Undercarboxylated osteocalcin, muscle strength and indices of bone health in older women

We investigated the association between undercarboxylated osteocalcin (ucOC) and lower-limb muscle strength in women over the age of 70years. The study also aims to confirm the association between bone turnover markers and heel ultrasound measures. A post-hoc analysis using data collected as part of a randomized placebo-controlled trial of vitamin D supplementation. An immunoassay was used to quantify total OC (tOC), with hydroxyapatite pre-treatment for ucOC. We determined associations of absolute and relative (ucOC/tOC; ucOC%) measures of ucOC with lower-limb muscle strength, heel ultrasound measures of speed of sound (SOS) and broadband ultrasound attenuation (BUA), bone turnover markers (BTMs; P1NP and CTx) and the acute phase protein alpha-1-antichymotrypsin (&alpha;-ACT). ucOC%, but not absolute ucOC concentration, was positively associated with hip flexor, hip abductor and quadriceps muscle strength (all p<0.05). ucOC% was negatively associated with &alpha;-ACT (β-coefficient=-0.24, p=0.02). tOC was positively associated with both P1NP and CTx (p<0.001). For each per unit increase in tOC (μg/L) there was a corresponding lower BUA, SOS and SI (β-coefficient = -0.28; -0.23 and -0.23, respectively; all p<0.04). In conclusion, ucOC% is positively associated with muscle strength and negatively associated with &alpha;-ACT. These data support a role for ucOC in musculoskeletal interactions in humans. Whilst tOC is associated with bone health, ucOC% and ucOC may also be linked to falls and fracture risk by influencing muscle function.

First isolation and structural determination of cyclic β-(1→2)-glucans from an alga, Chlorella pyrenoidosa

The aqueous extract of the edible green microalgae Chlorella pyrenoidosa is of interest because of its immunostimulatory activity. Some components in the extract have been identified previously, namely a unique type of arabinogalactan and a galactofuran. Further fractionation of this extract was accomplished by treating the aqueous solution of the fraction precipitated by addition of 1.5vol of 95% ethanol with cetyltrimethylammonium bromide. The residue obtained by concentration of the supernatant was fractionated further by anion-exchange chromatography and size-exclusion chromatography on Sephadex G-100. Two fractions from the latter column were retained, of which one was a starch-like alpha-(1-->4)-linked d-glucan with some alpha-(1-->6) branches, and the other contained a starch plus a mixture of beta-(1-->2)-d-glucans. ESI mass spectrometry was used to show that the mixture contained both cyclic and linear beta-(1-->2)-d-glucans in a cyclic:linear ratio of 64:36, based on intensities of mass spectral peaks. For the cyclic beta-(1-->2)-d-glucans, ring sizes ranged from 18 to 35 monosaccharides with the ring containing 21 glucose units (54% of the cyclic glucans) being greater than three times more abundant than the next most abundant component, the ring containing 22 glucose units (15%). No rings containing 20 glucose units were present. This is the first observation of cyclic beta-(1-->2)-d-glucans in algae, as far as we are aware. For the linear beta-(1-->2)-d-glucans, the component containing 20 glucoses was most abundant (35% of the linear glucans), while the component containing 21 glucose units was the next most abundant (17%). These relatively low-molecular-weight glucans had low immunostimulatory activity.

Brief intense interval exercise activates AMPK and p38 MAPK signaling and increases the expression of PGC-1α in human skeletal muscle

From a cell signaling perspective, short-duration intense muscular work is typically associated with resistance training and linked to pathways that stimulate growth. However, brief repeated sessions of sprint or high-intensity interval exercise induce rapid phenotypic changes that resemble traditional endurance training. We tested the hypothesis that an acute session of intense intermittent cycle exercise would activate signaling cascades linked to mitochondrialbiogenesis in human skeletal muscle. Biopsies (vastus lateralis) were obtained from six young men who performed four 30-s "all out" exercise bouts interspersed with 4 min of rest (<80 kJ total work). Phosphorylation of AMP-activated protein kinase (AMPK; subunits 1 and 2) and the p38 mitogen-activated protein kinase (MAPK) was higher (P  0.05) immediately after bout 4 vs. preexercise. Peroxisome proliferator-activated receptor- coactivator-1(PGC-1) mRNA was increased approximately twofold above rest after 3 h of recovery (P  0.05); however, PGC-1protein content was unchanged. In contrast, phosphorylation of protein kinase B/Akt (Thr³⁰⁸ and Ser⁴⁷³) tended to decrease, and downstream targets linked to hypertrophy (p70 ribosomal S6 kinase and 4E binding protein 1) were unchanged after exercise and recovery. We conclude that signaling through AMPK and p38 MAPK to PGC-1 may explain in part the metabolic remodeling induced by low-volume intense interval exercise, including mitochondrial biogenesis and an increased capacity for glucose and fatty acid oxidation.

