Impaired Fas Response and Autoimmunity in Pten+/− Mice

See allHide authors and affiliations

Science  24 Sep 1999:
Vol. 285, Issue 5436, pp. 2122-2125
DOI: 10.1126/science.285.5436.2122


Inactivating mutations in the PTEN tumor suppressor gene, encoding a phosphatase, occur in three related human autosomal dominant disorders characterized by tumor susceptibility. Here it is shown that Pten heterozygous (Pten +/−) mutants develop a lethal polyclonal autoimmune disorder with features reminiscent of those observed in Fas-deficient mutants. Fas-mediated apoptosis was impaired in Pten +/− mice, and T lymphocytes from these mice show reduced activation-induced cell death and increased proliferation upon activation. Phosphatidylinositol (PI) 3-kinase inhibitors restored Fas responsiveness inPten +/− cells. These results indicate thatPten is an essential mediator of the Fas response and a repressor of autoimmunity and thus implicate the PI 3-kinase/Akt pathway in Fas-mediated apoptosis.

The PTEN gene encodes a phosphatase homozygously mutated in a high percentage of human tumors (1, 2). Heterozygous inactivation of PTEN results in three human dominant disorders: Cowden disease, Bannayan-Zonana syndrome, and Lhermitte-Duclos syndrome (3). Disruption ofPten in the mouse results in early embryonic lethality (4). Pten heterozygous (Pten +/−) mice display hyperplastic-dysplastic features as well as high tumor incidence (4). The complete penetrance of the hyperplastic-dysplastic changes suggests that these features could be due to Pten haploinsufficiency. However, the specific biological consequences of Ptenhaploinsufficiency remain unclean as well as whether completePten inactivation must occur for full neoplastic transformation. A major substrate of PTEN is phosphatidylinositol trisphosphate (PIP-3), a lipid second messenger produced by PI 3-kinase (5). In the absence of Pten activity, PIP-3 concentrations are increased, leading to enhanced phosphorylation and activation of the survival-promoting factor Akt/PKB (6). Pten −/−embryonic stem cells and mouse embryonic fibroblasts are protected from some apoptotic stimuli (6), which suggests thatPten can inhibit Akt-dependent survival signals induced in response to PI 3-kinase activation. Here, we show that Ptenhaploinsufficiency results in a lethal autoimmune disorder and that Fas-mediated apoptosis is impaired in Pten +/−mice.

Almost 100% of Pten +/− females (44 of 45, in the C57BL6/129Sv background) developed, between 4 and 5 months of age, a severe lymphoadenopathy affecting mainly the submandibular, axillary, and inguinal lymph nodal stations (Fig. 1A). The mice died, most likely of renal failure (see below), before they were 1 year old. Male mutants were more mildly affected, with 83% of the animals (20 of 24) developing less severe lymph node hyperplasia by 8 months of age and surviving up to at least 15 months.

Figure 1

Histopathological analysis ofPten +/− mice. (A) Lymph node hyperplasia (arrow) in a 7-month-old Pten +/−female mouse. (B) Dissected spleen and lymph node from sex- and age-matched wild-type (left) and Pten +/−(right) mice. Scale bar, 1 cm. (C) Photomicrograph of a lymph node from a Pten +/− mouse, stained with H&E. Note the hyperreactive germinal centers and the partial effacement of the normal corticomedullary boundaries. Inset: H&E staining of a lymph node from a wild-type mouse, showing well-delineated microanatomical features. (D) Photomicrograph of a spleen from a Pten +/− mouse, stained with H&E. Note the hyperreactive white pulp and the partial effacement of the normal spleen histology. Inset: H&E staining of a spleen from a wild-type mouse, showing well-defined white and red pulp. (E) Consecutive section from the lymph node in (C), stained with anti-B220, showing intense staining in the hyperreactive germinal center. Inset: B220 staining of a lymph node from a wild-type mouse. (F) Consecutive section from the lymph node shown in (C) and (E), stained with anti-CD3, showing staining in the areas surrounding the germinal center and part of the medullary zone. Inset: CD3 staining of a lymph node from a wild-type mouse. (G) Photomicrograph of the lung from a Pten +/−mouse stained with H&E. Note the increased thickening of the interstitial alveolar spaces. (H) Photomicrograph of the kidney from a Pten +/− mouse stained with H&E, showing dilated tubules filled with proteinaceous material (arrowhead) and glomeruli displaying signs of focal proliferation and sclerosis (arrows). Scale bars, 200 μm (C to F); 50 μm (G and H).

