Proteinases of the mammary gland: developmental regulation in vivo and vectorial secretion in culture

Abstract

The extracellular matrix (ECM) is an important regulator of mammary epithelial cell function both in vivo and in culture. Substantial remodeling of ECM accompanies the structural changes in the mammary gland during gestation, lactation and involution. However, little is known about the nature of the enzymes and the processes involved. We have characterized and studied the regulation of cell-associated and secreted mammary gland proteinases active at neutral pH that may be involved in degradation of the ECM during the different stages of mammary development. Mammary tissue extracts from virgin and pregnant CD-1 mice resolved by zymography contained three major proteinases of 60K (K = 10(3) Mr), 68K and 70K that degraded denatured collagen. These three gelatinases were completely inhibited by the tissue inhibitor of metalloproteinases. Proteolytic activity was lowest during lactation especially for the 60K gelatinase which was shown to be the activated form of the 68K gelatinase. The activated 60K form decreased prior to parturition but increased markedly after the first two days of involution. An additional gelatin-degrading proteinase of 130K was expressed during the first three days of involution and differed from the other gelatinases by its lack of inhibition by the tissue inhibitor of metalloproteinases. The activity of the casein-degrading proteinases was lowest during lactation. Three caseinolytic activities were detected in mammary tissue extracts. A novel 26K cell-associated caseinase--a serine arginine-esterase--was modulated at different stages of mammary development. The other caseinases, at 92K and a larger than 100K, were not developmentally regulated. To find out which cell type produced the proteinases in the mammary gland, we isolated and cultured mouse mammary epithelial cells. Cells cultured on different substrata produced the full spectrum of gelatinases and caseinases seen in the whole gland thus implicating the epithelial cells as a major source of these enzymes. Analysis of proteinases secreted by cells grown on a reconstituted basement membrane showed that gelatinases were secreted preferentially in the direction of the basement membrane. The temporal pattern of expression of these proteinases and the basal secretion of gelatinases by epithelial cells suggest their involvement in the remodelling of the extracellular matrix during the different stages of mammary development and thus modulation of mammary cell function.

Fig. 1

Gelatinase activity at the various stages of mammary development. Extracts of freshly isolated mammary tissue from 11-to 12-week-old virgin (v), pregnant (p), lactating (L), and involuting (i) CD-1 mice and milk from 9-day lactating mice were assayed for gelatin-degrading activity by SDS–substrate gel zymography. The specific day within each stage of mammary development is indicated in arabic numerals. Lanes were loaded with 15 μg of protein per lane.

Fig. 2

Caseinase activity at the various stages of mammary development. Tissue samples, milk, conditions of the gel and all abbreviations are as described in the legend to Fig. 1.

Fig. 3

Quantitative analysis of caseinase activity across mammary gland development. Caseinolytic activity in whole tissue extracts from freshly isolated mammary glands from virgin, mid-pregnant (11–14 days), lactating, 1 day (1d Inv.) and 3 days (3d Inv.) involuting mice, was measured in a soluble [¹⁴C]casein-degradation assay. Values are the average caseinase activity of duplicate mammary tissue samples from two mice per stage of development.

Fig. 4

Schematic diagram of MMEC cultured on EHS-matrix. The cells aggregate into alveolar-like structures with their apical side toward the lumen and their basal side underlined by a basement membrane. Cells in this aggregate retract the EHS-matrix and are engulfed within it. The insert shows a Nomarski image of a thick transverse section of such an alveolus-like structure. With this cell arrangement, the cells secrete basally into the medium (basal compartment) and apically into the lumen (apical compartment). Diagram and inset courtesy of Sally Nash and Charles Streuli respectively.

Fig. 5

Effect of the ECM on the secretion of gelatinases by MMEC in culture. MMEC were grown on EHS-matrix, floating collagen (FG), or plastic under serum-free conditions, and conditioned medium was collected every 24 h and samples from day (d) 4 to 7 were analyzed by SDS–gelatin zymography. Lanes were loaded at equal cell number (5.5×10³ cells per lane).

Fig. 6

The distribution of the gelatinases and caseinases into the basal, apical and cellular compartments as resolved on SDS–gel zymograms. Conditioned medium (basal), EGTA-extractable fraction (apical), and cell fraction of MMEC cultured on EHS-matrix in serum-free medium were harvested on day 6 of culture and stored at −70°C. Lanes (basal and apical) were loaded with medium conditioned from 5.5×10³ cells. The cell fraction was derived from the same number of cells (5.5×10³ cells).

Fig. 7

Characterization of the gelatinases. Mammary tissue extracts from 13-day pregnant (pl3), 9-day lactating (L9), and 2- and 4-day involuting (i2 and i4 respectively) mammary gland were separated on SDS–gelatin gels and then incubated without (A) or with (B) TIMP at 10 μg ml⁻¹ in substrate buffer. (C) Activation of the 68K gelatinase after incubation of tissue extract with APMA. L9 and i4 tissue extracts were incubated for 16 h at ambient temperature in 1 mM APMA before loading on the gel. Lanes were loaded with 15 μg of protein.

Fig. 8

Caseinase activity in different tissues. Extracts of freshly isolated tissue from kidney (K), spleen (S), liver (L), heart (H), brain (B), and mammary gland (M) of 15 day pregnant CD-1 mice were assayed for casein-degrading activity by SDS–substrate gel zymography. Lanes were loaded with 15 μg of protein per lane.