During the peripartum period, mammary parenchyma undergoes intensive growth and remodelling (Knight & Wilde, Reference Knight and Wilde1993) and the vascular system is expanded to form a dense network of capillaries around the secretory and myoepithelial cells (Cvek et al. Reference Cvek, Ridderstråle and Dahlborn1998). As milk production ceases in late lactation, both secretory tissue (Anderson et al. Reference Anderson, Harness, Snead and Salah1981) and the capillary network regress (Tatarczuch et al. Reference Tatarczuch, Philip and Lee1997; Cvek et al. Reference Cvek, Ridderstråle and Dahlborn1998). Mammary blood flow increases markedly at the onset of lactation, and in ruminants mammary blood flow is closely correlated to milk yield during lactation and somatotropin treatment (Linzell, Reference Linzell, Larson and Smith1974; Nielsen et al. Reference Nielsen, Schleisner, Jakobsen and Andersen1995). Milk production and hence metabolic activity of the mammary gland may therefore be closely related to development and functionality of the mammary microvascular system. Metabolic activity of the mammary epithelial cells is associated with release of CO2 in proportion to milk production (Guinard-Flament et al. Reference Guinard-Flament, Delamaire, Lamberton and Peyraud2007). The local blood perfusion is important for provision of nutrients and for elimination of metabolic waste products, and it can be regulated in accordance with metabolic activity in body tissues through release of vasoactive substances (Prosser et al. Reference Prosser, Davis, Farr and Lacasse1996) such as vasodilatory CO2. Nielsen et al. (Reference Nielsen, Schleisner, Jakobsen and Andersen1995) demonstrated that changes in mammary metabolic activity and CO2 production during lactation in dairy goats were reflected in changes in HCO3− in the venous blood leaving the mammary gland. CO2 is converted to HCO3− in a reversible reaction: CO2+H2O↔H2CO3↔H++HCO3−, catalysed by the enzyme carbonic anhydrase (CA; EC 4.2.1.1).
Carbonic anhydrase is widely distributed in mammalian tissues, and is involved in local regulation of pH in body tissues and in the diffusion of CO2 from cell to blood. There are 16 known isozymes of CA, of which some are soluble, some are membrane bound and some are secreted (Krishnamurthy et al. Reference Krishnamurthy, Kaufman, Urbach, Gitlin, Gudiksen, Weibel and Whitesides2008; Supuran, Reference Supuran2008). In the lactating goat mammary gland, CA is located exclusively in the capillary membranes (Cvek et al. Reference Cvek, Ridderstråle and Dahlborn1998) and is most likely identical to the membrane bound isozyme CA IV, as in capillaries of other organs (Ridderstråle, Reference Ridderstråle1976, Reference Ridderstråle, Dodgson, Tashian, Gros and Carter1991). In lactating goats, CA can thus be used as a capillary marker (Cvek, Reference Cvek1997).
Cvek et al. (Reference Cvek, Ridderstråle and Dahlborn1998) followed goats in the peripartum period and found that when one mammary gland was dried off before parturition, CA activity gradually disappeared in the gland not milked, but remained unchanged in the milked gland. This observation confirms that CA activity in mammary gland is closely related to milk production. The number of mammary capillaries stained for CA changed during lactogenesis, and CA activity increased in parallel to the rise in milk production. The authors therefore suggested that CA is a prerequisite for milk secretory processes rather than for tissue growth, since CA staining was scarce during periods with extensive secretory tissue growth (i.e. during the dry period).
We hypothesized that the capillary network and the capillary enzyme CA are important regulatory factors involved in coordinating mammary blood flow according to changes in mammary metabolic activity and hence milk yield during established lactation. The main objective of the present study was to study morphological changes in the capillary network and CA activity during lactation and after induction of changes in mammary metabolic activity by 14 d of bovine somatotropin (BST) treatment in mid-lactation, and to relate these to changes in mammary blood flow.