Normal hypertrophy accompanied by phosphoryation and activation of AMP-activated protein kinase α1 following overload in LKB1 knockout mice

The activation of the AMP-activated protein kinase (AMPK) and inhibition of the mammalian target of rapamycin complex 1 (mTORC1) is hypothesized to underlie the fact that muscle growth following resistance exercise is decreased by concurrent endurance exercise. To directly test this hypothesis, the capacity for muscle growth was determined in mice lacking the primary upstream kinase for AMPK in skeletal muscle, LKB1. Following either 1 or 4 weeks of overload, there was no difference in muscle growth between the wild type (wt) and LKB1−/− mice (1 week: wt, 38.8 ± 7.75%; LKB1−/−, 27.8 ± 12.98%; 4 week: wt, 75.8 ± 15.2%; LKB1−/−, 85.0 ± 22.6%). In spite of the fact that the LKB1 had been knocked out in skeletal muscle, the phosphorylation and activity of the &alpha;1 isoform of AMPK were markedly increased in both the wt and the LKB1−/− mice. To identify the upstream kinase(s) responsible, we studied potential upstream kinases other than LKB1. The activity of both Ca2+–calmodulin-dependent protein kinase kinase &alpha;(CaMKK&alpha;) (5.05 ± 0.86-fold) and CaMKKβ (10.1 ± 2.59-fold) increased in the overloaded muscles, and this correlated with their increased expression. Phosphorylation of TAK-1 also increased 10-fold following overload in both the wt and LKB1 mice. Even though the &alpha;1 isoform of AMPK was activated by overload, there were no increases in expression of mitochondrial proteins or GLUT4, indicating that the &alpha;1 isoform is not involved in these metabolic adaptations. The phosphorylation of TSC2, an upstream regulator of the TORC1 pathway, at the AMPK site (Ser1345) was increased in response to overload, and this was not affected by LKB1 deficiency. Taken together, these data suggest that the &alpha;1 isoform of AMPK is preferentially activated in skeletal muscle following overload in the absence of metabolic adaptations, suggesting that this isoform might be important in the regulation of growth but not metabolism.

Reduced-order output feedback controllers for discrete-time systems

This article considers the stabilization by output feedback controllers for discrete-time systems. The controller can place all of the closed-loop poles within a specified disk D(-&alpha;, 1/β), centred at (-&alpha;,0) with radius 1/β, where | - &alpha;|  + 1/β < 1. The design method involves the decomposition of the system into two portions. The first portion comprises of all of the poles that are lying outside of the specified disk. A reduced-order model is constructed for this portion. The second portion comprises of all of the remaining poles of the system and is characterized by an H_∞-norm bound. The controller design is then accomplished by using H_∞-control theory. It is shown that, subject to the solvability of an algebraic Riccati equation, output feedback controllers can be systematically derived. The order of the controller is low, and can be as low as the number of the open-loop poles that are lying outside of the specified disk. A step-by-step design algorithm is provided. Numerical examples are given to illustrate the attractiveness of the design method.

Impaired activation of AMP-Kinase and fatty acid oxidation by globular adiponectin in cultured human skeletal muscle of obese type 2 diabetics

