Abstract
We generated monoclonal antibodies (MAbs) against mouse αs1- and β-casein and used them to survey casein immunochemically in mammary glands of mice at pericoitous and pregnant stages. Two MAb-producing hybridoma cells, designated MCα1 cell and MCβ1 cell, were established. Each antibody, when used in Western blotting, recognized specifically mouse αs1- and β-casein among a wide spectrum of proteins of both a lactating mammary homogenate and mouse skim milk. Immunohistochemistry revealed αs1- and β-casein in sections of lactating mammary glands. Staining was found in substances in the lumen and cytoplasm of duct and alveolar cells, particularly in rough endoplasmic reticulum and the Golgi apparatus. Mammary glands at Days 2, 4, 6, 8, and 14 of pregnancy showed positive staining specific to both αs1- and β-casein in the lumen and cytoplasm of duct cells, whereas the glands at estrus and Day 0 of pregnancy were positive mainly for αs1-casein. Semiquantitative Western blotting analysis of both casein components in epithelial cell fractions from glands during pregnancy confirmed that intra-epithelial αs1- and β-casein changed during three phases, elevated from trace levels to detectable levels during initial stages of pregnancy (Days 0, 2, and 4), declined to lower levels during mid-pregnancy (Days 6 and 8), and then rose to high levels during late pregnancy (Day 14).
THE FATE OF MAMMARY EPITHELIAL CELLS, including milk-secreting cells, is determined during embryogenesis in the mouse (Sakakura et al. 1976). Despite their determined fate, mammary epithelial cells in vivo need to experience pregnancy and parturition to differentiate functionally. A number of reports have dealt with mechanisms for induction of differentiation in mouse mammary epithelial cells. In particular, an approach using a primary cell culture technique and a radioimmunoassay for mouse γ-casein have confirmed that the combination of the three hormones, i.e., prolactin, insulin and glucocorticoids, stimulates casein secretion by the cells of mice in mid-pregnancy (Emerman et al. 1977). However, it is reported that mammary glands of mid-pregnant mice contain casein, whereas those of mature virgin mice rarely do (Smith and Vonderhaar 1981). Therefore, it was not clear whether these three hormones are truly sufficient for inducing differentiation in mammary epithelial cells. Recently, we have demonstrated, using a serum-free cell culture system and Western blotting, that mammary epithelial cells of cycling virgin mice secrete casein in response to these three hormones and that the hormone requirement for αs1-casein is different from that for other major casein components (unpublished observations). These observations imply that mammary epithelial cells of virgin mice are as capable of differentiating as are those of pregnant mice and that αs1-casein might be expressed differentially in vivo. However, our studies and those of others concerning casein synthesis have used polyclonal antibodies, and it is therefore difficult to exclude the possibility of detecting crossreactive noncasein proteins and to examine each casein component selectively. In the present study we generated two monoclonal antibodies (MAbs) against mouse αs1-and β-casein and have surveyed each casein in mammary glands of mice at peri-coitous and pregnant stages using IHC to clarify the timing of expression of each casein component. Furthermore, we have quantified casein components in the mammary epithelium by Western blotting to assess differentiated function during pregnancy.
Materials and Methods
Animals
Inbred KA strain mice, aged 10 to 24 weeks, were used for collecting milk and mammary glands. Sprague-Dawley strain rats and Balb/c strain nu/nu mice, used for generation of MAbs, were purchased from Charles River Japan (Tokyo, Japan). These animals were maintained with MF chunk diet (Oriental Yeast; Tokyo, Japan) and water ad libitum in a biohazard protection condition, which conformed to the Ethics Review Committee for Animal Experimentation of the University of Tokyo. The estrous cycle of the mice was determined by vaginal smear observation. The days of vaginal plug and parturition were counted as Day 0 of pregnancy and Day 1 of lactation, respectively.