Gross pathological analysis consistently revealed features that are typically observed in autoimmune disorders:

1) The spleen was enlarged and the lymph nodes were markedly hyperplastic (Fig. 1B). Histological examination (7) showed that these organs were characterized by hyperreactive features, with partial to total effacement of the normal architecture. Germinal centers displayed hyperplastic changes (Fig. 1, C and D). The expansion of T and B lymphocytes was polyclonal (Fig. 1, E and F), as verified by a flow cytometric analysis (see below).

2) Most other organs were congested, with inflammatory infiltrates. In particular, the lung showed increased thickening of the interstitial alveolar spaces with vascular congestion and, in some instances, complete collapse of the pulmonary lobes (Fig. 1G).

3) The kidneys showed proliferation of mesangial cells and increased extracellular matrix in the glomeruli, accompanied by vacuolization and dilation of proximal tubules, with the presence of proteinaceous material in the lumen. In the most severely affected mice, the glomeruli were segmentally or totally solidified (Fig. 1H), with focal thickening of the capillary walls, a picture of segmental glomerulosclerosis. Similar pathological features are observed in human autoimmune syndromes and in autoimmune-prone mouse strains (8).

To confirm that Pten +/− mice suffered from an autoimmune glomerulopathy, we analyzed the kidneys from our mutant animals for the presence of immune complexes deposited in the glomeruli. Immunofluorescence microscopy showed immunocomplex deposition displaying the classical diffuse global granular staining of the capillary walls (Fig. 2A) (9).

Figure 2

Autoimmune features inPten +/− mice. (A) Immunohistochemical analysis of a kidney from a 9-month-old female, showing deposition of immune complexes in the glomerulus detected by antibody to mouse IgG. A wild-type control is shown in the inset. Scale bar, 1 μm. (B) Serum concentrations of IgG in individual wild-type and Pten +/− mice between 6 and 10 months of age (n = 6 Pten +/+ and 8 Pten +/− males; 5Pten +/+ and 10 Pten +/−females). IgG concentrations are expressed in milligrams per milliliter. Means and SD are indicated for each group. (C) ANAs in Pten +/− mice. Serum was diluted 1:50 and tested by indirect immunofluorescence on HEp-2 cells. A wild-type control is shown in the inset. Scale bar, 1 μm. (D) Titers of anti-ssDNA autoantibodies in the serum of individual wild-type and Pten +/− mice (n = 7 Pten +/+ and 6Pten +/− males; 7 Pten +/+and 13 Pten +/− females). Sera were diluted 1:100 and tested by ELISA. Mean values are indicated for each group.

As the primary cause for deposition of immune complexes in the kidney is hypergammaglobulinemia with autoantibodies reacting against nuclear antigens [histones, double-stranded DNA, and single-stranded (ss) DNA], we analyzed the immunoglobulin concentrations of tumor-freePten +/− mice and found a marked increase in serum immunoglobulin G (IgG) concentrations relative to wild-type mice (9). This was particularly conspicuous for the females (P < 0.01) (Fig. 2B). We next tested the presence in the serum of nuclear antibodies (ANAs) (9). Half the affected females had high titers of ANAs (Fig. 2C). Moreover, almost all the Pten +/− female mice analyzed had higher titers of antibodies to ssDNA (anti-ssDNA) (9) relative to wild-type controls (P < 0.01) (Fig. 2D). In thePten +/− males the difference with respect to their wild-type counterparts was less striking (P < 0.05). Finally, urine analysis of Pten +/−mutants revealed the presence of additional features of an autoimmune disorder, such as proteinuria and high leukocyte and hemoglobin concentrations in the most severely affected animals (9).

