Copyright 0 1987 by the Genetics Society of America

The Putative Oncogene Pim-1 in the Mouse: Its Linkage and Variation Among t Haplotypes

Joseph H. Nadeau' and Sandra J. Phillips The Jackson Laboratory, Bar Harbor, Maine 04609

Manuscript received April 2, 1987 Revised copy accepted July 20, 1987

ABSTRACT Pim-1, a putative oncogene involved in T-cell lymphomagenesis, was mapped between the pseudo-

alpha globin gene Hba-4ps and the alpha-crystallin gene Crya-1 on mouse chromosome 17 and therefore within the t complex. Pim-I restriction fragment variants were identifKd among t haplotypes. Analysis of restriction fragment sizes obtained with 12 endonucleases demonstrated that the Pim-1 genes in some t haplotypes were indistinguishable from the sizes for the Pim-1' allele in BALB/c inbred mice. T h e r e are now three genes, Pint-1, Crya-1 and H-2 I-E, that vary among independently derived t haplotypes and that have indistinguishable alleles in t haplotypes and inbred strains. These genes are closely linked within the distal inversion of the t complex. Because it is unlikely that these variants arose independently in t haplotypes and their wild-type homologues, we propose that an exchange of chromosomal segments, probably through double crossingover, was responsible for indistinguishable Pim-1 genes shared by certain t haplotypes and their wild-type homologues. T h e r e was, however, no apparent association between variant alleles of these three genes among t haplotypes as would be expected if a single exchange introduced these alleles into t haplotypes. If these variant alleles can be shown to be identical to the wild-type allele, then lack of association suggests that multiple exchanges have occurred during the evolution of the t complex.

ENES affecting a variety of developmental and G genetic processes including male fertility, em- bryonic development and meiotic transmission are located within the t complex (BENNETT 1975; LYON 198 1; SILVER 1985). These genes are distributed over a 15 centiMorgan (cM) segment of chromosome (Chr) 17 representing about 1 % of the genome. One of the exceptional features of the t complex is that variant alleles of genes affecting each of these properties are usually inherited together as a single Mendelian unit rather than assorting through recombination (DUNN and CASPARI 1945; DUNN and GLUECKSOHN-SCHOEN- HEIMER 1950). Maintenance of this unique combina- tion of alleles results from recombination suppression (LYON and PHILLIPS 1959; HAMMERBERG and KLEIN 1975a; LYON et al. 1979; NADEAU, PHILLIPS and EGO- ROV 1985). Although the basis for suppression was poorly understood for many years, it is now well- established that at least two paracentric inversions are responsible for suppressing recombination in mice heterozygous for a t haplotype and its wild-type hom- ologue (SILVER and ARTZT 1981; ARTZT, SHIN and BENNETT 1982; PLA and CONDAMINE 1984; HERR- MANN et al. 1986; HERRMANN, BARLOW and LEHRACH 1987). Although suppressed, however, very rare re- combination between the t complex and its wild-type homologue does occur (LYON and PHILLIPS 1959;

' To whom correspondence should be sent.

Genetics 117: 533-541 (November, 1987)

LYON and MEREDITH 1964a-c). Crossovers producing these partial t haplotypes usually occur within the proximal inversion or between the proximal and distal inversions; crossovers within the distal inversiun have not been described (HERRMANN et al. 1986; HERR- MA", BARLOW and LEHRACH 1987; SCHIMENTI et al. 1987). Mice with these partial t haplotypes usually lose some but not all of the properties associated with complete t haplotypes (LYON and PHILLIPS 1959; LYON and MEREDITH 1964a-c; LYON 1984, 1986). Loss of properties essential for the maintenance of t haplotypes in natural populations (LYON 1984; SILVER 1985) probably accounts for the absence of partial t haplotypes in wild mice (M. ERHART, S. J. PHILLIPS and J. H. NADEAU, unpublished data).

One of the most important consequences of recom- bination suppression is that the combination of alleles producing the various phenotypic properties of t h e t complex and variant alleles of other genes within the t complex are preserved. Surveys of genetic variation among t haplotypes and their wild-type homologues have demonstrated that, with the exception of the t lethal genes, alleles are usually shared among all t haplotypes and that alleles associated with t haplotypes are either not found in mice that do not have a t complex or do not vary among wild mice, regardless of whether they have a t haplotype (HAMMERBERG and KLEIN 197513; HAMMERBERG et al. 1976; LEVIN- SON and MCDEVITT 1976; HAUPTFELD, HAMMERBERG