Preparation of MAbs Against Mouse Casein Components
Mouse whole casein was collected and polymerized using glutaraldehyde as described previously (Kanazawa et al. 1999). Polymerized mouse whole casein (2 mg) was emulsified with Freund's complete adjuvant. To immunize, an 8-week-old rat was anesthetized with ether and was palpated for popliteal lymph nodes. The emulsified antigen was injected into the popliteal lymph nodes using an injection needle. The same antigen, emulsified with Freund's incomplete adjuvant, was booster-injected three times at 2-week intervals into the lymph nodes and subcutaneous tissues of the same rat. On the fifth day after the last immunization, spleen cells of the immunized rat were collected and were fused with mouse myeloma cells, AgX63-6.5.3., using polyethylene glycol (BDH; Liverpool, UK) as a fusion agent. After selection in medium supplemented with hypoxanthine, aminopterin, and thymidine (HAT; Sigma, St Louis, MO), hybridoma cells producing antibodies against mouse casein components were screened by ELISA with an indirect method using polymerized mouse whole casein as an adsorbed antigen. Positive cells were further analyzed for their specificity to casein components by Western blotting using non-polymerized mouse whole casein as a standard. Antibody-producing hybridoma cells were cloned by the limiting dilution method, using rat thymus cells as feeder cells. Immunoglobulin subclass was determined by using a rat immunoglobulin subclass typing kit (The Binding Site; Birmingham, UK). To induce ascites, clonal hybridoma cells were injected into the peritoneal cavity of nude mice that had previously been injected with 2,6,10,14-tetramethylpentadecane (Pristane; Aldrich, Madison, WI). Immunoglobulins in ascites were precipitated by the addition of ammonium sulfate and were further purified by anion-exchange column chromatography as described (Bazin et al. 1986). Each purified antibody was analyzed by SDS-PAGE under both reducing and non-reducing conditions.
Western Blotting for Mouse Casein Components
A lactating mammary homogenate was prepared. A lactating mouse at Day 10 of lactation was separated from suckling pups immediately before sacrifice. The mammary tissues were removed and homogenized in 10 vol of ice-cold PBS, pH 7.4, containing 0.25 M sucrose, 1 mM phenylmethylsulfonyl fluoride (Nacalai; Kyoto, Japan) and pepstatin A (2 μg/ml; Sigma), and were centrifuged at 3000 × g for 30 min at 4C. The supernatant was collected (mammary homogenate). Epithelial cells were collected from mammary glands of mice by collagenase digestion, as previously described (Kanazawa et al. 1999). Cells were counted on a hemocytometer and 1 × 108 cells were lysed in 1 ml of a cell extraction buffer consisting of 20 mM HEPES (pH 7.2), supplemented with protease inhibitor cocktail (Roche Diagnostics; Mannheim, Germany). The supernatant was collected (cell extract) after centrifugation for 20 min at 12,000 × g at 4C. Tissue homogenate, cell extract, and mouse skim milk were resolved by SDS-urea-PAGE and the resolved proteins were transferred onto filter membranes by the procedure described previously (Kanazawa et al. 1999). The filter was blocked for 1 hr in 1% gelatin (EIA grade; Bio-Rad, Hercules, CA) in PBS. To detect casein with polyclonal antibodies (Kanazawa et al. 1999), the blocked membrane was incubated for 1 hr with rabbit anti-mouse whole casein serum (1:500 dilution with 0.5% gelatin in PBS). The filter was washed with PBS and then was incubated for 1 hr with peroxidase-conjugated goat anti-rabbit IgG (1:1000 dilution in 5% goat serum in PBS; Cappel, Malvern, PA). Peroxidase activity was detected by color reaction using 4-chloro-1-naphthol (Nacalai) as substrate. To detect casein components with MAbs the blocked membrane was incubated first with anti-mouse casein rat MAb (1 μg/ml in PBS-0.5% gelatin) for 1.5 hr, second with goat anti-rat IgG (10 μg/ml; TAGO, Camarillo, CA) for 1.5 hr, and third with rat PAP (20 μg/ml; Serotech, Oxford, UK) for 1 hr. Peroxidase localization was visualized by the same color reaction as described above. After the color reaction, optical densities of colored bands were quantified as previously described (Kanazawa et al. 1999), using an Imaging Densitometer (Model GS-670; Bio-Rad) with Molecular Analyist Software (ver. 2.1). After subtraction of background values, optical densities of each band were calculated as values of 0.1 μg mouse skim milk.