Materials and Methods
Experimental animals and feeding
All animal experimental procedures were conducted under protocols approved by the Danish Animal Experiments Inspectorate and complied with the Danish Ministry of Justice Law no. 382 (10 June 1987) and Acts 739 (6 December 1988) and 333 (19 May 1990) concerning animal experimentation and care of experimental animals. Six Danish Landrace goats (parity 3–5) were used. They were surgically prepared with both superficial epigastric caudal veins (milk veins) exteriorized in skin-covered loops. Goats were housed in individual pens, fed hay ad libitum and concentrate (barley, molasses and a commercial mix) according to Danish feeding standards and had free access to water. They were fed and milked twice daily at approx. 8.15 and 17.00, half the concentrate and hay ration being given at each milking.
Experimental procedures
The experiment was conducted during months 1–9 of lactation, and none of the goats were pregnant. As shown in Table 1, recordings and samplings were performed at six different time points during lactation: in early lactation (EL; between weeks 3 and 6 for individual goats), midlactation (ML1; between weeks 11 and 16 for individual goats; BST treatment was initiated the day after this biopsy sampling), by the end of a 14-d period of BST treatment (BST), 14 d after cessation of the BST treatment (ML2), late lactation (LL, between weeks 28 and 31 for individual goats) and by the end of lactation when milk yield dropped to <1 kg/d and they were about to be dried off (D; between weeks 37 and 39 for individual goats).
Table 1. Schedule for biopsy sampling, milk vein blood velocity (MBV) measurements, and bovine somatotropin (BST) treatment
† EL: Early lactation, ML1: Mid lactation prior to BST treatment, BST: mid lactation by the end of a 14-d BST treatment period, ML2: mid lactation 14 d after cessation of BST treatment, LL: Late lactation, D: Late lactation just before drying-off
BST treatment consisted of daily subcutaneous injections for 14 d of 10 mg Somidobove® (Lot no. 505 ALO, Lilly Research Centre Ltd., Windlesham, Surrey, GU20 6PH, UK) dissolved in saline. Injections were given subcutaneously in the shoulder region at 9.30. BST treatment was initiated the day after recordings and samplings in period ML1.
Cross-sectional areas of milk veins remain constant during lactation in multiparous goats, and changes in milk vein blood velocity (MBV) were determined as a measure of changes in mammary blood flow (MBF) (Christensen et al. Reference Christensen, Nielsen, Bauer and Hilden1989). MBV was determined in both milk veins in the morning (approx. 09.00) and afternoon (approx. 14.00) by the ultrasound Doppler technique. Averages of these four measurements were used in the statistical analyses. Vena pudenda externa was manually clamped during MBV measurements to reduce risk of drainage of blood from non-mammary origin through the milk veins.
Two mammary biopsies were collected from each mammary gland the day after MBV measurements were performed. Biopsies were sampled from the same spot on the mammary gland, approx. 6–8 cm from the base of the udder and approx. 2–3 cm from the midline at the point where the gland protrudes the most caudally. Biopsies in the BST period were collected the day after cessation of BST treatment. Prior to biopsy sampling, the mammary gland was milked, the biopsy area was disinfected and locally anaesthetized by subcutaneous administration of 20 mg lidocainhydrochloride (Lidokain, Sygehus apotekerne i Danmark, Denmark). Fine-needle biopsies were sampled at a depth of 2–3 cm with a 17G biopsy needle using a biopsy gun (Biopty®, Radiplast, 750 07 Uppsala, Sweden). Each collected biopsy was approximately 10–15 mm3. After biopsy sampling, the site of incision was disinfected and sealed with a wound spray, and any accumulated blood removed by milking. Biopsies were immediately frozen in liquid N2 and stored at −80°C pending analyses.
Level of milk production in each sampling period was expressed as average milk yield over the last 3 d prior to biopsy sampling.