534 J. H. Nadeau and S. J. Phillips

and KLEIN 1976; NADEAU 1983, 1986; SILVER et al. 1983; Fox, SILVER and MARTIN 1984; ARTZT et al. 1985; FIGUEROA et al. 1985; M. ERHART, S. J. PHIL- LIPS and J. H. NADEAU, unpublished data). H-2 I-E has recently been shown not only to vary among t haplotypes but to have indistinguishable alleles asso- ciated with some t haplotypes and their wild-type homologues (DEMBIC, SINGER and KLEIN 1984; DEM- BIC et al. 1985). Because H-2 genes appear to be subject to unusual evolutionary pressures and con- straints (KLEIN 1980; KLEIN and FIGUEROA 1986), the patterns of I-E variation may not be indicative of patterns for other genes within the t complex. Com- parable results, however, were recently obtained for Crya-1 suggesting that this pattern of variation may not be limited to genes within the H-2 complex (SKOW et al. 1987). Pim-1, a putative oncogene involved in T-cell lymphomagenesis (CUYPERS et al. 1984; SEL- TEN, CUYPERS and BURNS 1985; SELTEN et al. 1986), has been assigned to Chr 17 by using somatic cell hybrids (HILKENS et al. 1986). We now show that Pim- I is located between the pseudo-alpha globin gene Hba-4ps and the alpha-1 crystallin gene Crya-1 within the t complex. We also show that indistinguishable alleles of Pim-1 are shared between certain t haplo- types and their wild-type homologues, that genes showing this pattern of variation are located exclu- sively within the distal inversion of the t complex, and that there is no obvious association between genes at each of these three loci among t haplotypes. We pro- pose that this pattern of variation results from an exchange, probably involving double crossingover, within the distal inversion.

MATERIALS AND METHODS

Mice: Mice were obtained from the research and produc- tion colonies of the Jackson Laboratory.

Genomic DNA: Most of the DNAs from inbred, con- genic, and recombinant inbred strains were purchased from the Mouse DNA Resource of the Jackson Laboratory. Other genomic DNAs were prepared from spleen by using meth- ods described by PHILLIPS and NADEAU (1 984) and NADEAU, PHILLIPS and EGOROV (1 985).

Pim-I probe: Probe A is a 923-bp BamHI genomic frag- ment containing portions of exons 5 and 6 and the interven- ing intron (CUYPERS et al. 1984; SELTEN et al. 1986).

Southern analysis: DNA (4 pg/lane) was digested with a 1 0-fold excess of restriction endonuclease for 3-6 hr under conditions recommended by the enzyme supplier (Bethesda Research Labs, Gaithersburg, Maryland). DNAs were elec- trophoresed in horizontal 1 % agarose gels at 35 V for 18- 20 hr using E electrophoresis buffer (E buffer is 0.04 M Tris, 0.02 M sodium acetate, 0.001 M Na2EDTA, p H 7.2). T w o gels were run simultaneously in long-bed gel boxes (Dankar Corporation, Reading, Massachusetts). Gels were prepared for transfer by soaking in 0.2 M NaOH, 0.6 M NaCI, 0.15% thymol blue for 1 hr , followed by neutraliza- tion in 1 M Tris, p H 7.4, 0.6 M NaCl for 1 hr. Transfer to Zetabind nylon membranes (AMF-Cuno, Stamford, Con- necticut) was overnight in lOXSSC as described previously

(SOUTHERN 1975). Membranes were blotted dry for 5 min at room temperature and UV irradiated for 30 sec by using a 254 nm UV transilluminator.

Preparation of probes and hybridization: Total plasmid DNA (200 ng) was linearized and labeled using random hexamers (Pharmacia, Piscataway, New Jersey) as described previously (FEINBERG and VOGELSTEIN 1983). Newly pre- pared filters were prewashed in O.lXSSC, 0.5% SDS at 65" for 1 hr , followed by soaking in 1X Denhardt's solution, 3XSSC for 2-24 hr at 65". Filters were hybridized in GXSSC, 1% Denhardt's solution, 0.5% SDS in S256 (S256 is 8 pg/ml poly(A), 8 pg/ml poly(C), 200 pg/ml yeast tRNA, 10 pg/ml Escherichia coli DNA, and 50 rg/ml salmon sperm DNA in 3 mM Tris, p H 8.1) with 1-3 X lo6 cpm/ml of probe overnight at 70" . Filters were then washed exten- sively in O.lXSSC, 0.05% SDS at 52" , dried, and exposed to Kodak XAR film using DuPont Lightening-plus intensi- fying screens at -70" for 1-5 days.

Statistical analysis: T h e maximum likelihood method described by BISHOP (1985) was used to identify the most likely gene order and to calculate the LOD score for order. Maximum likelihood methods described by HALDANE and WADDINGTON (1931) and by TAYLOR (1981) were used to calculate recombination frequencies.

RESULTS

Pim-2 restriction fragment variation among in- bred strains: Three restriction endonucleases, BamHI, HincII and TaqI, were used to examine re- striction fragments among selected inbred strains. Three additional endonucleases, EcoRI, Hind111 and MspI, were used to examine restriction fragments among progenitors of recombinant inbred (RI) strains. Variation was detected with BamHI, HincII and TaqI, but not with any other endonuclease tested. The six combinations of restriction fragments ob- served define six alleles of Pim-1 (Table 1).

Analysis of recombinant inbred and congenic strains: Restriction fragment variants involving 69 strains from 3 sets of recombinant inbred strains were used to map Pim-1 (Table 2). Maximum likelihood methods (BISHOP 1985) were used to identify the most likely gene order and to calculate the LOD score for order (Table 3). The gene order Hba-4ps-Pim-1-H-2 was 43 times more likely than alternative orders and the order Pim-1-Cry-1-H-2 was 729 times more likely than alternative orders (Table 3). Analysis of congenic strains support this order (Table 4). Results for the B6.TC2/Rn and B6.TC3/Rn strains (RINCHIK and AMOS 1983) suggest that Pim-1 is located proximal to C y - 1 and results for the LT.MA-GEo-lb strain suggest that Pim-1 is located proximal to H-2, whereas results for the B10.D2 strain suggest that Pim-1 is located distal to D l 7Leh54. Recombination percent- ages were Hba-4ps-4.4 f 2.1-Pim-1-1.1 & 0.8-Crya- 1. Because both Hba-4ps and Crya-1 are located within the t complex (D'EUSTACHIO et al. 1984; Fox, SILVER and MARTIN 1984; Skow and DONNER 1985; SKOW et al. 1987), Pim-1 must also be located within the t complex.

