P2 purinoceptors regulate calcium-activated chloride and fluid transport in 31EG4 mammary epithelia

It has been reported thatsecretory mammary epithelial cells (MEC) release ATP, UTP, and UDP uponmechanical stimulation. Here we examined the physiological changescaused by ATP/UTP in nontransformed, clonal mouse mammary epithelia(31EG4 cells). In control conditions, transepithelial potential (apicalside negative) and resistance were 4.4 ± 1.3 mV (mean ± SD, n = 12) and 517.7 ± 39.4  · cm², respectively. The apicalmembrane potential was 43.9 ± 1.7 mV, and the ratio of apicalto basolateral membrane resistance (R_A/R_B) was 3.5 ± 0.2. Addition of ATP or UTP to the apical or basolateral membranescaused large voltage and resistance changes with an EC₅₀ of~24 µM (apical) and ~30 µM (basal). Apical ATP/UTP (100 µM)depolarized apical membrane potential by 17.6 ± 0.8 mV (n = 7) and decreasedR_A/R_B by a factor of3. The addition of adenosine to either side (100 µM) hadno effect on any of these parameters. The ATP/UTP responses werepartially inhibited by DIDS and suramin and mediated by a transientincrease in free intracellular Ca²⁺ concentration (427 ± 206 nM; 15-25 µM ATP, apical; n = 6). This Ca²⁺ increase was blocked by cyclopiazonic acid, by BAPTA,or by xestospongin C. 31EG4 MEC monolayers also secreted or absorbedfluid in the resting state, and ATP or UTP increased fluid secretion by5.6 ± 3 µl · cm² · h¹(n = 10). Pharmacology experiments indicate that 31EG4epithelia contain P2Y₂ purinoceptors on the apical andbasolateral membranes, which upon activation stimulate apicalCa²⁺-dependent Cl channels and cause fluid secretion acrossthe monolayer. This suggests that extracellular nucleotides could playa fundamental role in mammary gland paracrine signaling and theregulation of milk composition in vivo.

     

ENaC- and CFTR-dependent ion and fluid transport in mammary epithelia

Mammary epithelial 31EG4 cells (MEC) were grown as monolayers onfilters to analyze the apical membrane mechanisms that help mediate ionand fluid transport across the epithelium. RT-PCR showed the presenceof cystic fibrosis transmembrane conductance regulator (CFTR) andepithelial Na⁺ channel (ENaC) message, and immunomicroscopyshowed apical membrane staining for both proteins. CFTR was alsolocalized to the apical membrane of native human mammary ductepithelium. In control conditions, mean values of transepithelialpotential (apical-side negative) and resistance(R_T) are 5.9 mV and 829  · cm², respectively. The apical membranepotential (V_A) is 40.7 mV, and the mean ratioof apical to basolateral membrane resistance (R_A/R_B) is 2.8. Apicalamiloride hyperpolarized V_A by 19.7 mV andtripled R_A/R_B. AcAMP-elevating cocktail depolarized V_A by 17.6 mV, decreased R_A/R_B by60%, increased short-circuit current by 6 µA/cm²,decreased R_T by 155  · cm², and largely eliminated responses toamiloride. Whole cell patch-clamp measurements demonstratedamiloride-inhibited Na⁺ currents [linear current-voltage(I-V) relation] and forskolin-stimulated Clcurrents (linear I-V relation). A capacitance probe methodshowed that in the control state, MEC monolayers either absorbed orsecreted fluid (2-4µl · cm² · h¹). Fluidsecretion was stimulated either by activating CFTR (cAMP) or blockingENaC (amiloride). These data plus equivalent circuit analysis showedthat 1) fluid absorption across MEC is mediated byNa⁺ transport via apical membrane ENaC, and fluid secretionis mediated, in part, by Cl transport via apicalCFTR; 2) in both cases, appropriate counterions move throughtight junctions to maintain electroneutrality; and 3)interactions among CFTR, ENaC, and tight junctions allow MEC to eitherabsorb or secrete fluid and, in situ, may help control luminal[Na⁺] and [Cl].

     

Assembly of gag-beta-galactosidase proteins into retrovirus particles.

We studied the expression of beta-galactosidase (beta-gal) and 15 gag-beta-gal fusion proteins in the presence of Moloney murine leukemia virus wild-type core (gag) proteins. Analysis indicated that proteins retaining the amino-terminal portion of gag through the capsid protein-coding region were incorporated into retrovirus particles. Proteins which deleted portions of the capsid protein were assembled into virions at low efficiency, indicating the importance of capsid protein interactions in retrovirus assembly. Fusion proteins which retained the amino-terminal matrix protein of the gag polyprotein but which lacked the capsid protein were released efficiently from cells in a nonviral form. The nonviral form was characterized by a high sedimentation coefficient and a low density, suggestive of membrane vesicles. While beta-gal was present in the cytoplasm of expressing cells, all fusion constructs were associated with cellular membranes. gag-beta-gal proteins which were capable of release from cells demonstrated a two-component immunofluorescence staining pattern consisting of a circle of fluorescence around the nucleus and a punctate pattern of staining throughout the remainder of the cell. Interestingly, fusions within the matrix protein were trapped intracellularly and yielded distinct perinuclear staining patterns, possibly localizing to the rough endoplasmic reticulum and/or Golgi. This observation suggests that Moloney murine leukemia virus gag proteins travel to the plasma membrane by vesicular transport associated with the cytoplasmic face of intracellular vesicles.