NIAM-deficient mice are predisposed to the development of proliferative lesions including B-cell lymphomas

Abstract

Nuclear Interactor of ARF and Mdm2 (NIAM, gene designation Tbrg1) is a largely unstudied inhibitor of cell proliferation that helps maintain chromosomal stability. It is a novel activator of the ARF-Mdm2-Tip60-p53 tumor suppressor pathway as well as other undefined pathways important for genome maintenance. To examine its predicted role as a tumor suppressor, we generated NIAM mutant (NIAM(m/m)) mice homozygous for a β-galactosidase expressing gene-trap cassette in the endogenous gene. The mutant mice expressed significantly lower levels of NIAM protein in tissues compared to wild-type animals. Fifty percent of aged NIAM deficient mice (14 to 21 months) developed proliferative lesions, including a uterine hemangioma, pulmonary papillary adenoma, and a Harderian gland adenoma. No age-matched wild-type or NIAM(+/m) heterozygous animals developed lesions. In the spleen, NIAM(m/m) mice had prominent white pulp expansion which correlated with enhanced increased reactive lymphoid hyperplasia and evidence of systemic inflammation. Notably, 17% of NIAM mutant mice had splenic white pulp features indicating early B-cell lymphoma. This correlated with selective expansion of marginal zone B cells in the spleens of younger, tumor-free NIAM-deficient mice. Unexpectedly, basal p53 expression and activity was largely unaffected by NIAM loss in isolated splenic B cells. In sum, NIAM down-regulation in vivo results in a significant predisposition to developing benign tumors or early stage cancers. These mice represent an outstanding platform for dissecting NIAM's role in tumorigenesis and various anti-cancer pathways, including p53 signaling.

Figure 1. NIAM mRNA expression is downregulated in multiple human cancers.

NIAM mRNA levels are significantly reduced in primary tumor (T) samples relative to normal (N) tissue, according to RNA-seq data obtained from the TCGA project database. The number of samples analyzed for each tissue was as follows: lung (N = 58, T = 488), liver (N = 50, T = 147), bladder (N = 19, T = 211), and breast (N = 108, T = 992). Error bars represent a 95% CI as calculated using the standard error, and an unpaired Welch's T-test was used to calculate statistical significance (*, p<0.0001).

Figure 2. Generating and verifying NIAM gene disruption in mice.

A. The targeted NIAM gene locus for the conditional-ready mouse knockout model. Cre-recombinase LoxP sites and Flippase targeted FRT sites are shown, as are locations for the PCR primers to detect the endogenous NIAM allele (a+b) or the mutant allele (a+c). Ex, exon. B. PCR amplification of the mutant (mut) allele results in a 380 bp product whereas the 548 bp product reflects the wild-type (WT) NIAM allele. DNA standards are shown in lane 4. C. Number of mice with each genotype (+/+, +/m, m/m) from heterozygous mouse crossings. Chi-square analysis shows a statistically significant difference from the Mendelian distribution of 1∶2∶1.

Figure 3. NIAM protein expression in tissues of NIAM mutant mice.

NIAM protein levels were assessed in mouse tissues from two mice of each genotype (+/+, +/m, m/m). Equivalent amounts of protein lysates from spleen, bladder, lung, brain, and pancreas were analyzed by immunoblotting with antibodies to NIAM and γ-tubulin. The relative expression of NIAM in each sample, as compared to the first wild-type (+/+) sample in each tissue set, is denoted below the lanes. Relative NIAM levels were determined by Image J analysis of band intensities followed by normalization of those values to quantified intensities of the loading control (γ-tubulin or the non-specific band in brain lysates) in each sample.

Figure 4. NIAM^m/m mice are predisposed to proliferative lesions.

