Broad phylogenetic expression of heavy-chain determinants detected by rabbit antisera to VHa allotypes

Molecular immunology, Vol. 22, No. 10, pp. 1177-1183, 1985 Printedin Great Britain

c

0161-5890185 $3.00 + 0.00 1985 Pergamon Press Ltd

BROAD PHYLOGENETIC EXPRESSION OF HEAVY-CHAIN DETERMINANTS DETECTED BY RABBIT ANTISERA TO V,a ALLOTYPES IRA L. ROSENSHEIN and JOHN J. MARCHALONIS Department of Biochemistry, Medical University of South Carolina, 171 Ashley Avenue, Charleston,

SC 29425, U.S.A. (First received 18 ~eeembcr 1984; accepted in r~~sed~orm 30 ApriI 1985) Alit-Immunoglobulin molecules from diverse vertebrate species were examined, using an enzymelinked immunosorbent assay (ELBA), for the expression of determinants detectable by rabbit antisera to V,a allotypes. The data indicate that immunoglobulins of elasmobranchs, teleosts, amphibians and birds express determinants cross-reactive with those specified by the al, a2 and a3 alleles in the domestic rabbit. We localize V,a cross-reactive specificity to the denatured heavy chain of a primitive vertebrate, the Galapagos shark (Carcharhinusgalupagensis). Furthermore, the N-terminal amino acid sequence of the shark heavy chain shows significant homology with rabbit heavy chains of known V,a type at positions where allotype-correlated differences have been implicated. V,a-related determinants are shared by immunoglobulins of a wide range of vertebrates from sharks to man and thus seem to be epitopes which have been conserved during vertebrate evolution. The determinants detected on immunoglobulins of lower vertebrates by rabbit anti-V,a ailotype sera most probably are V,-subgroup rather than allotypic markers. Their distribution demonstrates a strong phylogenetic conservation of V,-regions.

V,a allotypes of domestic rabbits are present in the variable regions of heavy chains of all classes of immunoglobulins (Todd, 1963; Feinstein, 1963). The V,a allotypic system contains three phenotypes designated al, a2 and a3 which are controlled by allelic genes at the a-locus (Mage, 1981). The individual V,a allotypes correlate with clustered amino acid substitutions located with the first (FRI) and third (FR3) framework regions of the V, sequence (Strosberg, 1977; Tonnelle et al., 1983). Antisera to rabbit V,a allotypes cross-react with human IgG (Knight et al., 1975) and with characterized murine myeloma proteins ~Mackel-Vandersteenhoven et al., 1984). These observations led us to investigate the question whether defined determinants specifying V,a allotypic markers of mammalian heavy chains are expressed by heavy chains from phylogenetically distant species, including primitive vertebrates such as sharks and stingrays (elasmobranchs), as well as more advanced vertebrates such as amphibians. In this paper we report the reactions of antisera to V,a allotypic determinants with immunogIobulins of lower vertebrates. We show that definite %‘,a-related crossreactions occur in species as primitive as the Galapagos shark and that the N-terminal amino acid sequence of the shark heavy chain is consistent with allotypic identification. MATERIALSAND

METHODS

Animals

One male Galapagos shark (Curcharhinus galupagensis) which was large and allowed collection of a

liter of blood was caught off the coast of the Ascension Island. Sandbar sharks (C. plumbeus) were taken by us off the coast of South Carolina in approx. 20 m of water. All specimens were exsanguinated by severing the caudal vessels. The blood was kept at 4°C until arrival at the laboratory where the serum was separated and stored at -20°C. Isolation of vertebrate immunoglobulins

