Induction of Endoplasmic Reticulum Stress in Thymic Lymphocytes by the Envelope Precursor Polyprotein of a Murine Leukemia Virus during the Preleukemic Period

Abstract

Infection of thymic lymphocytes by a mink cell focus-forming murine leukemia virus induces apoptosis during the preleukemic period of lymphomagenesis. In this study, we observed that during this period, the viral envelope precursor polyprotein accumulated to high levels in thymic lymphocytes from mice inoculated with virus. Envelope accumulation occurred with the same kinetics as the induction of endoplasmic reticulum (ER) stress, which resulted in the upregulation of the 78-kDa glucose-regulated protein (GRP78). In thymic lymphomas, GRP78 levels were higher than those in virus-infected preleukemic cells, and GRP58 was upregulated. These results suggest that Env precursor accumulation induces ER stress, which participates in thymic lymphocyte apoptosis. The subsequent upregulation of ER chaperone proteins GRP78 and GRP58 may contribute to rescuing cells from virus-induced apoptosis.

Pathogenesis by certain retroviruses is associated with their ability to induce cytopathic effects (1, 2, 4, 12, 17, 19-21, 23-27). Murine leukemia viruses (MLVs) that generate thymic lymphoma, including the mink cell focus-forming (MCF) MLV, have been observed to markedly reduce thymus cellularity via apoptosis during the preleukemic period (27). To elucidate the mechanism by which oncogenic MCF MLVs induce cell killing, we developed an in vitro culture system in which virus infection induces apoptosis in certain cell types (26). We recently demonstrated that killing of mink epithelial cells by MCF13 MLV involved the accumulation of high levels of the envelope (Env) precursor polyprotein, gPr80^env, which resulted in the induction of endoplasmic reticulum (ER) stress (14). The significance of this in vitro observation for the development of MCF MLV-induced lymphoma in AKR mice was examined in this study.

ER stress is induced when high levels of either misfolded or wild-type protein accumulate in the ER, resulting in the unfolded protein response (UPR) or ER overload response, respectively (8, 16). The UPR results in the upregulation of ER chaperone proteins, such as the glucose-regulated proteins (GRPs) of 58, 78, and 94 kDa (named GRP58, GRP78, and GRP94, respectively), which are involved in rescuing cells from ER stress (9, 11, 16, 22). The role of these chaperone proteins is to prevent protein aggregation and facilitate the exit of properly folded proteins from the ER. However, when ER stress is prolonged or severe, apoptotic pathways that are dependent on or independent of mitochondria can be activated (3, 22). The mechanism by which a cell either undergoes apoptosis or is rescued from ER stress is not well understood.

In this report, we demonstrate that ER stress occurs in thymic lymphocytes in mice inoculated with MCF13 MLV and correlates with the accumulation of the envelope precursor polyprotein. Furthermore, we observed that there are differences in the type and level of ER molecular chaperones in thymic lymphomas compared with preleukemic thymic lymphocytes. These results suggest a mechanism by which tumor cells can be rescued from virus-induced apoptosis.

Upregulation of GRP78 in thymic lymphocytes by MCF13 MLV infection.

Newborn AKR mice were inoculated intraperitoneally with 10⁶ infectious units of MCF13 MLV. Thymus tissue was removed at various times during the preleukemic period, i.e., before thymic lymphomas begin to appear at about 10 weeks after virus inoculation. Single-cell suspensions were prepared from thymus tissue from which protein extracts were subsequently isolated as done previously (27). We performed Western blot analysis of GRP78, GRP58, and GRP94, the ER chaperone proteins that we observed to be upregulated in MCF13 MLV-infected mink epithelial cells in response to the induction of ER stress (reference 14 and unpublished data). Our analysis revealed that GRP78 was upregulated in thymic lymphocytes isolated from virus-inoculated mice compared with control mice starting at approximately 4 to 5 weeks postinoculation (p.i.) and remained elevated at 8 weeks p.i. (Fig. 1A). As determined by densitometric analysis, GRP78 upregulation corresponded to a two- to fourfold increase over the levels in age-matched control mice (Table 1). All calculations for protein ratios between virus-inoculated and control mice include corrections for band intensities of α-tubulin loading controls. In contrast, significant differences in GRP58 and GRP94 levels between virus-inoculated and control mice were not detectable (Fig. 1B and C). Because the upregulation of GRP78 alone is an indicator of ER stress (9), we conclude that ER stress is induced in thymic lymphocytes in mice inoculated with MCF13 MLV.

FIG. 1.