Histopathologic analyses of multiple mouse tissues. A. A hemangioma (asterisk, right panel) found in the uterus of a NIAM^m/m mouse compared to normal uterine tissue (left panel). (H&E, 100x) B. Pulmonary papillary adenoma in a homozygous NIAM^m/m mouse (arrow, right panel). A depiction of normal lung tissue from a control mouse is shown (left panel). (H&E, 20x) C. Harderian gland adenoma of a NIAM^m/m mouse (asterisks, right panel). A representative image of a normal Harderian gland (asterisk) from a control mouse is depicted (left panel). (H&E, 40x) D. A representative image of a normal liver from a control mouse (left panel) versus a focus of cellular alteration (arrows) in a NIAM^m/m mouse (right panel). (H&E, 40x) Controls represent tissues from similarly aged wild-type (panels B and C) and NIAM^+/m heterozygous (panels A and D) mice, which were found to be indistinguishable in our analyses.

Figure 5. Loss of NIAM promotes reactive lymphoid hyperplasia and B-cell lymphoma in the spleen.

A. Weights of spleens from +/+ and +/m mice (CON, n = 15) compared to NIAM^m/m mice (m/m, n = 12). Note: one WT (+/+) mouse with an incidental case of severe chronic pyometra was excluded from the analysis. Statistical significance was shown by a Mann-Whitney test (*, p<0.05) B. The percentage of splenic white pulp per total splenic area was calculated for each genotype. NIAM^m/m mice (n = 8) are statistically different from both wild-type (+/+, n = 5) and heterozygous (+/m, n = 6) mice as calculated by an unpaired two-tailed T-test. (*, p<0.05) Note: one WT (+/+) and one mutant (m/m) mouse were excluded from the analysis because extensive extramedullary hematopoiesis prevented accurate visualization of the white pulp C. Representative images of spleens from control wild-type (+/+) and homozygous m/m mice (H&E, 40x). Splenic white pulp (asterisks) was increased in NIAM^m/m mice. Note: these samples show a range of variation in the extent of red pulp extramedullary hematopoiesis that was seen for both groups. D. Eosinophilic crystalline pneumonia in the lungs of control NIAM^+/m heterozygous and NIAM^m/m mice. Loss of NIAM resulted in more severe disease with extensive pulmonary lymphoid inflammation (arrows) and large extracellular crystals (asterisks) in the affected NIAM^m/m mouse. (H&E, 40× top and 100× bottom panels) E. White pulp from control NIAM^+/m heterozygous and NIAM^m/m homozygous mice (H&E, 600x). A spleen from a heterozygous mouse with chronic ulcerative dermatitis and reactive hyperplasia of white pulp characterized by a germinal center (GC) and tingible body macrophages full of cellular debris (arrows) on a background of small lymphocytes (SL) (left panel). In two homozygous m/m mice, typical white pulp architecture was multifocally depleted and effaced by centroblasts and centrocytes with multiple mitotic figures (inset and arrows) (right panel). Note that NIAM^+/m heterozygous and NIAM^+/+ wild-type mice appeared identical in our analyses and were therefore used interchangeably as controls.

Figure 6. Marginal zone B cells are increased in NIAM^m/m mice.

Flow cytometric analyses of splenic B cell development. Splenic B cells were isolated from young, 6 month old tumor-free wild-type (+/+) and NIAM mutant (m/m) mice (three of each genotype). A. Representative flow cytometry plots of NIAM wild-type versus m/m IgM positive splenocytes. Transitional (CD21^lo/CD23-), follicular (CD21+/CD23+) and marginal zone (CD21^hi/CD23-) B cells are identified. B. Average frequency of follicular (top left), transitional (top right), and marginal zone (bottom left) B cells from mice of the indicated genotypes. Average marginal zone B cell numbers are also shown (bottom right). All p values were calculated using unpaired, two-tailed Student's T tests.

Figure 7. NIAM is not required for basal p53 activity in splenic B cells.

Splenic B cells from young, tumor-free wild-type NIAM^+/+ or NIAM^m/m homozygous mutant mice were isolated and expression of endogenous NIAM, p53, Mdm2 and p21 proteins was assessed by western blotting. Relative levels of p53, Mdm2 and p21 (normalized to GAPDH loading and compared to the untreated, wild-type control sample) were determined following quantification of bands using Image J. Representative results from 4 of 5 pairs of mice (both 8 week and 6 months of age) are shown in Set A, while Set B shows data for one pair of 6 month old mice. Set A and B samples were analyzed on separate gels, and lanes within each set were spliced together from the same autoradiogram for image clarity.