Adiponectin is an adipocyte-derived hormone associated with antidiabetic actions. In rodent skeletal muscle, globular adiponectin (gAD) activates AMP-kinase (AMPK) and stimulates fatty acid oxidation effects mediated through the adiponectin receptors, AdipoR1 and AdipoR2. In the present study, we examined the mRNA expression of adiponectin receptors and the effects of gAD on AMPK activity and fatty acid oxidation in skeletal muscle myotubes from lean, obese, and obese type 2 diabetic subjects. Myotubes from all groups expressed approximately 4.5-fold more AdipoR1 mRNA than AdipoR2, and obese subjects tended to have higher AdipoR1 expression (P = 0.052). In lean myotubes, gAD activates AMPK[alpha]1 and -[alpha]2 by increasing Thr172 phosphorylation, an effect associated with increased acetyl-coenzyme A carboxylase (ACC[beta]) Ser221 phosphorylation and enhanced rates of fatty acid oxidation, effects similar to those observed after pharmacological AMPK activation by 5-aminoimidazole-4-carboxamide riboside. In obese myotubes, the activation of AMPK signaling by gAD at low concentrations (0.1 [mu]g/ml) was blunted, but higher concentrations (0.5 [mu]g/ml) stimulated AMPK[alpha]1 and -[alpha]2 activities, AMPK and ACC[beta] phosphorylation, and fatty acid oxidation. In obese type 2 diabetic myotubes, high concentrations of gAD stimulated AMPK[alpha]1 activity and AMPK phosphorylation; however, ACC[beta] phosphorylation and fatty acid oxidation were unaffected. Reduced activation of AMPK signaling and fatty acid oxidation in obese and obese diabetic myotubes was not associated with reduced protein expression of AMPK[alpha] and ACC[beta] or the expression and activity of the upstream AMPK kinase, LKB1. These data suggest that reduced activation of AMPK by gAD in obese and obese type 2 diabetic subjects is not caused by reduced adiponectin receptor expression but that aspects downstream of the receptor may inhibit AMPK signaling.

Effects of endurance training status and sex difference on Na+, K+-pump mRNA expression, content and maximal activity in human skeletal muscle

Aim: This study investigated the effects of endurance training status and sex differences on skeletal muscle Na⁺,K⁺-pump mRNA expression, content and activity. Methods: Forty-five endurance-trained males (ETM), 11 recreationally active males (RAM), and nine recreationally active females (RAF) underwent a vastus lateralis muscle biopsy. Muscle was analysed for Na⁺,K⁺-pump &alpha;1, &alpha;2, &alpha;3, β1, β2 and β3 isoform mRNA expression (real-time reverse transcription-polymerase chain reaction), content ([³H]-ouabain-binding site) and maximal activity (3-O-methylfluorescein phosphatase, 3-O-MFPase). Results: ETM demonstrated lower &alpha;1, &alpha;3, β2 and β3 mRNA expression by 74%, 62%, 70% and 82%, respectively, than RAM (P < 0.04). In contrast, [³H]-ouabain binding and 3-O-MFPase activity were each higher in ETM than in RAM, by 16% (P < 0.03). RAM demonstrated a 230% and 364% higher &alpha;3 and b3 mRNA expression than RAF, respectively (P < 0.05), but no significant sex differences were found for &alpha;1, &alpha;2, β1 or β2 mRNA, [³H]-ouabain binding  or 3-O-MFPase activity. No significant correlation was found between years of endurance training and either [³H]-ouabain binding or 3-O-MFPase activity. Significant but weak correlations were found between the number of training hours per week and 3-O-MFPase activity (r = 0.31, P < 0.02) and between incremental exercise V O_2(peak) and both   [³H]-ouabain binding (r = 0.33, P < 0.01) and 3-O-MFPase activity (r = 0.28, P < 0.03). Conclusions: Isoform-specific differences in Na⁺,K⁺-pump mRNA expression were found with both training status and sex differences, but only training status influenced Na⁺,K⁺-pump content and maximal activity in human skeletal muscle.

Effects of cannabinoid receptors on skeletal muscle oxidative pathways

The endocannabinoids, a recently discovered endogenous, lipid derived, signaling system regulating energy metabolism, have effects on central and peripheral energy metabolism predominantly via the cannabinoid receptor type 1 (CB1). CB1 is expressed centrally in the hypothalamus and nucleus accumbens and peripherally in adipocytes and skeletal muscle. This study determined the effect of endocannabinoids on the expression of genes regulating energy metabolism in human skeletal muscle. Primary cultures of myotubes (lean and obese; n = 3/group) were treated with the cannabinoid receptor agonist, anandamide (AEA) (0.2 and 5 μM) and the CB1 specific antagonist AM251 (0.2 and 5 μM) separately and in combination for 24 h. The expression of mRNA for AMP-activated protein kinase (AMPK) alpha 1 (&alpha;1) and alpha 2 (&alpha;2), pyruvate dehydrogenase kinase 4 (PDK4) and peroxisome proliferator-activated receptor-γ co-activator-1&alpha; (PGC-1&alpha;) were determined using ‘Real Time’ RT-PCR. AMPK&alpha;1 mRNA increased in lean and obese myotubes in response to AM251 (P < 0.05). AEA inhibited the effect of AM251 on AMPK&alpha;1 mRNA levels in myotubes from lean and obese subjects (P < 0.05); the dose–response curve was shifted to the left in the obese. In response to AM251, irrespective of the presence of AEA, PDK4 expression was decreased in lean and obese myotubes (P < 0.05). Taken together these data suggest that endocannabinoids regulate pathways affecting skeletal muscle oxidation, effects particularly evident in myotubes from obese individuals.

