Inheritance of a germline mutation in one of the DNA mismatch repair genes predisposes human individuals to hereditary nonpolyposis colorectal cancer, characterized by development of tumors predominantly in the colon, endometrium, and gastrointestinal tract. Mice heterozygous for a mismatch repair–null mutation generally do not have an increased risk of neoplasia. However, mice constitutively lacking mismatch repair are prone to tumor development from an early age, particularly thymic lymphomas. Mismatch repair–deficient mice crossed to Apc+/− mice develop an increased spontaneous intestinal tumor incidence, demonstrating that the tumor spectrum can be genetically influenced. Here, we bred Msh2- and Msh6-deficient mice to athymic nude mice, hypothesizing that a broader tumor spectrum may be observed if mice are able to survive longer without succumbing to thymic lymphomas. However, Msh2−/−;Foxn1nu/nu and Msh6−/−;Foxn1nu/nu mice developed primarily early-onset lymphoblastic lymphomas. Using B-cell–specific markers, we found these tumors to be predominately B-cell in origin. The development of hematologic malignancy in the mouse, even in the absence of a thymus, parallels the development of B- and T-cell lymphoma and leukemia in the few rare mismatch repair–null human patients that have been identified. The persistent development of hematologic malignancy both in the mouse and in human patients deficient in mismatch repair leads us to implicate mismatch repair as an important repair mechanism in normal B- and T-cell development. Thus, mismatch repair–deficient mice may prove to be a good model to study human hematologic malignancy.

In mammalian cells, DNA mismatch repair is mediated by the proteins Msh2, Msh3, Msh6, Mlh1, Pms1, Pms2, and Mlh3. The initial recognition of mismatches in DNA is carried out by functional heterodimers of Msh2 bound with either Msh6 or Msh3, termed MutSα and MutSβ, respectively (reviewed in ref. 1). Following the binding of either MutSα or MutSβ to a base mismatch in the DNA, recruitment of additional proteins facilitates completion of the repair process. Although MutSα has a preference for single base mismatches and MutSβ preferentially binds to small loops, each heterodimer has some functional overlap (1). Despite this overlap in function, in the absence of either Msh2 or Msh6, repair is severely compromised resulting in accumulation of excessive mutations (2). Germline mutations in one or more of the human mismatch repair genes (i.e., hMSH2 or hMLH1) have been associated with the human disease hereditary nonpolyposis colorectal cancer (HNPCC). These individuals have a greatly increased lifetime risk of developing cancer of the colon, endometrium, and gastrointestinal tract. Greater than 90% of human HNPCC tumors show microsatellite instability, indicative of a loss of mismatch repair (3–6). This instability is associated with subsequent mutation of critical downstream genes, resulting in deregulated cell proliferation and tissue-specific tumorigenesis (reviewed in ref. 7).

Although human HNPCC patients primarily develop cancers of the gastrointestinal tract, lymphomas and leukemias have been observed in certain kindreds (8, 9). Analogous to the mouse models homozygous for a defect in mismatch repair, several human patients with germline mutations in both copies of any one of the mismatch repair genes, MLH1, MSH2, or PMS2, have presented with early-onset childhood T- or B-cell malignancies (10–16).

In contrast to humans, mouse models genetically engineered to be haploinsufficient for any one of the mismatch repair genes generally do not show an increased risk of cancer whereas mice that are homozygous null for Msh2, Msh6, Mlh1, or Pms2 are prone to tumorigenesis (17–22). Mice deficient in Msh2 or Msh6 most commonly develop early-onset thymic lymphomas although other tumors such as ovarian and small intestinal tumors occur with a lower frequency (17, 19, 20).

Immunohistochemistry with T- and B-cell markers, as well as histology, showed that murine mismatch repair–deficient thymic lymphomas are very homogeneous (23). These lymphomas are predominantly of T-cell origin, characterized by a “starry sky” appearance, enlarged nuclei, reduced cytoplasm, and numerous mitotic figures (17). Hematopoietic development appears normal in the Msh2-null mice (17). Msh2-deficient thymic lymphomas are thought to represent a single histopathologic entity and the tumor homogeneity suggests specific recurring genetic events may underlie the lymphocyte transformation and expression of a malignant phenotype.

In Msh2-deficient mice the onset of thymic lymphomas begins at 2 months of age and in Msh6-deficient mice tumor onset begins at 9 months; 50% survival times are 5 and 9 months, respectively (17, 19). The spontaneous tumor spectrum of Msh2, Msh3, Msh6, and Mlh1 null mice has been altered by breeding these mice to Apc+/−/Min mice resulting in increased intestinal tumor incidence, intestinal tumor multiplicity per animal, and reduced survival (24–26). This led us to believe that the spontaneous tumor spectrum of Msh2- and Msh6-deficient mice may be shifted by breeding them to athymic nude mice. We hypothesized that Msh2- and Msh6-deficient mice succumb to thymic lymphomas before the development of other cancers such as colonic tumors. We tested our hypothesis by breeding Msh2- and Msh6-deficient mice to athymic nude mice to determine if these mice would develop a tumor spectrum similar to the human HNPCC tumor spectrum and provide us with a novel mouse model.

The nude phenotype is characterized by a loss-of-function mutation in the Foxn1 gene (forkhead box N1), a winged helix/forkhead transcription factor, which results in abnormal morphogenesis of the epidermis, hair follicles, and thymus (27). Without Foxn1 expression, there is a basic defect in development of the embryonic ectoderm resulting in the absence of a thymus and these mice cannot properly attract hematopoeitic precursor cells essential for T-cell development. Consequently, they have severely reduced numbers of immature and mature T-cells and are therefore immunocompromised (28). Nude mice do have a minimal amount of extra-thymic T-cell maturation occurring in such organs as the spleen (29) but have a greatly reduced population of circulating mature T-cells.

Here we show that Msh2−/−;Foxn1nu/nu and Msh6−/−;Foxn1nu/nu mice develop lymphoblastic lymphomas, predominantly of B-cell origin. Furthermore, the life span of Msh6-deficient mice is reduced on a nude background, with 50% survival occurring at 9 months in Msh6−/−;Foxn1nu/nu mice compared with 11 months in Msh6−/− mice. Our findings support a critical role for the mismatch repair proteins in normal T- and B-cell development, and show that constitutive absence of mismatch repair contributes to development of both T- and B-cell malignancy.

Genotyping. Msh2 mice (17) and Msh6 mice (19) were bred to nude (Foxn1nu/nu) mice (Charles River Laboratories) to generate Msh2−/−;Foxn1nu/nu, Msh6−/−;Foxn1nu/nu mice and Msh2+/+;Foxn1nu/nu, Msh6+/+;Foxn1nu/nu mice. Msh6 genotyping was carried out using earclip DNA as previously described (19). Msh2 genotyping was carried out using a PCR assay with earclip DNA and the following three primers: U771 forward (5′-GCTCACTTAGACGCCATTGT-3′) and L926 reverse (5′-AAAGTGCACGTCATTTGGA-3′) amplifying the wild-type allele and U771 forward and neo reverse (5′-TGG AAG GAT TGG AGC TAC GG-3′) amplifying the targeted allele. The nude phenotype is the result of a single base pair deletion mutation in the Foxn1 gene that when present in the homozygous state leads to the nude phenotype (27). Genotyping of the Foxn1 gene was done using earclip DNA as described (30). Briefly, modified PCR primers were used to introduce an artificial restriction site in the Foxn1 gene PCR product based on the presence or absence of the nude mutation. Subsequently, PCR-RFLP analysis was carried out to distinguish between heterozygous (Foxn1+/nu) and wild-type (Foxn1+/+) mice.