One hallmark of the autoimmune diseases affecting autoimmune-prone mouse strains is the progressive accumulation in the periphery of activated T and B cells (8, 10). We analyzed spleen and lymph nodes from wild-type and Pten +/− mice for the expression of lineage and activation markers (11). The ratio of B and T cells as well as the ratio of CD4+ and CD8+ T cell subsets was normal in the lymph nodes, confirming the polyclonal nature of the lymphocyte expansion. In contrast, an expansion of the splenic T cell compartment (wild type, 33.3 ± 3.4%; heterozygous, 41.2 ± 0.9%) and an increased CD4+ population (CD4+/CD8+ ratio: wild type, 1.2 ± 0.07; heterozygous, 1.87 ± 0.05) was observed in about 50% of the cases. Contrary to what was found in thelpr mutants, Pten +/− mice did not show a B220+/Thy1+ population or a CD4/CD8 T cell subset. Analysis of the T cell activation markers in all cases showed highly increased populations positive for CD44, CD54, and CD69 (Fig. 3, A and B). Numbers of CD44-, B7-2–, and CD5-positive B lymphocytes were also doubled. These results show that both T and B cells are activated inPten +/− mice.

Figure 3

Expression of activation markers on T and B lymphocytes from gender- and age-matched wild-type andPten +/− mice. Spleen cells were double-stained as indicated (A to C) and analyzed by flow cytometry. Percentages of analyzed cells are shown in each quadrant.

The Fas/FasL system plays a crucial role in the maintenance of peripheral tolerance. Fas is up-regulated upon T and B cell activation, leading to the elimination of these cells through Fas-mediated apoptosis (12). Accumulation of Fashigh B cells in the spleen ofgld mice has been explained by the inability of these primed cells to undergo apoptosis (12). We analyzedPten +/− mice for the expression of Fas in peripheral lymphocytes. Strikingly, the number of B cells expressing Fas was increased by a factor of 3 (Fig. 3C). Moreover, the level of expression of Fas on the surface of T and B cells was increased by a factor of 2 to 3 (mean fluorescence intensity: 77 ± 1 and 132.3 ± 9 for wild-type and Pten +/− T cells, respectively; 47.4 ± 16 and 122 ± 6 for wild-type and Pten +/− B cells, respectively).

The expansion of activated lymphocytes in the presence of a concomitant increase in the expression of Fas cells is suggestive of a defect in Fas-mediated apoptosis. Thus, we analyzed the role ofPten in lymphocyte activation, activation-induced cell death (AICD), and Fas-mediated apoptosis. Splenocytes fromPten +/− mice proliferated in amounts equal to those in littermate controls in response to treatment with anti-CD3 or lipopolysaccharide (LPS) (13). Concanavalin A (Con A) stimulation, however, resulted in an about 70% increase in proliferation in Pten +/− cells (Fig. 4A). Therefore, an enhanced response ofPten +/− mature T cells to mitogens might account for at least part of the accumulation of activated cells in the periphery.

Figure 4

Increased response to Con A and reduced Fas-dependent apoptosis in Pten +/− mice. Error bars represent SD. (A) [3H]Thymidine incorporation of activated wild-type (black bar) andPten +/− (white bar) splenocytes. Cells were stimulated for 72 hours as indicated; [3H]thymidine was added for the last 16 hours. (B) Reduced AICD inPten +/− T lymphocytes. CD3+-enriched spleen cells were activated with plate-bound anti-CD3ɛ with or without Fas:Fc chimeric protein (10 μg/ml), and cell death was determined at day 4 by trypan blue exclusion and in situ TUNEL assay. Wild type, black bar;Pten +/−, white bar. (C) Impaired anti-CD3ɛ and anti-CD95/Fas–induced apoptosis in activated T and B cells. Total spleen lymphocytes or CD3+-enriched cells were stimulated for 4 days with cross-linked anti-CD3ɛ or for 3 days with LPS. Live cells were purified and plated in the presence of plate-bound anti-CD3ɛ or anti-CD95/Fas for 24 to 48 hours. Cell death was determined as in (B) and is expressed as specific apoptosis (spontaneous apoptosis from untreated control samples has been subtracted from the corresponding treated samples). Wild type, black bar; Pten +/−, white bar; lpr, gray bar. Act. st., activation stimulus; Ap. st., apoptotic stimulus. (D) Impaired anti-CD95/Fas–induced apoptosis inPten +/− PEFs. Fibroblasts were treated with anti-CD95/Fas for 18 hours. Cell death was determined by Annexin V staining and is expressed as specific apoptosis. Wild type, black bar;Pten +/−, white bar. (E) Protein immunoblot analysis of protein extracts from freshly isolated (-), LPS-activated (LPS), and anti-Fas–stimulated splenocytes in the absence (αFas) or presence (αFas/wm) of wortmannin. β-Actin was used to show equal loading. (F) Wortmannin restores anti-CD95/Fas–induced apoptosis in activated B cells. Total spleen lymphocytes were stimulated for 3 days with LPS. Live cells were purified and plated in the presence of plate-bound anti-CD95/Fas for 24 hours. Cell death was determined as in (B). Wild type, black bar;Pten +/−, white bar. Error bars represent SD.