Depressed Na+-K+-ATPase activity in skeletal muscle at fatigue is correlated with increased Na+-K+-ATPase mRNA expression following intense exercise

We investigated whether depressed muscle Na⁺-K⁺-ATPase activity with exercise reflected a loss of Na⁺-K⁺-ATPase units, the time course of its recovery postexercise, and whether this depressed activity was related to increased Na⁺-K⁺-ATPase isoform gene expression. Fifteen subjects performed fatiguing, knee extensor exercise at ~40% maximal work output per contraction. A vastus lateralis muscle biopsy was taken at rest, fatigue, 3 h, and 24 h postexercise and analyzed for maximal Na⁺-K⁺-ATPase activity via 3-O-methylfluorescein phosphatase (3-O-MFPase) activity, Na⁺-K⁺-ATPase content via [3H]ouabain binding sites, and Na⁺-K⁺-ATPase &alpha;1-, &alpha;2-, &alpha;3-, ß1-, ß2- and ß3-isoform mRNA expression by real-time RT-PCR. Exercise [352 (SD 267) s] did not affect [³H]ouabain binding sites but decreased 3-O-MFPase activity by 10.7 (SD 8)% (P < 0.05), which had recovered by 3 h postexercise, without further change at 24 h. Exercise elevated &alpha;1-isoform mRNA by 1.5-fold at fatigue (P < 0.05). This increase was inversely correlated with the percent change in 3-O-MFPase activity from rest to fatigue (%Δ3-O-MFPaserest-fatigue) (r = –0.60, P < 0.05). The average postexercise (fatigue, 3 h, 24 h) {alpha}1-isoform mRNA was increased 1.4-fold (P < 0.05) and approached a significant inverse correlation with %Δ3-O-MFPaserest-fatigue (r = –0.56, P = 0.08). Exercise elevated &alpha;2-isoform mRNA at fatigue 2.5-fold (P < 0.05), which was inversely correlated with %Δ3-O-MFPaserest-fatigue (r = –0.60, P = 0.05). The average postexercise &alpha;2-isoform mRNA was increased 2.2-fold (P < 0.05) and was inversely correlated with the %Δ3-O-MFPaserest-fatigue (r = –0.68, P < 0.05). Nonsignificant correlations were found between %Δ3-O-MFPaserest-fatigue and other isoforms. Thus acute exercise transiently decreased Na^+-K⁺-ATPase activity, which was correlated with increased Na⁺-K⁺-ATPase gene expression. This suggests a possible signal-transduction role for depressed muscle Na⁺-K⁺-ATPase activity with exercise.

Prolonged submaximal exercise induces isoform-specific Na+-K+-ATPase mRNA and protein responses in human skeletal muscle