Examination of mice. The mouse experimental protocol was approved by the Health Science Animal Policy and Welfare Committee of the University of Alberta, and all animals cared for according to the guidelines of the Canadian Council on Animal Care. Mice were monitored daily for signs of poor health/grooming or tumorigenesis. Mice were sacrificed by isoflurane overdose on signs of lethargy, weight loss, or physical demonstration of a tumor. Murine tissues were fixed in 10% buffered formalin and embedded in paraffin blocks. Tumor and control brain tissue were frozen at −80°C for DNA analysis. Histologic analysis was done on 5 μm sections stained with H&E using standard methods (P.N.).

Generation of Kaplan-Meier survival curves and their statistical analysis.Msh2−/−;Foxn1nu/nu, Msh6−/−;Foxn1nu/nu and Msh2+/+;Foxn1nu/nu, Msh6+/+;Foxn1nu/nu mice were sacrificed and necropsies were done to determine the cause of lethargy and/or to excise tumor samples. Time of death and gross pathology were recorded for each mouse. Using the R statistical analysis software package version 1.5.1. and the “survival library” function, we generated Kaplan-Meier survival curves and compared time of death due to tumorigenic causes in wild-type mice versus mismatch repair–deficient mice (http://www.R-project.org; ref. 31).

Immunohistochemistry. Tumor sections (5-7 μm) were dewaxed in xylene, incubated in block solution, and stained with the following antibodies: CD3 (DakoCytomation, Glostrup, Denmark), B220 (BD·PharMingen, San Diego, CA), Pax5 (BD·PharMingen), Ki67 (BD·PharMingen), and CD45 (Research Diagnostics, Inc., Flanders, NJ).

PCR assay for DHJH rearrangement status. DHJH rearrangement analysis was done as previously described (32). Briefly, two forward primers were used that are immediately 5′ of the DH elements (DSF: 5′-AGGGATCCTTGTGAGGGATCTACTACTGTG-3′ and Dq52: 5′-GCGGAGCACCACAGTGCAACTGGGAC-3′). These primers are based on a consensus sequence and together recognize all of the DH elements. The reverse primer (JH4: 5′-AAAGACCTGCAGAGGCCATTCTTACC-3′) is immediately 3′ of the JH4 element. The DSF/JH4 primer pair amplifies recombination products between DSP and DFL DH gene elements and all of the JH elements. The Dq52 and JH4 primer pair amplifies recombination products between Dq52 and all JH4 elements. If rearrangement of the DHJH locus has occurred, one of four PCR product sizes (0.12 to 1.3 kb) will be observed depending on which JH element was used. DNA from NIH-3T3 cells was used as a germline control and mouse spleen DNA was used as a positive control.

Microsatellite instability analysis. Matched normal brain samples and tumor biopsies were harvested from five Msh2−/−;Foxn1Foxn1nu/nu mice and were analyzed for microsatellite instability. DNA was extracted using Qiaquick DNA extraction columns (Qiagen, Inc., Valencia, CA). Microsatellite loci were amplified using fluorescently labeled Licor PCR primers and products were analyzed on a Licor electrophoresis gel system (Licor-Biosciences, Lincoln, NE). Ten loci were investigated for microsatellite status using primers as follows: (a) five primer pairs used amplified mononucleotide repeats: JH101, JH102, JH103, JH104 (19), and U12235 (5′ GCTCATCTTCGTTCCCTGTC-3′ and 5′-CATTCGGTGGAAAGCTCTGA-3′; ref. 33); (b) three primer pairs amplified dinucleotide repeats: D1mit83, D7mit17, and D7mit91 (Whitehead Institute: http://www-genome.wi.mit.edu/cgi-bin/mouse/sts_info); (c) one primer pair, Tcrb, amplified a trinucleotide repeat (GCT)12 Tcrb forward 5′-AGTTTTAGGCTATAGGTT-3′ and (Tcrb reverse 5′-TGATCTAGAGAAAGGGTAGGTCTA-3′; ref. 34); and (d) one primer pair, Cyp1a2, amplified a tetranucleotide repeat (CAAG)10 Cyp1a2 forward 5′-TGGCAGGACTGCACCTAAGCT-3′ and (Cyp1a2 reverse 5′-ACTGGAACCTTAGAGCATGAG-3′; ref. 34).

Msh2−/−;Foxn1nu/nu and Msh6−/−;Foxn1nu/nu mice have a significantly reduced life span. Msh2 and Msh6 mice were bred to athymic nude mice to generate Msh2−/−;Foxn1nu/nu and Msh6−/−;Foxn1nu/nu mice. Msh2+/+;Foxn1nu/nu and Msh6+/+;Foxn1nu/nu mice were bred as control mice.

First, a survival curve was generated to investigate differences in rates of tumorigenesis between Msh6−/−;Foxn1nu/nu mice and Msh6−/− mice (Fig. 1A; Msh6−/− survival data from ref. 19). Msh6−/− mice live significantly longer than Msh6−/−;Foxn1nu/nu with a median survival time to tumorigenesis of 11 months compared with 9 months in the Msh6−/−;Foxn1nu/nu; P = 0.007. At 9 months of age only 46% of the Msh6−/−;Foxn1nu/nu mice were still alive whereas 60% of the Msh6−/− mice were still alive. By 12 months of age only 12.5% of the Msh6−/−;Foxn1nu/nu mice were alive, however, 35% of the Msh6−/− mice were still alive. Although the genetic background of the two mouse models was not identical (the Msh6−/− mice were on a C57BL/6 background and the Msh6−/−;Foxn1nu/nu mice were on a C57BL/6/Balb/C background), different genetic backgrounds of Msh6−/− mice had previously been determined to have no effect on spontaneous survival of the mice.3

3

Edelmann, personal communication.

Figure 1.

Survival of Msh2−/−;Foxn1nu/nu and Msh6−/−;Foxn1nu/nu mice. Life span of the mice and cause of death were recorded and Kaplan-Meier survival curves were generated using the R statistical analysis software package version 1.5.1. For all curves, death due to tumorigenic causes was recorded and used in calculating the survival curves. A, survival of Msh6−/− mice compared with Msh6−/−;Foxn1nu/nu mice; P = 0.007 (Msh6−/− survival data obtained from ref. 19). B, survival of Msh6−/−;Foxn1nu/nu mice compared with Msh6+/+;Foxn1nu/nu mice; P = 0.005. C, survival of Msh2−/−;Foxn1nu/nu mice compared with Msh2+/+;Foxn1nu/nu mice; P = 0.0005. D, survival of Msh2−/−;Foxn1nu/nu mice compared with Msh2−/− mice; P = 0.36.

Figure 1.