The purification procedures for immunoglobulins used in this study have been previously described as follows: human (Marchalonis and Wang, 1981): mouse (Szenberg et al., 1977); trout (DeLuca et al., 1983); goldfish (Warr et af., 1976); turkey (Atwell and Marchalonis, 1975); and toad (Diener and Marchalonis, 1970). The F(ab’), fragment derived from a pool of human IgG and rabbit IgG was prepared using pepsin proteolysis as described by Edelman and Marchalonis (1967). Shark immunoglobulin was prepared by precipitating the globulin fraction of serum with saturated ammonium sulfate (40% final saturation) and dissolving the concentrate in Tris-buffered saline (TBS) which consisted of 0.05 M Tris-HCl, 0.15 M NaCl, 0.02% NaN,, pH 8.0. High and low mol. wt shark immunoglobulin was separated using a column (26 x 1OOcm) of Sepharose CLdB (Pharmacia Fine Chemicals, Uppsala, Sweden) equilibrated in TBS buffer. All preparations of immunoglobulin were assayed for purity using SDS-polyacrylamide gel electrophoresis (SDS-PAGE) according to the method of Laemmli (1970) as previously described by Atwell and Marchalonis (1975). Protein was determined by the method of Lowry et al. (1951).

1177

1178

IRA L. ROSENSHEIN and JOHN J. MARCHALONIS

Reduction, a~k.~~~ation and preparation and fight chains

of shark heavy

Purified shark immunoglobulin was extensively reduced in 6 M guanidine hydrochloride containing 0.1 M 2-mercaptoethanol for 12 hr at 25°C. Carbaminomethylation was performed at room temp by the addition of solid 2-iodoacetamide (Fisher Scientific, Fairlawn, NJ) at a two-fold molar excess over 2-mercaptoethanol. Unbuffered 1.0 M Tris was added as required to maintain the pH at 8.0-8.5. The reduced and alkylated heavy and light chains were separated by gel filtration through a column

‘A binding =

peroxidase (Sigma Chemical Co.) at a l/500 dilution in DPBS-Tween-BSA were added to each well and the plates were incubated at room temp for 1 hr. After 5 additional washings 100~1 of the substrate 2,2’-azino-di-(3-ethylbenzthiazoline-sulfonic acid) (Sigma Chemical Co.) were added to each well in 0.1 M citrate buffer, pH 4.0, with 0.01% hydrogen peroxide. The reactions were read at 414 nm in a Titertek Multiscan (Flow Laboratories) after half an hour. All assays were performed in duplicate and included enzyme-linked antibody controls and substrate controls. The degree of binding was calculated as follows:

C O.D. of RnVua vs test antigen - E O.D. of NRS vs test antigen C O.D. of RaV,a

vs pooled rabbit F(ab’), - C O.D. NRS vs pooled rabbit F(ab’),y

x 100.

I

(2.5 x 100cm) of Sepharose CL-6B equilibrated in 6 M guanidine hydrochloride. Fractions of about 5 ml were collected and the O.D. measured at 280 nm. Heavy and light chains were dialyzed extensively against water at 4”C, lyophilyzed and stored in a desiccator until used. Antisera to rabbit V& ai~otypes

Thirteen antisera raised in rabbits to the Vna allotypes of this species (Mage, 1984) were generously provided by Dr Rose Mage of the Laboratory of Immunobiology of the National Institute of Allergy and Infectious Diseases. These antisera were as follows: anti-Vual: D327-3 F246-3 (a3-3, bS-5); AH261-4 + 5 (a2-2, b4-4); (a2-2, b4-4); D242-3 + 4 f AH275-3 (a2-2, b4-4); H164-6 (a2-2, b5-6); anti-V,a2: 184-X (a3-3, b4-5); Q3-3 (b4-5, c7 + 21); CM629-3 (al-l, b4-4); CQ857-2 (al-l, b4-4); ZX-80 (a3-3,b4-4); BL229-1 (al-3b, b-6); anti-V,a3: Ap15-I (al-l, dS-5); Apl5-2 (al-l, b5-5); AK350-2 + EP274-2 (al-2,h=4-5). Enzyme-linked

immunosorbent

assay (ELBA)

The assay used has been described previously (Mackel et al., 1983) with minor modifications. The tests were performed in polystyrene microtiter plates (Linbro, Flow Labs, McLean, VA) coated overnight with 1 pg of antigen per 100 ~1 per well in 0.2 M carbonate buffer, pH 9.6, containing 0.02% NaN,. The plates were washed 3 times with Dulbecco’s phosphate-buffered saline (DPBS), pH 7.8, containing O.OSg/, Tween 20. The unreacted sites were quenched using 5% bovine serum albumin (BSA) (Sigma Chemical Co., St. Louis, MO}. The antisera were diluted to the concns indicated with DPBS-Tween containing l%BSA. One hundred microliters of each serum dilution were transferred to antigen-coated wells and the plates incubated at room temp for 1 hr on a rocking platform. After being washed 4 times with DPBS-Tween, 100~1 of peroxidase-conjugated goat-anti-rabbit immunoglobulin (Cappel, Cochranville, PA) or protein A

