Interleukin-4 up-regulates mouse mammary tumor virus expression yet is not required for in vivo virus spread

Abstract

The mouse mammary tumor virus (MMTV) superantigen induces T-cell production of cytokines, such as interleukin-4, which in turn increase MMTV transcription. However, interleukin-4 is not required for in vivo virus spread, because mice lacking interleukin-4 or the STAT6 transcription factor showed wild-type infection of lymphoid and mammary tissue. In spite of this, mammary tumor incidence was decreased in STAT6 null mice.

FIG. 1

Cytokine induction of endogenous MMTV transcripts. (A) RNase protection of RNA from BALB/c splenocytes cultured in medium alone (−) or with IL-2, LPS, or ConA. (B) RNase protection of RNA from BALB/c splenocytes cultured as for panel A in the presence of IL-2, IL-1, or LPS. + cont, positive control mammary gland RNA.

FIG. 2

IL-4 induction of MMTV RNA in BALB/c, IL-4^−/− (top), and STAT6^−/− (bottom) mice. Although the RNA sample for untreated IL-4^−/− splenocytes (−) was slightly degraded, there is clear induction of Mtv-9 by IL-4 treatment (compare IL-2 and IL-4 treatment). C, BALB/c RNA control for the β-actin reaction.

FIG. 3

IL-2 and IL-4 induction of LTR-CAT transgene in splenocytes. (A) Representative CAT assay from splenocytes isolated from one mouse. (B) Three LTR-CAT transgenic mice were sacrificed, and their splenocytes were individually cultured with medium alone, IL-2, or IL-4. The conversion of chloramphenicol to the acetylated form was averaged for each condition. P values were calculated using Student's t test, compared to controls with medium alone.

FIG. 4

Peripheral cognate T-cell deletion in IL-4^−/− and STAT-6^−/− mice. Data are representative of two to nine mice per group.

FIG. 5

MMTV infection of the mammary gland in STAT-6^−/− and IL-4^−/− mice. RNase protection was performed on RNA isolated from the milk of five STAT6^−/−, five IL-4^−/−, and three BALB/cJ mice foster nursed on MMTV(LA)-infected C3H/HeN mothers. Densitometry analysis was performed on the autoradiographs using Gel Doc 1000 with a conversion screen (Bio-Rad Laboratories, Hercules, Calif.), and the MMTV signal was normalized to the β-actin signal. Shown below each group are average normalized signals with standard deviations.

FIG. 6

Mammary gland tumor incidence in BALB/c, IL-4^−/−, and STAT6^−/− mice. Box plots of the time to tumor development are shown. Middle bars show medians, the top and bottom edges of the boxes show 75th and 25th percentiles, and the top and bottom bars show minimum and maximum values. A nonparametric Mann-Whitney test was used to generate the statistics using SPSS (SPSS, Inc., Chicago, Ill.). The P values are two-sided compared to BALB/c.