Survival of Msh2−/−;Foxn1nu/nu and Msh6−/−;Foxn1nu/nu mice. Life span of the mice and cause of death were recorded and Kaplan-Meier survival curves were generated using the R statistical analysis software package version 1.5.1. For all curves, death due to tumorigenic causes was recorded and used in calculating the survival curves. A, survival of Msh6−/− mice compared with Msh6−/−;Foxn1nu/nu mice; P = 0.007 (Msh6−/− survival data obtained from ref. 19). B, survival of Msh6−/−;Foxn1nu/nu mice compared with Msh6+/+;Foxn1nu/nu mice; P = 0.005. C, survival of Msh2−/−;Foxn1nu/nu mice compared with Msh2+/+;Foxn1nu/nu mice; P = 0.0005. D, survival of Msh2−/−;Foxn1nu/nu mice compared with Msh2−/− mice; P = 0.36.

Close modal

Second, a Kaplan-Meier survival curve was generated comparing death due to tumorigenic causes in Msh6−/−;Foxn1nu/nu mice with Msh6+/+;Foxn1nu/nu mice (Fig. 1B). Of the 64 Msh6−/−;Foxn1nu/nu mice studied, 44 (68.8%) developed tumors compared with 4 of 28 (14.3%) Msh6+/+;Foxn1nu/nu control mice (Fig. 1B). Msh6−/−;Foxn1nu/nu mice developed tumors at a significantly higher frequency than did the Msh6+/+;Foxn1nu/nu mice; P = 0.005. Of the 44 Msh6−/−;Foxn1nu/nu mice that developed tumors, 39 (88.6%) mice developed lymphoblastic lymphomas. Of the four Msh6+/+;Foxn1nu/nu mice that developed tumors, all developed lymphoblastic lymphomas, with three of the four mice concurrently presenting with Pneumocystis carinii pneumonia. During the course of this study, a P. carinii infection within the colony caused the development of pneumonia and death in several Msh6+/+;Foxn1nu/nu mice, likely associated with the immunocompromised state of the nude background.

Third, a Kaplan-Meier survival curve was generated comparing death due to tumorigenic causes in Msh2−/−;Foxn1nu/nu mice with Msh2+/+;Foxn1nu/nu mice (Fig. 1C). The Msh2−/−;Foxn1nu/nu mice developed tumors at a significantly higher frequency than did the Msh2+/+;Foxn1nu/nu control mice; P = 0.00005. Of the 14 Msh2−/−;Foxn1nu/nu mice studied, eight (57%) developed lymphoblastic lymphomas, one developed a basal cell skin tumor, four died unexpectedly and were not available for postmortem analysis, and one mouse died of pneumonia. Median survival time of Msh2−/−;Foxn1nu/nu mice was 5.54 months. By 7 months only 31% of the mice were alive and by 9 months only 15.5% of mice were alive. To investigate the time to tumorigenesis in Msh2−/−;Foxn1nu/nu mice versus Msh2−/− mice, we generated a Kaplan-Meier survival curve using survival data from our Msh2−/− colony housed at the University of Alberta Fig. 1D. Time to tumorigenesis in Msh2−/−;Foxn1nu/nu compared with Msh2−/− mice was not significantly different; P = 0.36.

Msh2−/−;Foxn1nu/nu and Msh6−/−;Foxn1nu/nu mice develop lymphoblastic lymphomas. H&E staining of sections from tumors arising in Msh2−/−;Foxn1nu/nu and Msh6−/−;Foxn1nu/nu mice was done and tumors were examined histologically. Of the tumors arising in the Msh6−/−;Foxn1nu/nu mice, the majority were lymphoblastic lymphomas (88.6% of tumors; Table 1). Of the nine mice that developed tumors in the Msh2−/−;Foxn1nu/nu colony, eight mice developed lymphoblastic lymphomas (88.9% of tumors; Table 1). Both Msh6−/−;Foxn1nu/nu and Msh2−/−;Foxn1nu/nu lymphoblastic lymphomas appear histologically as monotonous sheets of poorly differentiated lymphoid cells with macrophages scattered throughout giving a “starry sky” appearance. This parallels what is seen in Msh2−/− thymic lymphomas. In addition to the “starry sky” appearance, both mismatch repair–deficient thymic lymphomas and mismatch repair–deficient lymphoblastic lymphomas have cells with enlarged nuclei, reduced cytoplasm, and numerous mitotic figures.

Table 1.

Histopathologic findings in Msh6−/−;Foxn1nu/nu mice

Mouse IDAge (mo)HistologyLymph nodeColonSpleenLiverOther tissues
6N565 2.0 (a) Lymphoblastic, (b) salivary gland tumors − −  
6N658* 3.3 Lymphoblastic − Skin: normal 
6N619 3.5 (a) Basal cell tumor, (b) papilloma NI − − −  
6N347 5.1 Lymphoblastic NI  
6N643* 5.3 Lymphoblastic  
6N593* 5.5 Lymphoblastic NI  
6N641*, 5.6 Lymphoblastic −  
6N646*, 5.6 Lymphoblastic −  
6N627* 5.6 Lymphoblastic NI − −  
6N877 5.7 Lymphoblastic NI − − Lung: P. carinii infection 
6N621* 5.7 Lymphoblastic − Skin: lymphoma infiltration 
6N877 5.7 Lymphoblastic − − NI Bone marrow: normal 
6N870 7.0 Lymphoblastic NI  
6N626 7.4 Lymphoblastic − −  
6N605* 7.7 Lymphoblastic NI −  
6N608 7.7 Lymphoblastic − − −  
6N410 8.1 Lymphoblastic NI  
6N348 8.3 (a) Basal cell tumor, (b) sebaceous adenoma − −  
6N649 8.6 Lymphoblastic NI NI  
6N497 8.8 Lymphoblastic NI NI NI  
6N665 8.9 Lymphoblastic − NI Lung: P. carinii infection 
6N666 8.9 Lymphoblastic NI − Lung: normal 
6N718 9.9 Lymphoblastic NI NI NI NI  
6N629 10.2 Lymphoblastic − − −  
6N597 10.8 Lymphoblastic −   
6N722 11.1 Lymphoblastic −  
6N555 11.2 Bronchoalveolar tumor NI − − − Lung: bronchoalveolar tumor; lung: P. carinii infection 
6N559 11.2 Lymphoblastic − Lung: P. carinii infection 
6N645 11.3 Lymphoblastic NI  
6N653 11.8 Lymphoblastic − Lung: lymphoma infiltrate; bone marrow: normal 
6N501 12.7 Basal cell tumor NI − − − Skin: basal cell tumor; lung: P. carinii infection 
Mouse IDAge (mo)HistologyLymph nodeColonSpleenLiverOther tissues
6N565 2.0 (a) Lymphoblastic, (b) salivary gland tumors − −  
6N658* 3.3 Lymphoblastic − Skin: normal 
6N619 3.5 (a) Basal cell tumor, (b) papilloma NI − − −  
6N347 5.1 Lymphoblastic NI  
6N643* 5.3 Lymphoblastic  
6N593* 5.5 Lymphoblastic NI  
6N641*, 5.6 Lymphoblastic −  
6N646*, 5.6 Lymphoblastic −  
6N627* 5.6 Lymphoblastic NI − −  
6N877 5.7 Lymphoblastic NI − − Lung: P. carinii infection 
6N621* 5.7 Lymphoblastic − Skin: lymphoma infiltration 
6N877 5.7 Lymphoblastic − − NI Bone marrow: normal 
6N870 7.0 Lymphoblastic NI  
6N626 7.4 Lymphoblastic − −  
6N605* 7.7 Lymphoblastic NI −  
6N608 7.7 Lymphoblastic − − −  
6N410 8.1 Lymphoblastic NI  
6N348 8.3 (a) Basal cell tumor, (b) sebaceous adenoma − −  
6N649 8.6 Lymphoblastic NI NI  
6N497 8.8 Lymphoblastic NI NI NI  
6N665 8.9 Lymphoblastic − NI Lung: P. carinii infection 
6N666 8.9 Lymphoblastic NI − Lung: normal 
6N718 9.9 Lymphoblastic NI NI NI NI  
6N629 10.2 Lymphoblastic − − −  
6N597 10.8 Lymphoblastic −   
6N722 11.1 Lymphoblastic −  
6N555 11.2 Bronchoalveolar tumor NI − − − Lung: bronchoalveolar tumor; lung: P. carinii infection 
6N559 11.2 Lymphoblastic − Lung: P. carinii infection 
6N645 11.3 Lymphoblastic NI  
6N653 11.8 Lymphoblastic − Lung: lymphoma infiltrate; bone marrow: normal 
6N501 12.7 Basal cell tumor NI − − − Skin: basal cell tumor; lung: P. carinii infection 