Amino acid sequence analysis

Sequence analyses of shark heavy chains were performed on a Beckman 890C automatic protein sequencer by using 0.1 M Quadrol (Beckman Instruments Inc., Palo Alto, CA), program 345801 in combination with polybrene (Pierce Chemical Co., Rockford, IL) (3mg). Fifteen nanomoles of glycylglycine (Sigma Chemical Co.) was loaded and 4 precycles were run prior to loading of sample. The anilothiazolinone derivatives of amino acids were manually converted into their phenylthiohydantoins by treatment with 25% trichloroacetic acid (Aldrich Chemical Co., Milwaukee, WI) and water containing 0.01% (M/v) dithioerythritol (Aldrich Chemical Co.) for 30min at 50°C and identified by highperformance liquid chromatography on a DuPont system in combination with a Waters automatic sample processor (WISP). Separation of the derivatives was on a Zorbax column (8 x 0.62 cm), with a Whatman guard column. An exponential gradient (9 min) was generated from 10 mM sodium acetate (Fisher Scientific), pH4.90 (solvent A), and 100% acetonitrile (Pierce Chemical Co.) running from 83 to 61% of solvent A at 50°C. The phenylthiohydantoin derivatives were detected by their absorption at 254 nm at a flow rate of 1.7ml/min. The column requilibration time between runs was 9 min. RESULTS

We tested anti-ailotypic antibodies for reactivity with purified vertebrate immunoglobulin in an ELISA assay. A single sera representative of each of the V,a allotypic specificities and their reactivities with vertebrate immunoglobulin is shown in Fig. 1. Immunoglobulin from primitive vertebrate species such as shark (Fig. IA and B) and stingray (Fig. 1F) as well as pooled human F(ab’),y (Fig. ID) show reactivity with the anti-V,al and antiVua3 sera whereas the anti-V,a2 reagent displayed minimal reactivity above the normal rabbit serum control. The V,al and Vua2 sera exhibit significant binding to the

1179

Phylogenetic expression of V,

Pharkl8S

A

IgM

Shark 7S IgM

8

2.0-

1.5-

to-

o.!?

2.05 f

1.5-

f 0”

to-

P e 0.5-

Toad IgM 2.0-

1.5

to-

0.5-

20

80

128l

320

RECIPFIOCAC

20

80

160

1280

DILUTION

Fig. 1. Reactivity in an ELBA of D327-3 (anti-V,al, H); ZX-80 (anti-V,a2, 0); AK350-2 (anti-V,a3, A) and normal rabbit serum (*) with vertebrate immunoglobulins. Microtiter wells were coated with l.Opg of purified immunoglobulin. The results are expressed as the absorbance value at 414nm.

human (KP) myeloma protein zumac (Fig. 1C). The anti-a3 reagent also binds the human monoclonal protein but at lower antibody concns. All 3 of the anti-V,a anti-allotypic sera display a positive reactivity profile with amphibian macroglobulin (Fig. 1E). Not every anti-Vna serum cross-reacts with immunoglobulin from a particular species (Table 1) but, with 1 exception (anti-a3 and mouse IgG), at least 1 Vna antisera from each group cross-reacted with immunoglobulin from a species. These variattributable to the well-known abilities are subspecificities seen within the rabbit group a allotype system (Rodney and Braun, 1979).

The reactivity of a single anti-Vna3 (AK350-2) sera with pooled rabbit F(ab’),y, pooled mouse IgG and pooled goldfish macroglobulin is shown in Fig. 2. Ansari and Mage (1978) have previously shown that AK350-2 recognizes molecules displaying a3 allotypic determinants and does not bind to molecules carrying al or a2 allotypic markers. As expected at each antibody dilution AK350-2 reacts best with the homologous test antigen rabbit F(ab’),y. Approximately 41% cross-reactivity relative to the rabbit F(ab’),y profile is observed using goldfish IgM as the test antigen and 10% cross-reactivity is observed using mouse IgG.