To evaluate the sensitivity of Pten +/−mature T cells to AICD, we activated CD3ɛ +-enriched spleen cells (in the presence of a Fas:Fc chimeric protein, which neutralized mouse Fas ligand) and measured cell death 4 days later (14).Pten +/− T cells showed a 50% decrease in the number of dying cells relative to controls (Fig. 4B). Thus, in the absence of one Pten allele, AICD of peripheral T cells is impaired, leading to the expansion of an activated T cell compartment.

We then examined peripheral T and B cells for their sensitivity to apoptosis induced by CD3ɛ and CD95 (Fas) agonistic antibodies (14). Under all experimental conditions, the induction of apoptosis in Pten +/−-activated T and B lymphocytes by these stimuli was decreased, even though these cells overexpressed Fas (Fig. 4C). Moreover, we found thatPten +/− primary embryonic fibroblasts (PEFs) were also protected from Fas-dependent apoptosis (14) (Fig. 4D). Thus, Fas-mediated signaling is impaired in Pten +/− mice and cells.

To investigate the mechanisms underlying the protection from Fas-dependent apoptosis in Pten +/− mice, we first analyzed the expression levels of molecules involved in the death-inducing signaling complex (DISC) formation and function and found no difference between wild-type andPten +/− cells (15).

To evaluate whether the protection from Fas-dependent apoptosis could be attributed to a PI 3-kinase/Akt–dependent pathway, we analyzed the phosphorylation status of Akt in wild-type andPten +/− splenocytes (15). Akt was hyperphosphorylated in both freshly isolated and LPS-activated Pten +/− splenocytes (Fig. 4E). Similar results were obtained in splenocytes activated with Con A.

Akt is degraded early, along with Parp, upon the activation of the caspase proteolytic cascade (16). We therefore analyzed the amount of Akt and Parp after anti-Fas treatment and found that the caspase-dependent degradation of these proteins was severely impaired in Pten +/− cells. These differences were abrogated by wortmannin (Fig. 4E). Moreover, pretreatment of the cells with wortmannin completely restored the sensitivity ofPten +/− cells to Fas (Fig. 4F). These data indicate a pivotal role for Ptenhaploinsufficiency-dependent Akt activation in the protection from Fas-induced apoptosis observed in Pten +/−cells.

In summary, our findings lead to two major conclusions: (i)Pten function is crucial for Fas-mediated elimination of activated lymphocytes, including self reactive cells in the periphery, and (ii) inactivation of one Pten allele increases the survival and proliferation of certain cell types. In principle, this could lead to the accumulation of further mutations, including loss of heterozygosity at the Pten locus, which would ultimately result in full neoplastic transformation. Thus, Pten may exert its tumor suppressor function by facilitating programmed cell death upon DNA damage or neoplastic transformation. These data are consistent with the fact that Pten +/−mutants develop T and B cell lymphomas (4), which originate from cells that are particularly sensitive to the action of Fas (17), and provide physiological evidence suggesting that PTEN negatively regulates a PI 3-kinase/Akt–dependent pathway for the suppression of Fas induced apoptosis.

  • * To whom correspondence should be addressed. E-mail: p-pandolfi{at}


View Abstract

Navigate This Article