~]~EVIEWS
Minor hist0c0mpatibility antigens
H i s t o c o m p a t i b i l i t y antigens are cell surface molecules that lead to the rejection of grafted tissues and organs by the vertebrate immune system. A systematic genetic definition of minor histocompatibility antigens was undertaken by George Snell, w h o in the 1940s began breeding, by backcrossing and intercrossing, strains of mice that differed from each other by only a single histocompatibility antigenL He selected mice for their ability to reject a tumor graft from the inbred strain used in the backcrossing; the only mice that survived were those lacking, and therefore able to respond to, a minor histocompatibility antigen of the backcross strain. The sch6me was laborious and tended to select for differences at the stronger histocompatibility loci, including //-2, the major histocompatibility complex (MHC), because the tumors could outgrow the immune response to weaker minor histocompatibility antigens. Later, the selection was changed to skin graft rejection, a far more sensitive measure of incompatibility, which did not require homozygosity: progeny mice whose skin was rejected by the inbred backcross parent had retained a gene encoding a minor histocompatibility antigen. It was also found that congenic strains that had been bred by selection for a coat color marker, such as In or a t, often differed by a linked, minor histocompatibility locus. Today, more than 50 loci have been defined, and almost every chromosome of the mouse, including the mitochondrial genome, carries at least one (Fig. 1). Like mice, humans have a single MHC, the HLA complex on chromosome 6, and multiple minor histocompatibility loci. Typing for HLA antigens has
2
1
3
~.
5
7
8
9
KIRSTEN FISCHER LINDAHL
Histocompatibllity antigens have been studied for over 50years because they form a major obstacle to clinical transplantatiorL Human minor histocompatibility antigens remain ill-defined, but minor histocompatibillty loci have been mapped on nearly every mouse chromosome. Recent molecular definition of several transplantation antigens suggests that they are by-products of an immune system poised to present viral antigens, and a mutation in any gene may give rise to a new minor histocompatibility antige~
developed from a fine art into a high science e. and an HLA-identical donor is always preferred in clinical transplantation. However, with bone marrow: grafts, which are used for treatment of aplastic anemia and after eradication of lymphoid tumors, complications such as rejection or chronic graft-versus-host disease still arise in 20-70% of recipients of HLA-matched grafts, due to minor histocompatibility antigens:. It has been possible to recover lymphocytes from patients undergoing such a crisis and study the specificity of the response 4. The human minor histocompatibility antigen most easily defined is H-Y, the male antigen: lymphocytes from a female donor will attack the tissues ()f :i male recipient, all of which express the H-Y antigent
II
12
1/+
15
H-22 H'21+
H(IU
~t(In)
H-21 H(pi) H(go) H-27 H-I H-15 H-16
H-3 :H-42 .H-4 I+ H-45 -H-13
H-2~ H:d) _H-37 H-20 H -3 ~.
HOs) H-40
H-29
[
19
/-/-39
H-33
.H-8 H-4
17
Y
X
mf
iHya "~4:f H(ep)
H-2 H-31.,
.H-30
.H-43
.H(Eh)
H-7 H'19
_H-23
lH(fn)
H-I8 .H-28
Hxa
FIGI!
Genetic map of minor histoc()mpatibility loci in the mouse, based on Ref. i3. Thu mit()ch(~ndrial ot2i~(H]]ci Ill[) J', }LI[(){ ",({tit' ,
•ri(; Jt't'~' 1991 VOL 7 XO. 7 ] IULtd ({ K} I)L(~H 9 Ira9 91 502 I}0
E
i
i
[]~EVIEWS
Model I: T-cell receptors
Target cells
Model I1: T-cell receptors
f
Recognitionand lysis
(~ 9
H-Y antigen in MHC-dependentconformation and ~
MHC molecules
I k
J
HUE MHC restriction of the killer T-cell response to H-Y. Killer T cells from an MHC A female immunized with MHC A male cells can attack and lyse MHC A and AxB male target ceils, which present the male antigen (_~)with an A-type MHC molecule (AA), but they cannot lyse MHC B male target cells, which present the male antigen with a B-type MHC molecule ('B fiat'). Two models of T-cell receptors are illustrated: model t (top) assumes that the receptor interacts with both peptide and MHC molecule: model II (bottom) assumes that the receptor interacts only with peptide, which is presented in an MHC-dependent conformation.