180

IRA L. ROSENSHEINand JOHN J. MARCHALONIS

Fig. 2. Reactivity in an ELBA of AK350-2 (antiV,a3) with: (0) pooled rabbit F(ab’),y, (0) pooled goldfish IgM, and wells were coated with (W) pooled mouse IgG. Microtiter 1 pg of purified immunoglobulin. The absorbance value at 414nm of normal rabbit serum has been subtracted from each antibody dilution.

The data in Table 1 summarizes the reactivities of 13 anti-V,a anti-allotypic sera with purified vertebrate immunoglobulin. Data are calculated as percentages of cross-reactivity in ELISA binding, normalizing our results to the reactivity of pooled rabbit F(ab’),y as described in Materials and Methods. Elasmobranch immunoglobulins (shark and stingray) show reactivity with reagents directed against each of the three rabbit Vaa allotypes tested. Similarly, pools of fish (trout and goldfish), amphibian (toad), avian (turkey) and mammalian immunoglobulins also display V,a determinants cross-reactive with all 3 of the rabbit group a allotypes. These data show that an allotypic marker found on the V, domain of rabbit immunoglobulin and shared by man (Knight, 1973) is also expressed by at least a subset of immunoglobulin V, domains from evolutionarily distinct vertebrate species representing primitive classes. Furthermore, our data parallel Knight’s (1973) in that strong interspecies cross-reactions are often observed. Purified shark (C. galupagensis) 7-S immunoglobulin and its isolated heavy and light chain resolved by PAGE in the presence of SDS under reducing conditions is shown in Fig. 3 (insert). The preparations are free from detectable contamination. An anti-V,a3 (AK350-2) sera was observed to bind significantly to the immunoglobulin from this species of shark. The reactivity of AK350-2 to purified shark heavy chain (Fig. 3) is very similar to that shown by the intact molecule and the binding to the isolated light chain is negligible. These data show that rabbit antisera directed against V,a allotypes, notably Vna3, react with shark immunoglobulin and the reaction can be localized to the isolated heavy chain. Comparison of the shark heavy-chain aminoterminal sequence with rabbit heavy chains of known allotype shows that the greatest similarity exists between the shark and Vna3 (Table 2) where 8 identities are found in the first 20 residues. In the

1181

Phylogenetic expression of V, .6’

2

\ .2-

.8 -

A

.4 -

20

80

320

RECIPROCAL

1280

DILUTION

Fig. 3. Binding in an ELISA of rabbit antiserum (AK350-2) against the rabbit Vaa3 allotype to intact 7-S ~mmunoglobulin and to isolated heavy and light chains of C. gulu~uge~~. (A) Intact 7-S Ig, (0) isolated shark heavy chain, (a) isolated shark light chain, and (0) normal rabbit serum reactivity. Microtiter wells were coated with 1.0 pg of test antigen. Insert: a 12% SDS-polyacrylamide gel run under reducing conditions showing purified C. galupugensis 7-S IgM and its isolated heavy and light chains. Lane 1, intact 7-S IgM; lane 2, purified 7-S IgM heavy chain; lane 3, purified 7-S light chain; lane 4, mol. wt standards: (a) phosphorylase b (94,~), (b) BSA (67,~), (c) ovalbumin (43,~) (d) carbonic anhydrase (30,000), (e) soybean trypsin inhibitor (20,100), and (f) a-lactalbumin (14,400).

N-terminal 20 residues most of the sequence similarity is clustered throughout residues 13-20. This precise area of the first framework region (FRl) has been determined to be the site of a major determinant related to group a allotypy (Margolies et al., 1977; Mage et al., 1984). Throughout these 8 positions the shark shows 6 identities with V,a3 sequence, 2 of which are allotype-correlated positions (Mage et al., 1984; Strosberg, 1977).