NOTE: NI, not investigated; +, positive for lymphoma infiltration; −, negative for lymphoma infiltration.

*

Immunophenotyped with CD3 and PAX5 (Table 2).

Investigated for the presence of DHJH gene rearrangements (Table 3).

Lymphoblastic lymphoma characterization. Several of the Msh2−/−;Foxn1nu/nu and Msh6−/−;Foxn1nu/nu mice presented with clearly visible s.c. tumors commonly appearing on the neck and chest region. H&E staining showed many of these tumors to be lymph nodes that had undergone general infiltration and replacement by malignant lymphocytes. However, in several cases the precise tissue of origin could not be identified due to the destruction of tissue structure by malignant cells (Fig. 2). Widespread tumor infiltration into major organs was found to have occurred in the majority of mice. Organs commonly infiltrated by tumor cells include the liver, spleen, kidney, heart, and colon. In the liver and spleen there was sinusoidal tumor infiltrate as well as perivascular infiltrate (Fig. 2C). Dissemination to the heart occurred in several mice as well, and although the myocardium itself was normal, there was an infiltrate of tumor cells in the adventitia of the aorta and associated nerves. Tumor masses were composed of a uniform population of quite immature lymphocytes and presented with the signature “starry sky” appearance due to the scattered tingible body macrophages (Fig. 2B). As well, mitoses and aberrant mitotic forms were present at a high rate. Three Msh6−/−;Foxn1nu/nu mice were investigated for the presence of malignancy in the bone marrow. In all cases scattered malignant lymphocytes were present but it was not clear whether there was indeed bone marrow origin of tumor cells or whether this was simply an infiltration of metastasis.

Figure 2.

Histology of Msh6−/−;Foxn1nu/nu murine lymphoblastic lymphomas and tissue infiltration. A, Msh6−/−;Foxn1nu/nu murine lymphoblastic lymphoma stained with H&E (200×). Note the destruction of normal lymph node structure by the infiltrating malignant cells. B, Msh6−/−;Foxn1nu/nu murine lymphoblastic lymphoma stained with H&E (500×). Note the monotonous sheet of neoplastic cells interlaced with macrophage cells giving rise to the “starry sky” appearance. Arrows, mitotic figures. C, Msh6−/−;Foxn1nu/nu mouse demonstrating infiltration of malignant lymphocytes around the portal veins of the liver (50×). D, enlargement of the area of malignant infiltration around the portal veins in the liver (400×). Note the very similar appearance of the malignant infiltration in the liver to that of the malignant lymph node in B.

Figure 2.

Histology of Msh6−/−;Foxn1nu/nu murine lymphoblastic lymphomas and tissue infiltration. A, Msh6−/−;Foxn1nu/nu murine lymphoblastic lymphoma stained with H&E (200×). Note the destruction of normal lymph node structure by the infiltrating malignant cells. B, Msh6−/−;Foxn1nu/nu murine lymphoblastic lymphoma stained with H&E (500×). Note the monotonous sheet of neoplastic cells interlaced with macrophage cells giving rise to the “starry sky” appearance. Arrows, mitotic figures. C, Msh6−/−;Foxn1nu/nu mouse demonstrating infiltration of malignant lymphocytes around the portal veins of the liver (50×). D, enlargement of the area of malignant infiltration around the portal veins in the liver (400×). Note the very similar appearance of the malignant infiltration in the liver to that of the malignant lymph node in B.

Close modal

Cellular changes in colonic tissues. We hypothesized that in the absence of a thymus Msh6−/−;Foxn1nu/nu and Msh2−/−;Foxn1nu/nu mice may develop primary tumors of the gastrointestinal epithelium similar to human HNPCC patients. However, mice began to develop lymphoid tumors with no indication of primary intestinal tumor formation. Of note, colons from several mice did show lymphoid infiltration of the lamina propria of intestinal villi and presence of lymphoid follicles in the mucosa with increased prominence of submucosal lymphoid follicles. To further investigate the possibility of early stages of tumor formation in the colon, we undertook a hyperproliferation study of the colon hypothesizing that early cellular changes may be occurring that are not seen macroscopically. Msh6−/−;Foxn1nu/nu mice were sacrificed at 24 days, 8 weeks, and 3 months of age and each colon was divided into three sections. Three mice for each age group were used to account for individual variation between mice. In all colon sections examined there were no unusual cellular changes or thickening of the epithelial cell wall. We found no evidence of cellular hyperproliferation or precancerous lesions in the colon and conclude that the colon is not susceptible to tumor formation in these mice.

Lymphoblastic lymphoma immunohistochemistry characterization. To further characterize the tumors arising in the Msh6−/−;Foxn1nu/nu mice, we did immunohistochemistry using several antibodies on eight lymphoblastic lymphomas and one squamous cell tumor (Tables 1 and 2). The Msh6−/−;Foxn1nu/nu tumors were stained with CD45 (leukocyte common antigen), a marker found on hematopoietic cells with higher expression in lymphocytes then leukocytes (35). One of the eight Msh6−/−;Foxn1nu/nu lymphoblastic lymphomas tested showed weak positive staining (<10% of cells staining positive). The remaining seven tumors tested were negative (Fig. 3E). Negative staining for CD45 leads us to suggest that the lymphoblastic lymphomas arising in these mice are arrested at a very early stage in differentiation, having arisen before the expression of CD45. Tumors were stained with Ki67, a nuclear protein expressed in proliferating cells but absent in resting cells. Positive Ki67 staining in all nine tumors tested indicates a high proliferation index, leading us to conclude that these tumors were rapidly growing (Fig. 3F). To determine if the lymphoblastic lymphomas were T- or B-cell in origin, we stained the tumors with CD3 (T-cell–specific marker) and B220 (B-cell–specific marker, also known as CD45R). None of the tumors tested stained positive for CD3 (Table 2). Only two were positive for B220 (Table 2). Tumors were then stained with Pax5, a very early marker specific for the B-cell lineage (36). The Pax5 protein (also known as B-cell–specific activator protein) is a transcription factor expressed in the earliest B-cell precursors and, once activated, causes a cascade of other genes to be turned on and the cell to develop along the B-cell lineage (36). All of the Msh6−/−;Foxn1nu/nu lymphoblastic lymphomas tested were Pax5 positive, indicating these tumors were early B-cell in origin (Fig. 3F and G). The squamous cell skin tumor did not stain positive for either of the T- or B-cell markers.