Histological analyses
Biopsies were fixed in phosphate-buffered 2·5% glutaraldehyde (pH 7·2) for 6 h, subsequently rinsed in phosphate buffer (pH 7·2) and dehydrated through graded ethanols and embedded in water-soluble resin (Historesin®, Leica Instr., 69115–69126 Heidelberg, Germany). CA was localized using the cobalt precipitation method of resin embedded tissue described by Ridderstråle (Reference Ridderstråle1976, Reference Ridderstråle, Dodgson, Tashian, Gros and Carter1991), which results in a black precipitate at sites of enzyme activity, including all CA isozymes. From each biopsy, 2-μm thick sections were cut and subsequently incubated for 6 min floating on the surface of an incubating medium containing 157 mm-NaHCO3, 3·5 mm-CoSO4, 11·7 mm-KH2PO4, and 52·6 mm-H2SO4. After incubation, the sections were rinsed in 0·67 mm-phosphate buffer (pH 5·9) for 1 min, treated with 0·5% (NH4)2S for 3 min, and finally rinsed in two successive baths of distilled water, 1 min in each. Controls were run with the specific CA inhibitor acetazolamide (10−5m) in the incubation medium. Prior to mounting, some sections were weakly counter-stained with azure blue to stain alveoli for image analyses.
Image analysis
Photomicrographs were taken of the azure blue counter-stained sections using the Nikon Microphot-FXA imaging system, and image analysis performed using the TEMA image analysis data software (Bio-Rad Scan Beam A/S, 9560 Hadsund, Denmark). The images analysed were placed successively to cover the entire biopsy but avoiding the disrupted structures at the edges of the biopsy. Azure blue stains the different structures in the mammary gland. Thereby the alveoli could be outlined in the computer. The blackening of the capillaries and the azure blue staining were used to outline the capillaries for calculation of capillary area. The capillaries were manually marked with an individual click of the computer mouse on each capillary. The image analysis program then calculated the number of capillaries and alveoli, capillaries per alveolus and mean alveolar and capillary area (μm2). The mean capillary area was calculated by the program as total area covered by capillaries in the image divided by number of capillaries. At least 40 alveoli were analysed from each section, except in a few biopsies containing less secretory tissue. The procedure is described in detail by Cvek et al. (Reference Cvek, Ridderstråle and Dahlborn1998). Capillary membranes were detected by the black staining of CA. Blackened areas were assigned a pixel value between 0 and 256, which increased with decreasing level of blackening, i.e. the higher the pixel value, the lower the CA activity. To avoid any bias in the analysis of the sections, the identities of the individual slides were not disclosed until all the sections had been analysed.
Calculations and statistical analyses
Total number of alveoli and capillaries are standardized to (and will in the following refer to) a number per 100 000 μm2. Alveolar and capillary density (μm2) refer to the area covered by alveoli or stained capillaries, respectively, per 100 000 μm2 mammary section. The alveolar diameter is expressed as the mean Feret diameter, which is the separation between pairs of parallel lines that just enclose an object. The mean Feret diameter was calculated as the mean of the two diameters in parallel to the x- and y-axis.
Statistical analyses were performed using the MIXED procedure in SAS® (SAS Institute Inc., Cary 27512 NC, USA). All data were tested for normal distribution. Milk yield, MBV, total number of alveoli, alveolar density, mean Feret alveolar diameter, total number of capillaries, capillary density, capillary-to-alveolar density, number of capillaries per alveolus, total capillary pixel value and pixel value per capillary were analysed for fixed effect of period (EL, ML1, BST, ML2, LL, D) and random effect of goat with period as the repeated measure. Covariance structures CS and AR(1) were fitted as described by Littell et al. (Reference Littell, Henry and Ammerman1998). Effect of BST treatment was evaluated by comparing BST period against ML1 and ML2 periods, using them as pre- and post-treatment controls, respectively. Pearson correlation coefficients were calculated for selected traits. All results are presented as LSMEANS±sem.