Immunohistochemical Detection of Casein Components
Inguinal mammary glands (fourth glands) were removed from mice at various reproductive stages. Mammary glands at Day 10 of lactation were removed immediately after separation of the mother from suckling pups. The tissues were fixed in Bouin's solution for 1 hr and then embedded in paraffin using standard procedures. Tissue sections of 1–2 μm were dewaxed in xylene and then treated with 0.3% hydrogen peroxide in methanol for 30 min at room temperature (RT) to inactivate endogenous peroxidase activity. Sections were rehydrated in a descending series of ethanol and blocked in 1% gelatin in PBS (pH 7.4). Sections were then incubated with rat anti-mouse casein MAbs (1 μg/ml in blocking solution) for 1.5 hr, then with goat anti-rat IgG (10 μg/ml in blocking solution; TAGO) for 1.5 hr, and finally with rat PAP (20 μg/ml in blocking solution; Serotech) for 1 hr. The localization of peroxidase was detected by color reaction for 5 min at RT using 4,4′-diaminobenzidine tetrahydrochloride (Dojin; Osaka, Japan) as substrate. As controls, primary antibodies were substituted with normal rat IgG. When a mixture of the two types of anti-mouse casein MAbs was used as primary antibody, peroxidase-conjugated goat anti-rat IgG (10 αg/ml; TAGO) was used as the second antibody and staining was performed as described above.
Results
Production and Characterization of MAbs Against Mouse Casein Components
Antibody-producing hybridoma cells were screened by ELISA and two hybridoma clones, MCα1 and MCβ1, were established after three sequential clonings by the limiting dilution method. Each hybridoma clone successfully produced ascites antibody when injected into abdominal cavities of nude mice. MCα1 and MCβ1 produced rat IgG2a and IgG1, respectively, by typing based on Ouchterlony's gel double-diffusion method (data not shown). These two types of antibodies, purified from ascites by salting-out with ammonium sulfate and anion-exchange column chromatography, ran as single bands on SDS-PAGE under non-reducing conditions (data not shown). SDS-PAGE under reducing conditions revealed that MCα1 and MCβ1 are composed of heavy chains of similar molecular weights (52 kD) and distinctive light chains (29 kD and 32 kD, respectively) (data not shown).
Western blotting was performed to identify antigens recognized by these antibodies (Figure 1). As shown in our previous report (Kanazawa et al 1999), rabbit anti-mouse whole casein antiserum selectively recognized specific casein components, i.e., αs1-, β-, γ-casein, and other minor components, in mouse skim milk and a lactating mouse mammary gland homogenate (Figure 1, Lanes 6 and 7). Rat MAbs specifically recognized a single class of mouse casein component among many classes of proteins: MCα1 and MCβ1 stained αs1- and β-casein, respectively, in a homogenate of a lactating mammary gland as well as in mouse skim milk (Figure 1, Lanes 4, 5, 8, and 9). Two faint bands that ran above and below the major band of αs1-casein are presumably hyperphosphorylated and dephosphorylated products, respectively (Rocha et al. 1985). A faint band below the major β-casein band in mouse skim milk is also a dephosphorylated product of β-casein.
Western blotting analysis of anti-mouse casein rabbit antiserum and rat MAbs. Protein standard (Lane 1), a homogenate of lactating mammary glands (Lanes 2, 4, 6, and 8) and mouse skim milk (Lanes 3, 5, 7, and 9) were separated by SDS-urea-PAGE and transferred onto a membrane as described in Materials and Methods. Proteins were stained with either amido black (Lanes 1–3), anti-mouse whole casein rabbit serum (Lanes 6 and 7), anti-mouse αs1-casein rat MAb (Lanes 4 and 5), or anti-mouse β-casein rat MAb (Lanes 8 and 9).