The male antigen H-Y is the most studied of all minor histocompatibility antigens, p r o b a b l y because it is so easy to define: males have it and females do not. It was first observed when female mice rejected the skin of males of the same inbred strain s. As with all histocompatibility antigens, the rejection is caused by T lymphocytes, and it is possible to culture and study these cells in vitro. Antibodies do not detect H-Y, but helper and killer T cells provide reliable and efficient typing reagents for the antigen, which shows no polymorphism s. H-Y is e n c o d e d on the short arm of the mouse Y c h r o m o s o m e by the H y a gene, which is tightly linked to the testis-determining gene, Tdy ~'. In humans the gene encoding the testis-determining factor, 7DF, has been m a p p e d to the short arm of the Y c h r o m o s o m e r and the H - Y g e n e to the long arm s. Hya has not been
cloned, and its function is unknown; it may be involved in spermatogenesisg. It remains a riddle why the antigen should be expressed on cells, such as lymphocytes, that have no apparent sex-specific function. Alternatively, H-Y may represent a Y-specific isoform of a housekeeping protein like the human ribosomal protein $4, which is e n c o d e d by a gene located next to TDF t°.
T-cell recognition of antigen Minor histocompatibility antigens are recognized by T cells in an MHC-restricted fashion 11. Therefore T ('ells from a female immunized with MHC-compatible male cells can detect the male antigen only on cells of the same MHC type s (Fig. 2). For killer T cells, it is important that the polymorphic class I antigens of the MHC (in humans e n c o d e d by the HIA-A. HLA-B and H I A - C loci) are shared by the immunizing cells and
uc; ItaLY 1991 VOL. 7 :~0. 7
[]~EVIEWS the target; for helper T cells, it is TABLE1. Defined peptides that Entm|c minor histocompatibility and the class II antigens of the MHC tumor antigens (in humans e n c o d e d by the HLA D series of loci) that matter. Peptide Sequence Bound by MHC Recognition by MHC restriction also applies to class I molecule killer T cells T-cell responses to foreign proMTF tein antigens and viral antigens, ND 1oil-17 fMFFINI_LTLLVPIIJAM HMTa anti-MTFct and the basis is now v,cll ND 1[ 3 1 - 1 7 fMFFINALTLLVPILIAM HMT~ anti-MTF~ understood from studies in these systems. As a rule (but with fMet 1 2 - c z fMFFINILTLLVP HMT~ anti-MTF~ exceptions), cellular processing of Metl 2-0t MFFINILTLLVP No No antigen follows txvo pathways~-'. AcMetl 2 - 0 c AcMFFINILTLLVP No No External antigens, tk)r example fVa112-0~ fVFFINILTLLVP HMTa anti-MTF~ secreted bacterial {oxiu>', van 1)~ taken tip in endosomes, broken tllnld o w n to peptides and presented P91A wild type [STQNRRALDLVA No No P91A mutant ISTQNHRALDLVA H-2Ld anti-P91A with MHC class II antigens: proteins synthesized within the P35B wild type NGPPHSSNFGY H-2Dd anti-P35B cell, such as influenza hemagP35B mutant NGPPHSNNFGY H-2Dd anti-P35B glutinin, must be released and taken up again to be processed P198 wild type KYQAVTATLEE H-2Kd No and presented by MttC class 11 P198 mutant 14-24 KYQAV"_I~LEE H-2Kd anti-P198 molecules. MHC class I molecules P198 mutant 15-24 YQAVI~_ EE H-2Kd No bind peptides while tk)lding in P198 mutant 16-24 QAVI~_"LEE No No the endoplasmic reticulum 1~, and these peptides are derived Critical residues are underlined. The sixth residue of the mitochondrial ND1 protein determines the allelic forms of MTF21; the formyl (f) group on the aminofrom e n d o g e n o u s proteins; they terminal methionine is critical for binding to HMTa, the MHC class I molecule cannot be obtained by processing that presents this minor histocompatibiliW antigenS(,; Ac represents an acetyl of external proteins 12 but, luckily group. For the tumor transplantation (rum-) antigens P91A, P35B and P198, wild for immunologists, they can be type refers to the sequence found in the nonimmunogenic original tumor line supplied as synthetic peptides in and mutant refers to the immunogenic variants 17-22,23. the mediuml'L The peptides are 'mounted' in a groove formed by the restricting class I or class II Molecular def'mition of minor histocompatibility MHC molecule, which holds the peptide like a vise antigens (see front cover). The heavy chain of the MHC class I Three approaches have recentl} st_lccccdud in molecules is not stable in the properly folded contordc-fining transplantation antigens in molecular tc.rms. mation at 37°C unless it binds a peptide as well as [32- Mta, the maternally transmitted antigen of mice, conmicroglobulinL~. The peptides are so firmly b o u n d that sists of a mitochondrially e n c o d e d minor histocompatithey cocrystallize with the MHC molecules ~5. Natural bility antigen (MTF) that is presented by an MHC class antigens, such as influenza nucteoprotein 14, can be I molecule 2°. Four allelic forms of MTF are kl-lOwn. mimicked by synthetic peptides 10-15 amino acids and it was possible by sequencing the four representatong. The ability of these peptides to bind to MHC tive initochondrial g e n o m e s t() identify a ~,ingle c(xl()n molecules and be recognized by T cells is often in the N D I gene [encoding a 30 kl)a hydr()ph()bic subcritically d e p e n d e n t on single amino acicts I~,litimhm se~: H 220, 3146 6 McLaren. A. et al. (1988~ Proc .\?tt/Acad sci t ~ l ,vq. 644245445 7 Sinclair, A.tt. et al. (1990) Aature 3i6, 24(>244 8 Simpson. E. et al. (1987) A}~ture 320. 8"-(>8-8 9 Burgoyne, P.S., Levy, E.R. and McLaren, A. 172
rig JULY1991 VOL. 7 NO. 7
m
[~EVIEWS 10 Fisher, E.M.C. et al. (1990) Cell63, 1205-1218 11 Bevan, M.J. (1975) Nature 256, 419-421 12 Morrison, L.A. etal. (1986)J, Exp. Med. 163, 903-921 13 Townsend, A. el al. (1990) Cell 62, 285-295 14 Townsend, A.RM. et al. (1986) Cell 44, 959-968 15 Bjorkman, p.J. el al. (1987) Nature 329, 506-512 16 Shawar, S.M., Cook, R.G., Rodgers, J.R. and Rich, R.R. (1990) .L Exp. Med. 171,897-912 17 Sibille, C. etal. (1990) J. F_xp.Med. 172, 35-i5 18 Bjorkman, P.J. el al. (1987) Nature 329, 512-515 19 Falk, K., ROtzschke, O. and Rammensee, H-G. (1990) Nature 348, 248-251 20 Fischer Lindahl, K. (1985) Trend~ Genet. 1, 135-139 21 Loveland, B.E. etal. (1990) Ce1160, 971-980 22 Lurquin, C. et al. (1989) Cell 58, 293-303 23 Szikora, J-P. et al. (1990) EMBOJ. 9, 1041-1050 24 Wallny, H-J. and Rammensee, H-G. (1990) Nature 343,
32 Bailey, I).W. (1970) Transplant. Proc. 2, 32-38 33 Roopenian, D.C. and Davis, A.P. (1989) lmmunogenetics
30, 335-343 34 Davis, A.P. and Roopenian, D.C. (1990) Immunogenetics
31, 7-12
35 yon Boehmer, H., Haas, W. and Jerne, N.K. (1978) Proc. Natl Acad. Sci. USA 75, 2439-2442 36 Graff, R.J. (1978) Origins ¢flnbredMice (Morse, H.C., III,
ed.), pp. 371-389, Academic Press 37 Speiser, D.E. et al. (1990) Proc. Natl Acad. Sci. USA 87,
2021-2025 38 Simpson, E. (1987)lmmunol. Today8, 176-178 39 Wettstein, p.J. and Colombo, M.P. (1987).I. Immunol. 139,
2166-2171
40 Wraith, D.C. et al. (1990) Cel159. 247-255 41 von Boehmer, H., Teh, H.S. and Kisielow, P. (1990) lmmunol. Today' 10, 57451 42 VidoviO, I). and Matzinger, P. (1988)Nature 336, 222-225
275-278
43 Doolittle, D.P., Davisson, M.T., Roderick, T.H. and
25 Heath, W.R., Hurd, M.E., Carbone, ER. and Sherman, L.A. (1989) Nature 341,749-752 26 Schild, H., ROtzschke, O., Kalbacher, H. and Rammensee. H-G. (1990) Science 247, 1587-1589 27 ROtzschke, O. et al. (1990) Science 249, 283-287 28 ROtzschke, O. et al. (1990) Nature 348, 252-254 29 Van Bleek, G.M. and Nathenson, S.G. (1990) Nature 348,
Hillyard, A.L. (1990) Mouse Genome 87, 14-27
44 Garrett, T.EJ. el al. (1989) Nature 342, 6924596
•K.