Results
Milk vein blood velocity and milk yield
Recordings of MBV in early lactation (EL) were unfortunately lost. MBV (Fig. 1) was fairly stable during the mid-lactation period with a small but non-significant increase during BST treatment, and then decreased (P<0·001) by approximately 50% from late lactation till the time of drying-off (D: 23·4±5·06 cm/sec). Milk yield (Fig. 1) peaked at 4·17±0·31 kg/d and like MBV declined during lactation (P<0·001). Milk yield decreased from EL to ML1 (P<0·05), increased to the highest levels in BST (P<0·001) and then decreased through LL to reach minimum levels in D (P<0·001). MBV and milk yield were positively correlated (r=0·47; P<0·05).
Fig. 1. Milk vein blood velocity (MBV, cm/sec; •) and milk yield (kg/d; ○) of experimental goats. EL, ML1, BST, ML2, LL and D: see footnote to Table 1.
Alveolar number, area and density
Total number of alveoli per 100 000 μm2 (Fig. 2, upper panel) increased by 54% from EL to D (P<0·01). It tended to increase in BST compared with EL (P=0·055) and was reduced after cessation of BST treatment in ML2 (P<0·05), whereafter it increased to reach the highest levels in D (P<0·001).
Fig. 2. Upper panel: Total number of alveoli per 100 000 μm2 mammary section (•) and area of the individual alveolus (μm2) (○). Lower panel: Total number of capillaries per 100 000 μm2 mammary section (•) and area of the individual capillary (○). EL, ML1, BST, LL and D: see footnote to Table 1.
Alveolar density (area covered by alveoli in a 100 000 μm2 section; Fig. 3, upper panel) decreased by 48% over lactation (P<0·001), and this could be ascribed to a decrease occurring in the last part of lactation (LL and D), whilst no significant changes occurred over the early and mid-lactation periods, including BST. Alveolar density and total number of alveoli were negatively correlated (r=−0·57; P<0·001).
Fig. 3. Upper panel: Alveolar (•) and capillary (○) density (area in μm2 covered by alveoli and capillaries per 100 000 μm2 mammary section, respectively). Lower panel: Number of capillaries per alveolus (•) and capillary-to-alveolar density ratio in mammary sections (○). EL, ML1, BST, LL and D: see footnote to Table 1.
Mean Feret alveolar diameter (results not shown) had a maximum value of 110·5±5·07 μm in EL and a minimum value of 77·6±4·78 μm in D, and was unaffected by BST treatment. It was negatively correlated with total number of alveoli per 100 000 μm2 (r=−0·61; P<0·001), but strongly and positively correlated with alveolar density (r=0·95; P<0·001).
Area of the individual alveolus (Fig. 2, upper panel) was stable through EL to ML1 and BST, increased to the highest levels in ML2 (1760±241 μm2; P<0·05) and decreased thereafter to reach the lowest level in D (441±253 μm2; P<0·001), which was 75% lower than in EL. Area of the individual alveolus was negatively correlated to total number of alveoli per 100 000 μm2 (r=−0·60; P<0·001), but positively correlated to alveolar density (r=0·73; P<0·001) and mean Feret alveolar diameter (r=0·67; P<0·001).
Alveolar density (r=0·48; P<0·01) and Feret alveolar diameter (r=0·49; P<0·01) were positively correlated to MBV.
Capillary density, number and area
Total number of capillaries per 100 000 μm2 (Fig. 2, lower panel) was highest in EL (63·6±5·8), decreased significantly from ML1 to BST (P<0·05), where it stabilized at around 45 throughout the rest of the experimental period at a value 32% lower in LL compared with EL (P<0·01).
Area of the individual capillary was smallest (Fig. 2, lower panel and Fig. 4) in EL (65±19 μm2); capillaries underwent dilatation from EL to ML1 (P<0·001), and dilated further in LL (189±19 μm2) (P<0·001), at which point they had almost tripled in size (291%) compared with EL. From LL to D they decreased in size again (P<0·01).