Table 2.

Immunohistochemistry of Msh6−/−;nu/nu lymphoblastic lymphomas

Mouse IDTumor typeMarker
Tumor origin (T or B cell)
CD3B220PAX5Ki67CD45
6N586 Lymphoma − − − B cell 
6N593 Lymphoma − − − B cell 
6N621 Lymphoma − − − B cell 
6N627 Lymphoma − + (low) − B cell 
6N641 Lymphoma − − B cell 
6N643 Lymphoma − − − B cell 
6N646 Lymphoma − − − B cell 
6N658 Lymphoma − − + (low) B cell 
6N605 Squamous skin tumor − − − − − Unknown 
Mouse IDTumor typeMarker
Tumor origin (T or B cell)
CD3B220PAX5Ki67CD45
6N586 Lymphoma − − − B cell 
6N593 Lymphoma − − − B cell 
6N621 Lymphoma − − − B cell 
6N627 Lymphoma − + (low) − B cell 
6N641 Lymphoma − − B cell 
6N643 Lymphoma − − − B cell 
6N646 Lymphoma − − − B cell 
6N658 Lymphoma − − + (low) B cell 
6N605 Squamous skin tumor − − − − − Unknown 

NOTE: −, negative staining in the tumor; +, positive staining in the tumor; + (low), <10% of cell staining positive staining in the tumor.

Figure 3.

Immunohistochemical staining of Msh6−/−;Foxn1nu/nu murine lymphoblastic lymphomas. A, lymphoblastic lymphoma staining negative for B220 (400×). Note the positively staining normal cells but a lack of staining in the malignant cells. B, lymphoma-infiltrated liver tissue with the malignant infiltrate staining negative for B220 (arrow; 400×). C, lymphoblastic lymphoma staining negative for CD3. Note the occasional normal T cell staining positive for CD3 (400×). D, lymphoma-infiltrated liver tissue with malignant infiltrate staining negative for CD3 (arrow; 400×). E, lymphoma-infiltrated liver tissue with malignant infiltrate staining negative for CD45 (400×). F, lymphoma-infiltrated liver tissue with malignant infiltrate staining positive for Ki67 (400×). G, lymphoblastic lymphoma staining positive for Pax5 (400×). H, lymphoma-infiltrated liver tissue with malignant infiltrate staining positive for Pax5 (400×).

Figure 3.

Immunohistochemical staining of Msh6−/−;Foxn1nu/nu murine lymphoblastic lymphomas. A, lymphoblastic lymphoma staining negative for B220 (400×). Note the positively staining normal cells but a lack of staining in the malignant cells. B, lymphoma-infiltrated liver tissue with the malignant infiltrate staining negative for B220 (arrow; 400×). C, lymphoblastic lymphoma staining negative for CD3. Note the occasional normal T cell staining positive for CD3 (400×). D, lymphoma-infiltrated liver tissue with malignant infiltrate staining negative for CD3 (arrow; 400×). E, lymphoma-infiltrated liver tissue with malignant infiltrate staining negative for CD45 (400×). F, lymphoma-infiltrated liver tissue with malignant infiltrate staining positive for Ki67 (400×). G, lymphoblastic lymphoma staining positive for Pax5 (400×). H, lymphoma-infiltrated liver tissue with malignant infiltrate staining positive for Pax5 (400×).

Close modal

DHJH gene rearrangement analysis of lymphoblastic lymphomas. To further elucidate the stage of B-cell development of the lymphoblastic lymphomas, we assayed for the presence of DH to JH gene rearrangements. Six of the eight Msh6−/−;Foxn1nu/nu lymphoblastic lymphoma tumors tested were B220 negative but seven of the eight tumors tested were Pax5 positive, leading us to hypothesize that these tumors were arrested early in B-cell development and would likely not have undergone DH to JH gene rearrangements. We looked for the presence of DH to JH gene rearrangements in seven Msh6−/−;Foxn1nu/nu lymphoblastic lymphomas. Of the seven tumors analyzed, six showed DH to JH gene rearrangements in at least one locus (Table 3). On average, tumors displayed two (of a possible four) DH to JH gene rearrangement products. Four of the seven tumors investigated had matched normal brain control DNA. Interestingly, three of the four matched normal brain samples showed the same DH to JH gene rearrangement products found in the corresponding tumor tissue, leading us to suggest that tumor metastases to the brain have occurred as the brain is not expected to have undergone DH to JH gene rearrangement.

Table 3.

DHJH gene rearrangements in lymphoblastic lymphomas from Msh2−/−;Foxn1nu/nu and Msh6−/−;Foxn1nu/nu mice

DHJH rearrangements per tumor
DSF/JH4 primer pair
Dq52/JH4 primer pair
DJH1DJH2DJH3DJH4GermlineDJH1DJH2DJH3DJH4Germline
Msh2−/−;Foxn1nu/nu            
Tumor 40 ▪          
Tumor 2N109     ▪    ▪  
Tumor 2N134 ▪ ▪ ▪ ▪  ▪ ▪ ▪ ▪ ▪ 
Tumor 2N283   ▪ ▪      ▪ 
Tumor 2N319  ▪ ▪   ▪    ▪ 
Tumor 2N346  ▪        ▪ 
Msh6−/−;Foxn1nu/nu            
Tumor 6N501     ▪     ▪ 
Tumor 6N641   ▪ ▪      ▪ 
Tumor 6N646   ▪        
Tumor 6N649    ▪   ▪ ▪  ▪ 
Tumor 6N722  ▪ ▪ ▪      ▪ 
Tumor 6N870   ▪       ▪ 
Tumor 6N877 ▪ ▪ ▪ ▪   ▪ ▪   
DHJH rearrangements per tumor
DSF/JH4 primer pair
Dq52/JH4 primer pair
DJH1DJH2DJH3DJH4GermlineDJH1DJH2DJH3DJH4Germline
Msh2−/−;Foxn1nu/nu            
Tumor 40 ▪          
Tumor 2N109     ▪    ▪  
Tumor 2N134 ▪ ▪ ▪ ▪  ▪ ▪ ▪ ▪ ▪ 
Tumor 2N283   ▪ ▪      ▪ 
Tumor 2N319  ▪ ▪   ▪    ▪ 
Tumor 2N346  ▪        ▪ 
Msh6−/−;Foxn1nu/nu            
Tumor 6N501     ▪     ▪ 
Tumor 6N641   ▪ ▪      ▪ 
Tumor 6N646   ▪        
Tumor 6N649    ▪   ▪ ▪  ▪ 
Tumor 6N722  ▪ ▪ ▪      ▪ 
Tumor 6N870   ▪       ▪ 
Tumor 6N877 ▪ ▪ ▪ ▪   ▪ ▪   

NOTE: ▪, DHJH gene configurations that were present.

In addition, we analyzed for the presence of DH to JH gene rearrangements in six lymphoblastic lymphomas arising in the Msh2−/−;Foxn1nu/nu mice (Table 3). All six of the Msh2−/−;Foxn1nu/nu lymphoblastic lymphomas tested showed DH to JH gene rearrangements. Three of the six tumors had four, three, and two rearrangement products. Five of the six tumors had matching normal brain control DNA. Three of the five brain control samples had rearrangement products corresponding to those found in the tumor, leading us to suggest that metastasis has occurred to the brain in the Msh2−/−;Foxn1nu/nu lymphoblastic lymphomas as well as the Msh6−/−;Foxn1nu/nu lymphoblastic lymphomas.