Immunohistochemical Survey of Case in Components in Mouse Mammary Glands at Various Reproductive Stages
In the tissue sections prepared from lactating mammary glands, positive staining for both αs1- and β-casein was observed when MCα1 and MCβ1 were included in the first step of antibody incubation (Figures 2A and 2B). Staining was localized both on the granular substances in the lumen (asterisks) and in most cells surrounding the lumen of the glands, including alveoli and small ducts (small arrowheads) and large ducts (large arrowheads). In contrast, no positive staining was observed when primary antibody was excluded from the immunoreaction or when primary antibody was substituted with normal rat IgG (Figure 2C). The PAP method, in general, produces more sensitive staining but also tends to produce more diffuse staining. Therefore, we next examined localization of casein components in the casein-positive cells by a standard indirect method using a mixture of MCα1 and MCβ1 as a primary antibody and peroxidase-conjugated goat anti-rat IgG as the secondary antibody (Figure 2D). The strongest positive staining in the cytoplasm localized predominantly in a particular area between the apical region and nucleus (small arrows), which corresponds to endoplasmic reticulum, the Golgi complex, and/or secretory vesicles (Burgoyne and Duncan 1998). The apical regions of cytoplasm (large arrows) were less intensely stained.
It is important to clarify the timing of casein synthesis in vivo to understand the mechanism of differentiation of mammary epithelial cells. We therefore surveyed casein components in sections prepared from mammary glands at estrus of cycling mice and at Days 0, 2, 4, 6, 8, and 14 of first pregnancies. The PAP method, using MCα1 and MCβ1 as primary antibodies, produced αs1- or β-casein-specific positive staining at Days 2, 4, 6, 8, and 14 of first pregnancies (Figure 3). Positive staining was obvious both in the apical region of the luminal cytoplasm and in lumens of the glands at Days 2, 4, and 14, but the staining was more restricted to glandular alveoli and was less intense in the glands at Days 6 and 8. Granular substances in the lumens of alveoli and ducts were strongly stained, whereas the cytoplasm of luminal and duct cells was weakly or diffusely stained. In contrast to the glands at Day 2 or later, mammary glands at estrus and at Day 0 of pregnancy exhibited clearly positive staining for αs1-casein but only faintly positive staining for β-casein. Although casein-specific positive staining was detected, few distended lumens were observed in the glands of pregnant mice, which was in contrast to the lactating glands.
Immunohistochemical detection of casein in lactating mouse mammary gland. Mammary glands were removed from a lactating mouse, fixed in Bouin's fixative, and embedded in paraffin. Sections at 1–2 μm were immunostained for casein by the PAP method (A-C) or by the standard indirect method (D). Anti-mouse αs1-casein rat MAb (A), anti-mouse β-casein rat MAb (B), normal rat IgG (C), a mixture of antimouse αs1-casein rat MAb and anti-mouse β-casein rat MAb (D) were used as primary antibodies, respectively, and counterstained with hematoxylin. Note that luminal epithelial cells are least distended because the tissues were taken from a mouse immediately after separation from suckling pups. Bars = 20 μm.
