME 128,1988

JUST, J. J., SCHWAGER,J., and WEBER, R. (1977). Hemoglobin transition in relation to metamorphosis in normal and isogenic Xenopu,s laevis. Wilhelm Roux Arch. 183,307-323. JUST,J. J., SCHWAGER,J., and WEBER,R. (1980). Immunological analysis of hemoglobin-transition during metamorphosis of normal and isogenic Xenopus Wilhelm Roux Arch. Dev. Bid H&75-83. KAUFMAN, J. F., FLAJNIK, M. F., Du PASQUIER,L., and RIEGERT, P. (1985). Xmopus MHC class II molecules. I. Identification and structural characterization. J ImmunoL 134,3248-3257. KESSLER,S. W. (1975). Rapid isolation of antigens with a staphylococcal protein A-antibody absorbent: Parameters of the interaction of antibody-antigen interactions with protein A. J. Immud 115, 1617-1624. KNOWLAND,J., and WESTLY, B. (1979). The control of vitellogenin synthesis by oestrogen in Xenqpus laewis and conversion of vitellogenin into yolk platelets. In “Maternal Effects in Development” (D. B. Newth and M. Balls, Eds.), pp. 111-126. Cambridge University Press, Cambridge. KOBEL, H. R., and Du PASQUIER,L. (1975). Production of large clones of fully identical clawed toads (Xenopw). Immunogenetics 2,87-91. KOBEL, H. R., and WOLFF, J. (1983). Two transitions of hemoglobin expression in Xenopus: From embryonic to larval and from larval to adult. werentiation 24,24-26. LAEMMLI, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227, 680-687. NIUEWKOOP,P. D., and FABER, J. (1967). “Normal Table of Xaopus laevis.” North Holland, Amsterdam. O’FARRELL,P. H. (1975). High resolution two-dimensional electrophoresis of proteins. J. BioL Chem 250,4007-4021. PLOEGH,H. L., ORR, H. T., and STROMINGER,J. (1981). Major histocompatibility antigens: the human (HLA-A, -B, -C) and murine H-2K, H2-D) class I molecules. Cell 24,287~299. RATCLIFFE, M. J. H., and JULIUS, M. H. (1982). H-2 restricted T-B interactions involved in polyspecific B cell responses mediated by soluble antigen. Eur. J. ImmunoL 12,634-642. REEVES,0. R. (1975). Adult amphibian epidermal proteins: Biochemical characterization and developmental appearance. J. EmbryoL Exp. Morph. 34,55-73. ROVERA, G., MAGERIAN, C., and BORUM,T. W. (1978). Resolution of hemoglobin subunits by electrophoresis in acid urea polyacrylamide gels containing Triton X-100. AnaL Biochem. 85,506-518. SCHWAGER,J. (1985). “The Lymphocytes of the Amphibian Xenopus laevis: Studies on the Differentiation of the B Lymphocyte and the Biosynthesis of Immunoglobulins.” Ph.D. Thesis, University of Geneva, Geneva, Switzerland. VON BOEHMER,H. (1986). The selection of the a, @heterodimeric receptor for antigen. ImmunoL Today 7,333-336. WIDMER, H. J., ANDRE& A.-C., NIESSING, J., HOSBACH, H., and WEBER, R. (1981). Comparative analysis of cloned larval and adult globin cDNA sequences of Xeno.nu.sluevis. Dev. BioL 88,325-332. WIDMER, H. J., HOSBACH, H., and WEBER, R. (1983). Globin gene expression in Xencpus laevis: Anemia induces precocious globin transition and appearance of adult erythroblasts during metamorphosis. Dew. BioL 99,50-60.