Microsatellite instability analysis of lymphoblastic lymphomas. Two dinucleotide repeat microsatellite markers were used previously to investigate the presence of microsatellite instability in six Msh2−/− mice (17). Whereas normal tissues from the Msh2−/− mice were stable, all lymphomas and tumor-infiltrated organs displayed novel allele sizes at one or both of the loci tested (17). We therefore hypothesized that Msh2−/−;Foxn1nu/nu lymphoblastic lymphomas would also display high levels of microsatellite instability. Using five Msh2−/−;Foxn1nu/nu lymphoblastic lymphomas from five different mice, we investigated the stability of microsatellite sequences at 10 loci. Five primer pairs analyzed mononucleotide repeats; three primer pairs analyzed dinucleotide repeats; one primer pair analyzed a trinucleotide repeat; and one primer pair analyzed a tetranucleotide repeat. The size of the PCR product from tumor DNA was compared with the size of the PCR product from normal matched brain DNA. Tumors were considered microsatellite unstable low if 1 or 2 markers of 10 tested showed instability. Tumors were considered microsatellite unstable high if 3 or more markers of 10 tested showed instability. All tumors tested showed instability in at least one marker tested and were classified as microsatellite unstable. Two of five tumors tested showed instability at one or two markers and were classified as microsatellite unstable low whereas three of five tumors tested showed instability at greater than three markers tested and were classified as microsatellite unstable high. Overall, there were 14/50 unstable reactions corresponding to 28% instability.

As mismatch repair is thought to have a ubiquitous role in DNA repair and mutation avoidance, the tissue specificity of tumorigenesis in human HNPCC patients as well as in mismatch repair–deficient mice is unexpected. In contrast to the gastrointestinal tumors that develop in human HNPCC patients, mismatch repair–deficient mice develop predominately thymic lymphomas.

We hypothesized that mismatch repair–deficient mice succumb to tissue-specific lymphomagenesis at too young an age to allow for the development of a tumor spectrum similar to human HNPCC patients. If mice are prevented from or delayed in developing lymphomas, they may develop gastrointestinal tumors similar to human HNPCC patients and therefore present as a new model with which to study mismatch repair–deficient gastrointestinal tumor development. To test this hypothesis, we bred Msh2- and Msh6-deficient mice to athymic nude mice.

We have shown that breeding Msh2- and Msh6-deficient mice onto an athymic nude background promotes the development of predominantly B-cell lymphoblastic lymphomas. Moreover, these lymphomas arise from a pre-B-cell lineage at an early age, starting at 6.5 weeks of age. Thus, mice lacking mismatch repair seem destined to develop hematologic malignancy, even in the absence of a thymus and a greatly reduced T-cell component.

The median time to tumorigenesis in the Msh2−/−;Foxn1nu/nu mice, 5.5 months, was not altered compared with the Msh2−/− mice. This is in contrast to our original expectation that mice would live longer when prevented from developing thymic lymphomas. We propose that mismatch repair plays a crucial role in the development of hematologic cells and, in its absence, these cells are prone to malignant transformation. In contrast to the Msh2−/−;Foxn1nu/nu mice, the Msh6−/−;Foxn1nu/nu mice have reduced survival time to tumorigenesis (median survival, 8.9 months) compared with the Msh6−/− mice (median survival, 11 months). The earlier age of onset of Msh6−/−;Foxn1nu/nu malignancies may be due to differences in genetic background between the Msh6;Foxn1 and Msh6 animals although work by Edelmann4

4

Edelmann, personal communication.

would suggest that genetic background differences have no effect on the survival of Msh6−/−animals. The loss of the FOXN1 transcription factor may contribute to neoplasia in the absence of Msh6, however, nude mice themselves do not show an increase in spontaneous tumorigenesis. The early onset of B-cell malignancies in Msh6−/−;Foxn1nu/nu mice shows no survival advantage to mice lacking a thymus and thus a T-cell component. No primary tumors of other organs, such as intestine, endometrium, and gastrointestinal tract, occur when the onset of thymic lymphomagenesis is reduced in the Msh2−/−;Foxn1nu/nu or Msh6−/−;Foxn1nu/nu mice. Mismatch repair–deficient mice treated with genotoxic exposure have accelerated onset of lymphogenesis (37–39), also supporting the idea that in the constitutive absence of mismatch repair, mice are poised to develop predominantly hematologic malignancies, whether spontaneous or induced.

Although not seen macroscopically, early lesions in the colon may be occurring in the Msh2−/−;Foxn1nu/nu and Msh6−/−;Foxn1nu/nu mice. This would indicate that cellular changes are in fact altering the epithelium in the colon but that mice are succumbing to lymphomagenesis before further development of the colonic lesions. However, we found no cellular differences suggestive of a hyperproliferative or altered state in any of the colons studied regardless of age or region of colon investigated, leading us to suggest that mismatch repair–deficient nude mice are not prone to colonic lesions. The continued development of lymphomas and the absence of intestinal malignancies in Msh2−/−;Foxn1nu/nu and Msh6−/−;Foxn1nu/nu mice support the idea that mismatch repair is involved in the maintenance of hematologic cells.

Mismatch repair–deficient thymic lymphomas have been shown to comprise a single histopathologic entity derived from normal hematopoietic progenitors (17, 19). Thymic lymphomas from Msh2−/− mice closely resemble human precursor T-cell lymphoblastic lymphomas (23). Histologic examination of lymphomas from both Msh2−/−;Foxn1nu/nu and Msh6−/−;Foxn1nu/nu mice revealed a “starry sky” appearance and numerous mitotic figures indistinguishable from Msh2−/− thymic lymphomas. From this we suggest that similar underlying molecular pathways leading to tumorigenesis occur in these mismatch repair–deficient mouse models. This supports a role for mismatch repair in maintaining the genomic integrity of hematologic cell types and suggests that the absence of mismatch repair is associated with the accumulation of mutations in key tumor suppressor and oncogenes and the development of hematologic malignancies.

Of the 23 Msh2−/− thymic lymphomas in the literature that have been immunophenotyped, 22 of 23 express CD3, demonstrating their derivation from precursor T-cells (17, 23). Of the eight Msh6−/− lymphomas that have been immunophenotyped, five of eight are B-cell in origin and only three of eight are T-cell in origin (19). This may suggest a propensity for mice lacking Msh6 to develop B-cell malignancies whereas mice lacking Msh2 develop T-cell malignancies. If mice lacking Msh6 are indeed prone to develop B-cell malignancies, then it is not surprising that Msh6−/−;Foxn1nu/nu mice rapidly develop B-cell lymphoblastic lymphomas.