FISCHER LINDAHL IS IN THE HOWARD HUGHES MEDICAL INSTITUTE, DEPARTMENTS OF MICROBIOLOGY AND BIOCHEMISTRY, UNIVERSITY OF TEXAS SOUTHWESTERNMEDICAL CENTER,
213-216
5323 HARRY HINES I~)ULEVARD~DALLA& TX 75235-9050,
30 Steinmuller, D. (1984) Immunol. Today 5, 234-240 31 Forman, J. et al. (1984)J. £vp. Med. 159, 1724--1740
USA.
The fruit fly, Drosophila melanogaster, is able to learn inff)rmation presented in a number of different learning situations. Since the development of the first reliable experimental paradigm in 1974 ~, numerous others have been devised and we now know that flies can be sensitized, habituated, learn associations with positive or negative reinforcement, and can be classically conditioned (reviewed in Ref. 2). Each of the learning situations offers the opportunity to study insect behavior itself, but each one also provides a phenotype and the possibility of isolating mutants. Genetic screens for mutants defective in learning/memory have yielded three that have been studied sufficiently to allow firm conclusions to be drawn regarding their biology. These are dunce (dnc), rutabaga (rut) and amnesiac (amn). The dnc mutants were the first to be identified; the locus defined by these mutants has been studied most thoroughly and it therefore serves as the prototypic Drosophila 'learning/memory' gene. We review here the structure, function, regulation and evolution of dnc. In addition, we identify some of the critical issues regarding dnc and other Drosophila learning/memory genes.
The Drosophiladunce 10cus.. learning and memory genes in the fly RONALD L. DAVISAND BRIG1TrEDAUWALDER The dunce gene, one of several genes criticalfor normal learning and memory in Drosophila, is organized in a complex and bizarre way, with enormous introns containing several other unrelated genes. Recent studies have focused on the spatial expression pattern of the product, cAMPphosphodiesterase, and have provisionaUy identified the mushroom bodies as important sites of action of dunce within adult braig In addition, the recent cloning and characterization of dunce counterparts from mammals has revealed that these too may participate in animal behavior and in particular, in the regulation of mood
quantifies the memory of dnc, rut and a m n flies at various times alter training. All of the mutants display some learning activity, defined as the appropriate avoidance behavior when tested as soon as possible after training (t = 0), although it is apparently reduced relative to normal flies. The mutations may be affecting acquisition, or they may be inducing a forgetfulness so rapid that memory is lost before it is possible to test the flies experimentally. The fact that the mutants show" normal learning scores immediately after training but rapid memory loss in some
The dnc behavioral syndrome The most sophisticated learning situation developed for flies is a classical conditioning procedure 3 that couples the presentation of olfactory cues to the negative reinforcement of electrical shock. When flies are put in a test tube and trained by presenting the ~ o sensory stimuli simultaneously~ the flies develop a long-lasting association, as si~own by tt~eir subsequent avoidance of the shock-associated odor. Figure 1 ik; IILY 1991 >),)1 li>cvlcl >~lc,ii
Minor hist0c0mpatibility antigens