Pim-I Oncogene

TABLE 1

Variation of Pim-1 among inbred and wild mice and among t haplotypes

535

~~

Fragment sizes (kb)

Pin-I allele BamHI HinclI TaqI Strain, species or I haplotype

Pim-I" ~~

2.2

Pim-I

Pim-1' Pim- I

Pim- I e

Pim- I f Pim- 1 '

0.9

5.1, 12.0 2.2

2.2 2.2 2.2

13.8 2.8 A/J, A/HeJ, A/WySnJ, AKR/J, CBA/CaJ, CEIJ, C3H/HeJ, CSHeB/FeJ, FL/1 Re, PERU-ATTECK/CamEi, YBR/ KiEi

10.2 2.8 BALB/cBy, BALB/cJ, C57BL/KsJ, C57BR/cdJ, C57L/J, C58/J, DBA/lJ, DBA/2J, FS/Ei, LP/J, LT/Sv, SEA/ GnJ, SEC/lReJ, TF/GnLe, 129/Sv, tw', t", t", tWl2, tw", tu71 p 7 5

10.2 2.8 NZB~BINJ, RIIIS/J 10.2 2.8 AEJ/GnRk, BDP/J, BUB/BnJ, C57BL/6J, C57BL/6By,

C57BL/1 OJ, IS/CamEi, MA/MyJ, P/J, SF/CamEi, SJL/J, SK/CamEi, SM/J, ST/bJ, SWR/J

13.8 3.0 CAST/Ei, MOLF/Ei, RF/J 10.2 2.5 Mus sgretw 18.0 3.4 to, t 6 , t w 3 2

TABLE 2

Inheritance of Pin-1 and other Chr 17 genes in the AKXD, AKXL, and BXD RI strains"

AKXD strain

Geneb 1 2 3 6 7 8 9 I O 1 1 12 13 14 15 16 17 18 20 21 22 23 24 25 26 27 28

Pim-1 A D A A A D D A A D D A A D A D D A D D D A A A A

C r y - 1 A D A A A D D A A D D D A D A D D A D D D A A A A H-2 A D A - A D D A A D D D A D A D D A D D D A A - A

x

AKXL strain

Geneb 5 6 7 8 9 12 13 14 16 17 19 21 24 25 28 29 37 38

Hba-4ps L L L A L L A L L L L A A L A A L A

Pim-1 L A L A L L A L L L L A A L A L L A

H - 2 L A L A L L A A L L L A L L A L L A

x x

x x

BXD strain

Gene' 1 2 5 6 8 9 1 1 12 13 14 15 16 18 19 20 21 22 23 24 25 27 28 29 30 31

Hba-4ps D B D D B B B D B - B B D B B D D B D D B D B D D

Pima- 1 D B D D B D D B B B B D D B B D D B D D B D B D D

Crya-1 D B D D B D D D B B B D D B B D D B D D B D B D D

H-2s D B D D B D D D B B B D D B B D D B D D D D B D D

x x x x

x

x

a The progenitors of the AKXD, AKXL, and BXD RI strains are AKR/J and DBA/2J, AKR/J and C57L/J, and C57BL/6J and DBA/2J, respectively. A, B, D, and L are used as generic terms for alleles inherited from AKR/J, C57BL/6J, DBA/2J, and C57L/J respectively. An x is used to denote intervals where crossing over resulted in recombination.

Sources of data are the following: Crya-I [SKOW and DONNER (1985), except data for BXD-5, BXD-20 and BXD-27 which are from H. LEHRACH (personal communication)], H-2 in AKXD [SKOW and DONNER (1985), except data for AKXD-11, AKXD-21, and AKXD-26 which are from H. C. MORSE and B. A. TAYLOR (personal communication)], Hba-4ps (D'EUSTACHIO et al. 1984), H-2 in AKXL (CLAFLIN et al. 1978), Hba-4ps in BXD (MANN, ELLIOT and HOHMAN 1984), and H-2S (RUSSELL et al. 1981).

Pim-I variation among t haplotypes: To determine whether Pim-1 varies among t haplotypes, 10 inde- pendently derived t haplotypes were each digested with BamHI, BanII, BclI, EcoRI, HincII, HindIII, PstI, PvuII, Sad, StuI, Taql or XbaI. The to, t6 and tw3' haplotypes had a combination of BamHI, HincII, and TaqI restriction fragment variants (Figure 1) not

found in other t haplotypes or inbred or wild mice tested (Table 1). The 2.2-kb BamHI fragment in the Pim-I allele associated with these three t haplotypes was indistinguishable from that found in C57BL/6J and other inbred strains with the Pim-ld allele. By contrast, the 18-kb HincII and the 3.4-kb Tag1 frag- ments were found only in the Pim-1 allele associated

536 J. H. Nadeau and S. J. Phillips

TABLE 3

Maximum likelihood analysis of gene order for Hba-4ps, Pim-I, Cya-1, and H-2 in RI strains