DISCUSSION

Comparative studies of the immunoglobulins of vertebrates have shown that each immunoglobulin

possesses a basic structural unit which consists of heavy and light chains (Marchalonis, 1977). It appears likely that immunoglobulins of all species have originated from a common precursor ancestral gene (Singer and Doolittle, 1966). Insight into the evol-

Table 2. Comparison of N-terminal sequences (FRl) of shark and rabbit heavy chains”

* * 1 C. mlaooeensis

u

2

3

4

5

6

7

EVVnTnTOAEYGmVmKm

l 8

9

0 00 l 00 10 11 12 13 14 15 16 17 18 19 20

Identities 20120

‘Comparison of N-terminal sequences (FRl) of shark and rabbit heavy chains of different allotypes. l designates allotype-correlated residues, * designates allotype-associated residues; data taken from Mage er al. (1984); rabbit allotype sequence data taken from Tonnelle er al. (1983). Residues in boxes indicate regions of homology with shark p-chain. -indicates deletion used to maximize homology. %dicates position 1 is cyclized pyrrolidone carboxylic acid.

1182

IRA L. ROSENSHEINand JOHN J. MARCHALONIS

ution of immunoglobulins could be gained by comparison of amino acid sequences of heavy and light chains of different species but primary-sequence data for birds and lower vertebrates are scarce (Kabat et al., 1983). Cross-reactivity is an alternative to sequence determination for the demonstration of similar structures on 2 proteins. In addition, phylogenetic relationships between the species and proteins under study can also be derived by analyzing patterns of immunological cross reactivity. Numerous studies have reported cross-reactivities of anti-immunoglobulin with immunoglobulin of different species (Nash et al., 1969; Nash and Mach, 1971; Neoh et al., 1973; Hadge et al., 1980). Antisera in these reports were used either to detect immunoglobulin classes or to define immunoglobulin subclasses or used intact immunoglobulin as an immunogen, making it difficult to conclude where the cross-reacting epitopes were located. In contrast, we have used characterized antisera to rabbit Vna allotypic determinants to demonstrate conservation of defined variable-region markers on the V, domains of vertebrate immunoglobulin. The results summarized on Table 1 indicate that Vna determinants are displayed on immunoglobulin from such evolutionary distinct species as sharks, amphibians and mammals. Reagents raised against each of the group a allotypes show a wide reactivity profile. Immunoglobulin from some species, i.e. human and goldfish, were bound strongly by all 3 types of alloantisera whereas pooled immunoglobulins such as turkey and mouse showed a weaker overall cross-reactivity profile. These results show that the expression of V, domains bearing Vna determinants is markedly different in distinct vertebrate lines. Antisera to rabbit group a allotypes have previously been used to detect Vna determinants on human IgG (Knight et al., 1975) and murine myeloma proteins (Mackel-Vandersteenhoven et al., 1984); these observations have been independently confirmed in our study. Observations similar in nature to ours have been described by Eshhar et al. (1983) who found that monoclonal antibody to purified V, cross-reacts with heavy chains from a wide range of vertebrate species. The monoclonal antibody detected highly conserved common Vn determinants which were thought to be related to a conserved sequence stretch in the V, framework. Similarly, Parsons and Herzenberg (1981) describe a monoclonal antibody which recognizes a constantregion allotype on mouse IgG, cross-reacted with reptilian, avian and mammalian sera. Since we investigated pooled immunoglobulins of various species, we cannot conclude whether the Vna-related determinants detected are allotypic in the individual species. They most probably represent Vn-subgroup markers involving the V,III family which is known to show evolutionary conservation (Capra et al., 1973).

The rabbit a locus allotypes are known genetically controlled heavy-chain variable-region markers that have been studied in detail. The comparable localization of Vna-related determinants on lowervertebrate immunoglobulin was demonstrated by the reaction of an anti-a3 sera (AK350-2) with the shark heavy chain but not the light chain. Binding of the anti-Vna3 AK350-2 to the shark heavy chain apparthe presence of an intact ently requires disulfide-bonded loop within the V, structure because exhaustive reduction and alkylation of the heavy chain in 6 M guanidine prior to performing the ELISA binding assay destroys the cross-reactive determinant. Sequence comparisons of rabbit heavy chains of known group a type have shown that amino acid differences that correlate with Vna allotypes are found in the first and third framework regions (Strosberg, 1977; Mage et al., 1984). The study of 3-dimensional models of V, domains (Mage et al., 1977) and immunoelectron microscopy of complexes of immunoglobulin (Roux and Metzger, 1982) indicate that suspected stretches of the sequence appear to be on the surface of the molecule and are removed from the antigen binding site. The shark heavy chain shares a sequence stretch with rabbit Vna, notably Vna3 that is a major allotope of group a allotypy (Mage et al., 1984; Margolies et al., 1977). It appears likely that the cross-reactivity of the Vna allosera, particularly AK350-2, with the shark heavy chain and vertebrate immunoglobulin is directed against this Vna allotope. During the development of the vertebrates, the conserved nature of the function of V, domains has placed structural constraints within the framework regions. These conserved stretches are reflected as common Vu-domain determinants. Some of these conserved V, structures are only exposed on naked heavy chains whereas some are on the surface of the intact molecule and exposed to the solvent. These exposed V, determinants appear to be good candidates to have developed into an immunoregulation system controlling the expression of individual variable-region domains and thus have been maintained during the evolutionary development of the vertebrates. Acknowledgements-This