Fig. 4. Sections of azure blue counter stained lactating mammary tissue from the same goat, illustrating large alveoli surrounded by high numbers of heavily stained (high CA activity) small capillaries (arrow) in early lactation (left panel), and fewer but larger and less stained (lower CA activity) capillaries (arrow) surrounding smaller alveoli in late lactation (right panel).
Capillary density (Fig. 3, upper panel) was lowest in EL (3477±575 μm2), increased from EL to ML1 (P<0·001), remained at a constant level through ML1, BST and ML2, and subsequently increased to maximum levels (7619±536 μm2) in LL (P<0·01); more than two-fold higher (219%) compared with EL. A fairly strong positive correlation was found between capillary density and area of the individual capillary (r=0·76; P<0·001), whereas capillary density was negatively correlated to total number of capillaries (r=−0·48; P<0·001). MBV and milk yield were not correlated with either total number of capillaries or capillary density.
Total number of capillaries per alveolus (Fig. 3, lower panel) decreased across lactation by 61% from EL (21·24±2·45) to the lowest values in D (P<0·001) and was unaffected by BST treatment.
The capillary-to-alveolar density ratio (Fig. 3, lower panel) tripled from minimum values in EL to maximum levels in D (300%) (P<0·001) and was not affected by BST treatment. Thus high numbers of very small capillaries surrounded large alveoli in early lactation, and fewer but increasingly larger capillaries surrounded diminishing alveoli in the late lactating mammary gland.
Carbonic anhydrase activity
Total capillary pixel value per 100 000 μm2 (Fig. 5) increased 110% from lowest values in EL (65·9±7·9) to highest levels in D (138·2±10; P<0·001) (P<0·001) reflecting decreased CA activity in capillaries across the lactation period. Through ML1 to ML2 (including BST period) total capillary pixel values remained stable. A negative correlation was found between total capillary pixel value and milk yield (r=−0·39; P<0·05).
Fig. 5. Total capillary pixel value per 100 000 μm2 mammary section (•) and pixel value per individual capillary (○). EL, ML1, BST, LL and D: see footnote to Table 1.
Pixel value per individual capillary (Fig. 4) increased 146% from lowest levels in EL (1·4±0·36) to peak value (3·4±0·34) in LL (P<0·01). Pixel value per capillary was stable through ML1 to ML2 and was not affected by BST treatment. Pixel value per capillary was negatively correlated with milk yield (r=−0·39; P<0·05) and total number of capillaries (r=−0·66; P<0·001) but positively correlated with capillary density (r=0·47; P<0·01) and area of the individual capillary (r=0·82; P<0·001) but not correlated to MBV.
Discussion
Changes in the mammary alveoli and capillaries during lactation and BST treatment
Parenchymal volume in the mammary gland (Knight & Wilde, Reference Knight and Wilde1993) and number of mammary epithelial cells (Capuco et al. Reference Capuco, Wood, Baldwin, McLeod and Paape2001) have previously been reported to follow changes in milk production over the course of lactation. Our study supports this, as cross-sectional area of individual alveoli and the overall alveolar density in mammary sections decreased as lactation progressed. The gross morphological changes in the mammary capillary system were however more dramatic than the changes at the level of the alveoli, which is interesting since so little attention has been paid to functional changes in the mammary microvasculature in dairy animals, and the role in regulation of mammary function during lactation.