R I strain Log likelihood for order

AKXL

AKXD, BXD

BXD

Hba-4ps - Pim-1 - H-2 -6.4 1 -8.04 -8.04

LOD score for order: 1.63

Comment: The most likely order is 43 times more likely than the order with the next

Pim-I - Crya-I - H-2 -8.29 -11.15 -11.15

LOD score for order: 2.86

Comment: The most likely order is 729 times more likely than the order witK the

Hba-4ps - Pim-1 - Crya-1 -10.42 -1 1.04 -12.03

LOD score for order: 0.62

Comment: The most likely order is 4 times more likely than the order with the next

Hba-4ps - H - 2 - Pim-I Pim-1 - Hba-4ps - H-2

largest likelihood

Pim-1 - H-2 - Crya-I Crya-1 - Pim-I - H-2

next largest likelihood Hba-4ps - Crya-I - Pim-1 Pim-1 - Hba-4ps - Crya-I

largest likelihood

TABLE 4

Mapping of Pim-I in congenic strains"

Strain DI 7Leh54 Hba-4ps

C57BL/ 1 OJ B B

DBA/2J D D

LT/Sv L L L T . M A - G ~ O - ~ ~ ? M MA/MyJ M M C57BL/6J B B BG.TCZ/Rn - _.

B6.TC3/Rn - -

BALBIc C C

B10.D2 B x D

Pim-1

B D D

L M M B C C C

Glo-l Crya-1 H-2 ~~

B B B - D D D D D

L L L M x ? L M M M

B B B C x B B C x B B c c c

B, D, L, M and C are used as generic terms for alleles inherited from C57BL/1 OJ (C57BL/6J), DBA/2J, LT/Sv, MA/MyJ and BALB/c, respectively. References for typing data are the following: D17Leh54 (M.-B. TCHETGEN and J. H. NADEAU, unpublished obser- vations), Hba-4ps [D'EUSTACHIO et al. (1984), except B10.D2 and LT.MA-Glo-Ib which is D. BURKART and J. H. NADEAU (unpub- lished observations)], Glo-I (EPPIG and EICHER 1983; RINCHIK and AMOS 1983), Crya-I [SKOW and DONNER (1985), except for B10.D2 which is D. BURKART and J. H. NADEAU (unpublished observa- tions)], and H-2 (KLEIN, FIGUEROA and DAVID 1983).

with the t*, t 6 and t w j 2 haplotypes. This Pim-I allele, which was defined by the unique HincII and Tag1 restriction fragments, represents a seventh allele of Pim-1 and is designated Pim-1'. Pim-I variants among the 10 t haplotypes were not detected with any of the other endonucleases tested.

Organization of the Pim-1 gene associated with t haplotypes: Restriction fragment sizes observed in the previous analysis for each of the 10 t haplotypes were compared to the fragment sizes predicted by a se- quence analysis of the Pim-I gene in BALB/c mice (SELTEN et al. 1986). Results of this comparison are presented in Figure 2. The only restriction fragments

that were unique to the t complex were BumHI, HincII and Tug1 fragments observed in the Pim-1' gene as- sociated with the to, t 6 and tw3' haplotypes. Other fragments observed in these and the other seven hap- lotypes analyzed were indistinguishable from pre- dicted fragment sizes for the Pim-I allele (Table 1, Figure 2). These results suggest that the Pim-I genes in certain t haplotypes and inbred strains were indis- tinguishable.

Lack of association between variant alleles among t haplotypes: The occurrence of Pim-I variants among a panel o f t haplotypes was compared to the occur- rence of alleles of other genes known to vary among t haplotypes (Table 5 ) . It is readily apparent that there is no association between alleles of any of these genes. Lack of association suggests that a single exchange is not a sufficient explanation for the pattern of variation of these genes among t haplotypes.

DISCUSSION

Restriction fragment sizes suggested that the Pim-1 genes in t haplotypes and their wild-type homologues are organized in very similar ways. Analysis of ge- nomic DNAs with endonucleases such as BanII, PstI, PvuII , Sad, and StuI revealed fragments whose sizes were indistinguishable from those predicted by anal- ysis of the DNA sequence of the Pim-I gene from BALB/c mice (Figure 2). Although Probe A recog- nizes a relatively small portion of the Pim-I gene, the analysis of Pim-I organization extended well beyond the limits of the 923 bp homologous to the probe. For example, BclI, EcoRI, Hind11 and XbaI restriction fragments extended beyond the 5 ' , 3' or both bound- aries of the Pim-I gene (Figure 2). There was no evidence that deletions, insertions or other structural

Pim-I Oncogene 537

12.0 kb - 5.1 kb+

2.2 kb+

0.9 kb 4

A

FIGURE 1.-(A) Inbred strains. BamHl restriction fragments in CSH/HeJ (Pim-Ia), BALB/cJ (Pim-I*) and NZB/BINJ (Pim-1‘) mice are illustrated. Fragment sires (kb) are also given. (B) t haplotypes. The Pim-I‘ and Pim-lb alleles associated with the t’” and P” hap lotypes are illustrated. The 1”” and t” haplotypes belong to the t” and t”” complementation groups, respectively (BENNETT 1975; SILVER 1985). CSH-Tlfl+lf was included as a control. Genomic DNAs were digested with Tuql. Fragment sires (kb) are also given.

rearrangements of Pim-1 were responsible for the variation observed. Thus for all haplotypes except t”, t6 and tw3’, the Pim-I genes in t haplotypes and those in wild-type chromosomes were indistinguishable.