H i s t o c o m p a t i b i l i t y antigens are cell surface molecules that lead to the rejection of grafted tissues and organs by the vertebrate immune system. A systematic genetic definition of minor histocompatibility antigens was undertaken by George Snell, w h o in the 1940s began breeding, by backcrossing and intercrossing, strains of mice that differed from each other by only a single histocompatibility antigenL He selected mice for their ability to reject a tumor graft from the inbred strain used in the backcrossing; the only mice that survived were those lacking, and therefore able to respond to, a minor histocompatibility antigen of the backcross strain. The sch6me was laborious and tended to select for differences at the stronger histocompatibility loci, including //-2, the major histocompatibility complex (MHC), because the tumors could outgrow the immune response to weaker minor histocompatibility antigens. Later, the selection was changed to skin graft rejection, a far more sensitive measure of incompatibility, which did not require homozygosity: progeny mice whose skin was rejected by the inbred backcross parent had retained a gene encoding a minor histocompatibility antigen. It was also found that congenic strains that had been bred by selection for a coat color marker, such as In or a t, often differed by a linked, minor histocompatibility locus. Today, more than 50 loci have been defined, and almost every chromosome of the mouse, including the mitochondrial genome, carries at least one (Fig. 1). Like mice, humans have a single MHC, the HLA complex on chromosome 6, and multiple minor histocompatibility loci. Typing for HLA antigens has
2
1
3
~.
5
7
8
9
KIRSTEN FISCHER LINDAHL
Histocompatibllity antigens have been studied for over 50years because they form a major obstacle to clinical transplantatiorL Human minor histocompatibility antigens remain ill-defined, but minor histocompatibillty loci have been mapped on nearly every mouse chromosome. Recent molecular definition of several transplantation antigens suggests that they are by-products of an immune system poised to present viral antigens, and a mutation in any gene may give rise to a new minor histocompatibility antige~
developed from a fine art into a high science e. and an HLA-identical donor is always preferred in clinical transplantation. However, with bone marrow: grafts, which are used for treatment of aplastic anemia and after eradication of lymphoid tumors, complications such as rejection or chronic graft-versus-host disease still arise in 20-70% of recipients of HLA-matched grafts, due to minor histocompatibility antigens:. It has been possible to recover lymphocytes from patients undergoing such a crisis and study the specificity of the response 4. The human minor histocompatibility antigen most easily defined is H-Y, the male antigen: lymphocytes from a female donor will attack the tissues ()f :i male recipient, all of which express the H-Y antigent
II
12
1/+
15
H-22 H'21+
H(IU
~t(In)
H-21 H(pi) H(go) H-27 H-I H-15 H-16
H-3 :H-42 .H-4 I+ H-45 -H-13
H-2~ H:d) _H-37 H-20 H -3 ~.
HOs) H-40
H-29
[
19
/-/-39
H-33
.H-8 H-4
17
Y
X
mf
iHya "~4:f H(ep)
H-2 H-31.,
.H-30
.H-43
.H(Eh)
H-7 H'19
_H-23
lH(fn)
H-I8 .H-28
Hxa
FIGI!
Genetic map of minor histoc()mpatibility loci in the mouse, based on Ref. i3. Thu mit()ch(~ndrial ot2i~(H]]ci Ill[) J', }LI[(){ ",({tit' ,
•ri(; Jt't'~' 1991 VOL 7 XO. 7 ] IULtd ({ K} I)L(~H 9 Ira9 91 502 I}0
E
i
i
[]~EVIEWS
Model I: T-cell receptors
Target cells
Model I1: T-cell receptors
f
Recognitionand lysis
(~ 9
H-Y antigen in MHC-dependentconformation and ~
MHC molecules
I k
J
HUE MHC restriction of the killer T-cell response to H-Y. Killer T cells from an MHC A female immunized with MHC A male cells can attack and lyse MHC A and AxB male target ceils, which present the male antigen (_~)with an A-type MHC molecule (AA), but they cannot lyse MHC B male target cells, which present the male antigen with a B-type MHC molecule ('B fiat'). Two models of T-cell receptors are illustrated: model t (top) assumes that the receptor interacts with both peptide and MHC molecule: model II (bottom) assumes that the receptor interacts only with peptide, which is presented in an MHC-dependent conformation.