Immunohistochemical detection of casein in mammary glands of mice at estrus and pregnancy. Mammary glands removed from mice at estrus (A,B), Day 0 (C,D), Day 2 (E,F), Day 4 (G,H), Day 6 (I,J), Day 8 (K,L), and Day 14 (M,N) of first pregnancies were fixed in Bouin's fixative and embedded in paraffin. Sections of 1–2 μm were prepared and immunostained for casein with the PAP method as described in Materials and Methods. Anti-mouse αs1-casein rat MAb (
Western Blotting Survey of Casein Components in Mammary Epithelial Cells from Mice at Various Stages of Pregnancy
To confirm the casein components in mammary epithelial cells of mice at early stages of pregnancy, Western blotting was employed. Epithelial cell fractions were collected from mammary glands of individual mice at days 0, 2, 4, 6, 8, 10, 12, and 14 of first pregnancies and the respective cell extracts prepared from the same number of cells were subjected to SDS-urea-PAGE and were analyzed for casein by Western blotting using the PAP method. As shown in Figure 4, both αs1- and β-casein were detected in cells during pregnancy (data on cells at Days 10 and 12 are not shown). Intensities of both αs1- and β-casein bands at Days 0, 2, 6, and 8 were very faint, presumably due to lower sensitivity compared to the present IHC, although densitometrical scanning detected each casein band in these samples. Major casein bands exhibited the same mobility as did those from mouse whole casein. Minor bands below the major band were presumably due to degraded products of casein generated during the cell preparation procedure, because these bands were not detected in the homogenate samples as shown in Figure 1, Lanes 4 and 8. Densitometrical quantification of the casein bands revealed that intracellular casein components changed in good correlation with the results from IHC. Although intensities of both αs1- and β-casein bands at Day 0 were low as trace levels, both of the caseins were elevated from Day 0 to Days 2 and 4, decreased from Day 4 to Days 6 and 8, and then again elevated at Day 14 (Table 1). These three phases of change were statistically significant (Student's t-test, p<0.05). The relative value of αs1-casein tended to be higher than that of β-casein at Days 0 and 2 of pregnancy. This correlated well with the results from IHC showing that sections at Days 0 and 2 of pregnancy exhibited clearly positive staining for αs1-casein but only faintly positive staining for β-casein.
Western blotting analysis of casein components in mammary epithelial cells of mice at various stages of their first pregnancies. Cell extracts of mammary epithelial cells (1.5 × 106/lane), collected respectively from two individual mice at Day 0 (Lanes 1 and 2), Day 2 (Lanes 3 and 4), Day 4 (Lanes 5 and 6), Day 6 (Lanes 7 and 8), Day 8 (Lanes 9 and 10), and Day 14 (Lanes 11 and 12), were subjected to SDS-urea-PAGE and then casein components were detected by immunochemical colorization. Anti-mouse αs1- (A) and β-casein (B) MAbs were used respectively as primary antibodies of the PAP method as described in Materials and Methods.
Changes in casein components in mammary epithelial cells of mice at various stages of pregnancy a
Cell extracts, prepared from the same number (1.5 × 106 cells/lane) of mammary epithelial cells isolated from mice at indicated days in first pregnancy were analyzed for casein components by Western blotting, as shown in Figure 4. Colored bands corresponding to αs1- and β-casein were densitometrically analyzed as described in Materials and Methods. Data are expressed as mean ± SD (n=6) from three independent experiments, as multiples of the value of 0.1 μg mouse skim milk.
Discussion
We have generated anti-mouse αs1- and β-casein MAbs and have clearly demonstrated that both αs1-and β-casein components are present in mammary epithelium of mice at the estrous stage of cycling virgins and throughout stages of pregnancy.
It has been reported that primary mammary cells dissociated from unprimed mature mice require 3 to 4 weeks of preculture in medium (horse serum, FBS, and insulin) to synthesize casein in response to a lactogenic combination of hormones (insulin, prolactin, cortisol, and aldosterone) (Tonelli and Sorof 1982; Durban et al. 1985). However, we have recently observed that mammary epithelial cells isolated from cycling virgin mice can synthesize all the major casein components within 24 hr after similar hormone stimulation added on the second day to primary culture in a serum-free medium (Kanazawa and Kohmoto unpublished observations). This result suggests that mammary epithelial cells of cycling virgin mice are similar to those of pregnant mice in terms of the potential to synthesize casein in response to hormones. It has been reported, however, that casein mRNA reaches a detectable level only during mid-pregnancy in mice (Robinson et al. 1995). Similarly, casein-specific antigens are detected in mammary epithelium of mice at mid-pregnancy or later stages but are barely detected in virgin mice using an anti-mouse casein polyclonal antibody (Smith and Vonderhaar 1981; Barash et al. 1995). Because these earlier studies used a polyclonal antibody to probe casein, it is difficult to exclude the possibility of detecting crossreactive unrelated antigens. In contrast, the two MAbs generated in the present study were characterized as homogeneous IgG proteins distinctive from one another by electrophoresis and Ouchterlony's double-gel immunodiffusion tests. Moreover, these two MAbs were proved to be monospecific to αs1- and β-casein components by Western blotting, which enabled us to examine each casein component rigorously. Furthermore, we purified these antibodies before use as probes for casein. These technical improvements make the present results more convincing.