The organization of the Pim-1 gene associated with the t”, t6 and tw3’ haplotypes was very similar to the Pim-Ib allele in BALB/c mice. Base substitutions prob- ably account for the variable restriction sites. For example, gain (or loss) of a BamHI restriction site at position 5639 (SELTEN et al. 1986) accounts for the difference between the 0.9 kb and the 2.2 kb frag- ments (Figure 2). Similar arguments apply to the HincII and Tagl variants (data not shown). Although the variable HincII restriction site was probably lo- cated distal to the 3’ end of the gene, the BamHI

restriction site occurred in the translated portion of the last exon and the Tagl restriction site occurred in the 3’ untranslated portion of the Pim-1 gene (Figure 2) ( S . J. PHILLIPS and J. H. NADEAU, unpublished observations).

Three patterns of variation of genes among t h a p lotypes and their wild-type homologues can now be recognized. The two previously identified patterns are either variants unique to the t complex and shared by all t haplotypes or genes that vary among t haplo- types but with variants in all cases unique to the t complex. Examples of the former are t complex pro- teins-1 -9 (Tcp-I-Tcp-9) (SILVER, WHITE and ARTZT 1980; SILVER et al. 1983), glyoxalase-1 (Glo-I) (NA-

1984), and several randomly selected genomic se- quences within the t complex (ROHME et al. 1983; Fox, SILVER and MARTIN 1984; HERRMANN et al. 1986; J. H. NADEAU, M. ERHART, D. BURKART, S. J. PHILLIPS and M.-B. TCHETGEN, unpublished obser- vations), and examples of genes with unique alleles that vary among t haplotypes are the t lethal genes (BENNETT 1975), complement component4 gene C4 (GOLUBIC et al. 1984), and other class I and class I1 genes within the H-2 complex (HAMMERBERG and KLEIN 1975k LEVINSON and MCDEVITT 1976; SILVER 1982; SHIN et al. 1982).

The third pattern consists of genes that vary among t haplotypes and that have indistinguishable alleles associated with t haplotypes and some wild-type chro- mosomes. These alleles include Pim-1 (Table 1 and Figure 2), Crya-1 a and Crya-1 (SKOW et al. 1987), and I-E” (DEMBIC, SINGER and KLEIN 1984; DEMBIC et al. 1985). Extensive restriction fragment analysis for each of these alleles failed to reveal any differences between the allele associated with some t haplotypes and the corresponding alleles in inbred strains. These variant alleles are the only exceptions to the complete gametic disequilibrium associated with the t complex (HAM- MERBERG and KLEIN 1975b; NADEAU 1983).

Three hypotheses could account for variation of Pim-1 and other genes among t haplotypes: (a) t h a p lotypes (and therefore Pim-1 and other variants) could have independent origins, (b) indistinguishable Pim-I variants in t haplotypes and their wild-type homo- logues could have arisen independently through mu- tation and convergent evolution after t haplotypes originated, or (c) Pim-1 and other alleles could be exchanged between t haplotypes and their wild-type homologues. The first hypothesis does not seem very likely for two reasons. Although independent origins would account for the Pim-I variants associated with the to, t6 and tw3’ haplotypes, it would not account for the numerous variants of other genes such as Tcp-I- 9 (SILVER et al. 1983) that appear to be unique to and shared by all t haplotypes. Similarly, the second hy-

DEAU 1986). Hba-4pS (FOX, SILVER and MARTIN

538 J. H. Nadeau and S. J. Phillips

0

I I I I I I I I I I I I

CDI m

I I I I I I I I I I I I I

-I I I I I I I I I I

ml

ml

I I I I I I

Pim-1 Oncogene 539

TABLE 5

Variation of Pitn-I and other genes among t haplotypes"

t haplotype tcl Pim-1 Crya-1 H-2

to t6

t W X

tW' tw12

tw75

tW5 tW8

p 1 8

y 7 l

0 0 4 12 w l wl wl wl + w5 w5 SL

t a w29 t a w30 b N D Wild type t t w28 b t w30 b ND w30 b t w30 b t w31 b a w31 b t w29

~~

a Sources of data for genes other than Pim-1 are the following: tcl (BENNETT 1975), Cya-1 (SKOW et al. 1987), and H-2 (KLEIN, FICUEROA and DAVID 1983). ND indicates not done and SL indi- cates semilethal, i .e., no lethal gene. Pim-1, Crya-1, and H-2 are listed in order. Because t lethal genes are nonallelic (ARTZT, MC- CORMICK and BENNETT 1982; ARTZT 1984; SHIN, BENNETT and ARTZT 1984), they are listed here regardless of their location in each haplotype.