work was supported in part by NIH grant AI-17493 from the USPHS. We thank Dr Rose G. Mage for the generous gift of rabbit antisera to V,a allotypes and the assistance of Dr C. Schwabe and Mr H. Crow in the performance of the automatic sequence analysis. We are also grateful to Mr George Fam for expert technical assistance and to MS L. Seeker for preparation of the manuscript. REFERENCES Ansari A. A. and Mage R. G. (1978) Immunochemical studies of the a allotypes of rabbit heavy chain variable regions. I. Comparisons of a3 allotypic determinants on normal IgG and IgG of limited heterogeneity by radioimmunoassay with purified labeled anti-allotype antibodies. Immunochemistry 15, 561-568.

Phylogenetic Atwell J. L. and Marchalonis J. J. (1975) Phylogenetic emergence of immunoglobulin classes distinct from IgM. J. Immunogenet. I, 367-391. Capra J. D., Wasserman R. L. and Kehoe J. M. (1973) Phylogenetically associated residues within the Vu111 subgroup of several mammalian species. Evidence for a “pauci-gene” basis for antibody diversity. J. exp. Med. 138, 410-427. DeLuca D., Wilson M. and Warr G. W. (1983) Lymphocyte heterogeneity in the trout, Sulmo guirdneri, defined with monoclonal antibodies to IgM. Eur. J. Immun. 13, 54655 1. Diener E. and Marchalonis J. J. (1970) Cellular and humoral aspects of the primary immune response of the toad Bufo marinus. Immunology 18, 279-293. Edelman G. M. and Marchalonis J. J. (1967) Methods used in the study of the structure of immunoglobulins. In Methods in Immunology and Immunochemistry (Edited by Williams C. A. and Chase M. W.), Vol. 1. Academic Press, New York. Eshhar Z., Gigi 0. and Ben-Neriah Y. (1983) Monoclonal anti-V, antibodies recognize a common V, determinant expressed on immunoglobulin heavy chains from various species. Eur. J. Immun. 13, 533-540. Feinstein A. (1963) Character and allotypy of an immune globulin in rabbit colostrum. Nature, Lond. 199, 1197-l 198. Hadge D., Feibig H. and Ambrosius H (1980) Evolution of low molecular weight immunoglobulins. I. Relationship of 7s immunoglobulins of various vertebrate to chicken IgY. Devl camp. Immun. 4, 501-513. Kabat E. A., Wu T. T., Bilofsky H., Reid-Miller M. and Perry H. (1983) Sequences of proteins of immunological interest. U.S. Department of Health and Human Services, NIH. Kindt T. (1975) Rabbit immunoglobulin allotypes: structure, immunology and genetics. Adv. Immun. 21, 35-86. Knight K. L., Malek T. R. and Dray S. (1975) Human immunoglobulins with al allotypic determinants of rabbit immunoglobulin heavy chains. Nature, Lond. 253, 216-217. Laemmli U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, Lond. 227, 68&685. Lowry 0. H., Rosebrough N. J., Farr A. R. and Randall R. J. (1951) Protein measurement with the Folin phenol reagent. J. biol. Chem. 193, 265-275. Mackel A. M., Craddock G. R., Warr G. W., DeLuca D. and Marchalonis J. J. (1983) Xenoantisera against a murine T cell tumor product cross-react with immunoglobulin Fab determinants. J. Immun. 131, 1582-1590. Mackel-Vandersteenhoven A., Mosely J. M. and Marchalonis J. J. (1984) Partial characterization of T cell components related to defined V,(V,) markers. Cell. Immun. 88, 147-161. Mage R. G. (1981) The phenotypic expression of rabbit immunoglobulins: a model of complex regulated gene expression and cellular differentiation. Contemp. Topics molec. Immun. 8, 89-112. Mage R. G. (1984) Rabbit immunoglobulin allotypes. In Handbook of Immunology (Edited by Weir D. M., Her-