Only one of the Msh6−/−;Foxn1nu/nu lymphoblastic lymphomas stained positive for the marker CD45, confirming it was lymphoid in origin. Negative staining in seven of the eight tumors tested, however, suggested that these tumors were very early in development and did not yet present with the CD45 marker. B-cell development is characterized by the rearrangement and expression of immunoglobulin heavy and light chain genes. The first immunoglobulin gene rearrangement (occurring during the early pro-B-cell stage) is the joining of DH (diversity) gene elements to JH (joining) gene elements, followed in the late pro-B-cell stage by VH (variable) to DHJH joining. B220 (CD45R in humans) is a cell surface B-cell marker expressed in early pro-B-cells before or in conjugation with DH to JH gene rearrangement. After staining of the Msh6−/−;Foxn1nu/nu tumors found them to be B220 negative as well as CD3 negative, we investigated for the presence of Pax5. Negative staining for B220 and positive staining for Pax5 indicates that the majority of lymphoblastic lymphomas arising in Msh6−/−;Foxn1nu/nu mice are very early B-cell in origin. We hypothesized that these tumors are likely to be arrested in B-cell development before DH to JH joining. Surprisingly, we found all but one of the tumors showed some degree of DH to JH rearrangements, suggesting that B-cell development is abnormal in these tumors. It is possible that in the absence of mismatch repair, B-cell immunoglobulin gene rearrangement is incorrectly regulated, leading to tumorigenesis. We did find, however, that no particular class of DH or JH elements was overrepresented, suggesting that no particular element is responsible for the abnormal regulation. Further investigation for the presence of VH to DJH rearrangements may help determine the precise stage of development at which these lymphoblastic lymphomas arise.

We hypothesized that Msh2−/−;Foxn1nu/nu lymphoblastic lymphomas would display a high degree of microsatellite instability, an indication of a high level of overall genomic instability. All tumors tested showed microsatellite instability. Instability was seen almost exclusively in mono- and dinucleotide repeats; instability was negligible in the trinucleotide marker tested and nonexistent in the tetranucleotide marker tested. Particular mononucleotide markers (JH101, JH103, and U12235) and dinucleotide markers (D1mit83 and D7mit17) showed greater levels of instability than comparable markers, suggesting that some and not all microsatellites are informative when assaying for microsatellite instability. In addition, the length of a microsatellite may play a role in its susceptibility to instability. In Msh2−/−;Foxn1nu/nu lymphoblastic lymphomas, instability was not seen in the shorter mononucleotide repeats (JH102 and JH104, 8 and 10 bp, respectively) but was seen in the longer mononucleotide repeats (JH103 and U12335, 13 and 24 bp, respectively). Due to the lack of microsatellite instability seen in Msh6−/− tumor cells (19), likely due to the redundancy of Msh6 and Msh3 in forming a functional MutL heterodimer, microsatellite instability was not tested in tumors arising in Msh6−/−;Foxn1nu/nu mice.

In contrast to HNPCC patients with heterozygous mismatch repair mutations, patients with homozygous mutations in mismatch repair genes present with early-onset T- or B-cell malignancies. From this we suggest that the effect of a loss of mismatch repair is highly dependent on when in development loss of mismatch repair occurs. For example, constitutive loss of mismatch repair in either mice or humans leads to a predisposition to develop hematologic cancers. Alternatively, in HNPCC, loss of the wild-type allele that results in loss of mismatch repair occurs in specific tissues later in life. Those tissues, such as the colon, may be those with higher exposure to exogenous DNA damaging agents and may therefore result in a different tumor spectrum.

Previous reports have suggested that thymic lymphomas predominate in mismatch repair–deficient mice due to the high turnover of developing T-cells in the thymus and subsequent accumulation of mutations in key downstream genes (20, 40). Our finding that mismatch repair–deficient mice develop hematologic malignancies even in the absence of a thymus leads us to suggest that malignant transformation is independent of cell turnover rate. Altered mismatch repair gene expression, mismatch repair gene mutation, and/or microsatellite instability has been observed in varying proportions of human hematologic neoplasms, consistent with the idea that loss of mismatch repair has a significant role in human hematologic malignancy (41–45).

We have previously reported that constitutive loss of MSH2 in a human patient leads to T-cell acute lymphocytic leukemia and features consistent with neurofibromatosis type 1 (16). Recently, this patient developed B-cell lymphoma. This parallels the B-cell malignancies seen in the Msh2−/−;Foxn1nu/nu and Msh6−/−;Foxn1nu/nu mice and further strengthens our conclusions that mismatch repair plays an important role in the development of T- and B-cells. Our results support the hypothesis that mismatch repair has a key role in developing T- and B-cells. We implicate mismatch repair as an important repair mechanism in normal T- and B-cell development.

Grant support: Alberta Heritage Foundation for Medical Research and the Leukemia Research Foundation of Canada; studentships from Alberta Heritage Foundation for Medical Research and Alberta Cancer Board (M.R. Campbell); salary support from Alberta Heritage Foundation for Medical Research, Canadian Institutes of Health Research (S.E. Andrew).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Dr. Frank Jirik for initiation of this project, members of the University of Alberta Health Sciences Laboratory Animal Services for biotechnical assistance, and Dr. Mike Walter for critically reading the manuscript.