pothesis does not seem very likely. Analysis of the Pim-1 allele in certain t haplotypes and certain wild- type homologues with 12 different endonucleases re- vealed indistinguishable restriction maps (Figure 2b). The probability of obtaining indistinguishable restric- tion maps resulting from multiple changes in inde- pendent genes is negligible. The third hypothesis seems most likely. Evidence for this hypothesis is twofold: First, other genes that vary among t haplo- types are located in the same portion of the t complex. With the exception of the t lethal genes, genes known to differ among t haplotypes are closely linked to each other within the paracentric inversion associated with the distal portion of the t complex. Genes in other portions of the t complex have been studied, but variation among t haplotypes has not been observed (Fox, SILVER and MARTIN 1984; J. H. NADEAU, M. ERHART, D. BURKART, S. J. PHILLIPS and M.-B. TCHETGEN, unpublished observations). Second alleles of some of these genes are indistinguishable from alleles found in inbred strains. The Pim-1 allele asso- ciated with t haplotypes such as tw7' are indistinguish- able from the Pim-lb allele in BALB/cJ mice (Figure 2), the Crya-1 allele associated with to and several other haplotypes are indistinguishable from the Crya- 1 a allele in certain inbred strains of mice and the Crya- 1 allele associated with the tw'" haplotype is indistin- guishable from the C r y - 1 allele in certain inbred strains (SKOW et al. 1987). Most significant perhaps is the identical deletion within the I-E gene in certain t haplotypes and inbred strains (DEMBIC, SINGER and KLEIN 1984; DEMBIC et al. 1985). If exchange is responsible for alleles associated with t haplotypes that are indistinguishable from wild-type alleles, then lack of association between variant alleles among t haplo- types suggests that multiple exchanges have occurred.

Although heterozygosity for inversions associated with t haplotypes strongly suppresses recombination (LYON and PHILLIPS 1959; HAMMERBERG and KLEIN 1975a; LYON et al. 1979; NADEAU, PHILLIPS and EGO- ROV 1985; HERRMANN et al. 1986; HERRMANN, BAR- LOW and LEHRACH 1987; SCHIMENTI et al. 1987), rare crossovers do occur and invariably produce partial t haplotypes that have lost some but not all of the properties of the t complex (LYON and PHILLIPS 1959; LYON and MEREDITH 1964a-c; LYON 1984, 1986). These rare crossovers occur either within the proxi- mal inversion or between the proximal and distal inversions (HERRMANN et al. 1986; HERRMANN, BAR- LOW and LEHRACH 1987; SCHIMENTI et al. 1987). Recombination involving the distal inversion has not been observed. Because double crossingover within inverted segments in inversion heterozygotes can pro- duce viable progeny, however, recombination could exchange alleles of Pim-1 and other genes between the t complex and its wild-type homologue. Because t haplotypes in wild mice are usually heterozygous with wild-type homologues rather than with other comple- menting t haplotypes (BENNETT 198 l), there are many opportunities for recombination, even though recom- bination is strongly suppressed. We therefore propose that recombination between the t complex and wild- type homologues is responsible for the Pim-1, C r y - 1 , and some of the H-2 variation among t haplotypes. DEMBIC et al. (1985) proposed a similar argument to account for identical deletions involving the I-E gene in t haplotypes and wild-type homologues.

If exchange is responsible for wild-derived alleles associated with t haplotypes, then we might expect to find reciprocal exchanges with t-derived variants among wild-type chromosomes. Such t-derived var- iants have yet to be described for wild or inbred mice. Several properties of the t complex readily account for the absence of these variants in wild-type chro- mosomes. Although variants introduced into t haplo- types are preferentially preserved because of recom- bination suppression and transmission ratio distortion, variants resulting from reciprocal exchange will ex- perience the same fate as mutations and most will be lost.

We thank ANTON BERNS for his gift of the Pim-1 probe Probe A, LOREN SKOW for his gift of B6.TC2 and B6.TC3 DNAs, JEFFREY

L. CROSBY and CAROLYN DUNBAR for expert technical assistance, and MARK ERHART, JAMES STONE and BENJAMIN TAYLOR for criti- cally reading a draft of this paper. This work was supported by American Cancer Society grant CD273, National Institutes of Health grant GM32461 and National Science Foundation grants PCM-8215004 and DCB-8610297.

L I T E R A T U R E C I T E D

Gene mapping within the T/ t complex of the mouse. 111. t-Lethal genes are arranged in three clusters on mouse Chromosome 17. Cell 39: 565-572.

ARTZT, K., 1984

540 J. H. Nadeau and S. J. Phillips

ARTZT, K., P. MCCORMICK and D. BENNETT, 1982 Gene mapping within the T/t complex of the mouse. I. t-Lethal genes are nonallelic. Cell 2 8 463-470.

ARTZT, K., H.-S. SHIN and D. BENNETT, 1982 Gene mapping within the T/t complex of the mouse. 11. Anomalous position of the H-2 complex in t haplotypes. Cell 2 8 471-476.

ARTZT, K., H.3. SHIN, D. BENNETT and A. DIMEC-TALENTO, 1985 Analysis of major histocompatibility complex haplo- types of t-chromosomes reveals that the majority of variation is generated by recombination. J. Exp. Med. 162: 93-104.

BENNETT, D., 1975 BENNETT, D., 198 1

The T-locus of the mouse. Cell 6: 441-454. Population genetics of T/t complex mutations.

pp. 615-632. In: Origins of Inbred Mice, Edited by H. C. MORSE 111, Academic Press, New York.

The information content of phase-known matings for ordering genetic loci. Genet. Epidemiol. 2: 349- 361.

CLAFLIN, J. L., B. A. TAYLOR, M. CHERRY and M. CUBBERLEY, 1978 Linkage in mice of genes controlling an immunoglob- ulin kappa-chain marker and the surface alloantigen Ly-3 on T lymphocytes. Immunogenetics 6 379-387.