We have further examined mammary glands around coitous stages, using the PAP method, which is much more sensitive than the standard indirect method used for IHC in earlier studies (Smith and Vonderhaar 1981; Barash et al. 1995). In an earlier report, it was speculated that mammary epithelial cells synthesize casein during pregnancy in vivo and that certain differences in differentiative potential may exist in mammary epithelial cells of cycling virgin mice (Durban et al. 1985). Nevertheless, the present study has clearly demonstrated that casein is synthesized during the estrous stage of cycling virgin mice and the initial stage of pregnancy. The present data imply that mammary epithelial cells differentiate during the cycling virgin stage to synthesize casein. Moreover, the present results lead to the conclusion that differentiation of mammary epithelial cells does not entail qualitative changes but rather quantitative changes in differentiation-related functions of the relevant cells. The amount of casein synthesized in a certain number of cells, as assessed by the present Western blotting, underwent a change during the three phases of pregnancy. Intracellular casein was elevated from a trace level to a detectable level during the initial stage of pregnancy (Days 0, 2, and 4), declined to a lower level during early to mid-pregnancy (Days 6 and 8), and rose to a high level during late pregnancy (Day 14). This mode of change correlates with changes in intensity of casein-specific staining on tissue sections. These three phases of change in the amount of intracellular casein are presumably due to the hormonal conditions of mice and responsiveness of the cells to hormones and growth factors as described (Vonderhaar and Ziska 1989; Groner et al. 1994).
The present IHC study has also revealed that casein components are present in the cytoplasm of almost all luminal epithelial cells of lactating glands, including those lining large ducts. This observation contradicts an earlier observation that casein-specific staining localizes only to cells of alveoli and small ducts (Smith and Vonderhaar 1981). In general, positive IHC cannot exclude the possibility of translocation of the relevant antigens in vivo and during histological procedures and therefore does not directly prove synthesis at the sites of the signals. Nevertheless, the positive staining detected in the present study was localized predominantly to particular regions of the cytoplasm of the cells lining large ducts as well as those of alveoli or small ducts. Therefore, it is likely that such cells of large ducts also synthesize casein, as do alveolar epithelial cells. Moreover, the regions correspond to endoplasmic reticulum, the Golgi complex, and/or secretory vesicles (Burgoyne and Duncan 1998) gave the strongest and most dense staining for casein in the cytoplasm, whereas the apical region of secretory cytoplasm was weakly stained. According to Turner et al. (1992) there are two pathways for casein secretion by acinar mammary epithelial cells of the mouse, a Ca2+-dependent or -regulated pathway and a Ca2+-independent constitutive pathway. In the present study, lactating mammary glands were taken from mice immediately after separation from the pups. In addition, the present two MAbs apparently bind both micellary and non-micellary casein components. Therefore, the present observations may indicate that the last step of casein secretion under a physiologically suckled condition, defined as after the formation of casein secretory vesicles till exocytosis of the casein from the vesicles, occurs within a short time.
In conclusion, the present IHC study, using newly generated anti-mouse αs1- and β-casein MAbs, demonstrated that the casein level in mammary epithelium changed in three phases, by elevating from trace to a detectable level during initial stages of pregnancy (Days 0, 2, and 4), declining to a lower level during mid-pregnancy (Days 6, 8, and 10), and then rising to a high level during late pregnancy (Days 14 and later). Our present study also demonstrated both casein components in duct epithelium, which suggested the potentially similar ability of luminal epithelial cells in the mammary glands to synthesize casein.