In the mouse it has been demonstrated by Matsumoto et al. (Reference Matsumoto, Nishinakagawa, Kurohmaru, Hayashi and Otsuka1992) that the capillary network in the developed mammary gland forms basket-like structures around the alveoli. When the mammary gland is lactating, the capillary endothelial cells develop numerous folds and microvillous processes, which is particularly prominent in thin-walled capillaries lying adjacent to secretory epithelial cells with well-developed basal infoldings. This establishes a close contact between the capillary endothelium and the secretory epithelium, whereby diffusion distances are reduced and the surface area available for blood-tissue exchange is increased. Along this line of thought, small diameter capillaries would therefore be expected to favour a more efficient blood-tissue exchange of nutrients and waste products, even if perfused by the same volume of blood, because diffusion distances obviously would be smaller than in large diameter capillaries. But in addition to that, a small capillary could potentially also create closer proximity to the mammary epithelial cell and shorten diffusion distances even further by fitting better into the epithelial cell infoldings compared with a larger capillary. The substantial reduction (61%) we observed in our study in number of capillaries per alveolus with progressing lactation, in combination with an almost threefold increase in capillary diameter, could therefore contribute to explaining why extraction of most nutrients across the goat mammary gland becomes less efficient from early to late lactation (Madsen et al. Reference Madsen, Nielsen and Nielsen2005) so that a higher mammary blood flow is required to sustain synthesis of 1 kg of milk in late compared with early lactation (Linzell, Reference Linzell, Larson and Smith1974).
To our knowledge, adaptations in the mammary capillary system during the course of lactation have not previously been reported in dairy or other animals. Adaptation to exercise and metabolic activity in skeletal muscle of rats and man involves an increase in numbers of capillaries per muscle fibre (Kano et al. Reference Kano, Sampei and Matsudo2004; Terzis et al. Reference Terzis, Spengos, Manta, Sarris and Georgiadis2008). Likewise in the goat mammary gland, it appears that high metabolic activity (milk yield) in EL is associated with a higher number of capillaries per alveolus compared with LL, when metabolic activity (milk yield) has decreased. However, in skeletal muscle capillary luminal area is also increased in response to increased metabolic activity, and this was not the case in the mammary gland, where the smallest capillary areas were found in EL, when the mammary gland was most metabolically active. The reason for this discrepancy is not known.
Increased capillary area (diameter) with progressing lactation should theoretically be associated with reduced peripheral vascular resistance, and hence would favour a high rate of mammary blood flow. However, MBV declined across lactation despite the substantial increase in capillary diameter, confirming that MBF generally follows changes in milk production (Prosser et al. Reference Prosser, Davis, Farr and Lacasse1996). In the rat, mammary capillary blood flow is discontinuous, and occurs in stop-and-go movements (Davis et al. Reference Davis, Farr, Prosser and Thompson1993), demonstrating that peripheral resistance and capillary perfusion in the murine mammary gland (and vascular beds of other tissues) is not determined simply by the number of capillaries, but can be regulated at pre- or post-capillary sites (Fujiwara & Uehara, Reference Fujiwara and Uehara1984; reviewed by Prosser et al. Reference Prosser, Davis, Farr and Lacasse1996). Our results support the view that the capillary in itself may not be the major site that determines MBF in goats. However, the capillary changes over the course of lactation may be important determinants for the efficiency of nutrient and waste product exchange across the capillary-mammary epithelial cell membrane barrier. This could in turn determine the metabolic capacity of the mammary gland with advancing lactation, in addition to the metabolic and cell turnover events occurring at the level of the epithelial cell (e.g. Capuco et al. Reference Capuco, Wood, Baldwin, McLeod and Paape2001).
Future studies are needed to elucidate the regulatory mechanisms underlying remodelling of the microvascular system during lactation, and implications for mammary epithelial cell function and overall lactation performance.
BST stimulation of milk yield in our study coincided with increases in number and density of alveoli in mammary sections, but without changing the size of the individual alveolus. BST has been reported to stimulate mammary epithelial cell renewal in dairy cows (Capuco et al. Reference Capuco, Wood, Baldwin, McLeod and Paape2001) and to increase mammary glandular weight in dairy goats without however increasing alveolar size or number of epithelial cells within alvoli (Boutinaud et al. Reference Boutinaud, Rousseau, Keisler and Jammes2003). BST therefore appears to exert its galactopoietic effect by expanding the mammary epithelial cell population through formation of new alveoli rather than by increasing the cell number and size of already existing alveoli. Our results do not suggest, however, that BST is a major player in regulation of microvascular remodelling within the mammary gland during lactation, since none of the capillary measures were impacted by the BST treatment. This interpretation must however be taken with some caution. In several previous studies BST has been reported to stimulate MBF as part of its galactopoietic effect (Davis & Collier, Reference Davis and Collier1985; McDowell et al. Reference McDowell, Leenanuruksa, Niumsup, Gooden, van der Walt and Smithard1988; Mepham et al. Reference Mepham, Lawrence, Peters and Hart1984). But for some unknown reason the treatment failed to do so significantly in the present experment, although we did observe a modest numerical increase in MBV.