CUYPERS, H. T. , G. SELTEN, W. QUINT, M. ZIJLSTRA, E. ROBANUS- MAANDAG, W. BOELENS, P. VAN WEZENBEEK, C. MELIEF and A. BERNS, 1984 Murine leukemia virus-induced T-cell lym- phomagenesis: integration of proviruses in a distinct chromo- somal region. Cell 37: 141-150.

DEMBIC, Z., P. A. SINGER and J. KLEIN, 1984 E': a history of a mutation. EMBO J. 3: 1647-1654.

DEMBIC, Z., M. AYANE, J. KLEIN, M. STEINMETZ, C. 0. BENOIST and D. J. MATHIS, 1985 Inbred and wild mice carry identical deletions in their E MHC genes. EMBO J. 4 127-131.

D'EUSTACHIO, P., B. FEIN, J. MICHAELSON and B. A. TAYLOR, 1984 The a-globin pseudogene on mouse chromosome 17 is closely linked to H-2. J, Exp. Med. 159 958-963.

DUNN, L. C., and E. CASPARI, 1945 A case of neighboring loci with similar effects. Genetics 3 0 543-568.

DUNN, L. C., and S. GLUECKSOHN-SCHOENHEIMER (WAELSCH), 1950 Repeated mutations in one area of mouse chromo- somes. Proc. Natl. Acad. Sci. USA 36: 233-237.

Application of the ovarian teratoma mapping method in the mouse. Genetics 103: 797- 812.

FEINBERG, A. P., and B. VOGELSTEIN, 1983 A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal. Biochem. 132: 6-13.

Evo- lution of mouse major histocompatibility genes borne by t chromosomes. Proc. Natl. Acad. Sci. USA 82: 2819-2823.

An alpha globin pseudogene is located within the mouse t complex. Immunogenetics 1 9 125-130.

Re- striction fragment length polymorphism of C4 genes in mice with t chromosomes. Immunogenetics 21: 247-256.

HALDANE, J. B. S., and C. H. WADDINGTON, 1931 Inbreeding and linkage. Genetics 16: 357-374.

HAMMERBERG, C., and J. KLEIN, 1975a Linkage relationships of markers on chromosome 17 of the house mouse. Genet. Res.

Linkage disequilibrium between H - 2 and t complexes in chromosome 17 of the mouse. Nature 258: 296-299.

HAMMERBERG, C., J. KLEIN, K. ARTZT and D. BENNETT, 1976 Histocompatibility-2 system in wild mice. 11. H - 2 hap- lotypes of t-bearing mice. Transplantation 21: 199-2 12.

HAUPTFELD, V., C. HAMMERBERG and J. KLEIN, 1976 His- tocompatibility-2 system in wild mice. 111. Mixed lymphocyte

BISHOP, D. T. , 1985

EPPIG, J. T., and E. M. EICHER, 1983

FIGUEROA, F., M. GOLUBIC, D. NIZETIC and J. KLEIN, 1985

Fox, H. S., L. M. SILVER and G. R. MARTIN, 1984

GOLUBIC, M., F. FIGUEROA, M. TOSI and J. KLEIN, 1984

26: 203-2 1 1 . HAMMERBERG, C., and J. KLEIN, 1975b

reaction and cell-mediated lymphotoxicity with t-bearing mice. Immunogenetics 3: 489-497.

A large inverted duplication allows homologous recombination be- tween chromosomes heterozygous for the proximal t complex inversion. Cell 48: 813-825.

HERRMANN, B., M. BUCAN, P. E. MAINS, A.-M. FRISCHAUF, L. M. SILVER and H. LEHRACH, 1986 Genetic analysis of the prox- imal portion of the mouse t complex: evidence for a second inversion within t haplotypes. Cell 44: 469-476.

HILKENS, J., H. T . CUYPERS, G. SELTEN, V. KROEZEN, J. HILGERS and A. BERNS, 1986 Genetic mapping of Pim-I putative on- cogene to mouse chromosome 17. Somatic Cell Mol. Genet. 12: 81-88.

KLEIN, J., 1980 Generation of diversity at MHC loci: implications for T cell receptor repertoires. pp. 239-253. In: Immunology 80. Progress in Immunology IV, Edited by M. FOUGEREAU and J. DAUSETT. Academic Press, New York.

KLEIN, J., and F. FIGUEROA, 1986 Evolution of the major histo- compatibility complex. CRC Crit. Rev. Immunol. 6 295-386.

KLEIN, J., F. FIGUEROA and C. S. DAVID, 1983 H-2 Haplotypes, genes and antigens: second listing. 11. The H-2 complex. Im- munogenetics 17: 553-596.

Murine t factors: an association between alleles at t and at H-2. J. Exp. Med. 144:

LYON, M. F., 1981 The t complex and the genetical control of development. Symp. 2001. Soc. Lond. 47: 455-477.

LYON, M. F., 1984 Transmission ratio distortion in mouse t- haplotypes is due to multiple distorter genes acting in a re- sponder locus. Cell 37: 621-628.

Male sterility of the mouse t-complex is due to homozygosity of the distorter genes. Cell 4 4 357-363.

Investigations of the nature of t-alleles in the mouse. I. Genetic analysis of a series of mutants derived from a lethal allele. Heredity 19: 301-3 12.

Investigations of the nature of t-alleles in the mouse. 11. Genetic analysis of an unusual mutant allele and its derivatives. Heredity 1 9 313- 325.