Time course analysis of ER stress-associated chaperones in thymic lymphocytes after inoculation of MCF13 MLV into neonatal mice. Western blot analysis was carried out with protein extracts from thymic lymphocytes isolated from mice at different times after virus inoculation. C1 to C5, age-matched control mice inoculated with medium; M1 to M9, mice inoculated with virus at 2 to 4 days of age. ER chaperone proteins GRP78 (A), GRP58 (B), and GRP94 (C) were detected with specific antibodies (anti-GRP78, Santa Cruz Biotechnology, Inc.; anti-GRP58 and anti-GRP94, Stressgen Biotechnologies). α-Tubulin was detected as a control for loading.

TABLE 1.

Virus infection of thymic lymphocytes results in GRP78 upregulation

Time p.i. and mousea	Increase (fold)b in GRP78 level	% gp70+ cellsc
3 wk
C1		0
M1	NDd	1.1
4-5 wk
C2		0
C3		0
M2	3.1	35.5
M3	2.0	23.1
M4	4.2	29.3
M5	2.3	28.7
8 wk
C4		0.1
C5		1.5
M6	3.1	75.5
M7	2.7	75.0
M8	2.1	58.8
M9	2.2	62.6

^aNeonatal mice were inoculated with either 50 μl medium (C) or 10⁶ infectious units of MCF13 MLV (M). Numbers refer to different animals and correspond to those that were analyzed by Western blotting (Fig. 1). Thymic lymphocytes were isolated for analysis at the indicated number of weeks after inoculation.

^bIncrease in GRP78 for virus-inoculated mice compared with control mice by densitometric analysis of Western blot band intensities shown in Fig. 1.

^cPercentage of thymic lymphocytes expressing cell surface gp70 detectable by staining with MAb 83A25 and FACS analysis.

^dND, not detectable.

GRP78 upregulation correlates with virus infection and accumulation of envelope precursor polyprotein.

To determine whether the upregulation of GRP78 correlates with MCF13 MLV infection of thymic lymphocytes, we analyzed the cells used for the Western blots for cell surface envelope (Env) expression by fluorescence-activated cell sorter analysis (FACS) as described previously (27). Little or no MCF13 MLV envelope was detectable on thymic lymphocytes isolated from virus-inoculated mice at 3 weeks p.i. (Table 1). However, beginning at 4 to 5 weeks p.i., thymic lymphocytes expressing Env were detectable in virus-inoculated mice but not in control mice. This corresponded to the time when upregulation of GRP78 was detectable. However, although there was an approximately two- to threefold increase in percent virus-infected cells at 8 weeks p.i. over that at 4 weeks p.i., there was no comparable increase in GRP78 levels, suggesting that GRP78 is upregulated in a fraction of virus-infected cells. Nevertheless, the kinetics of the upregulation of GRP78 strongly correlates with that of MCF13 MLV infection of thymic lymphocytes.

Because our studies utilizing cultured mink epithelial cells demonstrated a strong correlation between ER stress and the accumulation of the envelope precursor polyprotein (14), we determined whether this also occurred in vivo. Western blotting was performed on protein extracts isolated from the thymic lymphocytes that were used to examine GRP78 and cell surface Env expression. To examine steady-state levels of envelope protein, we used the monoclonal antibody (MAb) 83A25, which detects Env of various MLVs (5). Our results indicate that the gPr80^env precursor was the major form of Env, with no detectable processed gp70 surface (SU) protein (Fig. 2A), which suggests that accumulation of the envelope protein occurs in the ER. Controls for gPr80^env and gp70 mobility are shown for protein extracts from productively infected mink cells. The faint band migrating between gPr80^env and gp70 may represent altered glycosylation of gPr80^env. These results support the idea that the accumulation of the envelope precursor in the endoplasmic reticulum of thymic lymphocytes results in the induction of ER stress.

FIG. 2.

Viral glycoprotein detection in thymic lymphocytes. Western blot analysis of protein extracts from preleukemic thymic lymphocytes at different times after virus inoculation (A) or from thymic tumors and a tumor-derived cell line (B). C1 to C7, age-matched control mice that were inoculated with medium. M1 to M9, mice inoculated with virus at 2 to 4 days of age. Mink, protein extracts from productively infected mink epithelial cells used as markers for the mobility of the Env precursor polyprotein gPr80^env and SU gp70. The ratio of these proteins in productively infected cells is different from that in acutely infected cells, in which gPr80^env predominates (14). T1 to T7, thymic lymphomas induced by inoculation of MCF13 MLV. T8 to T11, spontaneous tumors arising in uninoculated mice. 92316, T-cell line derived from a thymic lymphoma induced by MCF247 MLV. MCF MLV Env was detected with the MAb 83A25 (5). α-Tubulin was detected as a control for loading.