Carbonic anhydrase and mammary blood flow regulation and function
Our study confirmed that CA activity in the lactating goat mammary gland was associated exclusively with the capillary endothelium. This is in contrast to the rat, where mammary CA has been found both within the mammary epithelial cell cytoplasma as well as in the capillary endothelium (Cvek, Reference Cvek1997). In dairy cows, the only studies on CA we are aware of have focused specifically on the secretory form of the enzyme (CA VI), which is located mainly in epithelial cells and secreted into milk, with highest concentrations immediately post partum, and rapidly declining over the first 70 d of lactation (Keitaro et al. Reference Keitaro, Nishita, Yamato, Sakamoto, Hagino, Katoh and Obara2003; Nishita et al. Reference Nishita, Tanaka, Wada, Murakami, Kasuya, Ichihara, Matsui and Asari2007). Although the goat has been a much used model for dairy cows in lactation physiology, species differences do exist, such as different distributions of CA within the mammary gland and expression of different isoform(s) of CA. The implications of these differences are not known.
The decrease in mammary capillary CA activity in goats with decreasing milk production shows that there is coordinated adaptation also of the capillary network function as mammary metabolic activity decreases during lactation. Whether it is the changes in mammary metabolic activity that govern the changes in capillary morphology and CA or the other way around is not known. Although CA is a capillary enzyme in the caprine mammary gland, it may be coupled less to regulation of MBF than to metabolic activity and milk synthesis in mammary epithelial cells, by facilitating removal of metabolic CO2 and buffering of intracellular pH. Indications in support of this view are 1) the positive correlation observed between milk production versus CA activity (Cvek et al. Reference Cvek, Ridderstråle and Dahlborn1998; present study) and mammary CO2 production (Nielsen et al. Reference Nielsen, Schleisner, Jakobsen and Andersen1995) and 2) the ability of the mammary gland to regulate conversion of CO2 to HCO3− over the course of lactation, whereby partial pressure of CO2 in blood leaving the mammary gland is kept constant (Nielsen et al. Reference Nielsen, Schleisner, Jakobsen and Andersen1995).
BST treatment did not impact CA activity in line with our previous observations (Nielsen et al. Reference Nielsen, Schleisner, Jakobsen and Andersen1995) where increased mammary metabolic activity (i.e. CO2 production) during BST treatment affected venous blood pCO2, but not HCO3−, indicating that the conversion of CO2 to HCO3− by CA activity is not regulated by BST.
In conclusion, the early lactating mammary gland was characterized by large lactating alveoli surrounded by high numbers of small capillaries with high CA activity. As lactation progressed, capillaries almost tripled in size and CA activity more than halved. BST did not affect capillary traits, but increased number of alveoli and appeared to be targeted mostly towards the mammary epithelial cell. Capillary diameter and CA are apparently not major determinants of MBF, but the changes in capillary diameter and CA activity with progressing lactation could contribute to reduce efficiency of nutrient and waste product exchange across the capillary-mammary epithelial cell membrane barrier. Future studies are needed to improve our limited knowledge of mammary microvascular remodelling, and to assess whether this could be as important a factor in regulation of lactation performance as the mammary epithelial cell itself.
The Danish Research Council for Technology and Production provided financial support. We wish to thank the following technicians for their skillful help: K Søberg, R Jensen, J Randrup and DS Jensen.