LYON, M. F., and R. MEREDITH, 1964c Investigations of the nature of t-alleles in the mouse. 111. Short tests of some further mutant alleles. Heredity 19: 327-330.

LYON, M. F., and R. J. S. PHILLIPS, 1959 Crossing-over in mice heterozygous for t-alleles. Heredity 13: 23-32.

LYON, M. F., E. P. EVANS, S. E. JARVIS and I. SAYERS, 1979 t- Haplotypes of the mouse may involve a change in intercalary DNA. Nature 279 38-42.

Personal com- munication. Mouse Newslett. 71: 48.

Absence of detectable gametic disequilib- rium between the t-complex and linked allozyme-encoding loci in house mice. Genet. Res. 42: 323-333.

NADEAU, J. H., 1986 A glyoxalase-1 variant associated with the t- complex in house mice. Genetics 113: 91-99.

NADEAU, J. H., S. J. PHILLIPS and 1. K. EGOROV, 1985 Re- combination between the t6 complex and linked loci in the house mouse. Genet. Res. 45: 251-264.

Personal communication. Mouse Newslett. 7 0 83.

Recombination between two mouse t haplotypes (f""tfand P'): mapping of the H-2 com- plex relative to the centromere and tufted (tf) locus. Immu- nogenetics 20: 277-285.

Genetic analysis of the T- H-2 region in non-t chromosomes. Immunogenetics 17: 445- 455.

ROHME, D., H. Fox, B. HERRMANN, A.-M. FRISCHAUF, J.-E. ED-

HERRMANN, B. G., D. P. BARLOW and H. LEHRACH, 1987

LEVINSON, J. R., and H. 0. MCDEVITT, 1976

834-839.

LYON, M. F., 1986

LYON, M. F., and R. MEREDITH, 1964a

LYON, M. F., and R. MEREDITH, 1964b

MANN, E., R. W. ELLIOT and C. HOHMAN, 1984

NADEAU, J. H., 1983

PHILLIPS, S. J., and J. H. NADEAU, 1984

PLA, M., and H. CONDAMINE, 1984

RINCHIK, E. M., and D. B. AMOS, 1983

Pim-1 Oncogene 54 1

STROM, P. MAINS, L. M. SILVER and H. LEHRACH, 1983 Molecular clones of the mouse t complex derived from microdissected metaphase chromosomes. Cell 3 6 783-788.

RUSSELL, J. H., C. B. DOBOS, R. J. GRAFF and B. A. TAYLOR, 198 1 Genetic control of cross-reactive cytotoxic T-lympho- cyte responses to a BALB/c tumor. Immunogenetics 14: 263- 272.

An unstable family of large DNA elements in the center of the mouse t complex. J. Mol. Biol. 194.583-594.

Proviral activa- tion of the putative oncogene Pim-l in MuLV induced T-cell lymphomas. EMBO J. 4: 1793-1798.

SELTEN, G., H. T. CUYPERS, W. BOELENS, E. ROBANUS-MAANDAC, J. VERBEEK, J. DOMAN, C. VAN BEVEREN and A. BERNS, 1986 The primary structure of the putative oncogene pim-1 shows extensive homology with protein kinases. Cell 4 6 603- 611.

Gene mapping within the T/t complex of the mouse. IV. The inverted MHC is intermingled with several t-lethal genes. Cell 3 9 573-578.

SHIN, H.-S., J. STAVNEZER, K. ARTZT and D. BENNETT, 1982 Genetic structure and origin of t haplotypes of mice analyzed with H-2 cDNA probes. Cell 2 9 969-976.

SILVER, L. M., 1982 Genomic analysis of the H-2 complex region

SCHIMENTI, J., L. VOLD, D. SOCOLOW and L. M. SILVER, 1987

SELTEN, G., H. T. CUYPERS and A. BERNS, 1985

SHIN, H.-S., D. BENNETT and K. ARTZT, 1984

associated with mouse t haplotypes. Cell 2 9 961-968. Mouse t haplotypes. Annu. Rev. Genet. 1 9

179-208. Recombination suppression

of mouse t haplotypes is due to chromatin mismatching. Nature

SILVER, L. M., M. WHITE and K. ARTZT, 1980 Evidence for unequal crossing over within the mouse T/t complex. Proc. Natl. Acad. Sci. USA 77: 6077-6080.

SILVER, L. M., J. UMAN, J. DANSKA and J. I. GARRELS, 1983 A diversified set of testicular cell proteins specified by genes within the mouse 6 complex. Cell 35: 35-45.

SKOW, L. C., and M. E. DONNER, 1985 The locus encoding A- crystallin is closely linked to H-2K on mouse chromosome 17. Genetics 110 723-732.

SKOW, L. C., J. H. NADEAU, J. C. AHN, H A . SHIN, K. ARTZT and D. BENNETT, 1987 Polymorphism and linkage of the A-crys- tallin gene in t-haplotypes of the mouse. Genetics 116 107- 111.

Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol.

Recombinant inbred strains: use in gene mapping. pp. 423-438. In: Origins of Inbred Mice, Edited by H. C. MORSE. Academic Press, New York.

Communicating editor: R. E. GANSCHOW

SILVER, L. M., 1985

SILVER, L. M., and K. ARTZT, 1981

290 68-70.

SOUTHERN, E. M., 1975

9 8 503-517. TAYLOR, B. A., 1981