The male antigen H-Y is the most studied of all minor histocompatibility antigens, p r o b a b l y because it is so easy to define: males have it and females do not. It was first observed when female mice rejected the skin of males of the same inbred strain s. As with all histocompatibility antigens, the rejection is caused by T lymphocytes, and it is possible to culture and study these cells in vitro. Antibodies do not detect H-Y, but helper and killer T cells provide reliable and efficient typing reagents for the antigen, which shows no polymorphism s. H-Y is e n c o d e d on the short arm of the mouse Y c h r o m o s o m e by the H y a gene, which is tightly linked to the testis-determining gene, Tdy ~'. In humans the gene encoding the testis-determining factor, 7DF, has been m a p p e d to the short arm of the Y c h r o m o s o m e r and the H - Y g e n e to the long arm s. Hya has not been
cloned, and its function is unknown; it may be involved in spermatogenesisg. It remains a riddle why the antigen should be expressed on cells, such as lymphocytes, that have no apparent sex-specific function. Alternatively, H-Y may represent a Y-specific isoform of a housekeeping protein like the human ribosomal protein $4, which is e n c o d e d by a gene located next to TDF t°.
T-cell recognition of antigen Minor histocompatibility antigens are recognized by T cells in an MHC-restricted fashion 11. Therefore T ('ells from a female immunized with MHC-compatible male cells can detect the male antigen only on cells of the same MHC type s (Fig. 2). For killer T cells, it is important that the polymorphic class I antigens of the MHC (in humans e n c o d e d by the HIA-A. HLA-B and H I A - C loci) are shared by the immunizing cells and
uc; ItaLY 1991 VOL. 7 :~0. 7
[]~EVIEWS the target; for helper T cells, it is TABLE1. Defined peptides that Entm|c minor histocompatibility and the class II antigens of the MHC tumor antigens (in humans e n c o d e d by the HLA D series of loci) that matter. Peptide Sequence Bound by MHC Recognition by MHC restriction also applies to class I molecule killer T cells T-cell responses to foreign proMTF tein antigens and viral antigens, ND 1oil-17 fMFFINI_LTLLVPIIJAM HMTa anti-MTFct and the basis is now v,cll ND 1[ 3 1 - 1 7 fMFFINALTLLVPILIAM HMT~ anti-MTF~ understood from studies in these systems. As a rule (but with fMet 1 2 - c z fMFFINILTLLVP HMT~ anti-MTF~ exceptions), cellular processing of Metl 2-0t MFFINILTLLVP No No antigen follows txvo pathways~-'. AcMetl 2 - 0 c AcMFFINILTLLVP No No External antigens, tk)r example fVa112-0~ fVFFINILTLLVP HMTa anti-MTF~ secreted bacterial {oxiu>', van 1)~ taken tip in endosomes, broken tllnld o w n to peptides and presented P91A wild type [STQNRRALDLVA No No P91A mutant ISTQNHRALDLVA H-2Ld anti-P91A with MHC class II antigens: proteins synthesized within the P35B wild type NGPPHSSNFGY H-2Dd anti-P35B cell, such as influenza hemagP35B mutant NGPPHSNNFGY H-2Dd anti-P35B glutinin, must be released and taken up again to be processed P198 wild type KYQAVTATLEE H-2Kd No and presented by MttC class 11 P198 mutant 14-24 KYQAV"_I~LEE H-2Kd anti-P198 molecules. MHC class I molecules P198 mutant 15-24 YQAVI~_ EE H-2Kd No bind peptides while tk)lding in P198 mutant 16-24 QAVI~_"LEE No No the endoplasmic reticulum 1~, and these peptides are derived Critical residues are underlined. The sixth residue of the mitochondrial ND1 protein determines the allelic forms of MTF21; the formyl (f) group on the aminofrom e n d o g e n o u s proteins; they terminal methionine is critical for binding to HMTa, the MHC class I molecule cannot be obtained by processing that presents this minor histocompatibiliW antigenS(,; Ac represents an acetyl of external proteins 12 but, luckily group. For the tumor transplantation (rum-) antigens P91A, P35B and P198, wild for immunologists, they can be type refers to the sequence found in the nonimmunogenic original tumor line supplied as synthetic peptides in and mutant refers to the immunogenic variants 17-22,23. the mediuml'L The peptides are 'mounted' in a groove formed by the restricting class I or class II Molecular def'mition of minor histocompatibility MHC molecule, which holds the peptide like a vise antigens (see front cover). The heavy chain of the MHC class I Three approaches have recentl} st_lccccdud in molecules is not stable in the properly folded contordc-fining transplantation antigens in molecular tc.rms. mation at 37°C unless it binds a peptide as well as [32- Mta, the maternally transmitted antigen of mice, conmicroglobulinL~. The peptides are so firmly b o u n d that sists of a mitochondrially e n c o d e d minor histocompatithey cocrystallize with the MHC molecules ~5. Natural bility antigen (MTF) that is presented by an MHC class antigens, such as influenza nucteoprotein 14, can be I molecule 2°. Four allelic forms of MTF are kl-lOwn. mimicked by synthetic peptides 10-15 amino acids and it was possible by sequencing the four representatong. The ability of these peptides to bind to MHC tive initochondrial g e n o m e s t() identify a ~,ingle c(xl()n molecules and be recognized by T cells is often in the N D I gene [encoding a 30 kl)a hydr()ph()bic subcritically d e p e n d e n t on single amino acicts I~,litimhm se~: H 220, 3146 6 McLaren. A. et al. (1988~ Proc .\?tt/Acad sci t ~ l ,vq. 644245445 7 Sinclair, A.tt. et al. (1990) Aature 3i6, 24(>244 8 Simpson. E. et al. (1987) A}~ture 320. 8"-(>8-8 9 Burgoyne, P.S., Levy, E.R. and McLaren, A. 172
rig JULY1991 VOL. 7 NO. 7
m
[~EVIEWS 10 Fisher, E.M.C. et al. (1990) Cell63, 1205-1218 11 Bevan, M.J. (1975) Nature 256, 419-421 12 Morrison, L.A. etal. (1986)J, Exp. Med. 163, 903-921 13 Townsend, A. el al. (1990) Cell 62, 285-295 14 Townsend, A.RM. et al. (1986) Cell 44, 959-968 15 Bjorkman, p.J. el al. (1987) Nature 329, 506-512 16 Shawar, S.M., Cook, R.G., Rodgers, J.R. and Rich, R.R. (1990) .L Exp. Med. 171,897-912 17 Sibille, C. etal. (1990) J. F_xp.Med. 172, 35-i5 18 Bjorkman, P.J. el al. (1987) Nature 329, 512-515 19 Falk, K., ROtzschke, O. and Rammensee, H-G. (1990) Nature 348, 248-251 20 Fischer Lindahl, K. (1985) Trend~ Genet. 1, 135-139 21 Loveland, B.E. etal. (1990) Ce1160, 971-980 22 Lurquin, C. et al. (1989) Cell 58, 293-303 23 Szikora, J-P. et al. (1990) EMBOJ. 9, 1041-1050 24 Wallny, H-J. and Rammensee, H-G. (1990) Nature 343,
32 Bailey, I).W. (1970) Transplant. Proc. 2, 32-38 33 Roopenian, D.C. and Davis, A.P. (1989) lmmunogenetics
30, 335-343 34 Davis, A.P. and Roopenian, D.C. (1990) Immunogenetics
31, 7-12
35 yon Boehmer, H., Haas, W. and Jerne, N.K. (1978) Proc. Natl Acad. Sci. USA 75, 2439-2442 36 Graff, R.J. (1978) Origins ¢flnbredMice (Morse, H.C., III,
ed.), pp. 371-389, Academic Press 37 Speiser, D.E. et al. (1990) Proc. Natl Acad. Sci. USA 87,
2021-2025 38 Simpson, E. (1987)lmmunol. Today8, 176-178 39 Wettstein, p.J. and Colombo, M.P. (1987).I. Immunol. 139,
2166-2171
40 Wraith, D.C. et al. (1990) Cel159. 247-255 41 von Boehmer, H., Teh, H.S. and Kisielow, P. (1990) lmmunol. Today' 10, 57451 42 VidoviO, I). and Matzinger, P. (1988)Nature 336, 222-225
275-278
43 Doolittle, D.P., Davisson, M.T., Roderick, T.H. and
25 Heath, W.R., Hurd, M.E., Carbone, ER. and Sherman, L.A. (1989) Nature 341,749-752 26 Schild, H., ROtzschke, O., Kalbacher, H. and Rammensee. H-G. (1990) Science 247, 1587-1589 27 ROtzschke, O. et al. (1990) Science 249, 283-287 28 ROtzschke, O. et al. (1990) Nature 348, 252-254 29 Van Bleek, G.M. and Nathenson, S.G. (1990) Nature 348,
Hillyard, A.L. (1990) Mouse Genome 87, 14-27
44 Garrett, T.EJ. el al. (1989) Nature 342, 6924596
•K.