This study investigated effects of prolonged submaximal exercise on Na⁺-K⁺-ATPase mRNA and protein expression, maximal activity, and content in human skeletal muscle. We also investigated the effects on mRNA expression of the transcription initiator gene, RNA polymerase II (RNAP II), and key genes involved in protein translation, eukaryotic initiation factor-4E (eIF-4E) and 4E-binding protein 1 (4E-BP1). Eleven subjects (6 men, 5 women) cycled at 75.5% (SD 4.8%) peak O₂ uptake and continued until fatigue. A vastus lateralis muscle biopsy was taken at rest, fatigue, and 3 and 24 h postexercise. We analyzed muscle for Na⁺-K⁺-ATPase &alpha;1, &alpha;2, &alpha;3, β1, β2, and β3, as well for RNAP II, eIF-4E, and 4E-BP1 mRNA expression by real-time RT-PCR and Na⁺-K⁺-ATPase isoform protein abundance using immunoblotting. Muscle homogenate maximal Na⁺-K⁺-ATPase activity was determined by 3-O-methylfluorescein phosphatase activity and Na⁺-K⁺-ATPase content by [³H]ouabain binding. Cycling to fatigue [54.5 (SD 20.6) min] immediately increased {alpha}3 (P = 0.044) and {beta}2 mRNA (P = 0.042) by 2.2- and 1.9-fold, respectively, whereas {alpha}1 mRNA was elevated by 2.0-fold at 24 h postexercise (P = 0.036). A significant time main effect was found for &alpha;3 protein abundance (P = 0.046). Exercise transiently depressed maximal Na⁺-K⁺-ATPase activity (P = 0.004), but Na⁺-K⁺-ATPase content was unaltered throughout recovery. Exercise immediately increased RNAP II mRNA by 2.6-fold (P = 0.011) but had no effect on eIF-4E and 4E-BP1 mRNA. Thus a single bout of prolonged submaximal exercise induced isoform-specific Na⁺-K⁺-ATPase responses, increasing &alpha;1, &alpha;3, and β2 mRNA but only &alpha;3 protein expression. Exercise also increased mRNA expression of RNAP II, a gene initiating transcription, but not of eIF-4E and 4E-BP1, key genes initiating protein translation.

Chemical characterisation of the oligosaccharides in hooded seal (Cystophora cristata) and Australian fur seal (Arctocephalus pusillus doriferus) milk

Carbohydrates were extracted from hooded seal milk, Crystophora cristata (family Phocidae). Free oligosaccharides were separated by gel filtration and then purified by ion exchange chromatography, gel filtration and preparative thin layer or paper chromatography and their structures determined by ¹H-NMR. The hooded seal milk was found to contain inositol and at least nine oligosaccharides, most of which had lacto-N-neotetraose or lacto-N-neohexaose as core units, similar to those in milk of other species of Carnivora such as bears (Ursidae). Their structures were as follows: Gal(β1-4)Glc (lactose); Fuc(&alpha;1-2)Gal(β1-4)Glc (2′-fucosyllactose); Gal(β1-4)GlcNAc(β1-3)Gal(β1-4)Glc (lacto-N-neotetraose); Fuc(&alpha;1-2)Gal(β1-4)GlcNAc(β1-3)Gal(β1-4)Glc (lacto-N-fucopentaose IV); Gal(β1-4)GlcNAc(β1-3)[Gal(β1-4)GlcNAc(β1-6)]Gal(1-4)Glc (lacto-N-neohexaose); Fuc(&alpha;1-2)Gal(β1-4)GlcNAc(β1-3)[Gal(β1-4)GlcNAc(β1-6)]Gal(β1-4)Glc (monofucosyl lacto-N-neohexaose a); Gal(β1-4)GlcNAc(β1-3)[Fuc(&alpha;1-2)Gal(β1-4)GlcNAc(β1-6)]Gal(β1-4)Glc (monofucosyl lacto-N-neohexaose b); Fuc(&alpha;1-2)Gal(β1-4)GlcNAc(β1-3)[Fuc(&alpha;1-2)Gal(β1-4)GlcNAc(β1-6)]Gal(β1-4)Glc (difucosyl lacto-N-neohexaose); Gal(β1-4)GlcNAc(β1-3)Gal(β1-4)GlcNAc(β1-3)Gal(β1-4)Glc (para lacto-N-neohexaose); Fuc(&alpha;1-2)Gal(β1-4)GlcNAc(β1-3)Gal(β1-4)GlcNAc(β1-3)Gal(β1-4)Glc (monofucosyl para lacto-N-neohexaose). Milk of the Australian fur seal, Arctophalus pusillus doriferus (family Otariidae) contained inositol but no lactose or free oligosaccharides. These results, therefore, support the hypothesis that the milk of otariids, unlike that of phocids, contains no free reducing saccharides.