expression

of Vu

1183

zenberg L. A., Blackwell C. C. and Herzenberg L. A.), 4th edition. Blackwell Scientific Publications, Edinburgh. Mage R. G., Bernstein K. E., McCartney-Francis N., Alexander C. B., Young-Cooper G. O., Padlan E. A. and Cohen G. H. (1984) The structural and genetic basis for expression of normal and latent Vna allotypes of the rabbit. Molec. Immun. 21, 1067-1081. Mage R. G., Young-Cooper G. O., Reinek J., Ansari A. A., Alexander C. B.,Appeila E., Carta-Sorcini M., LanducciTosi S. and Tosi R. M. (1977) Rabbit immunonlobulin allotypes: complexities of their genetics, expression, structural basis and evolution. Cold Spring Harbor Symp. quant. Biol. 41, 677-686. Marchalonis J. J. (1977) Immunity in Evolution. Harvard University Press, Cambridge, MA. Marchalonis J. J. and Wang A. C. (1981) A marmoset T lymphocyte protein related to defined human serum immunoglobulin and fragments. J. Immunogenet. 8, 165-175. Margolies M. N., Cannon L. E., Kindt T. J. and Fraser B. (1977) The structural basis of rabbit V, allotypes: serologic studies on al H chains with defined amino acid sequence. J. Immun. 199, 287-293. Nash D. R. and Mach J. P. (1971) Immunoglobulin classes in aquatic mammals. Characterization by serologic crossreactivity, molecular size and bind of human free secretory component. J. Immun. 107, 14241430. Nash D. R., Vaerman J. P., Bazin H. and Heremans J. F. (1969) Identification of IgA in rat serum and secretions. J. Immun. 103, 145-148. Neoh S. H., Jahoda D. M., Rowe D. S. and Voller A. (1973) Immunoglobulin classes in mammalian species identified by cross-reactivity with antisera to human immunoglobulin. Immunochemistry 10, 805-813. Parsons M. and Herzenberg L. A. (1981) A monoclonal mouse antiallotype antibody reacts with certain human and other vertebrate immunoglobulins: genetic and phylogenetic findings. Immunogenetics 12, 207-219. Rodney L. S. and Braun 0. (1979) Rabbit allotype a locus subspecificities of homogenous antistreptococcal antibodies. Eur. J. Immun. 9, 379-384. Roux K. H. and Metzger D. W. (1982) Immunoelectron microscope localization of idiotypes and allotypes on immunoglobulin molecules. J. Immun. 129, 2548-2553. Singer S. J. and Doolittle R. F. (1966) Antibody active sites and immunoglobulin molecules. Science 153, 13-25. Strosberg A. D. (1977) Multiple expression of rabbit allotypes: the tip of the iceberg? Immunogenetics 4,499-5 13. Szenberg A., Marchalonis J. J. and Warner N. L. (1977) Direct demonstration of endogenous murine T-cell surface immunoglobulin. Proc. natn. Acad. Sci. U.S.A. 74, 2113-2117. Todd C. W. (1963) Allotypy in rabbit 19s protein. Biochem. biophys. Res. Commun. 11, 17C-175. Tonnelle C., Cazenave P.-A., Brezin C., Moinier D. and Fougereau M. (1983) Structural correlates to the immunoglobulin heavy chain a 100 allotype. Molec. Immun. 20, 753-761. Warr G. W., DeLuca D. and Marchalonis J. J. (1976) Phylogenetic origins of immune recognition: lymphocyte surface immunoglobulins in the goldfish. Proc. notn. Acad. Sci. U.S.A. 73, 2476-2480.