In a previous study, we demonstrated that there was a significant decline in thymus cellularity via apoptosis between 4 and 6 weeks after virus inoculation (27). It is notable that this decline begins at about the same time that Env precursor accumulation and upregulation of GRP78 are first detectable. Taken together, our results suggest that apoptosis of preleukemic thymic lymphocytes results from the induction of ER stress. This observation corroborates our previous results from virus infection of mink epithelial cells in which ER stress and apoptosis are induced (14, 26). We have shown for mink cells that superinfection by MCF13 MLV contributes to the high level of Env precursor accumulation, and we hypothesize that superinfection also occurs in AKR thymic lymphocytes.

Further increase in GRP78 and upregulation of GRP58 in thymic lymphomas.

We performed a similar analysis of thymic lymphomas that were generated by the inoculation of MCF13 MLV into neonatal mice. In the MCF13 MLV-induced tumors (T1 to T7) that were analyzed by Western blotting, we detected a 2- to 10-fold increase in GRP78 compared with thymuses from control mice (C1 and C2), as determined by densitometric analysis (Fig. 3A). To determine whether GRP78 was also upregulated in thymic lymphomas generated by other MCF MLVs, we analyzed spontaneous tumors (T8 to T11) induced by the de novo generation of MCF MLVs (15) and a cell line derived from a lymphoma induced by MCF247 MLV (92316, obtained from N. DiFronzo). In these tumors and this cell line, we detected a level of increase in GRP78 similar to increases induced by MCF13 MLV. Furthermore, the upregulation of GRP78 in nearly all of the lymphomas and the lymphoma-derived cell line was greater than that in preleukemic lymphocytes.

FIG. 3.

Detection of GRPs in thymic lymphoma cells. Western blotting was performed as described for Fig. 1. GRP78 (A), GRP58 (B), and GRP94 (C) were detected with specific antibodies. C1 and C2, age-matched control mice inoculated with medium; T1 to T7, thymic lymphomas induced by inoculation of MCF13 MLV into neonatal mice; T8 to T11, spontaneous tumors arising in uninoculated mice; 92316, T-cell line derived from a thymic lymphoma induced by MCF247 MLV. α-Tubulin was detected as a loading control.

In our analysis of GRP58, we detected a two- to ninefold upregulation of GRP58 in all of the thymic tumors and the cell line examined compared with control thymus cells (Fig. 3B). This was in contrast to preleukemic mice, in which no GRP58 upregulation was detectable (Fig. 1B). As in preleukemic thymic cells, however, upregulation of GRP94 was not detectable in the thymic tumors or cell line examined (Fig. 3C). Thus, our analysis of thymic tumors induced by different MCF MLVs indicates that a further increase in GRP78 levels compared with virus-infected preleukemic thymic lymphocytes and the upregulation of GRP58 commonly occur during tumorigenesis.

Western blot analysis of MCF MLV envelope protein showed that the precursor polyprotein was present at a four- to sevenfold excess over the gp70 SU form in most of the tumors induced by the MCF13 virus (T1 to T7) (Fig. 2B). Tumors T4 and T6 were exceptions, in which the amount of gp70 was about the same as or larger than the amount of gPr80^env. Nevertheless, significant amounts of gPr80^env were also present in these tumors. The Env precursor protein that was detectable with MAb 83A25 in the spontaneous tumors (T8 to T11) and the cell line 92316 migrated with lower mobility than the MCF13 gPr80^env protein, which could be due to a difference in protein size or glycosylation or both. We have thus observed that inefficient processing of the MCF MLV envelope precursor persists in frank lymphomas.

Our results support the idea that apoptosis of preleukemic thymic lymphocytes occurs as a result of ER stress that is induced by the accumulation of high levels of the MCF MLV envelope precursor polyprotein, similar to what we have observed for cultured mink epithelial cells. Upregulation of GRP78 occurs in preleukemic lymphocytes, which may contribute to their rescue from apoptosis. Studies have shown that GRP78 overexpression can rescue cells from apoptosis that is induced by different reagents, which may also contribute to tumorigenesis (10, 13, 18). Our results suggest that further thymic lymphoma progression requires an even greater increase in GRP78 levels and induction of GRP58, but not GRP94. This idea is consistent with studies showing that glucose-regulated proteins contribute to the growth and progression of other types of tumor (6, 7, 10).