Mutually clock-controlled feedback shift registers provide resistance to algebraic attacks

Algebraic attacks have been applied to several types of clock-controlled stream ciphers. However, to date there are no such attacks in the literature on mutually clock-controlled ciphers. In this paper, we present a preliminary step in this direction by giving the first algebraic analysis of mutually clock-controlled feedback shift register stream ciphers: the bilateral stop-and-go generator, A5/1, Alpha 1 and the MICKEY cipher. We show that, if there are no regularly clocked shift registers included in the system, mutually clock-controlled feedback shift register ciphers appear to be highly resistant to algebraic attacks. As a demonstration of the weakness inherent in the presence of a regularly clocked shift register, we present a simple algebraic attack on Alpha 1 based on only 29 keystream bits.

Antioxidant treatment with N-acetylcysteine regulates mammalian skeletal muscle Na+-K+-ATPase ∝ gene expression during repeated contractions

Exercise increases Na+–K+ pump isoform gene expression and elevates muscle reactive oxygen species (ROS). We investigated whether enhanced ROS scavenging induced with the antioxidant N-acetylcysteine (NAC) blunted the increase in Na+–K+ pump mRNA during repeated contractions in human and rat muscle. In experiment 1, well-trained subjects received saline or NAC intravenously prior to and during 45 min cycling. Vastus lateralis muscle biopsies were taken pre-infusion and following exercise. In experiment 2, isolated rat extensor digitorum longus muscles were pre-incubated without or with 10 mm NAC and then rested or stimulated electrically at 60 Hz for 90 s. After 3 h recovery, muscles were frozen. In both experiments, the muscles were analysed for Na+–K+ pump &alpha;1, &alpha;2, &alpha;3, β1, β2 and β3 mRNA. In experiment 1, exercise increased &alpha;2 mRNA by 1.0-fold (P = 0.03), but &alpha;2 mRNA was reduced by 0.40-fold with NAC (P = 0.03). Exercise increased &alpha;3, β1 and β2 mRNA by 2.0- to 3.4-fold (P < 0.05), but these were not affected by NAC (P > 0.32). Neither exercise nor NAC altered &alpha;1 or β3 mRNA (P > 0.31). In experiment 2, electrical stimulation increased &alpha;1, &alpha;2 and &alpha;3 mRNA by 2.3- to 17.4-fold (P < 0.05), but these changes were abolished by NAC (P > 0.07). Electrical stimulation almost completely reduced β1 mRNA but only in the presence of NAC (P < 0.01). Neither electrical stimulation nor NAC altered β2 or β3 mRNA (P > 0.09). In conclusion, NAC attenuated the increase in Na+–K+ pump &alpha;2 mRNA with exercise in human muscle and all &alpha; isoforms with electrical stimulation in rat muscle. This indicates a regulatory role for ROS in Na+–K+ pump &alpha; isoform mRNA in mammalian muscle during repeated contractions.

Bovine Muc1 is a highly polymorphic gene encoding an extensively glycosylated mucin that binds bacteria

The bovine Muc1 protein is synthesized by mammary epithelial cells and shed into milk as an integral component of the milk fat globule membrane; however, the structure and functions of this mucin, particularly in relation to lactation, are poorly defined. The objectives of this investigation were to investigate the Muc1 gene and protein structures in the context of lactation and to test the hypothesis that Muc1 has a role in innate immune defense. Polymerase chain reaction analysis of genomic DNA from 630 cattle revealed extensive polymorphism in the variable number of tandem repeats (VNTR) in the bovine Muc1 gene. Nine allelic
variants spanning 7 to 23 VNTR units, each encoding 20 AA, were identified. Three alleles, containing 11, 14, and 16 VNTR units, respectively, were predominant. In addition, a polymorphism in one of the VNTR units has the potential to introduce a unique site for N-linked glycosylation. Statistical analysis indicated weak associations between the VNTR alleles and milk protein and fat percentages in a progeny-tested population of Holstein-Friesian dairy cattle. No association with somatic cell count could be demonstrated. Bovine Muc1 was purified from milk fat globule membranes and characterized. The protein was highly glycosylated, primarily with O-linked sialylated T-antigen [Neu5Ac(&alpha;2–3)-Gal(β1–3)-GalNAc&alpha;1] and, to a lesser extent, with N-linked oligosaccharides, which together accounted for approximately 60% of the apparent mass of Muc1. Purified bovine Muc1 directly bound fluorescently labeled Escherichia coli BioParticles (Invitrogen, Mount Waverley, Australia) and inhibited their binding to bovine mammary epithelial cells grown in vitro.