FISCHER LINDAHL IS IN THE HOWARD HUGHES MEDICAL INSTITUTE, DEPARTMENTS OF MICROBIOLOGY AND BIOCHEMISTRY, UNIVERSITY OF TEXAS SOUTHWESTERNMEDICAL CENTER,
213-216
5323 HARRY HINES I~)ULEVARD~DALLA& TX 75235-9050,
30 Steinmuller, D. (1984) Immunol. Today 5, 234-240 31 Forman, J. et al. (1984)J. £vp. Med. 159, 1724--1740
USA.
The fruit fly, Drosophila melanogaster, is able to learn inff)rmation presented in a number of different learning situations. Since the development of the first reliable experimental paradigm in 1974 ~, numerous others have been devised and we now know that flies can be sensitized, habituated, learn associations with positive or negative reinforcement, and can be classically conditioned (reviewed in Ref. 2). Each of the learning situations offers the opportunity to study insect behavior itself, but each one also provides a phenotype and the possibility of isolating mutants. Genetic screens for mutants defective in learning/memory have yielded three that have been studied sufficiently to allow firm conclusions to be drawn regarding their biology. These are dunce (dnc), rutabaga (rut) and amnesiac (amn). The dnc mutants were the first to be identified; the locus defined by these mutants has been studied most thoroughly and it therefore serves as the prototypic Drosophila 'learning/memory' gene. We review here the structure, function, regulation and evolution of dnc. In addition, we identify some of the critical issues regarding dnc and other Drosophila learning/memory genes.
The Drosophiladunce 10cus.. learning and memory genes in the fly RONALD L. DAVISAND BRIG1TrEDAUWALDER The dunce gene, one of several genes criticalfor normal learning and memory in Drosophila, is organized in a complex and bizarre way, with enormous introns containing several other unrelated genes. Recent studies have focused on the spatial expression pattern of the product, cAMPphosphodiesterase, and have provisionaUy identified the mushroom bodies as important sites of action of dunce within adult braig In addition, the recent cloning and characterization of dunce counterparts from mammals has revealed that these too may participate in animal behavior and in particular, in the regulation of mood
quantifies the memory of dnc, rut and a m n flies at various times alter training. All of the mutants display some learning activity, defined as the appropriate avoidance behavior when tested as soon as possible after training (t = 0), although it is apparently reduced relative to normal flies. The mutations may be affecting acquisition, or they may be inducing a forgetfulness so rapid that memory is lost before it is possible to test the flies experimentally. The fact that the mutants show" normal learning scores immediately after training but rapid memory loss in some
The dnc behavioral syndrome The most sophisticated learning situation developed for flies is a classical conditioning procedure 3 that couples the presentation of olfactory cues to the negative reinforcement of electrical shock. When flies are put in a test tube and trained by presenting the ~ o sensory stimuli simultaneously~ the flies develop a long-lasting association, as si~own by tt~eir subsequent avoidance of the shock-associated odor. Figure 1 ik; IILY 1991 >),)1 li>cvlcl >~lc,ii