1
Buermeyer AB, Deschenes SM, Baker SM, Liskay RM. Mammalian DNA mismatch repair.
Annu Rev Genet
1999
;
33
:
533
–64.
2
Mark SC, Sandercock LE, Luchman HA, Baross A, Edelmann W, Jirik FR. Elevated mutant frequencies and predominance of G:C to A:T transition mutations in Msh6(−/−) small intestinal epithelium.
Oncogene
2002
;
21
:
7126
–30.
3
Thibodeau SN, Bren G, Schaid D. Microsatellite instability in cancer of the proximal colon.
Science
1993
;
260
:
816
–9.
4
Ionov Y, Peinado MA, Malkhosyan S, Shibata D, Perucho M. Ubiquitous somatic mutations in simple repeated sequences reveal a new mechanism for colonic carcinogenesis.
Nature
1993
;
363
:
558
–61.
5
Aaltonen LA, Peltomaki P, Leach FS, et al. Clues to the pathogenesis of familial colorectal cancer.
Science
1993
;
260
:
812
–6.
6
Shoemaker AR, Haigis KM, Baker SM, Dudley S, Liskay RM, Dove WF. Mlh1 deficiency enhances several phenotypes of Apc(Min)/+ mice.
Oncogene
2000
;
19
:
2774
–9.
7
Peltomaki P. Deficient DNA mismatch repair: a common etiologic factor for colon cancer.
Hum Mol Genet
2001
;
10
:
735
–40.
8
Hirano K, Yamashita K, Yamashita N, et al. Non-Hodgkin's lymphoma in a patient with probable hereditary nonpolyposis colon cancer: report of a case and review of the literature.
Dis Colon Rectum
2002
;
45
:
273
–9.
9
Rosty C, Briere J, Cellier C, et al. Association of a duodenal follicular lymphoma and hereditary nonpolyposis colorectal cancer.
Mod Pathol
2000
;
13
:
586
–90.
10
Bougeard G, Charbonnier F, Moerman A, et al. Early onset brain tumor and lymphoma in MSH2-deficient children.
Am J Hum Genet
2003
;
72
:
213
–6.
11
De Vos M, Hayward BE, Picton S, Sheridan E, Bonthron DT. Novel PMS2 pseudogenes can conceal recessive mutations causing a distinctive childhood cancer syndrome.
Am J Hum Genet
2004
;
74
:
954
–64.
12
Gallinger S, Aronson M, Shayan K, et al. Gastrointestinal cancers and neurofibromatosis type 1 features in children with a germline homozygous MLH1 mutation.
Gastroenterology
2004
;
126
:
576
–85.
13
Ricciardone MD, Ozcelik T, Cevher B, et al. Human MLH1 deficiency predisposes to hematological malignancy and neurofibromatosis type 1.
Cancer Res
1999
;
59
:
290
–3.
14
Vilkki S, Tsao JL, Loukola A, et al. Extensive somatic microsatellite mutations in normal human tissue.
Cancer Res
2001
;
61
:
4541
–4.
15
Wang Q, Lasset C, Desseigne F, et al. Neurofibromatosis and early onset of cancers in hMLH1-deficient children.
Cancer Res
1999
;
59
:
294
–7.
16
Whiteside D, McLeod R, Graham G, et al. A homozygous germ-line mutation in the human MSH2 gene predisposes to hematological malignancy and multiple cafe-au-lait spots.
Cancer Res
2002
;
62
:
359
–62.
17
Reitmair A, Schmits R, Ewel A, et al. MSH2 deficient mice are viable and susceptible to lymphoid tumours.
Nat Genet
1995
;
11
:
64
–70.
18
Edelmann W, Cohen PE, Kane M, et al. Meiotic pachytene arrest in MLH1-deficient mice.
Cell
1996
;
85
:
1125
–34.
19
Edelmann W, Yang K, Umar A, et al. Mutation in the mismatch repair gene Msh6 causes cancer susceptibility.
Cell
1997
;
91
:
467
–77.
20
de Wind N, Dekker M, Berns A, Radman M, te Riele H. Inactivation of the mouse Msh2 gene results in mismatch repair deficiency, methylation tolerance, hyperrecombination, and predisposition to cancer.
Cell
1995
;
82
:
321
–30.
21
Baker SM, Bronner CE, Zhang L, et al. Male mice defective in the DNA mismatch repair gene PMS2 exhibit abnormal chromosome synapsis in meiosis.
Cell
1995
;
82
:
309
–19.
22
Prolla TA, Baker SM, Harris AC, et al. Tumour susceptibility and spontaneous mutation in mice deficient in Mlh1, Pms1 and Pms2 DNA mismatch repair.
Nat Genet
1998
;
18
:
276
–9.
23
Lowsky R, DeCoteau JF, Reitmair AH, et al. Defects of the mismatch repair gene MSH2 are implicated in the development of murine and human lymphoblastic lymphomas and are associated with the aberrant expression of rhombotin-2 (Lmo-2) and Tal-1 (SCL).
Blood
1997
;
89
:
2276
–82.
24
Smits R, Hofland N, Edelmann W, et al. Somatic Apc mutations are selected upon their capacity to inactivate the β-catenin down-regulating activity.
Genes Chromosomes Cancer
2000
;
29
:
229
–39.
25
Edelmann W, Yang K, Kuraguchi M, et al. Tumorigenesis in Mlh1 and Mlh1/Apc1638N mutant mice.
Cancer Res
1999
;
59
:
1301
–7.
26
Kuraguchi M, Yang K, Wong E, et al. The distinct spectra of tumor-associated Apc mutations in mismatch repair-deficient Apc1638N mice define the roles of MSH3 and MSH6 in DNA repair and intestinal tumorigenesis.
Cancer Res
2001
;
61
:
7934
–42.
27
Nehls M, Pfeifer D, Schorpp M, Hedrich H, Boehm T. New member of the winged-helix protein family disrupted in mouse and rat nude mutations.
Nature
1994
;
372
:
103
–7.
28
Balciunaite G, Keller MP, Balciunaite E, et al. Wnt glycoproteins regulate the expression of FoxN1, the gene defective in nude mice.
Nat Immunol
2002
;
3
:
1102
–8.
29
Palacios R, Samaridis J. Rearrangement patterns of T-cell receptor genes in the spleen of athymic (nu/nu) young mice.
Immunogenetics
1991
;
33
:
90
–5.
30
Hirasawa T, Yamashita H, Makino S. Genetic typing of the mouse and rat nude mutations by PCR and restriction enzyme analysis.
Exp Anim
1998
;
47
:
63
–7.
31
Ihaka R, Gentleman R. R: A Language for Data Analysis and Graphics.
J Comput Graph Stat
1996
;
5
:
299
–314.
32
Chang Y, Paige CJ, Wu GE. Enumeration and characterization of DJH structures in mouse fetal liver.
Embo J
1992
;
11
:
1891
–9.
33
Edelmann W, Umar A, Yang K, et al. The DNA mismatch repair genes Msh3 and Msh6 cooperate in intestinal tumor suppression.
Cancer Res
2000
;
60
:
803
–7.
34
Hearne CM, McAleer MA, Love JM, et al. Additional microsatellite markers for mouse genome mapping.
Mamm Genome
1991
;
1
:
273
–82.
35
Kurtin PJ, Pinkus GS. Leukocyte common antigen—a diagnostic discriminant between hematopoietic and nonhematopoietic neoplasms in paraffin sections using monoclonal antibodies: correlation with immunologic studies and ultrastructural localization.
Hum Pathol
1985
;
16
:
353
–65.
36
Nutt SL, Heavey B, Rolink AG, Busslinger M. Commitment to the B-lymphoid lineage depends on the transcription factor Pax5.
Nature
1999
;
401
:
556
–62.
37
Jansen L, Claij N, Dekker M, et al. Acceleration of lymphomagenesis in mismatch-repair deficient mice by exposure to genotoxic agents.
Toxicol Lett
2000
;
112–3
:
245
–50.
38
Colussi C, Fiumicino S, Giuliani A, et al. 1,2-Dimethylhydrazine-induced colon carcinoma and lymphoma in msh2(−/−) mice.
J Natl Cancer Inst
2001
;
93
:
1534
–40.
39
Qin X, Liu L, Gerson SL. Mice defective in the DNA mismatch gene PMS2 are hypersensitive to MNU induced thymic lymphoma and are partially protected by transgenic expression of human MGMT.
Oncogene
1999
;
18
:
4394
–400.
40
de Wind N, Dekker M, van Rossum A, van der Valk M, te Riele H. Mouse models for hereditary nonpolyposis colorectal cancer.
Cancer Res
1998
;
58
:
248
–55.
41
Wada C, Shionoya S, Fujino Y, et al. Genomic instability of microsatellite repeats and its association with the evolution of chronic myelogenous leukemia.
Blood
1994
;
83
:
3449
–56.
42
Kotoula V, Hytiroglou P, Kaloutsi V, Barbanis S, Kouidou S, Papadimitriou CS. Mismatch repair gene expression in malignant lymphoproliferative disorders of B-cell origin.
Leuk Lymphoma
2002
;
43
:
393
–9.
43
Gu L, Cline-Brown B, Zhang F, Qiu L, Li GM. Mismatch repair deficiency in hematological malignancies with microsatellite instability.
Oncogene
2002
;
21
:
5758
–64.
44
Seedhouse CH, Das-Gupta EP, Russell NH. Methylation of the hMLH1 promoter and its association with microsatellite instability in acute myeloid leukemia.
Leukemia
2003
;
17
:
83
–8.
45
Matheson EC, Hall AG. Expression of DNA mismatch repair proteins in acute lymphoblastic leukaemia and normal bone marrow.
Adv Exp Med Biol
1999
;
457
:
579
–83.